267 131 654MB
English Pages 288 [289] Year 2018
MASS CUSTOMIZATION AND DESIGN DEMOCRATIZATION EDITED BY BRANKO KOLAREVIC & JOSÉ PINTO DUARTE
00 Front Matter.indd 1
2018-11-04 5:12 PM
First published 2019 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 52 Vanderbilt Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business © 2019 selection and editorial matter, Branko Kolarevic and José Pinto Duarte; individual chapters, the contributors The right of Branko Kolarevic and José Pinto Duarte to be identified as the authors of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Publisher’s Note This book has been prepared from camera-ready copy provided by the editors. Designed and typeset in Bell Gothic by Branko Kolarevic. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record has been requested for this book ISBN: 978-0-8153-6060-5 (hbk) ISBN: 978-0-8153-6061-2 (pbk) ISBN: 978-1-351-11786-9 (ebk)
00 Front Matter.indd 2
2018-11-04 5:12 PM
CONTENTS ACKNOWLEDGMENTS................................................................................................................... v PREFACE...................................................................................................................................... vii 1 FROM MASSIVE TO MASS CUSTOMIZATION AND DESIGN DEMOCRATIZATION Branko Kolarevic and José Pinto Duarte............................................................................................ 1 2 CUSTOMERING: THE NEXT STAGE IN THE SHIFT TO MASS CUSTOMIZATION B. Joseph Pine II ............................................................................................................................ 13 3 CREATING A SUSTAINABLE MASS CUSTOMIZATION BUSINESS MODEL Frank Piller.................................................................................................................................... 29 4 A CRAFT IT YOURSELF FUTURE Virginia San Fratello and Ronald Rael.............................................................................................. 41 5 MASSIVE CUSTOMIZATION Marc Fornes.................................................................................................................................... 55 6 CONTINUING TOWARD EXTREME MASS PRODUCTION Greg Lynn ...................................................................................................................................... 69 7 DEMOCRATIC DESIGN AND DAILY OBJECTS Philippe Starck .............................................................................................................................. 83 8 LEARNING AS IT GROWS: THE HUMANIZATION OF OBJECTS Assa Ashuach ................................................................................................................................. 87 9 CUSTOMIZING PROCESS, DEMOCRATIZING DESIGN Fabio Gramazio and Matthias Kohler ............................................................................................ 101 10 METADESIGNING CUSTOMIZABLE HOUSES Branko Kolarevic .......................................................................................................................... 117 11 CUSTOMIZING MASS HOUSING: TOWARD A FORMALIZED APPROACH José Pinto Duarte ......................................................................................................................... 129 12 INTERPLAY OF DESIGN, TECHNOLOGY, MANUFACTURING, AND BUSINESS Karl Daubmann ............................................................................................................................ 143 13 THE MODERN MODULAR: THE MASS CUSTOMIZATION OF THE SINGLE FAMILY HOME Joseph Tanney............................................................................................................................... 157 14 MASS PREFABRICATION: INVESTIGATING THE RELATIONSHIP BETWEEN PREFABRICATION AND MASS CUSTOMIZATION IN ARCHITECTURE Ryan E. Smith .............................................................................................................................. 175 15 FUTURE ADAPTIVE BUILDING: MASS-CUSTOMIZED HOUSING FOR AN AGING POPULATION John L. Brown .............................................................................................................................. 185 16 DEMOCRATIZING CREATIVITY Christopher Sharples .................................................................................................................... 197 17 SENTIENCE AND THE SPECIFICITIES OF CITIES Tom Verebes ................................................................................................................................. 215 18 CITY SCIENCE: TOWARD A NEW PROCESS FOR CREATING HIGH-PERFORMANCE, ENTREPRENEURIAL COMMUNITIES Kent Larson ................................................................................................................................. 231 19 THE PHYSICAL IMPLICATIONS OF A MASS-CUSTOMIZATION ECONOMY Thomas Fisher .............................................................................................................................. 249 AUTHOR BIOGRAPHIES............................................................................................................. 262 PHOTO CREDITS......................................................................................................................... 273 PROJECT CREDITS..................................................................................................................... 275 INDEX......................................................................................................................................... 276
00 Front Matter.indd 3
2018-11-04 5:12 PM
iv
00 Front Matter.indd 4
2018-11-04 5:12 PM
ACKNOWLEDGMENTS
We offer the deepest gratitude to the contributors to this book – their remarkable ideas and innovative, pioneering projects were the main inspiration for both this endeavor and the eponymous symposium that preceded it. In May of 2017, we gathered in Philadelphia to explore over two days the past, present, and future of mass customization and design democratization in business, industry, design, architecture, and urbanism. The symposium presentations, enriched by the ensuing discussions and email exchanges, have been further refined and captured in this book. Not all of the authors were able to join us at the symposium, but have kindly agreed to contribute to the book. The “Mass Customization and Design Democratization” symposium was organized by the Penn State Stuckeman School and held in Philadelphia on May 12 and 13, 2017. We are most grateful to the symposium sponsors: DIRTT and the Stuckeman Center for Design Computing (SCDC), the Center for Research in Design and Innovation (CRDI), Stuckeman School, the College of Arts and Architecture, School of Engineering Design, Technology, and Professional Programs (SEDTAPP) at the Pennsylvania State University; without their support, the symposium and this book would not have been possible. Initial formative research was supported with grants from the University Research Grants Council
(URGC) and with funding through the Chair for Integrated Design and Laboratory for Integrative Design (LID) at the University of Calgary. We are profoundly appreciative of our home institutions, the University of Calgary and the Pennsylvania State University; both are among top research universities in Canada and the United States, respectively. They continue to provide a stimulating and supportive context for our teaching, research, and scholarly work. Finally, and most importantly, we want to thank our students and our colleagues for the stimulating conversations about the subject of this book. In particular, we want to acknowledge students and staff who were involved with the ideas, content, and organization of the symposium, respectively: Eduardo Costa and Flávio Craveiro, and Jamie Perryman, Barbara Cutler, Janejira Kalsmith, and Scott Tucker. We are grateful to Francisca Ford at Taylor and Francis for her enthusiastic support of this project and Trudy Varciana for patiently navigating the editing of the manuscript and production of the book. Special thanks are due to Susan Dunsmore for her painstaking copyediting of the manuscript. At a personal level, we want to thank our families for putting up with us as we spent many evenings and weekends working on this project. The book is dedicated to them.
Branko Kolarevic and José Pinto Duarte
v
00 Front Matter.indd 5
2018-11-04 5:12 PM
vi
00 Front Matter.indd 6
2018-11-04 5:12 PM
PREFACE This book stems from our shared interest in mass customization in design and architecture and in the digital technologies that should make customizable designs broadly accessible. We both started with algorithmic design three decades ago, which opened unprecedented opportunities for exploring shape, form, and space in architecture, in a fluid, dynamic fashion. We both then engaged enthusiastically in exploring digital fabrication and the possibilities that opened up for producing economically highly differentiated components in a variety of materials. These two technologies – algorithmic design and digital fabrication – made customization of components possible on a massive scale. Repetition was no longer an economic and technological necessity. The consequence was that massive customization is a reality today in the building industry, as demonstrated in numerous projects completed over the past two decades. Our embrace of the technologies that made massive customization possible – to which interactive websites were added later – led us to believe that the design and making of products, from daily objects to cars and homes, would be made accessible to the masses shortly after those technologies came into use in the early 2000s. Such design customization by the masses was to be a social and cultural corollary to the massive customization that emerged in the context of the building industry and, naturally, architecture. This mass customization – and the design democratization it implied – have not happened as quickly and as broadly as we had anticipated, which raised the questions as to why that was the case. In an attempt to get some answers, we decided to bring together a number of prominent thinkers, designers, and researchers and convene a symposium on “Mass Customization and Design Democratization.” Our idea was to discuss the state of mass customization in design, architecture, and urbanism (i.e. at a variety of scales), its impact – or, rather, a lack of it – in contemporary culture, and whether design democratization has any future as a viable cultural and social construct. The symposium took place in May of 2017 in Philadelphia,
under cloudy skies, as an apt metaphor for our undertaking. The idea for the symposium that was born out of such initial examination of the current state of mass customization quickly evolved into discussions about design democratization as an overarching theme within which any discussion of mass customization should be situated. As we engaged these themes more deeply, we discovered that little had been written about them in the context of design and architecture. Thus, one of our goals was to address this perceived lack of adequate coverage in the literature. Our main motivation, however, was to explore what mass customization meant broadly in the context of the contemporary economy from a design perspective, and how it is manifested. That eponymous symposium led, thus, to this book. Its contents largely mirror the presentations and the discussions that took place over two days in Philadelphia. The thought leaders from the worlds of business and industry and the well-known, highly regarded designers, architects, technologists, and theoreticians we brought together became the contributors to this book, with the aim of providing informed views related to mass customization and design democratization. Their chapters offer a diverse and divergent set of ideas about mass customization in design and architecture and design democratization in our contemporary culture. The projects discussed in the pages that follow provide snapshots of emerging ideas, grounded in actual practices already taking place. We, as editors and contributors to this volume, obviously believe that mass customization and design democratization matter, but, as readers will notice, they mean different things to different people. This variety of perspectives should not be seen as negative; the meanings of both mass customization and design democratization are indeed multiple, intertwined, and sometimes contradictory. We didn’t seek a simple, coherent and succinct definition of these still evolving concepts. Their true meaning should become clear in the future, as design customization becomes an embedded cultural notion.
vii
00 Front Matter.indd 7
2018-11-04 5:12 PM
viii
00 Front Matter.indd 8
2018-11-04 5:12 PM
1 FROM MASSIVE TO MASS CUSTOMIZATION AND DESIGN DEMOCRATIZATION
BRANKO KOLAREVIC AND JOSÉ PINTO DUARTE 1
01 Kolarevic Duarte.indd 1
2018-09-08 11:35 PM
In this introductory chapter we address the current state of mass customization in architecture and design and the implicit promise of design democratization that has yet to find full social and cultural resonance. We discuss what mass(ive) customization means in architecture and the building industry and how it can be manifested in its most commoditized sector – suburban housing. We present parametric design and digital fabrication as the technological foundation for mass customization: through using parametric definition of the overall geometry and spatial layout, unique, customized houses could be designed. We use the term parametric design in a broad sense to include topological and dimensional variation, as well as variation of other shape attributes, such as material, color, and texture. Depending on the chosen systems for structure, enclosure, and partitions, their components could be fabricated with different degrees of automation by relying on digital fabrication and robotic assembly. We then examine what kind of mass customization is possible, useful, or desirable, and the conditions that could make design democratization a promising social and cultural construct. We explore whether design democratization, enabled by mass customization, is a viable proposition in the contemporary context, from technological, economical, cultural, and social perspectives. MASSIVE CUSTOMIZATION Various digital design and production technologies introduced in the late 1990s opened up unprecedented opportunities for architects to engage complexity and variability in formal articulation and the material realization of buildings.1 In the conceptual realm, digitally-driven, parametric design processes characterized by an inherent capacity for continuous transformation of three-dimensional forms gave rise to new architectonic possibilities. Digital fabrication technologies, such as computer numerically controlled (CNC) cutting and milling and three-dimensional (3D) printing, allowed the production and construction of uniquely shaped
components that until then had been very difficult and expensive to produce. More importantly, the new digitally driven processes of design and fabrication offered a direct link from the designed geometry seen on the computer screen to the material production of components using CNC machines. The consequence of these technological advances is that building projects today are not only born digitally, but they are also realized digitally through “file-to-factory” processes of CNC fabrication technologies. The processes of describing and constructing a design can be now more direct and more complex because the information can be extracted, exchanged, and utilized with far greater facility and speed. Thanks to parametric design and digital fabrication, it is now possible to mass-produce non-standard, highly differentiated products, from shoes and tableware to furniture and building components, and now even houses. Variety no longer compromises the efficiency and economy of production. Furthermore, parametric definitions of products’ geometry are made accessible via interactive websites to anyone, who could then design their own, unique versions of the product. (That at least was the optimistic vision; the reality is more nuanced, as discussed elsewhere in this book and later in this chapter.) The digital technologies of design and production certainly left their mark on the architecture of the early twenty-first century. The sparse geometries of twentieth-century Modernism were in large part driven by Fordist paradigms of industrial manufacturing, imbuing the building production with the logics of standardization, prefabrication, and on-site installation. The rationalities of manufacturing dictated geometric simplicity over complexity and repetitive use of low-cost, mass-produced components. But these rigidities of production are no longer necessary, as digitally-controlled machinery can fabricate unique, variably shaped components at a cost that is no longer prohibitively expensive. Variety, in other words, no longer compromises the efficiency and economy of production.
2
01 Kolarevic Duarte.indd 2
2018-09-08 11:35 PM
1.1 Today’s consumers have the option of designing their own, highly customized yet industrially produced products (Reebok, 2014).
The ability to mass-produce one-off, highly differentiated building components with the same facility as standardized parts introduced the notion of massive customization2 into building design and production – it is almost as easy and costeffective for a CNC milling machine to produce 1000 unique objects as it is to produce 1000 identical ones. In buildings, individual components could be customized to allow for optimal variance in response to differing local conditions, such as uniquely shaped and sized structural components that address different structural loads in the most optimal way, variable window shapes and sizes that correspond to differences in orientation and available views, and even functionally-graded material compositions of building parts in response to specific environmental conditions. The digitallydriven production processes introduced a different logic of seriality in architecture, one that is based on local variation and serial differentiation. TOWARD MASS CUSTOMIZATION Massive customization of components in the building industry should not be confused with mass customization. Massive customization is a contemporary technological capacity that
is afforded by advances in digital design and production technologies, as described earlier. Mass customization, on the other hand, is a contemporary business and marketing capacity that is aimed at meeting the unique needs of individual customers; it requires social and cultural conditioning so that customers – buyers of products, whether they are furniture, cars, or even houses – can demand and expect something unique, as opposed to a standard product that was mass-produced. In massive customization, we can randomly vary the characteristics of products; in mass customization, we need to link such variation to features of the context, be it physical features of the environment or social and individual features of product users, to obtain a product with higher performance. Mass customization, defined by Joseph Pine as the mass production of individually customized goods and services, offered the promise of a tremendous increase in variety and customization without a corresponding increase in costs.3 Over the past two decades, almost every segment of the economy, from services and consumer products to industrial production, has been affected by mass customization.4 For example, today’s consumers could create their own unique, non-standard, industrially produced shoes (figure 1.1) and jackets,
3
01 Kolarevic Duarte.indd 3
2018-09-08 11:35 PM
1.2 The “3D Configurator” by Blu Homes lets any visitor to the website choose and customize a selected house design.
1.3 Housebrand has conceived a complete design and delivery system (FAB House) based on modular house design that is highly customizable.
choosing materials, colors, and finishes as they please, at the same or marginally higher cost as the standard products made by the same manufacturer. As we entered the twenty-first century, there was an expectation that the introduction of parametric design and digital fabrication into a variety of design fields would lead to a proliferation of websites that would enable everyone to interactively manipulate the shapes and forms of a wide range of products, whose geometry could be changed on the fly and, when fixed, digitally fabricated. The customized designs could be produced relatively quickly by the manufacturer and then shipped to the customer,
or even better, inexpensively manufactured using locally available digital production facilities. The principal idea was that good design would become globally accessible, with production eventually taking place locally. Contrary to the expectations, that didn’t happen; the technology is there – parametric design, digital fabrication, integrated design and production processes, interactive websites – but what is missing is a broader cultural shift in society in how products are acquired. Turning customers into co-designers is an educational, cultural, and social challenge. It could be that the rather large majority of customers have no interest whatsoever in becoming designers.
4
01 Kolarevic Duarte.indd 4
2018-09-08 11:35 PM
1.4 House designs by Resolution: 4 Architecture can be customized by creating different relationships between pre-defined modules.
CUSTOMIZING HOUSES Websites by architects and/or builders currently exist that enable anyone to choose the house design they like, explore available options, and then customize it (within some carefully imposed limits). The “Design Your Own Home” website5 by Toll Brothers offers hundreds of stylistically very different house designs; each comes with several options that do not affect the overall geometry of the house, and the possibility to choose various material finishes. Blu Homes’ website6 features several different designs of prefabricated houses (figure 1.2), with various customization options; the process starts with “preliminary delivery assessment,” followed by
“conceptual design” (online or in person) and “code and zoning research” as the final step. The “FAB” house (figure 1.3) by Housebrand from Calgary is a complete design and delivery system based on modular design7 that provides a considerable degree of freedom in how each house “shell” could be configured as a one-, two-, three-, or four-bedroom house, with considerable cost and time savings. The website by Resolution: 4 Architecture8 from New York offers “modern modular” homes (figure 1.4), with “predefined typologies … formed from a series of standard modules, minimizing cost of production and maximizing possible combinations available for the consumer.” They all offer ways to customize predefined house designs, but none of them offer 5
01 Kolarevic Duarte.indd 5
2018-09-08 11:35 PM
1.5 Bernard Cache, Objectiles, 1997. The online interface for designing objectiles, nonstandard objects, features interactive manipulation of the parametrically defined geometry.
interactive manipulation of the house’s overall geometry or its internal layout. Mass customization is a particularly suitable production paradigm for the housing sector of the building industry, since houses (and buildings in general) are mostly one-off, highly customized products. It also offers a promise that a truly “customized” house – with a unique geometry, i.e. shape and form – could eventually become available to a broader segment of society. The technologies to deliver economically mass-produced, yet highly customized houses are there: parametric design, digital fabrication, interactive websites for design, visualization, evaluation, and estimating (and automatic generation of production and assembly data). The challenges for wider adoption of house design that can be interactively customized are not technological; as we argue below, they are largely social, i.e. cultural. A mass-customizable house could be parametrically defined, interactively designed (via a website or an app), and digitally prefabricated,
using file-to-factory processes. While this is technologically possible and economically attainable, it is nevertheless socially and culturally questionable. After all, how many of us have designed our own shoes or jackets? How many of us would dare design our own parametrically-variable car (if such an option were available)? How many of us would have the confidence that such a car would have good performance characteristics and be esthetically pleasing? Finally, how many of us are prepared to become designers instead of mere customers? DEMOCRATIZING DESIGN The implied “democratization” of design through mass customization raises additional questions such as the authorship of design and the functional and esthetic quality of products (shoes, tableware, furniture, houses) designed by non-designers. Bernard Cache was one of the first designers to “democratize” design by making his parametrically defined furniture and paneling designs publicly
6
01 Kolarevic Duarte.indd 6
2018-09-08 11:35 PM
1.6 Gramazio and Kohler, mTable, 2002. Various mTable designs created by customers using a mobile phone app.
accessible over the Internet in 1997. Cache’s objectiles, as he referred to his designs, were conceived as non-standard objects, procedurally calculated in modeling software and industrially produced using CNC machines. It was the modification of parameters of design that allowed the manufacture of different shapes in the same series, thus making the mass customization, i.e. the industrial production of unique objects, possible.9 Anyone could change online the parameter values that control the geometry of the objectiles simply by manipulating the sliders (figure 1.5) and could immediately see the effects of the changes.10 Fabio Gramazio and Matthias Kohler from Zurich went a step further in 2002 with their mTable parametrically variable table design (with holes) “that customers can co-design.”11 They created an
interactive application for mobile phones so that customers could easily specify the size, dimensions, material, and color of the table (figure 1.6). Next, by placing “deformation points” on the underside of the table and by “pressing” them, the customers could create holes with very thin edges. If satisfied with the design, the customer then transmits the parameters that define the table as a simple series of numbers to a website (at www.mshape.com), where the designed table is rendered in high resolution. A final step is the placement of the production order, with the table fabricated by a CNC milling machine. Gramazio and Kohler weren’t just interested in interactive, customizable design. They wanted to “examine the consequences of customer interaction when designing non-standard products.” The project raised interesting questions, such as the extent of 7
01 Kolarevic Duarte.indd 7
2018-09-08 11:35 PM
responsibility that a customer was able and willing to assume in making certain design decisions, and more importantly, who ultimately is the author of the final design – the designer of the parametric system, or the “customer” who chose the parameter values for the design? Such democratization of the design process has interesting implications for the building industry, especially in its most commoditized sector – the commercial provision of suburban housing, as discussed earlier. It is possible that we will soon see the emergence of websites where customers could fully customize the overall spatial layout and appearance of the chosen house design, selecting, for example, the size of the living room, location of the entry door, etc., down to the number of mullions in the windows (let alone the materials and finishes on the houses, which are already on offer). CHALLENGES TO MASS CUSTOMIZATION As mentioned earlier, the difference between massive customization and mass customization relies on the ability to link product features to contextual features. In other words, to randomly generate design variations is not enough; it is also necessary that such variation stems from careful interpretation of the design context. This includes the physical, technological, cultural, and social context, as much as the individual context of the user. The aim is to obtain products with increased functional performance and high customer satisfaction. There are several steps toward this goal, each with its own challenges. Some steps are connected to the development of the customization system, whereas others are connected to its use in the generation of a customized product. In developing the system, it is necessary to define the precise scope of the system. In the realm of housing, this means choosing an appropriate house type to encode.12 Habraken defined type as the result of three vectors: functional organization, building system, and decorative motives.13 Thus, the system’s developer will have to decide not just whether to encode multi-family or single-family
housing, but also to capture dwelling layouts that reflect the existing ways of living and to choose technological means that can be available to materialize them. To encode a house type or any other object type, one needs to foresee dimensional variation, but also topological variation. This can be achieved by recourse to rule-based design systems, such as shape grammars, as parametric design in the strict sense can only account for dimensional, color, and texture variation. Rule-based design systems are more difficult to implement as they may run into difficult shape recognition problems.14 Nonetheless, it is possible to convert a rule-based design system into several parametric design models, thereby facilitating computer implementation. To choose an appropriate building technology to materialize design solutions means deciding on the level of automation. This might signify avoiding fully automated systems, due to current technological limitations, the costs involved, or ethical issues. For instance, in some cases, it might be better to engage the community in the construction of its own houses, providing them with work at the same time, and using local materials and construction techniques, than simply adopting “alien” technology. Another important challenge in the development of the customization system is to set the balance between what is defined by the designer and what is left for the product user to decide. Increasing the amount of choice available to the user increases the opportunity for mass customization, but it also may represent a risk to design quality. In systems that are too open, it may be difficult or even impossible to foresee or detect combinations of variable values that may result in bad solutions from a functional or esthetic viewpoint. These systems will have to rely on the judgment of users who are design experts, to avoid such solutions. In addition, some users might feel overwhelmed with the burden of choice and not be interested in becoming the designers of their houses.15
8
01 Kolarevic Duarte.indd 8
2018-09-08 11:35 PM
Regarding the application of the system in the generation of customized products, there is the challenge of optimizing functional performance. This means developing sophisticated simulation systems that can assess and compare the performance of different alternative solutions from the environmental and structural viewpoints and find those configurations that have better performance. This has an impact on the degree of automation of the design system with different scenarios being possible, from fully automation to situations where the user of the systems makes the decisions with the system providing feedback by indicating the performance of the corresponding solution. There is also the challenge concerning the user interface. If the decision to empower the user with a high degree of freedom is made, one still has to find appropriate ways for the user, potentially a non-designer, to understand what is at stake and make informed decisions. This requires developing effective ways of communicating the meaning of more complex design variables to the user and then showing the impact of selected variable values. This becomes more important as one moves away from customization limited to the selection of surface finishings and embraces the selection of variables concerning topological and dimensional variation. Finally, even if one finds ways to overcome all the challenges above, one still has to address the problem of the user being ready or willing to make design decisions. It is perfectly acceptable if the user prefers to hire a designer to conceive the product, an option that might be available only to a wealthy few. The solution here might be to configure the system to work as a digital designer, with the ability to engage in a dialogue with the users, helping them to understand the needs, providing advice, and offering solutions. In the chapters that follow you will find answers to these challenges by contributors from different fields, including design, engineering, and business. They have developed significant work in the field, either in research or in practice, which required them to develop theoretical models or practical applications to tackle these challenges.
THE PAST, PRESENT, AND FUTURE OF MASS CUSTOMIZATION After Alvin Toffler introduced the idea of mass customization in his seminal book, Third Wave,16 and Stan Davis defined the concept in Future Perfect,17 Joseph Pine, a business scholar, systematized the field and is considered by many as the founding father of mass customization.18 In Chapter 2, he writes about the next stage for companies in the process of adopting customization, in which they should replace marketing by customering. In marketing, companies target groups of users considered to have similar consumer profiles, whereas in customering, customers are individuals with unique and distinct needs and desires. Frank Piller is another business strategist considered a pioneer in the field, like Pine. After introducing the concept of mass customization, in Chapter 3, he describes the challenges companies are facing when shifting to this new business model and details the capabilities they need to build for a successful shift. Virginia San Fratello, Ronald Rael, and Marc Fornes approach the topic from a design perspective focused on making. San Fratello and Rael argue, in Chapter 4, that by using digital fabrication it is possible to endow mass production with qualities usually associated with handcrafted processes. This includes giving some of the control of the making process back to the designer, overcoming the separation of design from making that emerged after industrialized processes. In Chapter 5, Fornes, on the other hand, is interested in exploring the potential of digital fabrication to go beyond the boundaries of individual choice sought in mass customization, to achieve what he dubs massive customization. In mass production, industrialized processes used repetition to produce affordable products in large quantities. In massive customization, digital fabrication is used to produce differentiated products on a large scale. The challenge then is how to design when the outcome is a high number of different objects. Mass customization of housing is the topic of contributions by Karl Daubmann, Joseph Tanney, Ryan Smith, and John Brown. In Chapter
9
01 Kolarevic Duarte.indd 9
2018-09-08 11:35 PM
12, Daubmann describes vividly how mass customization results from an interplay between design, manufacturing, and business supported by technology. He bases his account on his experience at Blu Homes and explains how technology can integrate the processes of designing and making houses, overcoming current limitations to the provision of houses, including customization. He completes his description with an account of the technological and organizational difficulties. In Chapter 13, Tanney describes the methodology developed by his office, Res4, to enable the mass customization of the single-family home. The methodology takes advantage of existing methods of prefabrication to develop a modular strategy that permits different spatial configurations in response to user needs. In Chapter 14, Ryan Smith examines the relationship between prefabrication and mass customization to explain that the former, rather than conflicting with the latter, can represent a viable way to achieve it in progressive continuity from existing industrialized systems. The underlying idea is that the housing industry might already have the seeds to attain design democratization through mass customization based on combinatorial variation of standard components. In Chapter 15, Brown focuses on the opportunity provided by mass customization to satisfy the needs of an aging population and describes the concept of Future Adaptive Building (FAB). Stemming from the Ecological Theory of Aging, based on the relation between older people and their environment, FAB’s goal is to increase their functional, emotional, and physical resilience by adapting living environments to the changing needs of people as they age. Mass personalization goes beyond massive and mass customization. While both strategies try to produce unique products with near mass production efficiency, “mass personalization aims at a market segment of one while mass customization at a market segment of few.”19 Greg Lynn, Assa Ashuach, Fabio Gramazio, Matthias Kohler, and Philippe Starck focus on mass personalization. Delving into his experience with the car industry and then with the design of the Embryological
House, Lynn argues in Chapter 6 that the architects’ dream to bring mass customization to the building industry might not be matched by consumers’ desire for unique products. Instead, he offers an alternative that he describes as extreme mass production, which entails the production of more standardized types at a greater frequency. Ashuach has focused his design practice on the development of methodologies and tools to support a cooperative partnership between designers and consumers of products. He has developed a technology called UCODO™ that enables end-users to add a new level of information to a digitallydesigned object, which, as he argues in Chapter 8, will change how the products are designed, made, and distributed. In Chapter 9, Gramazio and Kohler describe how access to computers at ETH during their architectural education and the realization that programming could enable them to go beyond drafting and modeling set them on the path to customization. The ability of programming to encode design instructions and generate varied outputs gained a new dimension when it was later matched by the ability to materialize the output of design programs through digital fabrication. They describe how they used these two abilities to create mTable, a pioneering project that brought them closer to the principles of mass customization. They use this project to explain basic principles, where we stand today and future challenges. In Chapter 7, Starck explains how his goal has always been to democratize design by making quality design increasingly affordable to a wider sector of society. He describes how TOG (All Creators TOGether), a company he co-founded in Brazil, has tried to solve the paradox of mass-producing high-quality yet unique products that customers could personalize using simple strategies. Christopher Sharples, Tom Verebes, Kent Larson, and Thomas Fisher write about mass customization at the city scale. In Chapter 16, Sharples criticizes the current separation between design and construction and the conservatism of the building industry, which constitutes a barrier to the adoption of methodologies and technologies
10
01 Kolarevic Duarte.indd 10
2018-09-08 11:35 PM
that could overcome such dualism and lead to greater design democratization. He uses his own projects to explain how digital technologies of design and fabrication could support this democratization, freeing workers from the boredom of repetitive tasks. Verebes argues, in Chapter 17, that mass production in the twentieth century led to the uniformity of cities, and that the same mistake is occurring today in East Asia on an even larger scale, when we already have the means to pursue alternative pathways. He goes on to show what these alternatives are in urbanism by describing three projects by Ocean Consultancy Network. The result is greater organizational and spatial complexity of cities. Larson defends a similar argument in Chapter 18, but introduces additional issues related to an aging population, sustainability challenges, and creativity as a productive force in the modern economy anchored in cities. He proposes a fourstep methodology toward a new approach to urban design and city planning, with technology being an important aspect in all of them and illustrates these steps with concrete projects. In Chapter 19, Fisher addresses the implications of a mass customization economy for the physical structure of architecture and cities. He points out that in the new economy consumers can also become producers of goods and services, who no longer consider it important to own the property they occupy and who work close to home or at home, avoiding commuting. He sees this trend, to a certain extent, as a return to the past when people were nomadic, owned little, and shared much.
seems to be the norm today, to dimensional variation, which could be seen as the next frontier. We also argue that even greater customization could be achieved by enabling topological variation. This poses increased technical and organizational difficulties but would permit spatial variation and lead to increased customer satisfaction. Within a framework based on rulebased design systems that supports dimensional and topological variation alike, research-oriented designers can use abstraction to infer, first, specific and then generic design systems from existing designs; practice-oriented designers can use concretization to conceive new specific systems and then end-users may use these to create customized designs. We strongly believe that mass customization is a promising direction for the housing sector of the building industry – and architectural design as a key part of it – that could make good design accessible to more people than is presently the case, and that more people will take an active role in the design of their homes, thus becoming co-designers instead of mere customers. This could lead to more heterogeneous cities in which difference and variety are embraced, instead of endless repetition of the same that unfortunately today characterizes many new urban (and suburban) environments around the world. We see design democratization as an empowering, enriching social and cultural construct that is enabled by mass customization as its technological and business foundation.
CONCLUSION Acknowledging the limited mass customization achieved in practice so far, we would argue for greater customization, particularly in the housing industry, and advocate for design democratization as its ultimate social and cultural outcome. The technological capacity already exists for the industries to go beyond color and material variation (i.e. cosmetic customization), which
11
01 Kolarevic Duarte.indd 11
2018-09-08 11:35 PM
NOTES 1 Kolarevic, Branko (ed.), Architecture in the Digital Age: Design and Manufacturing, London: Spon Press, 2003. 2 The term “massive customization” was first used by Marc Fornes in his public presentations about the work done by his firm THEVERYMANY. See Chapter 5 in this book for more information. 3 Pine, Joseph B. II, Mass Customization: The New Frontier in Business Competition, Boston: Harvard Business School Press, 1993. 4 Ibid. 5 See www.designyourownhome.com. 6 See www.bluhomes.com. 7 See www.housebrand.ca for more information. 8 See www.re4a.com. 9 For Bernard Cache, “objects are no longer designed but calculated,” allowing the design of complex forms and laying “the foundation for a nonstandard mode of production.” See Cache, Bernard, Earth Moves: The Furnishing of Territories, Cambridge, MA: MIT Press, 1995. 10 The website www.objectile.com is no longer functional. 11 Gramazio, Fabio, and Kohler, Matthias, “Towards a Digital Materiality,” in Kolarevic, B. and Klinger, K. (eds.), Manufacturing Material Effects, London: Routledge, 2008, pp. 103–118. 12 Duarte, José Pinto, “Customizing Mass Housing: A Discursive Grammar for Siza’s Malagueira Houses,” PhD
dissertation, Massachusetts Institute of Technology, 2001, pp. 19–20. 13 Habraken, John, “Type as a Social Agreement,” paper presented at Asian Congress of Architects, Seoul, Korea, 1988. 14 Piller, Frank, Schubert, Petra, Koch, Michael, and Möslein, Kathrin, “Overcoming Mass Confusion: Collaborative Customer Co-Design in Online Communities,” Journal of Computer-Mediated Communication 10, vol. 10, no. 4 (2005); available at: https://doi.org/10.1111/j.1083-6101.2005.tb00271.x. 15 Barros, Mário, Duarte, José P., and Chaparro, Brubo, “Digital Thonet: An Automated System for the Generation and Analysis of Custom-Made Chairs,” in Proceedings of the 29th Conference on Education in Computer Aided Architectural Design in Europe, Ljubljana: Faculty of Architecture, University of Ljubljana, 2011, p. 523. 16 Toffler, Alvin, The Third Wave, New York: Bantam Books, 1980. 17 Davis, Stanley M., Future Perfect, Reading, MA: Addison-Wesley, 1987. 18 Pine, op. cit. 19 Kumar, A., “From Mass Customization to Mass Personalization: A Strategic Transformation,” International Journal of Flexible Manufacturing Systems, vol. 19, no. 4 (2008), p. 536.
12
01 Kolarevic Duarte.indd 12
2018-09-08 11:35 PM
2 CUSTOMERING: THE NEXT STAGE IN THE SHIFT TO MASS CUSTOMIZATION
B. JOSEPH PINE II 13
02 Pine.indd 13
2018-09-08 11:37 PM
There are no markets, only customers. Markets, as commonly conceived of in business, simply do not exist. They are a convenient fiction for companies that do not want to treat customers as the unique individuals they truly are. A customer is not part of a market, nor a segment, nor a niche, nor a generation, nor a persona, nor any other agglomeration of anonymous buying units of indeterminate size. A customer is a living, breathing individual person – or, if you sell to other businesses, an active, corporeal individual enterprise. We must therefore ascend to the proposition that all customers are unique; undeniably, unremittingly, unalterably unique. And that means we must stop “marketing,” and instead put into practice the principles of customering. I first thought along these lines way back when I was a strategic planner at IBM. My job for the AS/400 computer system (announced in June 1988) was to bring customers and business partners into the development process of the system. That experience made me realize a simple truth: every customer is unique. Every single customer that came through our Early External Involvement program, as we called it, wanted to use the system in different ways, connect it to different hardware, load on different software, achieve different objectives, and so on. Each customer was unequivocally unique. But we at IBM did not take that into account as we designed the AS/400 for this large, homogeneous, general-purpose minicomputer market that simply did not exist. MASS CUSTOMIZATION So when I shifted into strategic planning, I went in search for how we could resolve this dilemma, and that is when I read Stan Davis’s book, Future Perfect.1 When I read the chapter on “Mass Customizing,” it was like the heavens opened up and the angels sang, for Stan described exactly what was going on in my world of minicomputers
and what we could do about it. As he wrote in coining the term “mass customization”: What is the final step, the unitary building block for the market whole in the new economy? It is the “individual” customer. Units of one, whether a consumer or a corporation. But these are not the single consumers and firms who were reached with customized goods and services in the limited, preindustrial markets. Rather, in the same way that segments and niches are reached on a mass basis, individuals can now be reached on a basis that is simultaneously mass and customized.2 Individual customers! That’s the building block of the entire economy, the literal unit of analysis for all of business. We again need, therefore, to treat these individual customers as the unique individuals that they truly are. Mass customizing enables us to meet the co-equal imperatives of giving customers exactly what they want (customization) at a price they’re willing to pay (mass). I worked diligently to get that insight into the plans and strategies of the AS/400 division of IBM, after which the company sent me to MIT to get my Master’s degree in the Management of Technology. When it came time to write my thesis, I took the opportunity to begin writing a booklength treatment of this subject, resulting in my 1993 book, Mass Customization: The New Frontier in Business Competition.3 In writing it, I defined mass customization as “variety and customization through flexibility and quick responsiveness.”4 Which was wrong! My research had led to me to a wealth of data that showed how the old ways of the system of mass production were falling apart as more and more companies across a wide breadth of industries seemed to be abandoning the “any color car you want as long as it was black” mentality by greatly increasing the variety of their output. But eventually I realized that variety was managers’ last-ditch effort to preserve the old paradigm. Variety is still producing to inventory in
14
02 Pine.indd 14
2018-09-08 11:37 PM
2.1 Market development. Source: Davis, Future Perfect.
the hope that a customer would come along and want it. True customization only results from ondemand operations, where the final product is not produced until the customer indicates this is what he or she wants. And so, let me give you this much better, certainly more concise and simple, definition of mass customization: efficiently serving customers uniquely. That is what the system of mass customization is all about, for you must meet the co-equal imperatives of mass – high-volume, lowcost, efficient operations – and customization – serving that individual, living, breathing customer. It is about giving each customer exactly what he or she wants at a price he or she is willing to pay. MARKET DEVELOPMENT Let us go back to the very first framework on mass customization that Stan Davis published in his 1987 book, where he wrote about how markets develop. In the beginning were local markets, where everything was developed locally, was bought and sold locally, was produced locally. Then with the advent of the Industrial Revolution, business was able for the first time to create a truly mass market. It was, of course, Henry Ford who put it all together with the system of mass production, where everyone was part of the single mass market. That huge innovation – which was not about just marketing per se, but encompassed how the entire corporation was organized – diffused through industry after industry as mass production
became not only the paradigm but the paragon of business competition. Make no mistake, it was mass production that made the United States the number one economic power in the world. But then along came a gentleman by the name of Alfred Sloan, who put together General Motors. Sloan segmented that mass market and treated different segments differently. General Motors organized around five brands for five market segments (his not-as-famous phrase was “a car for every purse and purpose”), putting in separate factories, distribution outlets, marketing functions, and so forth for each segment, from Cadillac at the top, then to Buick and Oldsmobile, and on down to Pontiac and Chevrolet at the bottom. And that focus on market segmentation diffused through industry after industry. Over time, however, those segments kept getting smaller and smaller and smaller and eventually became mere niches. We coined a phrase at IBM that “the niches are the market” because segments of any size were nowhere to be found; niches provided all the activity. What Stan Davis foresaw was that if you take this progression to its logical conclusion, you get what he called “mass customized markets” (figure 2.1),5 or markets of one, where every customer is his or her own market. He realized that technology was bringing down the cost of customization and would eventually get to the point where companies could mass customize unique offerings to each individual customer. Here we finally reach that unitary building block of all economies, where the individual customer can get exactly what he or she wants
15
02 Pine.indd 15
2018-09-08 11:37 PM
at a price he or she is willing to pay. Mass customization breaks down the barriers standing in the way of addressing customers as the unique individuals they are. That is the way it was back in the days of local markets when buyers and sellers would come physically together to exchange wares for money (or sometimes barter for each other’s goods). But every customer was unique, and if the supplier didn’t have exactly what you wanted, for additional time and money you could work individually with your cobbler, your tailor, your toolmaker, or whoever it might be to get a product that better met your personal requirements than their standard, off-the-shelf (or in-the-cart) offerings – but then you had to pay far more for the privilege. But guess what? Even in the height of mass production, every customer was unique. But people subsumed their individual desires to buy standardized products that they otherwise could not afford at all. A car that was whatever color you wanted and matched whatever other specifications you personally thought important mattered little if you could not afford for the manufacturer to customize it to your particular wants and needs. But that doesn’t mean those personal wants, needs, and desires weren’t there, that people actually were all the same. That is just how enterprises necessarily had to treat them as they embraced the system of mass production and went down the old learning curve of cost versus volume. Eventually, it seemed that enterprises forgot that customers were unique and began to believe that there actually was a mass market of pretty much identical customers, and Madison Avenue did everything in its power to reify this myth. The same was true as producers introduced more variety through market segmentation. Still every customer was unique, and yet enterprises treated them as part of an amorphous agglomeration of anonymous buying units, just of lesser size than with mass markets. And the same with niche markets – they were smaller yet, but as far as any enterprise was concerned, every
customer within any particular niche was the same. But that wasn’t true. For every customer is unique. Always has been. Always will be. Undeniably, unremittingly, unalterably unique. CUSTOMERING That is why we must stop marketing and start customering. Just as one day in the past the term “marketing” did not fall as trippingly off the tongue as it does now, so too the term “customering” will take some getting used to. (Just saying the word “customering” out loud every time you encounter it in this chapter will go a long way to getting your tongue accustomed to saying it and your brain wrapped around it.) Behavior change starts with vocabulary change, however, and it is important to embrace this new language, or the organization will slip back into old practices that first developed in the days – and within the mindset – of mass marketing and the system of mass production. So what does customering mean? The American Marketing Association (AMA) defines marketing as “the activity, set of institutions, and processes for creating, communicating, delivering, and exchanging offerings that have value for customers, clients, partners, and society at large.”6 If we accept that definition, we can readily construct this one:
customering is the activity, set of institutions, and processes for creating, communicating, delivering, and exchanging customized offerings that have unique value for individual customers, clients, and partners while creating economic value for society at large. You can see the emphasis of customering on creating customer-unique value through (mass) customized offerings.7 As the AMA recognizes for marketing, customering is not just about “talking” to individual customers, recognizing them as individuals and understanding their
16
02 Pine.indd 16
2018-09-08 11:37 PM
2.2 Ways of interacting with customers.
unique requirements, so you can use that information to develop new economic offerings that meet those needs somewhere down the line. Customering is about creating a customized offering that meets the individual wants, needs, and desires of this particular customer at this moment in time. Think of this as a more straightforward definition: interacting directly with individual customers to create customer-unique value within them. Meeting that definition requires an aboutface from all of the core practices that have grown up around the concept of marketing for the past century, ever since the halcyon days of mass marketing that accompanied the shift to mass production. In this chapter I will compare and contrast the key marketing practices with their corresponding vital customering practices, using “From/To” statements. All told, the practices comprising these “To” statements provide an overall picture of customering. FROM PUSH/TO PULL Marketing, as it is commonly carried out in most enterprises, is push, push, push. It takes an offering that is generally already hanging on a rack, lying in a lot, or sitting on a shelf somewhere and seeks to push what it already made onto a customer, often without caring whether the offering really meets the needs of whoever buys it. There is a reason people dread the push-y salesperson, whether in a store, on a lot, or over the phone!
Customering is the exact opposite; it’s all about pull. You start with the customer – not the product – and pull intelligence about the wants, needs, and desires of this individual customer before you determine what to offer to him or her. In fact, the ideal is to do it before you even create the final offering for the customer. In fact, there are four different ways that you can interact with customers (figure 2.2), depending on whether you push or pull your offerings – the economic value you create for customers in exchange for their money – and push or pull the intelligence about those offerings and/or customers. Marketing once again is all about push: pushing what information you have on your own offerings out to customers through all the means at marketing’s disposal, most notably (as well as most annoyingly) through advertising. And then, again, it pushes the offerings already produced out to the customer as well. Even in industries such as automobiles where tremendous customization capabilities exist, salespeople do whatever it takes to push you to buy something that is already produced and sitting there on the lot taking up space and adding up depreciation. Customer Relationship Management, or CRM, seeks to pull intelligence from customers about what they need – often customizing messages to do so – but still largely pushes the same offerings (or variations thereof) out to those customers. CRM can be a giant leap forward for enterprises that previously never talked with or even interacted directly with individual customers, but at the end of
17
02 Pine.indd 17
2018-09-08 11:37 PM
the day, it rarely results in changing operations to make offerings customized to those customers. Vendor Relationship Management, or VRM, is a still uncommon but intriguing concept advocated by Doc Searls in his book, The Intention Economy,8 specifically designed as the exact opposite of CRM. Here, customers push back to enterprises their intent to buy particular offerings, including the specifications they are looking for individually or as a group, and then any company can respond with offers that meet those specifications. Searls calls it a “personal RFP”9 (request for proposal), where customers are in control, pushing out their needs and pulling from those companies connected to their personal VRM systems the offerings that meet those needs. Finally, then, there is customering. This way of interacting is pull-pull: pulling intelligence from individual customers – where in its ideal form companies only ever ask a customer a question when the answer will benefit that particular customer – and then pulling the offerings most exactly responding to what the company learns back through its own operations (and that of its suppliers) to meet an individual customer’s needs. In terms of interacting with customers, customering realizes the concept of “1:1 marketing” that Don Peppers and Martha Rogers first wrote about in The One to One Future.10 A company embraces 1:1 marketing, and satisfies this practice of customering, if it enters into dialogue with customers to learn about their individual wants, needs, and desires (pulling intelligence) and then customizes what it offers to those same customers in response to what it learns (pulling offerings). FROM TARGETING/TO ENTICING … Perhaps the most pernicious practice of marketing is targeting customers that the company thinks – based on what little it knows (usually mere demographics) – might be in the market for one of its offerings, and then bombarding them with messages to buy, buy, buy that offering. Think of
the online ads that follow you around from site to site because you once upon a time looked at a particular product and, in all probability, decided you didn’t want it. You have been targeted. And no one likes a bullseye on their back. Customering, on the other hand, means enticing your customers to want to talk with you, to want to share intelligence about themselves and their desires, to be open to offerings you have that would in fact fulfill those desires. Part of the allure, of course, is your reputation and promise to keep what you learn private and secure, but companies also must surround their offerings with experiences that draw potential customers in, engage them in the process of discovery, and help them see the possibilities. … AND TO CULTIVATING LEARNING RELATIONSHIPS Customering goes further, however, by recognizing that interacting with customers is not a oneperiod game – in fact, it is not a game at all, but a relationship that grows and deepens over time. You must cultivate a learning relationship with each individual customer, predicated on a virtuous cycle (figure 2.3) where you understand that every interaction – not just transaction, but all the ways you can interact with customers physically and virtually – provides an opportunity to learn. And the more you learn about this individual, living, breathing customer, then the better you can customize to this particular customer. (And it may not just be offerings you customize; you also can customize how you represent your offerings, how you interact with customers, the content you provide to them, the recommendations, and so forth.) The better you customize to this customer, then, the more that customer is going to benefit – customization has value, and experiencing that value means clearly understanding the benefits. And the more he or she benefits, the more willing he or she will be to interact with you again, and every interaction is an opportunity to learn.
18
02 Pine.indd 18
2018-09-08 11:37 PM
2.3 The learning relationship.
This virtuous cycle exists around each individual customer and can become virtually impregnable to competitors who do not have the capabilities for learning and customizing. And even when competitors embrace customering themselves and entice customers into entering into such learning relationships, it is still an incredibly powerful competitive advantage. Why? Because customers teach you so much about themselves that even if they were to go to somebody else with the selfsame capabilities, they would have to teach that company all over again what you already know today. When you cultivate learning relationships – with the provisos that you do not try to then gouge your customers or fail to innovate new offerings that they will desire – your customers will come back to you time and time and time again, whenever they need what your capabilities provide. FROM PRODUCT-CENTRIC/TO CUSTOMERCENTRIC Marketing tends to be exceedingly productcentric, which follows directly from its practices of targeting customers so it can push its offerings on them. Product-centric companies are like the snake-oil salesmen of old; they always seem to have exactly what you need for whatever ails you. Customering must be customer-centric. To understand what this means, let us start with proper definitions. If you look up the word
“customer” in The Oxford English Dictionary you will see that it is “One who frequents any place of sale for the sake of purchasing; one who customarily purchases from a particular tradesman; a buyer, a purchaser.”11 In other words, the customer is the one who pays you money. And “centric” is quite simply “That is in or at the centre, central.”12 Putting them together yields this clear definition: being customer-centric means placing the one who pays you money at the center of everything you do. Now there is one word in this definition that I want to emphasize, and the one word I want to emphasize is the word “one”: being customercentric means placing the one who pays you money at the center of everything you do. Again, a customer is not a market, nor a segment, nor a niche, nor a generation, nor a persona, nor any other agglomeration of anonymous buying units of indeterminate size. A customer is a living, breathing individual person – or, if you sell to other businesses, an active, corporeal individual enterprise. And you must place that customer – each and every customer – at the center of everything you do. That doesn’t mean you oblige every possible customer out there rather than choose which customers to serve. It doesn’t mean you pander to customers and just do exactly what they say they want rather than discern what they truly need. It doesn’t mean you try to lock customers into a learning relationship that enables you to dominate them to gain the lion’s share of the relationship’s economic benefits. It doesn’t mean you need to follow their every whim rather than think on their behalf about what they are going to need in the future. It doesn’t mean you ignore new technologies and industry disruptors rather than continue to innovate at least as much, as fast, and as effectively as the ecosystem to which you belong. Rather, being customer-centric means you recognize the uniqueness of each customer you choose to serve so that you can in fact efficiently serve customers uniquely. It means you endeavor
19
02 Pine.indd 19
2018-09-08 11:37 PM
to provide what they need, as the individuals they are, now and into the future. It means you innovate for, with, and on behalf of your individual customers. For only then will you place the one who pays you money at the center of everything you do. FROM FINISHED PRODUCTS/TO MODULAR CAPABILITIES The product centricity of marketing arises from the fact that marketers want to generate demand for finished products, already in inventory, complete, and ready to be pushed out to customers. Even if an enterprise’s offerings are intangible services for which there is no inventory, marketers generally predefine the offerings, with no chance of changing them because of the needs of any particular customer. When practicing customering, you need to shift from finished products to modular capabilities. Modularity is the key to mass customizing your offerings, to efficiently serving customers uniquely, to giving every customer exactly what he or she wants but with low-cost, high-volume, efficient operations. When you think of modularity, think of LEGO building bricks. What can you build with LEGOs? Anything you want! Why? Because you have a large number of bricks of different sizes, different shapes, different colors, plus a simple and elegant linkage system for snapping them together. That is modularity – modules plus linkage system. If you have any LEGO bricks handy, take six 2 x 4 bricks. (If you have to imagine them, they have eight studs on the top and underneath have three tubes that enable the studs of another brick to couple, or link, the two elements together.) Examine them and play with them, putting them together in different ways, and see if you can figure out the answer to this question: How many different ways can you put these six LEGO bricks together? (We will ignore color.) If you are an engineer type, the phrase “six factorial” probably jumped immediately into mind. For those unfamiliar with the math concept, this is 6 x 5 x 4 x 3 x 2 x 1 = 720 different
combinations, which answers the question “How many different ways can you stack six colored LEGO bricks immediately on top of each other?”. But that is nowhere close to the numbers of ways that six such LEGO bricks can connect; there are so many different ways to connect the bricks that the actual number that LEGO Group itself published is 102,981,500!13 (By way of comparison, astronomers estimate there are approximately 100 million black holes in the Milky Way galaxy.) In fact, however, even that astronomical number is way too low. After LEGO published this number, a couple of Danish mathematicians tried to figure out how the company arrived at that figure, and in analyzing the situation realized that LEGO only counted the 6-high combinations, those with one of the 2 x 4 bricks as the base, a second one placed on top of that in various different ways, the third somewhere atop that, and so forth. But you can put the six bricks together with multiple ones at the same level (such as a base of two or three bricks), and so can also make 5-high, 4-high, 3-high, and 2-high combinations. (There are no 1-high combinations; that is just the six unconnected bricks lying on the floor.) So adding up the total of all of these combinations, the mathematicians arrived at the even more astronomical number of 915,103,765.14 And that is with just six bricks! It shows the power of modularity, for it doesn’t take very many modules or a very robust linkage system before you, too, could have millions or even billions of different, customized offerings for each and every one of your potential customers. But by no means do you want to expose your customers to that many choices! That is the easiest mistake mass customizers make, overwhelming their customers with so many choices in a way that makes their eyes glaze over. Understand that customers don’t want choice. They just want exactly what they want. You therefore need some sort of design tool, a technology that enables you to draw out of each customer what his or her individual needs are – even if he or she doesn’t know what that is or can’t articulate it – and specifies the
20
02 Pine.indd 20
2018-09-08 11:37 PM
offering meeting those needs. The design tool lets customers visually and ideally viscerally experience what the final offering would be, while simultaneously ensuring you know it is something you can produce efficiently. Once you have that perfect order for this particular customer at this moment in time, you can get that information back into operations to produce that offering by linking together exactly the right set of pre-defined modules in your portfolio of capabilities. It is again modularity that makes all this efficient, that enables you to mass customize your economic offerings. Even if you can’t modularize your offerings, however, you could still modularize your processes, where you perform different activities for different customers. And even if you can’t do that, you can modularize information so you present yourself and your offerings differently to different customers. All information today can be represented in digital technology, and anything you can digitize, you can customize. Digital technology presents the ultimate in modularity, for once something exists in bits of zeroes and ones, you can instantaneously change a zero into a one and vice versa at no cost. And then you can combine those bits into modular bytes, and those bytes into modular strings, and those strings into modules of information of increasing meaning and value, precisely because of this nested architecture of modularity. Many industries today have essentially information-based offerings, including news, music, entertainment and media of all stripes, games across all digital platforms, telecommunications, banking, and insurance. And so all such industries have gained greater levels of customization following on every stride in digitization. (Few companies in these industries, unfortunately, have embraced all the practices of customering, still treating many of their customer interactions through the lens of marketing.) As with other forms of modularity, too, even if you can’t digitize your product, you may be able to digitize your processes to make them more efficient and easier to link together physical modules on
behalf of individual customers, or information about your products, including how you present your offerings to each customer on your website or app, say, or in your design tool. FROM PRODUCING ECONOMIC WASTE/TO DOING ONLY AND EXACTLY There is another benefit to this whole schema of interaction that modularity enables: eliminating economic waste. With marketing’s dependence on the system of mass production, customers must buy standardized offerings produced for mass markets, or (with increasing variety) segmented markets and even niche markets. Until you get down to markets of one – what Stan Davis originally called “masscustomized markets” – enterprises always want to sell something they have already produced. Sure, sometimes that off-the-shelf product really does meet your needs perfectly fine. In the vast majority of cases, however, buying standard means it doesn’t meet your needs across all the dimensions that are important to you, sometimes by a little, often by a lot. Whenever a company puts something it has already produced on sale, it is in effect trying to pay customers to take something that doesn’t fully meet their needs, since they certainly didn’t want it at the normal price. In the extreme case of the apparel industry, I am told that over half the products produced aren’t purchased at full price, and the lion’s share of those are never bought at all, ending up cut to pieces in a landfill somewhere. That is pure economic waste. In fact, whenever a company produces something that doesn’t meet a customer’s exact needs, it has, as a by-product, produced economic waste. It has wasted the Earth’s resources – whether the limited commodities it extracted from the earth, the costly goods that were used as components in creating the offerings, or the precious supply of human capital required in production, delivery, fulfillment, and, yes, marketing. The only way to eliminate such economic waste is to follow this fifth practice of customering and do only and do exactly what each individual customer
21
02 Pine.indd 21
2018-09-08 11:37 PM
wants. And the key to making that happen is to focus relentlessly on customer sacrifice. Most marketers measure customer satisfaction, a measure of what customers perceive they received from an offering relative to their expectations of that offering. It is in fact a good way for those with a push mentality to begin focusing externally, outward toward the customers. But there are two primary problems with it. First, the measure is misnamed, for it measures market satisfaction, not customer satisfaction. Any information gleaned from an individual customer is never used to benefit that customer, but instead is rolled up into one giant “CustSat” number of average customer satisfaction. (Net Promoter Score, or NPS, is similar in this respect.) Second, it is relative to expectations, which often have nothing to do with what customers truly want, for most marketers spend not inconsiderable resources to lower their customers’ expectations to make them easier to hit. Customer sacrifice, on the other hand, is a true customer measure, seeking the difference between what a customer settles for and what he or she wants exactly. Forget about their expectations. What is the ideal offering for this particular customer, even if, again, the customer himself or herself doesn’t know what that is or can’t articulate it? What can I do to reduce this individual customer’s sacrifice gap by getting closer and closer and closer to doing only and exactly what he or she wants? Then use your mass-customizing capabilities to reduce that gap, maybe not all at once, but gradually and persistently over time. Contrast this relentless focus on customer sacrifice with the old ways of gaining competitive advantage. With the old learning curve – the very basis of mass production – costs come down with volume. It has its own virtuous circle: the more volume you have, the lower your costs. The lower your costs, the lower you reduce your prices. The lower your prices, the more customers buy. And the more customers buy, the more volume you have. Mass producers wanted to operate on the flat, low-cost part of that old curve until the cash
cows came home and mooed with delight. And it worked perfectly well for as long as we could treat customers as part of a mass market, as long as customers were willing to forego their individual desires to get standardized offerings that were cheap enough to pay for all that sacrifice they encountered, and as long as competitors didn’t themselves figure out that they could gain an advantage by getting closer themselves to meeting individual need. Those days are gone. With the new learning curve – the basis of customering by cultivating the virtuous circle of learning relationships – customer sacrifice comes down with interactions. As discussed earlier, every interaction is an opportunity to learn and then better customize to that individual, living, breathing customer. You want then to operate on the flat part of this new curve where you know this particular customer so well he or she will never go to another company and teach it all over again what you already know today. Here, of course, there is a different learning curve for each customer, with every one predicated on getting closer and closer to doing only and exactly what each customer wants. FROM “PRODUCT & SERVICES”/TO ECONOMIC OFFERINGS So much does marketing focus on pushing out what they have already made (goods) or defined (services), that many companies refer to their offerings with the phrase “products & services.” This is so endemic in banking that it often comes out as one six-syllable word: productsandservices. But of course the financial services industry has no physical products (goods); it only has intangible services. But to them their “products” are sitting on the shelf waiting to be sold as if they were inventoried like physical goods. (Plus it is now generally accepted to use “products” as a synonym for “economic offerings” as I often do, reserving “goods” for tangible offerings. So “productcentric” as used above can apply to other than manufacturers.)
22
02 Pine.indd 22
2018-09-08 11:37 PM
2.4 The progression of economic value.
Such confusion will not do for those who want to move from marketing to customering, for you must understand exactly what kind of economic offerings your customers desire, and consider the genre(s) of offerings that will create the most value for – and within – individual customers. So, recognize, as seen in figure 2.4,15 that there are five and five only genres of economic offerings. To briefly summarize the history of economic progress, the agrarian economy (based on commodities) was supplanted by the industrial economy (based on goods), which in turn was superseded by the service economy, and now that economy based on services has been unseated by the experience economy. Today, experiences are fast becoming the predominant economic offering as well as the source for growth in jobs and GDP in all developed economies. An experience is a distinct economic offering, as distinct from services as services are from goods. Experiences are when you use goods as props and services as the stage to engage each and every individual in an inherently personal way, and thereby create a memory, which is the hallmark of an experience. There is no country and no industry in the world that has not been touched by this shift into the experience economy. Enterprises must figure out how they incorporate experience staging into their businesses – or end up commoditized. Even those
who have always been experience stagers – think sporting events, concerts, plays, movies, museums, games, theme parks, and the like as well as many hotels, restaurants, retailers, and so forth – must up their game, because now everyone else is getting into their business. Note the two dynamics in this progression of economic value (figure 2.4). First, as we shifted into the experience economy, goods and services were commoditized – treated like undifferentiated commodities that people only want to buy on the basis of price and convenience. The second dynamic, customization, is the antidote to commoditization. You cannot help but be differentiated if you work directly with individual customers and create offerings just for them. That is in fact how I discovered the progression of economic value, for I realized that mass customizing a good automatically turned it into a service. Think about the classic distinctions between the two: goods are standardized while services are customized – done on behalf of an individual customer; goods are inventoried after production while services are delivered on demand – when the customer says this is exactly what he or she wants; and goods are tangible, while services are intangible – and part and parcel of mass customization is the intangible service of helping customers figure out exactly what they want. So mass customizing a good means defining,
23
02 Pine.indd 23
2018-09-08 11:37 PM
making, and delivering an exact item that fits each individual customer’s needs at a particular moment in time – and that, by definition, is a service! I next realized that mass customizing a service turned it into an experience. If you design a service that is so appropriate for this particular person, exactly what he or she needs at this moment in time, then you can’t help but make him or her go “Wow!” and turn it into a memorable event. But as you can see from the progression of economic value (figure 2.4), experiences can also be commoditized (as indicated by that hallmark phrase “been there, done that”). And so the progression of economic value has one more step for customizing an experience – designing and staging exactly the right experience that this particular person needs at this moment in time – turns it into what we often call a “life-transforming experience,” an experience that changes us in some significant way. Here companies use experiences as the raw material to guide customers to change, to help them in achieving their aspirations, to become what they want to become. That is, economically, a transformation. With transformations – think fitness centers, healthcare, universities, management consulting, and so forth – it does not matter what inputs the company provides, the only thing that matters is the outcomes the customer achieves. In the final analysis, there is no more economic value you can provide than helping someone achieve his or her aspirations. Besides getting out of the “products & services” trap, another reason why this framework is so important for customering is that we can summarize it with one word: individualization. Each successive economic offering gets closer and closer and closer to what each customer really truly wants, needs, and desires as an individual, living, breathing human being (or, again for B2B companies, an active, corporeal individual enterprise). Think about it. Commodities are some arm’s-length stuff you hardly ever touch and feel anymore. Goods are tangible objects that we own like our cars, our clothes, and so forth. Services
are activities performed on those objects, like changing the oil in our car or cleaning our clothes, or on ourselves, like cutting our hair or providing an analysis of our finances. Experiences for the first time reach inside of us. Experiences happen within us as individuals, in reaction to the events that are staged in front of us. And transformations change us inside. With transformations the customer is the product – it is a changed being that individuals seek from transformation elicitors. So enterprises that embrace customering should think beyond “products & services” to seek out opportunities of providing much greater value to their customers through experiences and transformations via not just customizing but individualizing their offerings. FROM MULTIPLE CUSTOMERS IN A MARKET/ TO MULTIPLE MARKETS IN A CUSTOMER There is a second sense in which customering companies (some day we will be able to parallel “marketers” with “customerers,” but that day is yet to come) should reach inside of their individual customers to create customer-unique value. If you harken back to Stan Davis’ original framework for mass customizing, the market development framework given earlier, note that with mass markets, marketers viewed all customers as belonging to the one mass market. As business competition shifted to segmented markets, only a portion of the total number belonged to each segment, but still a segment was comprised of many, many customers. Fewer customers belonged to niche markets, but still every market was comprised of multiple customers. Then, finally, with what Stan Davis called “mass customized markets,” we reached the point where every customer is his or her own market. Every customer should be treated as the individual he or she is, deserving again to get only and exactly what he or she wants at a price he or she is willing to pay. It seems like that is as far as this progression of market development can go, from every customer being a part of the one mass market to fewer customers being successively part of smaller
24
02 Pine.indd 24
2018-09-08 11:37 PM
2.5 Multiple markets within model.
segments and niches, to finally each customer being his or her own market, a market of one. But it is not the end of this progression, for it inverses, where we realize that every customer is multiple markets (figure 2.5). Think about it. When I travel on business, I want one thing from my airline, my hotel, my rental car, the restaurants that I go to. If I bring my wife along, all that changes. Same thing if I travel for vacation versus for business – or a combination of both. If we bring the kids along, my needs in all these areas (and more) change yet again. I am the same customer, but I am in different markets in each of these situations and many more that could be explored. To employ customering, we must recognize these multiple markets within each individual customer. This, in fact, restores the original conception of the term “market” before businesspeople broadened it beyond measure and understanding. Originally – at the time of the “local markets” in Stan Davis’ market development framework – a “market” was a physical place where buyers and sellers came together to exchange money for wares (almost always commodities or goods back then). That is the same sense in which I am using “market” here, just expanded a bit: it is the physical or virtual place where buyer and seller come together to exchange money for this particular offering that meets the
needs of this individual customer at this moment in time. We must restore this meaning of the term marketplace. The same is true for business-to-business (B2B) customers: there are multiple markets within each customer, and here we can further acknowledge the multiple people who are involved with B2B purchases. You may come to terms with a purchaser, who has to meet the requirements of one or more functional executives, and who then have to meet all the individual needs of the end users that use your goods, benefit from your services, undergo your experiences, or change as a result of your transformations. As time goes by, as more and more enterprises cater to these markets within each customer, people will themselves more readily recognize how much they want particular offering categories differently at different times, and demand more and more that companies reach inside of them and access these multiple markets within. If you want to transform yourself from being a marketer to a customerer (still too soon?), don’t stop at serving markets of one – as great a leap as that may seem from the standpoint of current practice in your firm – but continue on to fulfill the wants, needs, and desires of the individual markets within your individual customers.
25
02 Pine.indd 25
2018-09-08 11:37 PM
FROM SMART PRODUCTS/TO GENIUS PLATFORMS Perhaps the easiest way to reach these multiple markets within is to infuse your offerings with digital technology so that they become what are known as “smart products” – products embedded with sensors and intelligence that enable them to sense and respond to their environment. We discussed earlier how digitization enables companies to efficiently serve customers uniquely through its inherent modularity; smart products amp this up to a whole new level by being able to modify their functionality to meet changing customer needs. As the infusion of digital technology across all facets of modern life continues apace and its cost continues to plummet, marketers already offer smart products across a wide array of industries. If they maintain, however, the other marketing practices discussed here – and especially the mass marketing mindset behind them – then they will never gain the full economic value of the capabilities available today and into the future. But by embracing a customering mindset focused not only on individual customers but the multiple markets within them, they will be able to go beyond mere smart products to finding their role in one or more genius platforms. Let me explain. Because of the existence of smart products, offerings without embedded intelligence are increasingly being called “dumb products” (a linguistic backformation, like the term “acoustic guitars,” which became necessary only after the invention of electric guitars). The distinction between dumb and smart products is by now well documented, but less understood is the effect on customers of being dumb in an increasingly smart world: it makes them feel downright stupid. How many times have you placed your hands in front of a faucet only to realize that it is not sensing your hands and you actually have to touch the handle to get it to turn on? Stupid. How often do you wait for a stoplight at an empty intersection, for no reason? Stupid.
Even smart offerings that have embedded intelligence can degenerate into stupidity. Think of the times you have called a contact center, keyed in your account number, and then reached a real person who promptly asks for your number again. Stupid. Or the mental contortions you have to go through to first make up a unique password for some site, app, or program – and then to remember it when asked for it over and over again. Stupid. And just because the company’s offering isn’t as smart as it should be, it makes you feel stupid. In a world increasingly filled with the smart, the dumb, and the stupid, you must strive for the smart and resist deteriorating into the stupid. And you should also consider becoming a budding genius, for as the research of my colleague Dave Norton of Stone Mantel (who greatly contributed to the notion of genius platforms) shows,16 the advent of intelligence in offerings has preconditioned people to expect more. More connections among all their devices. Better integration across their lives. Greater capabilities to do what before could only be imagined. For what Dave and I call genius platforms effectively endow people with superpowers. They give customers capabilities that to the average person living 20, 10, even 5 years ago would have seemed to be magic. What does it take to put this final practice of customering in place and go from the merely smart to the pure genius? Smart products again sense and respond to your needs, while genius platforms anticipate what you are trying to accomplish before you even have to say anything; genius is prescient. Smart products customize to the individual, while genius platforms individualize to the job to be done. Again, every customer is multiple markets, and fulfilling the need I have – the job I want done – at this moment in time, well, that is genius. And, finally, smart products enable customers to control their circumstances, while genius platforms understand and support digital context – all the collective intelligence residing across the entirety of a customer’s smart products – to go far beyond what any single smart product can do – and effectively enable superpowers!
26
02 Pine.indd 26
2018-09-08 11:37 PM
2.6 Marketing versus customering.
CUSTOMERING IN YOUR ENTERPRISE That is the full promise that customering holds for your enterprise. But it all comes down to mindset. Are you going to abandon the old mindset of marketing and put on the new mindset of customering – interacting directly with individual customers to create customer-unique value within them? And then will you cast off these old practices of marketing and embrace these nine new practices of customering (figure 2.6)?
For competition ever intensifies, business gets tougher year after year, the possibility of new value creation continually expands, and what customers value changes constantly. It will not be enough to continue giving customers exactly what they want today. You will have to continually renew your capabilities and better individualize your offerings. Recognize how great the challenge is that lies before you, and do not hesitate to rise to that challenge, to ascend to the proposition that all customers are unique, and to uniquely value each and every one as the precious lifeblood of your business that they truly are.
27
02 Pine.indd 27
2018-09-08 11:37 PM
NOTES 1 Davis, Stanley M., Future Perfect, Reading, MA: Addison-Wesley, 1987. 2 Ibid, p. 169. 3 Pine, B. Joseph, II, Mass Customization: The New Frontier in Business Competition, Boston: Harvard Business School Press, 1993. 4 Ibid, p. 44. Emphasis in original eliminated here. 5 Davis, op. cit, pp. 168–169. 6 “Definition of Marketing,” American Management Association, available at: www.ama.org/AboutAMA/ Pages/Definition-of-Marketing.aspx. 7 Every enterprise of any size can embrace customering, even if they cannot create the scale for truly mass customizing and instead must craft customize. 8 Searls, Doc, The Intention Economy: When Customers Take Charge, Boston: Harvard Business Review Press, 2012. 9 Ibid, p. 11.
10 Peppers, Don, and Rogers, Martha, The One to One Future: Building Relationships One Customer at a Time, New York: Currency, 1993. 11 The Oxford English Dictionary, “Customer”, 2nd edn, Oxford: Clarendon Press, 1989, vol. IV, p. 169. 12 Ibid., Volume II, p. 1038. 13 The Ultimate LEGO Book, London: Dorling Kindersley, 1999, p. 8. LEGO rounded down the actual number of 102,981,504. 14 Eilers, Søren, Abrahamsen, Mikkel, and Durhuus, Bergfinnur, “A LEGO Counting Problem,” Institut for Matematiske Fag, available at: www.math.ku.dk/~eilers/ lego.html. 15 Pine, B. Joseph, II, and Gilmore, James H., The Experience Economy, updated edn, Boston: Harvard Business Review Press, 2011, p. 245. 16 Norton, David, Digital Context 2.0: 7 Lessons in Business Strategy, Consumer Behavior, and the Internet of Things, Colorado Springs: Gifted Press, 2015.
28
02 Pine.indd 28
2018-09-08 11:37 PM
3 CREATING A SUSTAINABLE MASS CUSTOMIZATION BUSINESS MODEL
FRANK PILLER 29
03 Piller.indd 29
2018-10-18 9:28 PM
In the last two decades, mass customization has emerged as a dominant business strategy. The aim of mass customization is to profit from people’s differences by enabling the efficient provision of goods and services that best serve the personal needs of customers. This chapter introduces the concept, its background, and the fundamental capabilities that companies have to develop when building a mass customization offering. We also speculate about future developments in the field.1 CUSTOMIZATION BECOMES THE NEW STANDARD In many industries, we can observe today an uninterrupted trend toward heterogeneity of demand.2, 3 Explanations may be found in a growing number of single households, a changing demographic structure, an orientation toward design, and a new awareness of quality and functionality that demands durable and reliable products corresponding exactly to the specific needs of the purchaser.4 This trend is reinforced by the entry of Generations Y and Z5 as consumers in the marketplace, generations characterized by considerable experience in customizing their personal communications and media consumption (curating one’s own Facebook page, personalizing a music stream). It is likely that these young consumers transfer these experiences to other product categories. Customization thus becomes the new standard. Because of this, manufacturers are forced to create product portfolios with an increasing wealth of variants, right down to the production of units of one. As a final consequence, many companies have to process their customers’ demand individually. Since the early 1990s, mass customization has emerged as one leading idea for achieving precisely this objective. Similar to Joseph Pine, we define mass customization as developing, producing, marketing, and delivering affordable goods and services with enough variety and customization so that nearly everyone finds exactly what they want.6 In other words, the aim
is to give customers what they want, when they want it. However, to apply this apparently simple statement in practice is quite complex. As a business paradigm, mass customization provides an attractive business proposition to add value by directly addressing customer needs and, in the meantime, utilizing resources efficiently without incurring excessive cost. This is particularly important at a time where competition is no longer based just on price and conformance of dimensional quality. Today, many companies have employed mass customization successfully. Consider the following examples: •
BMW: Customers can use an online toolkit to
design Mini Cooper roofs with their very own graphics or pictures, which are then reproduced with an advanced digital printing system on special foils. The toolkit has enabled BMW to tap into the custom aftersales market, which was previously owned by niche companies. In addition, Mini Cooper customers can also choose from among hundreds of options for many of the car’s components, as BMW is able to manufacture all cars on demand according to each buyer’s individual order.7 •
mymuesli: Food customization is one of the fastest developing industries in the field of mass customization.8 The German company mymuesli is a pioneer in this market. It offers an Internet-based toolkit to customize cereal out of myriad options. Via interaction with the toolkit, customers can combine their preferred options from the components of “Base,” “Fruit,” “Nuts & Seeds” and “Extra Ingredients” into a mixed package. Interestingly, mymuesli also could successfully establish a large assortment of cereals in German supermarkets with its ability to produce efficiently in small batches; by offering special retail editions and large variety, it could conquer the market, utilizing its customization brand promise to win consumers from established mass-market brands.
30
03 Piller.indd 30
2018-10-18 9:28 PM
3.1 Overview of the three strategic capabilities for mass customization.
Regardless of product category or industry, these and many other companies have all turned customers’ heterogeneous needs into an opportunity to create value rather than regarding heterogeneity as a problem that has to be minimized.9 To achieve this objective, mass customization demands a specific set of processes and capabilities for aligning an organization with its customers’ needs. In the following, we will introduce three fundamental capabilities organizations need to implement to benefit from mass customization. THREE CAPABILITIES OF MASS CUSTOMIZATION The key to profiting from mass customization is to see it as a set of organizational capabilities that can supplement and enrich an existing system. While the nature and characteristics of these additional capabilities are clearly dependent on industry context or product characteristics, there are three fundamental groups of capabilities that determine the ability of a firm to mass customize (figure 3.1). Companies that successfully master the proposition of mass
customization have built competencies around these three core capabilities. In an earlier paper, my co-authors and I called these core capabilities: (1) “solution space development”; (2) “robust process design”; and (3) “choice navigation.”10 I will introduce these capabilities briefly in the following section and discuss them in more detail in the remaining ones. 1. Solution space development: First and foremost, a company seeking to adopt mass customization has to be able to identify what the idiosyncratic needs of its customers are. This is in contrast to the approach of a mass producer, where the company focuses on highlighting the “central tendencies” among its customers’ needs and targets them with a limited number of standard products. Conversely, a mass customizer has to identify the product attributes on which customer needs diverge the most. Once this is understood, the firm knows what is required to accurately meet the needs of its customers. Consequently, it can draw up the so-called solution space, clearly defining what it is going to offer and what it is not.
31
03 Piller.indd 31
2018-10-18 9:28 PM
2. Robust process design: A second critical requirement for mass customization is related to the relative performance of the value chain. Specifically, it is crucial that providing for increased variability in customers’ requirements does not lead to significant deterioration in the firm’s operations and supply chain. This demands a robust value chain design – defined as the capability to reuse or re-combine existing organizational and value chain resources to fulfill differentiated customer needs in a limited time frame and at acceptable costs. With robust process design, customized solutions can be delivered with close to mass production efficiency and reliability. 3. Choice navigation: Finally, the firm must be able to support customers in expressing their individual requirements and creating their own solutions while minimizing complexity and burden of choice.11 When a customer is exposed to too many choices, the cognitive cost of evaluation can easily outweigh the increased utility from having more choices.12 As such, offering more product choices can easily prompt customers to postpone or suspend their buying decisions. Therefore, the third requirement is the organizational capability to simplify the navigation of the company’s product assortment from the customers’ perspective. In the following sections, we will discuss the three fundamental capabilities of mass customization in more detail and also look into the approaches and practices connected with these capabilities. SOLUTION SPACE DEVELOPMENT A mass customizer must first identify the idiosyncratic needs of its customers – specifically, the product attributes on which customer needs diverge the most.13 This is in stark contrast to a mass producer having to focus on serving universal
needs, ideally shared by all the target customers. Once these key product characteristics are fixed, the company can define its “solution space,” that is, a statement of all the possible permutations of design parameters that are offered to prospective customers.14 This space determines what universe of benefits the manufacturer is willing to offer to its customers. In the case of mass customization, the solution space is precisely delimited and delivery conditions can be associated with any option without uncertainty relative to price, quality levels, and manufacturability. Setting an appropriate solution space is a major challenge for a mass customization company as it directly affects customers’ perceptions of the utility of the customized product and determines the efficiency of downstream processes in the fulfillment system.15 But despite this importance, there is surprisingly little research in the mass customization literature on how to develop a solution space. Hence, we want to comment briefly on some of the potential methods. The first option for solution space development is to engage conventional market research techniques. The manufacturer selects and studies a group of representative customers to obtain information on needs for new products, analyzes the data, develops a responsive product idea, and screens this idea against customer preferences (needs) and purchasing decisions. This model is dominant in the world of consumer goods, where market research methodologies such as focus groups, conjoint analysis, customer surveys, and analyses of customer complaints are used regularly to identify and evaluate customer needs and desires. However, dedicated market research methods have not been developed especially for the development of a solution space. What is needed is a method that distinguishes the customer heterogeneities that allow a firm to identify the components of an offering customers would like to have choice on, how much choice, and its range. From our experience of working with companies, most existing methods do not provide answers in appropriate detail to these questions.
32
03 Piller.indd 32
2018-10-18 9:28 PM
A second approach for solution space development for mass toolkits could be to provide (advanced) customers with an expert toolkit for user innovation. When Fiat was developing its retro Fiat 500, for example, the automaker created Concept Lab, an innovation toolkit that enabled customers to freely express their preferences regarding the interior of the car long before the first vehicle had been built. The company received more than 160,000 designs from customers – a product development effort that no automaker could replicate internally. And Fiat allowed people to comment on other people’s submissions, providing initial evaluation of those ideas. Of course, mass producers can also benefit from innovation toolkits, but the technology is particularly useful for mass customization because it can be deployed at low cost for large pools of heterogeneous customers. Third, in developing their solution space, companies can employ some form of “customer experience intelligence,” that is, applying methods to continuously collect data on customer transactions, behavior, or experiences and analyzing that information to determine preferences. This also includes incorporating data not just from existing customers, but also from others, who have taken their business elsewhere but might have been customers. Consider, for example, information about products that someone has evaluated but did not order. Such data can be obtained from the log files generated by the browsing behavior of people using online configurators. By systematically analyzing that information, managers can learn much about customer preferences, ultimately leading to a refined solution space. A company could, for instance, eliminate options that are rarely explored or selected, and it could add more choices for the popular components. Hence, it is important to note that solution space development is not a one-off activity but, rather, a continual, iterative improvement process. What customers want today may be different tomorrow. Companies would thus be well advised
to implement a formal process to revise, trim, or extend their solution space at regular intervals. There is a fourth path to solution space development, which is following one’s own need. The Customization 500 study, an investigation of 500 online mass customization companies in the consumer market,16–18 has shown that the practice of mass toolkits is dominated by startup companies. Research on entrepreneurship has investigated the nature of opportunity recognition – the process through which ideas for potentially profitable new business ventures are identified. This kind of opportunity recognition also plays an important role in developing an initial solution space for a mass customization company.19 In fact, many entrepreneurs started their businesses simply by translating their own unsatisfied needs into a custom product offering – mass customization meets lead users. In the era of mass production, customers implicitly agreed to trade off less customization for lower prices. If customers take a mass-produced product and adjust it to their own needs, it shows the potential opportunity, as there are other customers out there who would also prefer a similarly customized item. For example, indicustom.com and smart-jeans.com built their business model on the realization that many customers take their jeans to a tailor after purchasing them. Also, the first customized chocolate bar by Chocri was an attempt to create an original last-minute birthday gift. The tricky task in determining what should be customized, or not, is detecting the sacrifices made by most customers, not just one. ROBUST PROCESS DESIGN The core idea in mass customization is ensuring that an increased variability in customers’ requirements will not significantly impair the firm’s operations and supply chain.20 This can be achieved through robust process design – the capability to reuse or recombine existing organizational and value chain resources to deliver customized solutions with high
33
03 Piller.indd 33
2018-10-18 9:28 PM
efficiency and reliability. A successful mass customization system is hence characterized by stable, yet flexible, responsive processes that provide a dynamic flow of products.21, 22 Value creation within robust processes is the major differentiation of mass customization versus conventional (craft) customization. Traditional (craft) customizers reinvent not only their products but also their processes for each individual customer. Mass customizers use stable processes to deliver high-variety goods,23 which allows them to achieve “near mass production efficiency,” but also implies that the customization options are somehow limited. Customers are being served within a set of predefined options or components, the company’s solution space. Yet these options are what the customer wants.
Cost Drivers of Variety The core objective of robust process design is to prevent or counterbalance the additional cost resulting from the flexibility a company needs to build in order to serve its customers individually. We can differentiate two sources of additional cost in relation to flexibility: (1) increased complexity, and (2) increased uncertainty in business operations, which by implication results in higher operational cost.24 A higher level of product customization requires greater product variety, which in turn entails greater numbers of parts, processes, suppliers, retailers, and distribution channels. A direct consequence of such proliferation is increased complexity in managing all aspects of the business from raw material procurement to production and eventually to distribution. Further, an increase in product variety has the effect of introducing greater uncertainty in demand realizations, increased manufacturing cycle times, and increased shipment lead times. Increased system complexity and uncertainties (in demand and lead times) drive the operational cost upward, due to more complex planning,
greater hedging, increased resource usage, more complex production setups, diseconomies of scope, and higher distribution costs spread throughout the supply chain. Finally, integral to mass customization strategy is the offer of choice navigation for customers, which entails a sizeable increase in costs. This includes, for example, implementing a configuration system on a website or in a physical store.
Methods to Establish Robust Processes A number of different methods can be employed to reduce these additional costs or even to prevent their occurrence at all. A primary mechanism for creating robust processes in mass customization is the application of delayed product differentiation (postponement). Delayed product differentiation refers to partitioning the supply chain into two stages.25 A standardized portion of the product is produced during the first stage, while the “differentiated” portion of the product is produced in the second stage based on customer preferences that are expressed in an order. The success of delayed product differentiation is a direct manifestation of the fact that most companies offer a portfolio of products that consists of families of closely related products which differ from each other in a limited number of differentiated features. An example of delayed product differentiation in the automotive industry would be to send a standard version of a car (a stripped or partially equipped version) to dealers and then allow the dealer to install, on the basis of customer-specific requests, options like a CD/DVD player, the interior leather or fabric, the cruise control system, etc. Prior to the point of differentiation, product parts are re-engineered so that as many of the parts or components as possible are common to each configuration. Cost savings result from the riskpooling effect and reduction in inventory stocking costs.26 Additionally, as common functionality performance levels are selected by a number of customers, economies of scale can be achieved at
34
03 Piller.indd 34
2018-10-18 9:28 PM
the modular level for each version of the module, generating cost savings not available in pure customization-oriented production systems.27 While postponement starts at the design of the offerings, another possibility to achieve robust processes is through flexible automation. Although the words “flexible” and “automation” might have been contradictory in the past, that is no longer the case. In the auto industry, robots and automation are compatible with high levels of versatility and customization. Even process industries (pharmaceuticals, food, and so on), once synonymous with rigid automation and large batches, nowadays enjoy levels of flexibility once considered unattainable. Similarly, many intangible goods and services also lend themselves to flexible automated solutions, often based on the Internet. In the case of the entertainment industry, increasing digitalization is turning the entire product system over from the real to the virtual world. A complementary approach to flexible automation is process modularity, which can be achieved by thinking of operational and value chain processes as segments, each one linked to a specific source of variability in customer needs.28 As such, the company can serve different customer requirements by appropriately recombining the process segments without the need to create costly ad hoc modules.29 BMW’s Mini factory, for instance, relies on individual mobile production cells with standardized robotic units. BMW can integrate the cells into an existing system in the plant within a few days, thus enabling the company to quickly adapt to unexpected swings in customer preferences without extensive modification of its production areas. Process modularity can also be applied to service industries. IBM, for example, has been redesigning its consulting unit around configurable processes (called “engagement models”). The objective is to fix the overall architecture of even complex projects while retaining enough adaptability to respond to the specific needs of a client.
CHOICE NAVIGATION Finally, a mass customizer must support customers in identifying their own needs, expressing their individual requirements and creating solutions while minimizing choice complexity and the burden of choice during the customization process. This is what choice navigation refers to. When a customer is exposed to myriad choices, the cost of evaluating those options can easily outweigh the additional benefit from having so many alternatives. The resulting syndrome, in which too many options can actually reduce customer value instead of increasing it,30 has been called the “paradox of choice.” In such situations, customers might postpone their buying decisions and, worse, classify the vendor as difficult and undesirable. Recent research in marketing has addressed this issue in more detail, finding that from the consumer perspective, the perceived cognitive cost is one of the largest hurdles toward greater adoption of mass customization. To avoid this, companies have to provide means of choice navigation to simplify the ways in which people explore their offerings.
Configuration Toolkits Co-design activities are a prerequisite for mass customization to be able to fulfill the needs of individual customers. However, these activities are also a major driver for complexity, effort, and perceived risk from the customer’s perspective, limiting the success of mass customization strategies. Pine coined the term “mass confusion” to describe the stress and drawbacks for the consumer caused by mass customization interaction processes.31 We see mass confusion as one major explanatory factor for the delay in adoption of mass customization technologies in business practice. The traditional approach to navigate the customer’s choice in a mass customization system has been product configuration systems,
35
03 Piller.indd 35
2018-10-18 9:28 PM
also referred as “co-design toolkits,” configurators, choice boards/menus, design systems, or co-design platforms. Configurators are knowledge-based software tools that support a potential customer in specifying a product solution within a company’s product space and guide the customer through the elicitation process.32 Whenever the term “configurator” or “configuration system” is used in the literature, it is mostly applied in a technical sense, usually addressing a software tool. The success of such an interaction system, however, is by no means defined solely by its technological capabilities; success is also determined by its integration into the sales environment, its ability to allow for learning, its ability to deliver immediate (simulated) feedback on the potential outcome of the design ideas,33 its ability to provide experience and process satisfaction34 and its integration into the brand concept. Tools for user integration in a mass customization system contain much more than arithmetic algorithms for combining modular components. Taking up an expression from von Hippel,35 the more generic term “toolkits for customer codesign” might better describe the diverse activities taking place.36 In a configuration toolkit which is interpreted as a learning instrument, different product components are represented, visualized, assessed, and priced with an accompanying learning-by-doing process for the user.37 Via the toolkits, customers engage in the co-design process by trying out different combinations, matching these with their preferences or modifying their individual solutions from a list of options and predefined components. The core idea is to engage customers into fast-cycle, iterative, trial-and-error learning processes. Thanks to this mechanism, customers can engage in trial-and-error experimentation in order to self-design products that best fit their preferences.
Advanced Methods for Choice Navigation The application of toolkits for customer codesign may be the commonest approach to help
customers navigate choice in a mass customization system, but a number of other approaches exist too. One effective approach is what we have labeled “assortment matching.”38 Here, software automatically builds configurations for customers by matching models of their needs with characteristics of existing solution spaces (i.e. sets of options). Then customers only have to evaluate the predefined configurations, thus saving considerable effort and time in the search process. For example, consider Zafu.com. The startup has created a very profitable business model by taking customers’ body measurements and then recommending the bestfitting pair of jeans from the existing assortments of many major brands. From their users’ perspective, Zafu is offering a product that fits like tailormade jeans. But from the fulfillment perspective, Zafu is just matching standard inventory with individual needs. I consider this to be a true mass customization strategy: giving every customer what he or she wants – at high efficiency. But customers might not always be ready to make a decision after they have received recommendations. They might not be sure about their real preferences, or the recommendations may not appear to fit their needs. In such cases, combining a recommendation system with a codesign toolkit is a perfect solution.
From Toolkits to Smart Products and Services A number of companies are engaging in even more innovative and radical approaches to choice navigation. Choice navigation has been completely automated in some recent products, with the products “understanding” how they should adapt to users and then reconfiguring themselves accordingly. Equipped with so-called “embedded configuration capability,” the products paradoxically become standard items for the manufacturer while the user experiences a customized solution. Such is the case with Adidas One, a running shoe equipped with a magnetic sensor, a system to adjust cushioning, and a microprocessor to control the process. When the shoe’s heel strikes the ground, the sensor
36
03 Piller.indd 36
2018-10-18 9:28 PM
measures the amount of compression in its midsole and the microprocessor calculates whether the shoe is too soft or too firm for the wearer. A tiny motor then shortens or lengthens a cable attached to a plastic cushioning element, making it more rigid or more pliable. With this system, the sport shoe is continuously adapting to different user needs during its usage stage – without any customization before the point of sale. This corresponds to one of the four customization types proposed by Gilmore and Pine,39 i.e. adaptive customization. In adaptive customization, customization space is embedded in the standard goods or services, allowing customers to modify or customize according to different needs without direct interaction with the company, after the product has been purchased. The configuration toolkits to facilitate user co-design with products in use can be either embedded toolkits or connected toolkits (apps) installed in smart devices. The Philips Hue, a family of internet-connected LED lights, is a good example of this approach. Customers can either customize the light setting by directly pressing the smart control on the product or fully control it with the smart device in portable mode. The design of this lighting system enables clients to create different color effects and adjust brightness for reading, romantic moments, or lively parties. We believe that one of the most exciting future developments of mass customization lies in the exploitation of customization opportunities for such smart products. COMPLEMENTARITIES AMONG THE THREE CAPABILITIES We have described three strategic capabilities that companies need to build and use to make a mass customization business model work. But how do the three capabilities relate to each other? Data from the Customization 500 study allowed us to test several alternative models of the relationships between the three capabilities (and their antecedents). First, we find that the three capability dimensions are empirically distinguishable, emphasizing the multidimensional nature of mass customization.40
Second, and more interesting in the context of this chapter, we found in our research the three strategic capabilities do not (significantly) improve performance on their own, but show strong complementarities with each other. Complementarity theory implies that the magnitude of the effect of overall mass customization capability is greater than the sum of marginal effects from the development of each capability individually.41 We modeled the mass customization capability of a manufacturer as a reflective second-order construct to capture complementarities arising from the three capabilities.42 This second-order construct accounts for multilateral interactions between the three capabilities and can be shown to be statistically superior compared to a conceptualization of mass customization as three distinct yet correlated capabilities. It is the complementary and synergistic effects of the three distinct but highly interrelated capabilities that enable firms to achieve multiple performance goals, such as (in our sample) market success and financial success. Unfortunately, transforming a business along all three dimensions subsumed by the three capabilities is difficult, both for startups and for incumbents, although for different reasons. In the case of startups, the major challenge for the entrepreneur is to develop a vision that does not overlook key capabilities. Consider the case of mymuesli (not a failure, by any means). This provider of customized cereal had a well-designed toolkit that enabled users to overcome the challenge of simulating taste online. Moreover, mymuesli was rather proficient in solution space development, predominantly building on the lead user characteristics of their founders and continuously learning from past interactions. However, mymuesli neglected robust processes, especially with regard to flexible automation. Whenever a press report in their favor was published, their manufacturing capacity (plenty of human labor mixing muesli from about 65 different option categories by hand, using precision scales) was overwhelmed by demand, and the company had to set daily order limits and turn down prospective customers. Only after investing in
37
03 Piller.indd 37
2018-10-18 9:28 PM
a fully automated, scalable mixing system, could the complementarities of the business model be exploited, and the company began to generate profits. In many other examples we investigated, it was the capability to develop the solution space in particular that was lacking, leading to failure when toolkit options did not represent the heterogeneity of demand that mattered for users in the particular market. In the case of incumbents, the core problem associated with the implementation of sustainable business models leveraging mass toolkits is that the changes associated with the development of the three above-mentioned capabilities are pervasive and opposed by powerful inertial forces. Take, for instance, John Deere, one of the world’s largest manufacturers of garden equipment. To keep up with its market for premium lawn tractors, which had been evolving toward greater fragmentation and customization for more than a decade, the company began to offer more products, but this then resulted in a proliferation of parts and processes. Divisional managers were aware of this, and they knew that they could save millions of dollars every year by simplifying their product platforms. Yet it took years for the divisions to implement the needed changes, which required overhauling accounting and performance measurement systems and product development as well as manufacturing and supply chain processes. John Deere realized that these changes had to be done simultaneously because mass customization capabilities relate to all the major areas of the value chain. CONCLUSION
or user personas – all concepts that are built on the idea of finding the common denominator of a group – instead of focusing on where people differ. In manufacturing, we still focus on rampup, scalability, and market share – again, concepts from a world dominated by mass production – and not on mass customization. Also, as consumers or citizens, we often think how we can fit a general trend, how we have to adapt to an existing solution, and how we can meet the expectations of our peers, and we are uncomfortable with neighbors who are different and live a different lifestyle from our own. We all want personalization and customization but do rather little to reach this objective. Current digital technological developments, however, are making customization the new standard. Consider, for example, additive manufacturing (colloquially, 3D printing), a technology that breaks with the established laws of economies of scale and provides customized products without a cost penalty.43 Smart, connected products, based on IoT44 technology, allow the continuous adaption of designs and services to individual, context-specific user needs. Data analytics allows us to explore our inner self and helps us to match existing offerings to our personal needs, overcoming an old truth in personalization that people themselves don’t know what they want and require. At the same time, new technologies also challenge personalization and ask us to consider how privacy and personalization relate to each other, how personal recommendations balance between one’s own interests and those who provide the recommendations, and why we still like to hide in the crowd and share mass experiences on a large scale. The debate on personalization has just begun.
Despite all the talk about personalization and individualization in academia and the business press alike, and despite all the success stories and hundreds of startups that built a successful mass customization business, it is remarkable how much conventional “mass production” thinking continues to dominate our perception of the world. We talk of customer segments, market clusters,
38
03 Piller.indd 38
2018-10-18 9:28 PM
NOTES 1 This chapter builds on arguments developed in Salvador, F., de Holan, M. and Piller, F., “Cracking the Code of Mass Customization,” MIT Sloan Management Review, vol. 50, no. 3 (2009), pp. 70–79, and Piller, F., and Salvador, F., “Design Toolkits, Organizational Capabilities, and Firm Performance,” in Harhoff, D. and Lakhani, K. (eds.), Revolutionizing Innovation: Users, Communities, and Open Innovation, Cambridge, MA: MIT Press, 2016, pp. 483–509. 2 ElMaraghy, H., Schuh, G., ElMaraghy, W., Piller, F., Schönsleben, P., Tseng, M., and Bernard, A., “Product Variety Management,” CIRP Annals, vol. 62, no. 2 (2013), pp. 629–652. 3 Gruel, W. and Piller, F., “A New Vision for Personal Transportation,” MIT Sloan Management Review, vol. 57, no. 2 (2016), pp. 20–23. 4 Franke, N., Schreier, M., and Kaiser, U., “The ‘I Designed It Myself’ Effect in Mass Customization,” Management Science, vol. 56, no. 1 (2010), pp. 125–140. 5 Generation Y refers to people born in the 1980s and early 1990s. The name is based on Generation X that preceded them. Generation Z refers to people born from the mid-1990s to the early 2000s. 6 Pine, B. Joseph, Mass Customization, Boston: Harvard Business School Press, 1993. 7 Gruel and Piller, op. cit. 8 Kolb, M., Blazek, P., and Streichsbier, C., “Food Customization: An Analysis of Product Configurators in the Food Industry,” Proceedings of the 7th World Conference on Mass Customization, Personalization, and Co-Creation, 2014, pp. 229–239. 9 See www.configurator-database.com for a large list of customizable products in all industries. 10 The derivation of these capabilities builds on work by Salvador et al., op. cit. 11 Ibid. 12 Huffman, C. and Kahn, B.E., “Variety for Sale: Mass Customization or Mass Confusion?” Journal of Retailing, vol. 74, no. 4 (1998), pp. 491–513. 13 Salvador et al., op. cit. 14 von Hippel, E., “User Toolkits for Innovation,” Journal of Product Innovation Management, vol. 18, no. 4 (2001), pp. 247–257. 15 Tseng, M. and Piller, F., “The Customer Centric Enterprise,” in Tseng, M. and Piller, F. (eds.), The Customer Centric Enterprise: Advances in Mass
Customization and Personalization, New York: Springer, 2003, pp. 1–18. 16 Walcher, D. and Piller, F., The Customization 500: An International Benchmark Study on Mass Customization and Personalization in Consumer E-Commerce, Raleigh, VA: Lulu Inc., 2012. 17 Harzer, T., “Value Creation Through Mass Customization: An Empirical Analysis of the Requisite Strategic Capabilities,” dissertation thesis, RWTH Aachen University, 2012. 18 Piller, F., Harzer, T., Ihl, C., and Salvador, F., “Strategic Capabilities of Mass Customization-Based E-Commerce,” Proceedings of the 47th Hawaii International Conference on System Science, vol. 47, no. 1 (2014), pp. 3255–3264. 19 Harzer, op. cit. 20 Pine, op. cit. 21 Kortmann, S., Gelhard, C., Zimmermann, C., and Piller, F.T., “Linking Strategic Flexibility and Operational Efficiency: The Mediating Role of Ambidextrous Operational Capabilities,” Journal of Operations Management, vol. 32, nos. 7–8 (2014), pp. 475–490. 22 Tu, Q., Vonderembse, M.A. and Ragu-Nathan, T.S., “The Impact of Time-Based Manufacturing Practices on Mass Customization and Value to Customer,” Journal of Operations Management, vol. 19, no. 2 (2001), pp. 201–217. 23 Pine, op. cit. 24 ElMaraghy et al., op. cit. 25 Yang, B., Burns, N.D. and Backhouse, C.J., “Postponement: A Review and an Integrated Framework,” International Journal of Operations & Production Management, vol. 24, no. 5 (2004), pp. 468–487. 26 Ibid. 27 Steiner, F. and Hergenröther, I., “Modular Product Architectures as an Enabler of the Simultaneous Application of a Mass Customization Strategy and Efficient Ramp-Up Management,” International Journal of Product Development, vol. 19, no. 4 (2014), pp. 231–253. 28 Pine, B.J., Victor, B., and Boynton, A.C., “Making Mass Customization Work,” Harvard Business Review, vol. 71, no. 5 (1993), pp. 108–119. 29 Zhang, Q., Vonderembse, M.A., and Lim, J-S., “Manufacturing Flexibility: Defining and Analyzing Relationships among Competence, Capability, and Customer Satisfaction,” Journal of Operations Management, vol. 21, no. 2 (2003), pp. 173–191.
39
03 Piller.indd 39
2018-10-18 9:28 PM
30 Huffman and Kahn, op. cit. 31 Ibid. 32 Trentin, A., Perin, E., and Forza, C., “Overcoming the Customization-Responsiveness Squeeze by Using Product Configurators,” Computers in Industry, vol. 62, no. 3 (2013), pp. 260–268. 33 von Hippel, op. cit. 34 Franke et al., op. cit. 35 von Hippel, op. cit. 36 Franke, N. and Piller, F., “Value Creation by Toolkits for User Innovation and Design: the Case of the Watch Market,” Journal of Product Innovation Management, vol. 21, no. 6 (2004), pp. 401–415. 37 Franke, N. and Hader, C., “Mass or Only ‘Niche Customization’? Why We Should Interpret Configuration Toolkits as Learning Instruments,” Journal of Product Innovation Management, vol. 31, no. 6 (2014), pp. 1214–1234. 38 Salvador et al., op. cit. 39 Gilmore, J.H. and Pine, B.J., “The Four Faces of Mass Customization,” Harvard Business Review, vol. 75, no. 1 (1997), pp. 91–101. 40 For detailed data and statistical evidence relating to these findings, see Harzer et al., op. cit. and Piller et al., op. cit.; for a related study, see Kortmann et al., op. cit. 41 Milgrom, P. and Roberts, J., “The Economics of Modern Manufacturing: Technology, Strategy, and Organization,” The American Economic Review, vol. 80, no. 3 (1990), pp. 511–528. 42 Piller et al., op.cit. 43 Weller, C., Kleer, R., and Piller, F., “Economic Implications of 3D Printing: Market Structure Models in Light of Additive Manufacturing Revisited,” International Journal of Production Economics, vol. 164 (2015), pp. 43–56. 44 Internet of Things.
40
03 Piller.indd 40
2018-10-18 9:28 PM
4 A CRAFT IT YOURSELF FUTURE
VIRGINIA SAN FRATELLO AND RONALD RAEL 41
04 San Fratello Rael.indd 41
2018-10-18 4:11 PM
4.1 The Picoroco Block™, a modular 3D-printed building block for wall fabrication.
The definition of craft often refers to custom work performed by hand. However, a new genre of craft has emerged, one that takes lessons from both the industrial and pre-industrial eras, enlightened by new methodologies that have arisen from fabrication technologies. There is a renewed interest in building knowledge through repetition and practice of a skilled trade employing the aid of digital tools in order to mass customize and to create mass complexity. Architects were originally craftsmen, however, the age of industrialization removed the architect from being directly involved in the shaping of materials that comprise buildings. Computer-aided design (CAD) and computeraided manufacturing (CAM) have brought about the return of craft to the built environment by the architect – a computer-aided craft (CAC). While several CAD/CAM tools have facilitated new ways of thinking about assembly and material manipulation, 3D printing represents a future of crafting or making by architects and designers – it embodies the concepts of democratic design, mass customization, and complexity, on demand production, waste-free manufacturing, and craft it yourself material innovation, engaging the craftsperson at every level.
DEMOCRATIC CRAFT Democratic craft means that good products can be available to everyone. Decreased costs associated with desktop 3D printers and increased access to them in public libraries, grade schools, and universities now mean that access to 3D printing has become the norm rather than the exception. Bio-plastic filament used in desktop printers is also very inexpensive and does not require any post-processing and is made from renewable resources such as cornstarch or sugar cane. Most desktop printers are fused deposition modelers (FDM), which is an additive manufacturing technology commonly used for prototyping, but rarely for making final products. FDM works on the “additive” principle by depositing plastic filament along a path. For this to occur, plastic filament is unwound from a coil and supplied to an extrusion nozzle. The metal nozzle is heated and melts the plastic, which is then extruded through the nozzle and deposited onto a build platform. Printed objects using FDM methods are fabricated from the bottom up, one layer at a time. Emerging Objects1 selected FDM printing and bio-plastic to produce the Picoroco Block™, a modular 3D-printed building block for wall fabrication (figure 4.1). There are three unique blocks that have
42
04 San Fratello Rael.indd 42
2018-10-18 4:11 PM
4.2 The Star Lounge dome is composed of almost 3000 3D printed hexagonal blocks.
a dimension of 5.75” x 5.75” x 2” designed to fit in the build bed of a small desktop 3D printer. The blocks have a range of openings from two to four, and can be arrayed both vertically and horizontally and freely rotated to make a continuous and variable pattern across the surface of the Picoroco wall. The blocks are held in place with four-prong, 3D-printed clips. The Picoroco wall takes advantage of the translucency of the bioplastic, giving it a softly scalloped diaphanous quality when used as a screen or partition in front of a window or a light. The wall requires no specialized labor or skills for assembly; anyone can easily put it together. Building on the knowledge gained in the Picoroco wall, Emerging Objects fashioned one of the largest bio-plastic 3D-printed structures to date, the Star Lounge (figure 4.2). The Star Lounge is unique among many other large-scale 3D-printed structures in that it demonstrates the architectural potential of using small desktop printers in order to 3D print a building. The hexagonal blocks that make up the large star and hexagon panels were printed using a “Bot Farm” of over 100 3D printers. The freestanding dome structure is 8.5 feet tall with a footprint that measures approximately 11 ft by 11 ft. The dome is composed of almost 3000 hexagonal
blocks printed in various translucent colors that correspond to custom block types, which helps simplify the construction process and creates a beautiful and logical pattern of 21 larger stars and hexagons that are riveted together. The very instructions for assembly are encoded in the colors used to define specific modular components. To facilitate file management, only 58 unique block types make up the doubly curved dome structure and each block has a number printed on the interior surface that locates the block in the assembly. The design of the domed structure maximizes the efficiency of the printer and the print volume; two blocks can be printed per printer without support material in just over an hour. The Star Lounge demonstrates a possible democratic future for rapid production and manufacturing of unique and custom parts. Specialized and expensive tools and equipment are not required; materials are inexpensive, ubiquitous and easily accessible and come in a range of material and color choices. While the patterning of the Star Lounge is directly tied to its geometry and assembly, it is reminiscent of Arabic star motifs and mid-nineteenth-century American quilts. 43
04 San Fratello Rael.indd 43
2018-10-18 4:11 PM
4.3 The Wave Curtain is a passive solar curtain.
MASS CUSTOMIZATION: CRAFTING SPECIFICITY Mass customization in architecture and design relies on manufacturing techniques, such as 3D printing, which can combine the flexibility and personalization of custom-made products with the low unit costs associated with mass production. Unique, one-of-a-kind building components can be generated quickly and economically from advanced 3D modeling software; a 3D printer can print thousands of mass-customized, or unique parts, just as economically and quickly as it could print massproduced, or same parts. Some advantages of mass customization in architecture and design for the built environment include: building façades that are designed specifically for their solar orientation and climatic conditions, interior partitions and curtains that can be designed for specificity related to programmatic or occupant needs and dimensions, and furniture that can be customized to suit an individual’s precise proportions, dimensions, or specialized needs. The Wave Curtain is a passive solar curtain (figure 4.3) that is designed to admit the low winter sun into the building interior and restrict
the direct, intense summer sun in order to help keep the interior cool. The curtain does this through the use of 3D-printed bio-plastic cylindrical tubes that vary in width and depth along the length and height of the window. Because the cylindrical tubes are hollow, one always has access to exterior views – even when the sun is being blocked – unlike a typical shade or curtain. The Seat Slug (figure 4.4) is a case study examining how unique, one-of-a-kind building components can be generated quickly and economically from advanced 3D modeling software. The seat is a biomorphic interpretation of a bench. The form is inspired by Flabellina Goddardi, the newest species of sea slugs discovered off the coast of California, and by the infinite tessellations of Japanese karakusa patterns. The Seat Slug blurs the lines between biology, technology, and furniture, and is a new twist on function and form. It is constructed from 230 unique, 3D-printed cement blocks that are coated with organic resins to create a reflective, finished surface. The Seat Slug demonstrates the potential for mass customization as 3D printers don’t care if they are printing 1000 identical or unique parts – the cost of the material is the same and the labor needed to make each part is the same. Custom parts can be manufactured without additional costs.
44
04 San Fratello Rael.indd 44
2018-10-18 4:11 PM
4.4 The Seat Slug is a biomorphic interpretation of a bench.
MASS CUSTOMIZATION: CRAFTING VARIATION 3D printing allows one to design and manufacture unique parts for highly specific functions as well as customized parts for purposes related to individuality, expression, and variation. Using a simple form and allowing users to modify the form through the deployment of adjustable parameters allows for the creation of new forms of expression. One can use the G-code2 itself to make exact modifications that radically change the outcome of the scale, density, texture, or surface quality of an object or building component. Simply changing materials but keeping forms also allows for the expression and individual preference. The Starlight (figure 4.5) combines the intention of creating a spherical structure out of the minimum number of parts with that of generating diffuse light through various woven patterns created by custom G-code. The Starlight comprises 20 hexagons and 12 pentagons and is a buckminsterfullerene, or “buckyball.” The difference between a buckyball and the Starlight is that the geometry of the Starlight extends beyond the ball to make stunning, light-filled cones. A fine-woven mesh
4.5 The Starlight consists of hexagonal and pentagonal cones made from fine-woven mesh created by custom G-code.
generated by the G-code creates a loopy, textile-like surface with a soft texture on each cone. The object glows because of the translucency of the material but also permits direct light to shine through the surface of the cone itself. The Starlight can be 3D printed in many different materials, such as copper, aluminum, and white bio-plastic. The GCODE.clay objects are fabricated using various clay bodies (b-mix with grog, paper clay, porcelain, basaltic clay with manganese, recycled clay, and local clays) to explore the creative potential of crafting digital code to produce vessels and tiles (figure 4.6). The exploration concerns itself less with the object’s shape or profile and more with the path that defines the movement of the 3D printer. Through this exploration, the 3D printer is pushed beyond the boundaries of what would typically define the printed object; a series of new objects with new expressions in clay have been created that are defined by the plasticity of the material, gravity, and machine behavior. One outcome of this experiment was the creation of textured surfaces that are reminiscent of textile knitting patterns. Typically, extrusionbased 3D-printed ceramic objects are defined by the striations of the clay layers on the object’s surface, 45
04 San Fratello Rael.indd 45
2018-10-18 4:11 PM
4.6 The GCODE.clay objects are produced by crafting digital code to produce vessels and tiles.
but in this case, the surface takes on the appearance of a knitted textile, with clay being looped, purled, and knotted as it droops away from the surface. Occasionally a “dropped stitch” causes a loop to pull away from the surface, making every print unique. ON-DEMAND PRODUCTION One advantage of 3D-printing architectural products is that they can be made on demand, so there is no surplus, no storage, and no shipping products around the world – printed parts, or digital files, can be sent to job sites where components can be fabricated in situ. In an era of disposable products, over-consumption, excessive energy use, and toxic materials, architects have a responsibility to the public, and the planet, to change our mindset about what our buildings are made of and how they function, by engaging directly with the manufacturing processes used to construct architecture.
WASTE-FREE MANUFACTURING The process of powder-based 3D printing consists of the spraying, or jetting, of a liquid binder material onto a thin layer of powder. The liquid binder solidifies the powder, and another thin layer of powder is rolled out over the top of the previous layer and the process continues hundreds, if not thousands, of times. Ultimately, after the 3D-printed part is complete, the three-dimensional object must be excavated from the loose powder surrounding it. This loose powder also serves as support material, which allows for overhangs, undercuts, and complex forms to be created. The object is then cleaned with a brush to remove the excess loose powder and the remaining powder is blown away or vacuumed off. The remaining powder can then be recycled and reused in subsequent prints, which means there is little to no waste when compared to subtractive methods of production. Also, recycled materials or materials that are already in the waste stream can be used in 3D printing.
46
04 San Fratello Rael.indd 46
2018-10-18 4:11 PM
4.7 The Sawdust Screen is made of pulverized walnut shells and sawdust.
In 2013, in the United States alone, over 42 million tons of wood waste was generated on construction sites. The sawdust from wood waste can be used to create 3D printable wood materials. 3D printing with sawdust has some similarities to recycled engineered wood products, but in many ways, it is also quite different. Whereas many current applications for upcycling sawdust into building products use equal parts sawdust and polymers, our 3D printed sawdust begins with nearly 85 percent of the recycled wood and cellulose particles. It is only after a 3D-printed object emerges that a polymer coating is applied, which gives the printed object a materially rich texture and surface in addition to its strength. The color and texture are ultimately a product of the wood species that is printed from. Pine flour produces objects that are lighter in color, and softer, compared to hardwood fillers such as maple or walnut. Surprisingly, the layers that are a product of the additive manufacturing process impart a grain similar to natural wood, as if the wood wanted to
return to its original state and express its internal growth. Sawdust isn’t the only material that can be used to 3D print wood-like objects. Nutshells, husks, and seeds are all agricultural waste products that can be ground into fine powders and flours and used to make 3D-printed objects that have similar colors and properties. The Sawdust Screen is made of pulverized walnut shells and sawdust, and retains the layering effect from the additive manufacturing process, simulating natural wood grain. The screen is comprised of individual 3D-printed wood components that are affixed together to form a variably dimensional enclosure and surface (figure 4.7). Its porous pattern is inspired by the vessels found in a microscopic analysis of wood anatomy in hardwoods. When viewed from the end grain, vessels demonstrate the porosity of wood. In a live tree, vessels serve as the pipelines within the trunk, a transportation system for water and sap. In the Sawdust Screen, the vessels serve as an opportunity 47
04 San Fratello Rael.indd 47
2018-10-18 4:11 PM
4.8 The Saltygloo is made from 330 3D-printed salt tiles.
for visual porosity. The Sawdust Screen is evidence that not only the process of 3D printing reduces construction waste but also that building components such as tiles and blocks can be made out of waste. CRAFT IT YOURSELF MATERIAL INNOVATION 3D printing is still in its infancy and there is room for innovation and development. Space to develop one’s own materials for 3D printing exists and Emerging Objects has taken advantage of this opportunity to develop materials for 3D printing that are sustainable, durable, and affordable. Materials developed for 3D printing by Emerging Objects include chardonnay grape skins and seeds, coffee, tea, curry, cotton candy, sawdust, salt, cement, calcium carbonate, and recycled rubber tires. Sustainable materials for 3D printing include materials that are recycled such as rubber tires, or ecologically derived materials such as salt, which require only sun and wind for production. Even cement can be made more sustainable through 3D printing. The concrete industry is one of the largest producers of carbon dioxide in the world because of the chemical processes required for the production of cement, and the burning of fuel. Concrete also uses unrivaled natural resources, such as sand and water and the formwork used to mold concrete is often thrown away. 3D printing with cement reduces energy consumption and completely eliminates formwork, reducing carbon dioxide emissions and raw material consumption. CIY (Craft It Yourself) materials, such as cement and sawdust, are also incredibly durable and strong and suitable for use in the building industry. 3D printed cement is stronger and lighter than typical concrete. Typical concrete cures to 3000 psi but the powder-based cement polymer cures to 4700 psi in compression when mixed with fiber reinforcement. Emerging Objects sawdust has a compressive strength of 902 psi, stronger than a typical 2 x 4, which has a compressive strength of around 600 psi depending on the wood species. These preliminary strength tests indicate that 3D-printed materials for building components have the potential for longevity and can withstand the forces applied to them in
applications such as building cladding, structural interior partitions, and as surface finishes on walls and ceilings. Proprietary materials for powder printing can cost up to $1200 for 20 lbs. Materials that are in the waste stream are almost free, and local materials that can be harvested in abundance such as salt result in a 99 percent cost savings. Salt The Saltygloo is an experiment in 3D printing using locally harvested salt from the San Francisco Bay to produce a large-scale, lightweight, additively manufactured structure. In the landscape of the San Francisco Bay Area, salt is a locally available building material. The salt is harvested from 109-year-old salt crystallization beds in Newark, California. From this landscape, a new kind of saltbased architecture was realized, created through the lens of 3D printing and computer-aided design. Inspired by traditional cultures that employ the building material found directly beneath their feet, Emerging Objects embarked upon a similar process. Named Saltygloo, because it is made of salt y glue, it is made of a combination of salt harvested from the San Francisco Bay and glue derived from natural materials, which makes for an ideal 3D printing material, one that is not only strong and waterproof but also lightweight, translucent, and inexpensive. The form of the Saltygloo (figure 4.8) is drawn from the forms found in the Inuit igloos, but also the shapes and forms of tools and equipment found in the ancient process of boiling brine. Additionally, each tile is based upon the microscopic forms of crystallized salt. The 330 3D-printed salt tiles that make up the surface of the Saltygloo are connected together to form a rigid shell that is further strengthened with lightweight aluminum rods flexed in tension, making the structure extremely lightweight and able to be easily transported and assembled in only a few hours; in many ways, it is salt. The translucent qualities of the material, a product of the fabrication process and the natural properties of salt, allow for light to permeate the enclosure and highlight its assembly and structure revealing the unique qualities of one of humankind’s most essential minerals.
48
04 San Fratello Rael.indd 48
2018-10-18 4:11 PM
4.9 The Rubber Pouf can be used as a low seat or footstool.
Rubber Recycled rubber tires can be also used to make 3D-printed parts. Recycled rubber is not a foreign material to the building industry. There are many very refined building products such as floor tiles, roofing tiles, and waterproof membranes that are made of recycled rubber crumb. The Emerging Objects rubber used for 3D printing comes from crumb that is ground into a fine rubber powder. The Rubber Pouf is a playful piece of furniture that can be used as a low seat or footstool (figure 4.9). The form of the Rubber Pouf resembles a sixpointed star with rounded heads on the ends of all six points. The pouf is printed in eight parts that are stuck together to make one solid object. The detailed, beveled texture on the surface of the pouf gives the appearance of button tufting, which makes the piece look padded and soft, like an upholstered piece of furniture. 3D printing has the potential to turn endof-life tires into high value sustainable products for the built environment, such as outdoor furniture. 49
04 San Fratello Rael.indd 49
2018-10-18 4:11 PM
4.10 Bloom is an experimental pavilion made from 840 customized 3D-printed Portland cement blocks.
Cement Bloom is an experimental pavilion that employs 3D-printed Portland cement at an architectural scale (figure 4.10). It is a 9-ft-tall freestanding tempietto with a footprint that measures approximately 12 ft by 12 ft and is composed of 840 customized 3D-printed cement blocks. The floral motif embedded in the surface of Bloom is derived from traditional Thai flower patterns and is mapped cylindrically onto the surface of the structure. This creates a figural pattern comprised of openings in the surface that produce stunning visual effects of light, shade. and shadow on the
exterior and interior (figure 4.11). On the interior one can find an internal structural grid that carries the forces of the weight of the cement blocks to the ground. The individual blocks were printed on a printfarm of 11 powder 3D printers, with a special cement composite formulation comprised of iron oxide-free Portland cement. Iron oxide imparts a gray color to cement, and its removal makes these 3D printed blocks much lighter in color. The curvilinear form of the overall structures provides added stiffness to the thin, lightweight shell. In plan, Bloom is a curved cruciform shape
50
04 San Fratello Rael.indd 50
2018-10-18 4:11 PM
4.11 Interior of the Bloom.
that rises 9 feet to meet the same shape rotated 45 degrees, creating a torqued “x” shape with an entrance 45 degrees off its primary axis. The undulating form and spaces recall an elephant’s foot or, when coupled with the flower pattern on the surface, the traditional mud houses of the Tiebele people in Ghana – a reference to the earliest inspirations for 3D printing by Emerging Objects. Bloom is an excellent example of how Portland cement can be used with powder-based printing and very little water to create intricate and complex cement structures that have strength comparable to more traditional concrete constructions.
CRAFT IT YOURSELF (CIY) ARCHITECTURE Digital craft and 3D printing allow every tile, block, and panel to be unique and customized, but a well-crafted future doesn’t stop there; it also includes the crafting of hardware, software, and tools that allow for the production of personal and meaningful buildings and experiences. The San Francisco Bay Area is experiencing a housing crisis; never before has the cost of real estate, new construction, and rent been so high. Because of this emergency situation, cities around the Bay Area have relaxed their zoning laws, design review 51
04 San Fratello Rael.indd 51
2018-10-18 4:11 PM
4.12 The Cabin of 3D Printed Curiosities is composed of 3D-printed Planter Tiles.
process and permitting requirements in order to allow home owners to craft secondary structures on their lots. These relaxed laws are opening the door to experiments in Craft It Yourself (CIY) Architecture. The Cabin of 3D Printed Curiosities (figure 4.12) takes advantage of these relaxed codes and laws and brings many of our material, software, and hardware experiments together to demonstrate the architectural potential of additive manufacturing on a weathertight, structurally sound building. All of the cabin’s componentry are produced in a micro-factory, the print farm, which is located near the site of the cabin. The front façade has been described as a box of exquisite chocolates – it is composed of 3D-printed
4.13 The 3D-printed Planter Tiles create a living wall of succulents.
Planter Tiles to create a living wall of succulents that naturally thrive in the Northern California climate (figure 4.13). Several different materials are used to craft the tiles, including different shades of Portland cement, sawdust, chardonnay, and combinations thereof. The roof gable and east and west façades are clad in 3D-printed ceramic tiles that serve as a rain screen (figure 4.14). Designed for easy assembly, these tiles are made to be hung on a building façade or interior. The surface of each ceramic tile visually emulates a knitting technique called the seed stitch. G-code is used to control each line of clay as it is 3D printed to create a loopy texture that looks
52
04 San Fratello Rael.indd 52
2018-10-18 4:11 PM
4.14 The roof gable and east and west façades are clad in 3D-printed ceramic tiles that serve as a rain screen.
4.15 Interior of the Cabin of 3D Printed Curiosities.
like seeds scattered across the surface. While all ceramic tiles are printed from the same file, each tile is intentionally unique as a product of fabrication, during which the tiles wave back and forth, causing the printer to pull at the line of clay and creating longer and shorter loops toward the end of each tile, producing a distinct machine-made texture that is different every time. The interior is clad with 3D-printed bio-plastic panels that are backlit with LED lights (figure 4.15). The bio-plastic is inherently translucent and glows to illuminate the interior. The interior also contains an assortment of chairs and tables that have been printed using different materials such as bio-plastic
and nylon. Other objects in the interior, including the coffee set and wine goblets, are 3D printed using coffee grounds and chardonnay grape skins respectively. The Cabin of 3D Printed Curiosities represents the first steps to a crafted future that we can embrace for both its functionality and beauty. It solves problems around architectural issues that we face every day, it brings value to the occupants’ lives and it has a strong connection to its locale technologically, geologically, and agriculturally. The cabin speaks to the possibility of crafting a future that is just emerging, a future which inventors, designers, and creators have the power to customize and produce, which is both exciting and important. 53
04 San Fratello Rael.indd 53
2018-10-18 4:11 PM
NOTES 1 Emerging Objects, co-founded by the authors, is an independent, creatively driven, 3D Printing MAKE-tank specializing in innovations in 3D printing architecture and building components. For more information, see www. emergingobjects.com. 2 G-code is the common name for the most widely used numerical control (NC) programming language that is used primarily in computer-aided manufacturing to control automated machine tools.
54
04 San Fratello Rael.indd 54
2018-10-18 4:11 PM
5 MASSIVE CUSTOMIZATION
MARC FORNES 55
05 Fornes.indd 55
2018-10-18 9:35 PM
Computation provides one avenue to customization. That has been the point of approach for THEVERYMANY,1 where a computational and prototypical methodology drives many other aspects of design practice, as a way to iterate through many potential built outcomes. Mass customization – often understood as the ability of the consumer to choose to the nth degree – is not a goal of this process of design and fabrication. Scaling beyond the possibilities of individual choice, we aim at something that might be called massive customization. What can be designed when one cannot conceive all possible outcomes? How does computation allow one both to approach the undefined bounds of iteration and to build from unforeseen combinations? How does this work meet a public that may not recognize it as architecture?
its total output. It is not a “black box” out of which some unforeseen form materializes, but between these explicit steps is a phenomenon of resonance, an ambiguity that is due to the sheer number and shared dependencies of many explicit steps. Taking code literally, we can understand it as both a cipher and a de-cipherer. It writes and defines our actionable elements, but it also reads everything in sequence: from the top to the bottom and sometimes back and forth. In the end, it is very difficult to anticipate the effect of all the accumulated and ordered conditional statements and the exceptions that are written into a code sequence. The tension between precise, explicit protocols and the relative indeterminacy of their results is how we have come to understand THEVERYMANY: as the curious sum of many integral parts.
EXPLICIT PROTOCOLS, PRECISE INDETERMINATION
CUSTOMIZING ARCHITECTURAL TOPOLOGIES
All the experiments described here are a result of explicit and encoded protocols. Explicit, clearly written protocols enact transformations to geometry through algorithmic instructions or even mathematical expressions. Protocols are also encoded because they must be formatted in a particular way: they are compiled in a text file in specific, computational syntax. That is how these transformations are made executable. An objective of this type of protocol is not an expected outcome, but what might be qualified as precise indetermination. While precision and control are integral to a rigorous process of experimentation, its ability to produce unforeseen results lies first in the ability to execute consistently. One should be able to run the same code twice and get the same result in both instances. That is a requirement that one must be able to debug. Eventually, the protocol can implement a solution. But indetermination emerges from a succession of clear, deterministic steps that are prescribed in the code. The author of the code must be in control of the logical process of the code itself down to the last comma, and yet cannot anticipate
One element of architecture that can be isolated for computational inquiry is the surface. In a few lines of code, one can get something that appears to be quite complex. For example, using a simple process of inflation or relaxation on a mesh, one can easily generate a complex, doubly curved geometry. For an architect, especially one with a constrained budget, the looming question remains: how do you manufacture that out of simple, flat elements? This is a consideration that must be made at the outset of a project, rather than as a post-rationalization. By negotiating the segmentation and tessellation of these parts early on, we can establish the ability to customize how a complex mesh can be manufactured and assembled. Approaches to decomposing computational forms have changed from the parametric experiments of the 1990s and early 2000s. The 1990s paradigm reduced for construction the doubly curved surface as a sum of triangles. At the end of the 1990s and into the 2000s, when computation became more broadly accessible, designers and specialists were interested in solving the issue of the sharp aspect ratio of triangles in the case of glass panels, for
56
05 Fornes.indd 56
2018-10-18 9:35 PM
5.1a and b n|Edge was installed at Galerie Roger Tator as part of “parcours raisonance” of the 2009 Lyon Art Biennale.
example, by approximating complex surfaces into arrays of planar quads or quad meshes. There will be n corners. The n-sided polygon is one of the very first descriptions in an early project of ours, n|Edge (figures 5.1a, 5.1b), which describes a surface into sets of panels with three, four, five and up to n edges. The premise of the project was simple. There was no budget; it was a self-commissioned exhibition in a small gallery in Lyon. But it has come to define a way of working with a topology through parts – parts that are described on a mesh, unrolled, and ultimately nested. In the gallery, we arrived with a suitcase of parts, cut from nested profiles on flat sheets of very thin aluminum, and reassembled them on site. As an interrogation of process, the project can be described as a sum of parts and operations: 2 surfaces (top & bottom) 2 people traveling with four suitcases from New York to Lyon 81.6 kg – 38.274 m2 – 22 4’ x 8’ sheets of gold reflective aluminum 6 days to CNC cut 2796 individual panels (from 2 to n edges) 2796 tags (3–5 digits) 5375 holes 6500 rivets (or 13 boxes) 5 rivet guns 10 days assembly 10 people assembling
What looks like a success (or at least successfully assembled) is actually the first failure in our attempts at customization. For n|Edge, we were able to recompose the surface computationally, design it with endless variation, describe it with multiple parameters, unroll it, and digitally fabricate it. However, it still took ten days and ten people to assemble the nearly 3000 components. This excess of parts requires an impractical labor of reassembly, a condition that inhibits the ability to grow this kind of topology to the architectural scale. Our research from that point on acknowledged a misapplication of the studio’s name: we needed to go from the very many to the very least. At some point, you encounter a limit: labor time, quantity of materials, crates for shipping, and cost are all familiar constraints. The project of realization has to give in to these limits, of what can be produced and ultimately appear for public enjoyment. In the face of massive singularities, these limits might also be productive. How can we try to produce such a complex geometry while reducing the number of constituent parts and collapse the time and space between digital and physical? We first considered a protocol of recombination. Once a mesh is described or tessellated according to selected criteria (for example, locating smaller elements where there is the tightest curvature), those singular parts can be recombined into larger unique entities rather than accepting the sheer number of
57
05 Fornes.indd 57
2018-10-18 9:35 PM
singularities. Linear recombinations of individual panels as chains quickly evolved into stripes. The sum of developable singularities can produce potentially developable linear elements if there are no issues of self-intersection when unrolling. Moving toward recombination was an initial “revolution” that not only dramatically reduced our number of parts, and with it assembly time, but had the secondary effect of increasing the structural properties of the assembled topology. Linear recombination increases the instances of connection from one part to another as continuous elements, producing redundancies that add rigidity necessary for self-support. In pseudocode, the process reads: start at one point, get its neighbor, stop wherever there are too many, and when all the parts are taken, the process stops. Instead of 3000 unique parts, you can define 300–400 elements, which dramatically reduces the time to move between the digital and physical. It is a process of refinement and control. However, even as we institute more extensive control through the use of parameters – typically referred to as parametric design – we are limited by what we can describe explicitly. The architect’s education often does not include a comprehensive training in descriptive geometry and mathematics. But we can compensate for this with numbers – not with one or several numerical inputs, but by unleashing a huge population of variables. Computation and search protocols handle the manipulation of many, many elements to do this job. SPEED = DESIGN Imagine a set of ants crawling along a surface. They are what we call agents, or encapsulated sets of rules. Those agents proceed according to a precise behavior, and when they find no way forward, they “die” and that trajectory becomes geometry. One wants to push the conversion to geometry as far into the workflow as possible, because computation is much faster at executing mathematical expressions than geometric instruction. With multi-agent behavior, one can maintain speed. If a
protocol can be executed swiftly, one can get many iterations for a single problem, and therefore, make an informed design decision. Defining rule sets is easy. However, in trying to debug a protocol of description for a complex geometry, a specific behavior may solve a problem locally only to create new ones elsewhere in the geometry. The author of a protocol may be able to define best-fit behavior at a specific instance, but there is no global understanding of the entire system. Instead, parallel computing by multiple agents can be deployed to evaluate the whole system at each loop and exchange feedback. So rather than a single set of rules, one can describe a geometry through multiple, competitive rule sets – agents with occasionally conflicting motives – which produce a kind of schizophrenic behavior. In our project for the Centre Pompidou in Paris (figure 5.2), the geometry is not black and white for aesthetic purposes, but rather expresses the very nervous behavior of those stripes.2 In this case, protocols feed agents local parameters in reverse order of best-fit. The solution that first passes the test is not necessarily the best, but rather the first acceptable one, which then triggers a best average solution. MASSIVE CUSTOMIZATION The dual concerns of speed and search bring us closer to the topic of mass customization. From this perspective, the interest in the issue of customization has less to do with a variably choosing mass. Rather, the search for customizable rule sets may produce something that the public has not yet seen and could not possibly have selected from a defined set of options. A basic understanding of “mass customization” is couched in consumption: you go to BMW or Porsche, for example, and you are presented with a number of features. You have a choice of color, the leather type, the stitches, and so on. These choices have, in the end, a finite range, defined by very basic math: the number of parameters is exponential to the number of variables. You just need to identify
58
05 Fornes.indd 58
2018-10-18 9:35 PM
5.2 Commissioned for the Centre Pompidou’s permanent collection in 2012, Y/SURF/STRUC is a non-linear morphology with a non-linear description.
your parameters and your variables; the logic of the relationship is always the same. You can determine quite easily how many species of the Nike running shoe exist because it is written within the topology. Between the car or the shoe and all the possible combinations of their constituent design features, the topology is constant. But what we are able to define through massive customization is the customization of rule sets. In that case, we don’t have to know
the maximum number of permutations there are to describe a surface, for example. Still we run the search protocol, and it returns a number of discrete parts that we have not necessarily anticipated. Of course, when we consider the number of sheets that these parts will have to be cut from, we start to care about the quantity of species. Whether we reduce or maximize these singularities, we end up with a sum of parts that can assume other kinds of variation.
59
05 Fornes.indd 59
2018-10-18 9:35 PM
5.3 A mixed cheshire gradient plays across the shingles of this permanent folly in Edmonton, Alberta (2014).
PART-TO-PART VARIATION There is a quote from the art director Paul Rand that looms large in the design field: “If you can’t make it interesting, make it big. If you can’t make it big, make it red.” The world of THEVERYMANY is a colorful one. And though the public might be divided on bold choices like red, we are living in the age of millennial pink. Our pink is a hot pink, but we use it in concert with other, variable palettes to achieve what we call coloration: a procedural art of applying color across a set of parts (figures 5.3, 5.4). You might perceive such a surface morphology to be green or blue, or maybe you can’t discern the exact color. You approach it and then your eyes focus, and you start to understand the result of a gradient. Linear gradients, non-linear gradients, or one set merging into the other might include a dozen separate shades. Together, they render the surface ambiguous, indeterminate, up for debate. The viewer is not asked to choose or decide what it is or might be, but is invited to spend time trying to understand it. At some point, it becomes an issue of resolution. 5.4 This scheme for a pavilion in Miami integrates juxtaposed gradients across sets of non-linear stripes. Points of intensity emerge, which make it difficult to read its overall logic.
60
05 Fornes.indd 60
2018-10-18 9:35 PM
5.5 Labrys Frisae, in assembly at The Rotunda Space, Art Basel Miami (2011).
RESOLUTION Resolution, a further concern of massive customization, is registered in a project in white: Labrys Frisae, a project for Art Basel Miami, tested our ability to scale systems with a fine resolution of many unique parts (figures 5.5, 5.6). It consisted of 10,300 unique parts that took three people six weeks to reassemble. The initial base mesh, decomposed into flat, machinable elements, was put back together on site in a parallel process of cutting and assembly. Every four days we received another package and continued building up. This structure
topped out at 18 ft tall and barely 30 ft long. It is simply not possible to apply this kind of customization to the architectural scale at that resolution or pace. Again, it comes down to speed and singularities. Thus, that project initiated a shift in our practice of describing systems of parts. Rather than defining a set of unique parts that produce intricacy and idiosyncrasy through their sum, what if the part itself takes on other variations? What if we cross the protocols? What if there are two sets of agents? One set of agents finds the parts for the purposes of fabrication and assembly, and another set of agents describes a new behavior within the parts.
5.6 Labrys Frisae, completed.
61
05 Fornes.indd 61
2018-10-18 9:35 PM
5.7 Situation Room at the Storefront for Art and Architecture in New York (2014).
INTRICACY WITHIN AND THROUGH PARTS This layering of protocols was pursued in Situation Room, a collaborative project for the Storefront for Art and Architecture in New York (figure 5.7).3 Spherical shells of incremental diameters were aggregated within the gallery’s narrow plan to envelope an interior in sound, color, and light. Across its surface of conjoined spheres, stress flows are rendered as a secondary condition across the parts that make up the continuous surface. They result in porosity, in cuts that help transmit light and give visual cues to what is happening on the surface. A structural engineer might be able to tell that these aren’t stresses on the surface alone, but they flow in relation to transducers mounted on it. Sound is transmitted from them directly onto the surface, and by vibrating the aluminum, sounds actually are emitted from the surface. Although it is composed of many unique, interconnected stripes, the objective of this
experiment was not to express the sum of many parts. The fluorescent pink attempts to eliminate the apparent differences between parts. The contrast is reduced to the extent that the eye is not able to assess distances or distinguish edges. When you are in the space, it is difficult to discern its depth, which elements project toward you, and which surfaces recede. Between the curvature, the color, and this blending of elements, you arrive in an environment that cannot be described by conventional architectural vocabulary. There are no walls, ceilings, or floors, per se, but something less familiar. Alongside the development of these morphologies, it becomes necessary to engage a new lexicon to convey what this architecture is and how it is made. There are elements of pleating and splitting and recombination, moments of porosity, of density and intensity, and the component parts might be shingles or stripes or something not yet defined.
62
05 Fornes.indd 62
2018-10-18 9:35 PM
5.8 A close-up of the Tour de Force(s), Harmony of the Seas, Royal Caribbean Cruises (2016).
SCALE As the protocols and effects start to naturalize a new architectural vocabulary, the question remains: how can a massively customized structure grow in scale? An opportunity arose to develop an architectural skin for the interior of The Harmony of the Seas, the largest cruise ship in the world, when it was released in 2016. On this floating city, which hosts 7500 guests and 3500 crew members, the space in which we were asked to intervene vertically spanned 16 of the ship’s decks, a columnar atrium about 40 m or 120 ft high. For the first time, we designed a structure held in tension, where we typically design for compression. The more interesting driver of the Tour de Force(s) project was its constraints: we had four days to assemble a very large structure that merged two morphological systems. We used the old paradigm of the singular element – unified into a continuous surface of 2-mm-thick aluminum – and a system of stripes which are alternately networked
across the surface to connect those smaller parts which are not joined in any other way. Hexagonal parts are held in place, edge-to-edge, while nonlinear stripes snake across both sides of the torqueing skin to hold them in place, while also allowing the surface to assume the necessary double curvature (figure 5.8). Rivets reveal the pattern of stripes on the opposite side. At this resolution of many small elements, the surface easily accommodated the desired double curvature without any manual folding and at a thickness that resisted the cold bending we depend on in other projects with larger and thinner pieces. If you look very carefully, each of those parts is a singular, hexagonal element that meets edge to edge, with a slight gap in between. That tolerance allows for subtle repositioning from part to part to assume the overall curvature. The linear connectors join them, and also take the torque necessary to span the surface in tension. 63
05 Fornes.indd 63
2018-10-18 9:35 PM
5.9 Marquise, self-supporting architectural skin in El Paso (2018).
we can still bend them and not increase our radius of curvature – otherwise we lose our structure – but thick enough to resist this pin problem, among other stresses of production. PART-TO-PART RELATIONSHIPS: FOLDS, OVERLAPS, CELLS
Tour de Force(s) was conceived as a potentially scalable architectural skin. Because it was designed in tension, its thickness does not need to scale with it. Artificially thickening it, however, provides resistance to the stresses of transportation and the fast-paced marine riggers in the ship’s atrium. The piece had to be resilient in both its assembled and modular form. Of course, growing in scale through this system means that we had to deal with many more parts: 47 pre-assembled sections of 10,813 parts arrived on site in 1 x 1.25 x 2 m crates. Retaining our usual thickness of a few millimeters, large-scale structures remain extremely light and thin. Beyond the concerns of their delivery and assembly, we encounter something in very thin structures called the pin problem: If you concentrate all your force onto a single point, that point of connection can actually puncture the surface. Across these experiments in custom parts, we must still calibrate the performance of the whole system at the level of the part itself. Parts must be thin enough that
Another self-supporting architectural skin, Marquise, completed in El Paso in 2018, demonstrates a different relationship between parts (figure 5.9). Approaching from the front, the surface of the structure appears to be made of shingles – what we have called structural shingles in other projects, i.e. parts that overlap to connect. That individual piece appears to be a continuous stripe from the interior, with integral tabs that bridge across others from the front. That strategy of overlap allows us to multiply the material thickness locally, without increasing the thickness of a single stripe. At any point, there is twice the material: one stripe in the back and a target cell overlapping it to resist concentrated forces. The overlap is facilitated by a fold, an additional process in the fabrication and assembly. There is no primary structure in this project; the structural performance is accomplished through the geometry and aided in small ways by manipulating parts. These maneuvers add to the time and labor of processing, but the fold is another condition that is negotiated to produce variation, to lend to the structural performance or impact the construction. These customizations are engineered to work with the demands or advantages of a given project. The connection detail, which is also structural, provides an opportunity to introduce and manipulate coloration. Across two directions of tabbed stripes, we have assembled a gradient of five and six colors, a detail that contributes to several scales of perception. From afar, it strikes you as a flattened pattern, but closer examination reveals a gradient, one made through a system of parts. Zooming even further in, the mechanical connector becomes apparent, and lends to a much different reading of the structure on its interior.
64
05 Fornes.indd 64
2018-10-18 9:35 PM
5.10 Pleated Inflation, a large, airy pavilion in Argeles-sur-Mer (2015), takes shape through mesh inflation, and is re-composed through a system of shingles.
Coming back to the issue of scale, these parts and the connections between them must also selfsupport through the process of assembly. We have developed a strategy of cell-scaffolding, another advantage of a system of parts. Easing the process of assembly, self-supporting components eliminate the need for supplemental supports or molds. The parts not only take on strength as a whole, but they are also able to stand up if the arches are not yet closed. They are even able to carry the load of the installation team, who can stand or walk on the surface as it evolves. In a large project that was not assembled by our team – a first for us – we had to use a substructure – another first. Building on a smaller shell structure for a school in France, we designed an amphitheater based on an inflated mesh, with intensified pleats to maintain structural performance across a much larger area of double curvature (figure 5.10). This bandshell was designed to be self-supporting, but as a system of shingles, it
appears as the cladding over a steel structure that is only there to hold lights and other theatrical rigging. The project nonetheless can reveal other forms of customization. In The Chrysalis, a bandshell in Columbia, Maryland (figure 5.11), we were intentionally ambiguous about the program of the structure, so that it can take on many different guises and occupations. The aim was to stage the experiences of the morning jogger, the Sunday walker, the afternoon stroller, and someone there for a show too. It is an amphitheater, but also a pavilion in the park, a tree house, a public artwork, ready to be engaged in any number of possible ways. In the end, we are interested in experience. How does someone perceive a surface of intricately linked parts? How does the color strike them? What does the porosity reveal about the geometry and its analysis and what other effects does it produce in situ? All those motives are derived from computational protocols, but we must consider to what end they are applied. The outcomes of these
65
05 Fornes.indd 65
2018-10-18 9:35 PM
5.11 The Chrysalis, a bandshell in Columbia, Maryland (2017), is fully outfitted for high-production concerts and events.
efforts must amount to something greater than the sum of parts, or the unfamiliar result of many coincidental rule sets. It has to activate space and experience and play. THE MASSIVE, THEVERYMANY, THE VERY LEAST Despite the name of the studio, we do not have a fascination with the quantity of parts. They are necessary to describe the radius of curvature, which in turn defines the structure. The aluminum material and the mechanical rivet are also a reliable medium to produce this geometry in the most economical way, and not necessarily the desired medium. Ultimately, the client determines what is acceptable for the site and its use, and what is the most economical protocol for achieving that. In the case of our collaboration with Louis Vuitton and Yayoi Kusama for a Pop Up store at Selfridges in London, the client had very specific
needs but we used their constraints to develop our research on the materiality of these unique parts. We competed to do a series of pop-up stores and we were able to produce an environment of 1 mm shells out of carbon fiber – the first architectural application of a completely carbon fiber structure. We decomposed the carbon fiber shells into slices (figure 5.12) that could be fabricated and carried into the store. So how did we end up there? People often dismiss carbon fiber because of its cost. What we were able to do for Louis Vuitton was to simplify the production to the extent that we could fabricate, transport, and assemble a room-scaled structure in very compressed timeline. What makes carbon fiber so complicated and expensive to produce as a composite is that it requires a mold; the primary challenge is to get rid of the formwork, reuse it, or adapt it. Our strategy was very simple: we used the cheapest mold ever. It is a table, 7-m-long with a very standard finish on the top. On it, we infused a 7-m-long carbon fiber sheet, which was water-jet cut
66
05 Fornes.indd 66
2018-10-18 9:35 PM
5.12 Fabrication of the carbon fiber shells for the Yayoi Kusama x Louis Vuitton Pop-Up at Selfridges London (2012).
67
05 Fornes.indd 67
2018-10-18 9:35 PM
into slices, with a motif in the language of Yayoi Kusama’s dots. It was placed onto an inexpensive, low-density foam to cure, which also lowered the cost dramatically. The ruled surface is bonded together at the edges and embedded within the bond are cables that power the lighting, cameras, and all of the electronic equipment needed in this retail environment. In total, five subshells were brought together as a set of merged, corrugated spheres. Each slice is unique because they were digitally cut, but very manageable to cure and join. The particular advantage of these carbon fiber sections is that they were small enough to fit through Selfridge’s doors, rigid enough to maintain their shape pre-assembly, and light enough to be carried in by a single person, which had to be done after hours over the course of four nights. Everything is made using that process. There is no additional steel, no other furniture. The shells, tables, lamps are all products of the same material system. From thousands of individual parts, we arrived in London with just 178 carbon fiber slices and only 144 unique types.
NOTES 1 Marc Fornes leads THEVERYMANY, a New York-based studio specializing in large-scale, sitespecific structures that unify skin, support, form, and experience into a single system. 2 Commissioned for the Centre Pompidou’s permanent collection in 2012, Y/SURF/STRUC, is a non-linear morphology with a non-linear description. While the structural performance of hyper-thin surfaces or volumes relies on their extensive or overall double curvature, intensive curative process constrains the maximum radii of that curvature. This constraint produces recombinations, curls or splits in a morphology. Those network elements are needed in order to properly structure a self-supporting shell. 3 Commissioned by the Storefront for Art and Architecture in New York, Situation Room was a collaboration with the sound artist Jana Winderen to create a disorienting environment that enacts a dialogue between sound and structure, the familiar and the unfamiliar.
IN CONCLUSION Massive customization is not a process of multiplication. We are not trying to produce exponentially within a finite number of possibilities. The very specific procedures of customization lead to something greater than any quantity – something most adaptable to changing goals or project constraints. Moving beyond parametrics and mass customization, models that are now long established, we try to use computational protocols of search to unify many diverse concerns. In doing so, we don’t need to enumerate all of the possible solutions, but rather, conjure something we would not have foreseen, and quickly. Massive customization is a way to produce an unknowable future, but also one we have the power and resources to realize.
68
05 Fornes.indd 68
2018-10-18 9:35 PM
6 CONTINUING TOWARD EXTREME MASS PRODUCTION
GREG LYNN 69
06 Lynn.indd 69
2018-10-18 9:44 PM
Since the early 1990s, when architecture adopted both procedural modeling design tools and the ability to speak to manufacturing machines from 3D printers to CNC-controlled lathes and mills, there has been an attack on the modular and a desire for the digital bespoke. Architecture is one of the few industrial professions where tools are employed to make generic components but the actual construction of industrialized buildings is one-off. Therefore, architects were the first to believe they could apply this knowledge of one-of-a-kind industrial processes to commercial products such as fashion, jewelry, housewares, vehicles, furniture, and athletic apparel. I vividly remember being invited by Volvo, along with Sanford Kwinter and Lindy Roy, to discuss how to place architects between car dealerships and factories to leverage exactly this model of massproduced diversity. I have been preaching this vision for more than 25 years when I first 3D printed more than three dozen models and exhibited them at Artists Space in New York. At the ANY Conference at the Guggenheim in New York, when I first showed the Embryological House (see figure 6.2), the first question was from Peter Eisenman: “Greg, please tell me which one is the best and what are your criteria for discrimination?” I have tried the mass-produced userconfigured bespoke with major lifestyle and design brands such as Alessi, Nike, Swarovski, and even companies I own and creatively direct and, much to my chagrin, the reality is that people want what I think is the “best one” rather than “their personal one.” What I never considered is that, more often than not, consumers do not desire bespoke products. Architects retooling mass production are supplying a need that has not yet been identified as a market. It may very well be that democratized design, let alone the design of one-of-a-kind mass-produced objects by well-known designers, is not what consumers want today. This does not invalidate the innovations in architecture that allow for an even greater degree of industrialized variation in building componentry, but it may be that the dream of the architect of shoes or cars to interpret the desires of individuals
and translate them into the instructions and documentations for the factory to produce bespoke consumer objects is only that, a dream. MY QUEST FOR THE PERSONALIZATION MARKET Mass customization is connected to the digital ecology of design and production. The first issue may be production and the desire to output things from the screen into the physical world. The combination of a desire for output, which is different than manufacture, and the desire for mass customization, is not accidental. In 1995, I had an exhibition at Artists Space in New York that was composed entirely of stereolithography prints. The show traveled to the Henni Onstad Arts Center in Oslo, Norway, where every single one of them was stolen on the night of the opening. They were insured and we 3D printed more. That was a new experience for me at many levels. In the early 1990s, I was very involved with 3D printing very small, expensive, resin models; that opened up the idea of labor-free, identically reproducible objects made by machines. It was like an industrial process, only on a miniature scale and without requiring tooling, a factory, a workforce, etc. Because the workings were on a small scale, I invested very quickly in a CNC milling machine that is still running 27 years later. I also bought a laser cutter several years later, which I came to very late, but realized it is very practical compared to the 3D printer and CNC mill. All these things got me talking to power tools with digital models. Every year the job of architects becomes more and more about talking to machines; soon the machines won’t need as much instruction. This is an exciting time to be designing with new possibilities to talk to machines without having to go through the painful process of industrializing objects, or what most people would call “design for manufacture.” 3D printing and CNC machining avoid many of those problems. Architects don’t know what they don’t know about the pain of that process. So, this is how production of models and prototypes changed for me with the digital ecology that was emerging in the 1990s. The next issue
70
06 Lynn.indd 70
2018-10-18 9:44 PM
6.1 A half-scale model of the Embryological House.
contributing to mass customization is the changes in the design process that involve procedural operations and tools. It was fun to advocate for architects and the experience and sophistication they bring to design processes that are procedurally based. When I started using digital tools from the animation industry, I became concerned as it was then that I realized that to be successful in architecture, you need a signature and the software was providing the signature; this was threatening, or at least concerning. I am very curious how technology impacts the signature. For example, with a manufacturing technology, when Adidas gets into 3D knitting, how does that affect Nike and how can you tell the difference on the shelf at Footlocker between the two brands, when they both move to 3D knitting technologies for shoe uppers? I wanted to make sure that the software I was using and the technology I was using didn’t dilute my creative role in the design. As a design exercise in what would have
to be considered mass customization I initiated a project called the Embryological House (figure 6.1). By the way, it was never meant to be a house; it was a response to Vitruvius’ statement that “an ideal house is a house to which no element can be added or subtracted without violating the perfection of the whole.” It was intended as a test of the variation and rigor possible with software; it was a test of signature using the extreme digital capabilities that were available at that time. Now I would use arrays of GPUs1 and machine learning, but back then I was using what was the state-of-the-art: two Silicon Graphics workstations (one Indigo and one Indigo 2) running overnight as design tools. The Embryological House begins with a controlling geometry of seven compatible spline curves that were interchangeably tested with lofted surfaces together to define the exterior envelope of a house. These curves were drawn in Microstation because its spline curves were better than those
71
06 Lynn.indd 71
2018-10-18 9:44 PM
6.2 Versions of the Embryological House produced by procedural variation.
in Alias and Maya at the time. From there, these curves were defined in combinations in a Microsoft Excel spreadsheet that was used to automate the lofting of more than 50,000 surfaces by using Maya’s MEL Script Editor. Each of the resulting surfaces is unique, yet each comes from the same library of curves designed to be compatible with one another both formally and spatially. Procedural operations to offset the surface to produce ribbon windows, shading, and primary structure could then be applied to this controlling geometry in an automated procedural fashion. Each house had the same number and arrangement of parts, yet they were all unique in their specific size and shape (figure 6.2). Because of the variety of results, I decided to make a website for the project2 that lets a user configure their own version of a house from very limited variables by picking colors, forms, windows, and cladding. Decades later I learned
this is called a “configurator.” In order to do this, we had to render tens of thousands of images and program their relationships in HTML. My guess is that I spent more time programming the website than I did designing the project. I was fortunate enough to get a grant from the Wexner Center who wanted to include the project in an exhibition. I picked up the project again and decided to produce models of all 50,000 versions of the house because I had only seen, at most, a few thousand versions on the screen and I had never seen them physically. The design process of procedural variation led directly to the desire I mentioned previously to produce physical artifacts from the digital ecology on the screen. We started 3D printing with stereolithography; we started CNC cutting molds and vacuum-forming plastic sheets; and we produced larger and larger models, finally we produced one for the Italian Pavilion at the Venice Biennale of
72
06 Lynn.indd 72
2018-10-18 9:44 PM
Architecture that weighed more than a ton and was several meters long and tall (see figure 6.1). The entire project is at the Canadian Center for Architecture now; luckily for them, we only managed to produce a few more than 2000 of the roughly 50,000 potential designs. I got into some trouble with the architectural community because people were asking me: “Which Embryological House is best? Or, how do you decide when to stop?” My response was always: “They are all like my kids, all 50,000 are beautiful; although I have seen all my kids and I have never seen all 50,000 house designs, I can say with absolute confidence that they are all perfect in their own way.” For me, it was a way of authoring the design and having a signature with an extreme number of unique objects, all of which were designed using procedural tools linked with methods of mass-producing physical models. I felt like I had mass-customized something for which
I had a consistent and perceptible signature. Still, you would identify an Embryological House when you see one and you would also recognize copies and buildings that have been influenced by the design. For me, this is important as it makes this a design problem and not an industrialization problem. So, I had a technique for mass customization that involved both design and production skills. I was also satisfied that I had an attitude and signature related to this, so that it wasn’t just mindless variety without intelligence or talent. I was already seeing lots of mindless variety – and even publications dedicated to mindless variety – and I wanted to steer clear of vocational digital design that was becoming popular at the time. I assumed it would be easy to find an audience and desire for serial differentiation with coherent family relationships and consistency. It could still work with a brand and a collection and was not just a mess of options. 73
06 Lynn.indd 73
2018-10-18 9:44 PM
6.3 Different designs of a coffee pot for Alessi.
WORKING FOR CUSTOMERS VERSUS WORKING FOR CLIENTS Very soon thereafter, I was one of approximately 20 architects asked to design a coffee service for Alessi. I had never designed a product; my mentors and heroes, however, all had designed products for Vitra, Alessi, Swid Powell, and others. I was given a budget of $70,000 per service set; the assumption was that I would 3D print a model and this would be given to artisans in Crusinallo in Italy, who would make them one-byone in sterling silver. The first thing I said was that for $70,000 I could make carbon molds and explode the vessels in super-formed titanium. Because it was one-use tooling, I designed a collection where no two would be the same, just as I had done with the Embryological House (figure 6.3). I designed
a series of curves and studied their compatibility without generating inflections, self-intersections, and other forms I found unacceptable. I then built a series of CNC-manufactured tools and vacuum-formed plastic shells to test their ability to be manufactured as well as their ergonomics while holding them. Alberto Alessi was an important early mentor of mine because of the great things he does with designers when introducing a brief to a project. He takes you to a little museum next to his office and walks you through all the projects he has ever done, while explaining his five principles of successful design, of which he says you only need to try and satisfy three in any one project. (I am not sure if they are secret, so I will not repeat his five principles.) Related to one of the principles, he told me that architects never put handles on things and that people burned their hand on coffee pots designed
74
06 Lynn.indd 74
2018-10-18 9:44 PM
6.4 A coffee and tea service set for Alessi made from titanium.
by architects. I got the message not to forget the handle. In fact, I made a design that was all handles (figure 6.4). The surfaces were dictated by grip; every surface was studied in physical prototypes, maybe 40 or 50 of them, for ergonomics. I showed the designs to Alberto and I said: “It’s amazing; instead of your collectors paying $70,000 for a teapot, they are going to get a one-of-a-kind tea service; every one is going to be unique.” He replied: “I think that’s a bad idea, but let me ask our dealer.” The dealer responded: “No collector wants a one-of-a-kind anything, because somebody else’s could be a little bit better than theirs.” It turns out that collectors trust designers; they are also very vain and want to know they have the best one and so they expect a designer to tell them what is best. With prints, silk screens, sculptures, and photography, you don’t get a variation of a Rodin, you get one of ten in
an edition of Rodins. I was told: “Pick your favorite one and we are going to make ten of them, otherwise it makes no sense.” The luxury market does not want bespoke, especially when it comes to collectible art. For me, that was failure number one of mass customization. We converted the ideas from this haute couture product into a commercial product for Alessi called the Supple Cups. For production, we needed to make more than one set of slip molds to make the volumes in an industrial manner. I suggested each mold be slightly different, and then I was introduced to the quality control and distribution divisions. Every dealer there had a different inventory with increased skews; I found out, however, the channel to the consumer was not appropriate for unique mass-produced cups. I was naïve and was forcing a concept that was slowly being beaten down. 75
06 Lynn.indd 75
2018-10-18 9:44 PM
6.5 Flatware design for Alessi.
There is a big difference between working for consumers (on products) and working for clients (on buildings). It is important that we acknowledge that architects work with clients and industrial designers work with businesses involved in mass production that sell to consumers. It is a very different world for the client and the consumer. Clients expect that no two buildings are alike. Consumers expect products to be mass-produced and retailers want to minimize the number of skews, whether in shape, size, pattern, or color.
The last project I tried to do with Alessi was a failure, sadly. The first two were more than successful on their own terms. For the third project I decided to use new methods of mass production to produce very high quantities of repeatable units, with very high functional specificity. I designed a collection of flatware (figure 6.5). At the outset, Alberto told me to do five pieces; if they were successful, then we would design all the other pieces. Instead, I designed them all at once and 3D printed them in metal using sintering. I started with
76
06 Lynn.indd 76
2018-10-18 9:44 PM
a primitive thing, with all the traits we could unfold into a fork, a spoon, or knife. I then took those elements in different directions to get all the basics. I did some research into the history of flatware, in particular, Victorian era flatware, and found pickle forks, fish spoons, oyster knives, olive spears, soft cheese knives, and all that stuff, and came up with 48 different pieces of flatware. Like the previous projects, they were all related as a family, from the same primitive collection of curves; unlike the others, however, they weren’t
just different ergonomic variations, but instead they were driven by more specific functional typologies. They were manufactured in 3D-printed bronze and tool steel, and then polished and silverplated. There was no supply chain for 3D-printing metals even though the costs for bills of materials would be acceptable for luxury retail flatware. Forging the designs was attempted unsuccessfully and now the prototypes are in the permanent collection of LACMA (the Los Angeles County Museum of Art). 77
06 Lynn.indd 77
2018-10-18 9:44 PM
EXTREME MASS PRODUCTION: MORE STANDARDIZED TYPES AT A GREATER FREQUENCY I believe this is the direction in which mass customization is going: more standardized types at greater frequency. Every car manufacturer now has a dozen models – not two – and these are launched with more frequency. But there are not bespoke cars made for an individual and designed by an individual. That is retardataire. It was an unexpected surprise when it came to the desire for higher frequency and greater numbers of standardized types. I was invited by Volvo, together with Lindy Roy and Sanford Kwinter, to go to Gothenburg. The Volvo design team had just bought licenses from the Alias Auto Studio and they were very excited that Volvos were going to get curves. They wanted to bring us in to talk to us about this new language of voluptuous Volvos.3 We met with their head of design and with some of their engineers and interestingly I was told the following: “We want you here, because we want to learn what software you are using, because our boss told us we need to become car architects.” I met with their boss at dinner who said: “We believe that the volume of models that we have to produce will continue to increase. We used to have a sedan, a wagon, and maybe a third model. Now we have close to twelve models. If I extrapolate, we are going to end up having hundreds of models of cars in our showrooms. That is more inventory than a showroom can afford. Eventually, our customers may want specified cars, but nobody wants to design their own car on the Internet, so we need architects at all of our showrooms to talk to our customers, interpret their desires, and configure a car for them using standardized off-the-shelf components, like architects do. Because architects have been doing mass customization for 120 years, that’s what you do. You take off-the-shelf windows, off-the-shelf steel, off-the-shelf wood panels, you could figure all the stuff and every building is different, so you’ve been in the mass customization business for over a century.” If Joe Tanney was
where I was then, he would be the CEO of Volvo now; what they wanted was exactly what Joe Tanney is doing with free-standing, single-family houses.4 The auto industry, by the way, is the most backward, lost, and unimaginative industry out there. No architect will learn anything from the auto industry; there is next to zero innovation there at all. There are electric cars: congratulations. How they are built, designed, engineered, and distributed is all so broken that it is unbelievable. The electric car everyone loves is more than two-and-a-half tons of rolling metal. That is nothing to be proud of. The auto industry is looking to architects to save them because we are mass customizers. The irony of today for me is that people are saying: “Hey, architects, you should get on the mass customization bandwagon.” But the architects are all saying: “How do I mass produce?” I have been interested in getting into the mass production game because it is so broken and all I want to do is get out of the mass customization game. You can have a much wider impact if you work with consumers, not clients. What I tell every industry is, if you want to learn how to do mass customization, hire an architect. Chris Bangle is one of my heroes for what he did with BMW, because he tackled the challenges he faced as if he was an architect. He understood that “BMW needs a signature.” He was looking at a $20,000 car and a $100,000 car and trying to figure out how to manage the fact that they needed a signature without simply scaling the car in the x, y, and z axes as Porsche and Audi were doing. You need a sedan, stretch a coupe in length; you need an SUV, stretch the sedan in height. Instead, BMW defined that the 3 Series would have flame surface styling, the 7 Series would have intersected surface styling, and everything in between those would be a blend. BMW kept their brand with more and more cars to serve more and more markets. I thought that was very smart; it was also very similar to what I had done with combinations of curved surfaces as a family language. For me, this is on a
78
06 Lynn.indd 78
2018-10-18 9:44 PM
6.6 Trimaran designed by Greg Lynn and Fred Courouble.
spectrum of mass customization; it is not about the bespoke but what I would call extreme mass production. I have had the opportunity to work closely with Nike and I would encourage anyone looking at true one-of-a-kind manufacture to look at the data on how many of their shoes are sold through NIKEiD (Nike’s custom shoes, trainers, and bags)5 and ask what is being designed by consumers with their NIKEiD shoes. What you find is the majority of the customer-configured shoes are exactly what is on the shelves in stores. Everybody thinks they are doing something individual, but what they are doing is whatever they saw on somebody else’s feet and they just think they are expressing their individual design sense or they are getting what isn’t available on the shelves quickly. Nike also had a hard time selling the user-configured product until they tightened up the variables, reduced the number of colors and designed the
parameter space for their customers so they were doing much less design in the configurator; the selections were pre-configured much like the curves of the Embryological House were pre-designed as an ensemble. Nike learns a lot about their customers with NIKEiD but I would guess that they are learning a lot less than they thought they would. I am certain they learn more from what their designers produce and what gets purchased than they learn from what people design, which is certainly generations behind what Nike is thinking. Most people I know don’t want to design their own shoes. They want to have a Nike designer design their shoes. They want to have an expert do it. They want to have someone with vision design their clothing. They want something high performance. They don’t want to look stupid. So again, I think you have to keep design expertise and brand identity in the equation if you are talking about consumers, and probably even clients, for that matter. 79
06 Lynn.indd 79
2018-10-18 9:44 PM
HIGH PERFORMANCE AND OPTIMIZED – NOT BESPOKE My design focus has gone through a fairly big change in the last five or so years and that has to do with performance. I think we should discuss the kinds of manufacturing technologies that allow much higher level of performance than what could be achieved previously. What is more important than the bespoke is the high performance and the optimized. A couple of years ago, I jumped into a totally different industry and designed a boat (figure 6.6) for myself, along with Fred Courouble, a naval architect. It was like going back to my animation design days in the 1980s as we used very sophisticated computational fluid dynamics to design the boat. We also manufactured the most critical components in the office using 3D printing processes. The bespoke in this case is about bespoke structure and bespoke performance; it is not about the fact that the boat was especially for me. It is also about the ability to manufacture a large complex vehicle like a boat using recyclable/ disposable tooling and 3D printing, that previously would have involved expensive tooling or artisanal and time-consuming manufacture. The ability of contemporary agile manufacturing to speed up access to the market cannot be overemphasized. In 2016, Jeffrey Schnapp and I started a company with Piaggio Group called Piaggio Fast Forward.6 We are designing and building lightweight transportation devices to change the way people move through cities. We had our first working prototype in less than a year. Before we even had an idea for a design, we set a policy that we wanted to rethink inventory, shipping, point of manufacture versus point of sale, and that every vehicle we did would be as streamlined in that production process as possible. So, rather than build a factory to make aluminum parts, we used MakeTime (on-demand CNC machining with upfront pricing and no request for quote)7 and we had the whole factory
of manufacturing at our disposal without the overhead of a plant. That is one of the ways we are addressing inventory, with this new thinking about a supply chain. We invested in MarkForged 3D printers before we even had furniture. We were able to get out of the aluminum prototype business and move into the plastic and composite prototype business. In five or six months, we were able to produce a vehicle that had 78 3D printed carbon mounts and fixtures. The knowledge all came from experiments in mass customization but the dividends were collected when that knowledge was applied to the problem of rethinking mass production at higher velocity and in higher volumes of types and models. I believe this is where the real action is for both design and commerce. The last project I will discuss is the microclimate chair (figure 6.7), only because it is yet another anecdote about the bespoke. We are collaborating with Nike to develop an intelligent piece of furniture that will give athletes a competitive advantage. We are looking to basketball since Nike has just taken over the NBA for the next few years. The first chair was custom-molded more or less to a specific player’s body so that it could modulate body temperature for specific muscle groups while cooling other parts of his body. The chair is primarily heating, although in local areas, especially on the spine, it cools. There is a microprocessor that can change every point on the chair where you see the 3D-printed elements, so that they can either heat or cool. We can generate a microclimate for the athlete in an arena where there is no control over the climate at all. That chair is intended for players who get paid tens of millions of dollars a year to perform at the absolute limit of their sport, so making each chair bespoke for them is meaningless in cost compared to the value of keeping them prepared and uninjured. We were ready to make custom, form-fitting chairs, but when we received the data, we were shocked to find out that all but a few elite players were pretty much the same size and shape. We asked what happened to the really big ones and were told they weren’t paid as much. We asked about the shorter ones and were told there were only a few of them. We made molds and we
80
06 Lynn.indd 80
2018-10-18 9:44 PM
6.7 The microclimate chair.
brought players in; they got into them and what they all said was that particular chairs were too snug. We found out that by molding the chairs to specific players that they actually didn’t like them. We ended up removing much of the ergonomic intricacies for comfort based on athletes’ feedback and found, when we looked at our data, that sub-centimeter variations between elite athletes were taken out by the smoothing away of intricacies. The vast majority of NBA players are 6’6” to 6’7” tall. Their legs are the same length as mine; the extra 3 inches is all in their upper body and they have an enormous “wing span.” We found out that NBA players are a typology. We started rethinking bespoke and found out that we didn’t even need three different sizes. Now we are able to manufacture and place all of the electronics and climate control and modification exactly where we need them. We are making custom heating elements with custom patterns; we are printing circuit boards that are made overnight and delivered in three business days. We are making the chairs rapidly and affordably, but there is no desire for many variations. So, in terms of mass customization, there is a lot of innovation in a product like this that wouldn’t have happened ten years ago. In this case, we are dealing with a customer and not a client. I was hoping for 14 NBA superstars as my clients and I realized
that I have a single customer: an NBA player. What makes this project interesting is that we are doing a chair for a particular customer group, a chair with the smartest, highest technology – the highest performance chair on earth. Because of advances in design and manufacturing technology, we delivered a high performance and optimized product, but not bespoke. The real advantage of these technologies was the ability to build close to a dozen foam models for ergonomic testing, three different prototype typologies, and three beta chairs all in roughly a year’s time with a team of three people. The desire for one-of-a-kind is almost always being pushed at the market by designers and design consultants; more often than not, there is no desire for it by consumers. I am very suspicious of the bespoke as it smacks of nostalgia and wannabe privilege. When I asked at the symposium how many people were wearing bespoke clothing, not one hand went up. With our clothing, and the things that we surround ourselves with, we are very tribal; we want to belong. I am suspicious if the bespoke is really a human desire; I would be curious to see the data on this and the market opportunity for bespoke things before pushing the concept further. As mentioned earlier, it is not about the bespoke but about extreme mass production – more standardized types at a greater frequency.
81
06 Lynn.indd 81
2018-10-18 9:44 PM
NOTES 1 Graphics Processing Unit. 2 See www.glform.com/embryonic/embryonic.htm. 3 Sanford was devastated; he drove a Volvo because it was the only car that was a box. 4 See Chapter 13 by Joseph Tanney in this volume. 5 See www.nike.com/ca/en_gb/c/nikeid. 6 See www.piaggiofastforward.com/. 7 See www.maketime.io/.
82
06 Lynn.indd 82
2018-10-18 9:44 PM
7 DEMOCRATIC DESIGN AND DAILY OBJECTS
PHILIPPE STARCK 83
07 Starck.indd 83
2018-09-08 11:42 PM
From the beginning, my designs were never intended for the elite, but for society as a whole. I have always worked toward democratic design, improving the quality while striving to make it accessible to the greatest number of people, at the right price. Sincere, modern elegance comes from the multiplication of an object, as opposed to the ideology of limited editions, where premeditation on rarity leads to a selection through money rather than necessity. I have followed this approach – aiming to provide the largest number of people with the best quality – in all domains of design, from tableware and furniture to houses. I have also tried to steer the craft of design toward political and social action, complicit and yet denouncing, to generate action and reaction. I often seek to convey a political, subversive message, by associating humor and poetry with my design undertakings. A good design should surprise, provoking love or rejection. An object that is instantly accepted doesn’t have the right to exist.
TOG – ALL CREATORS TOGETHER We all need mass production because only the multiplication made by ethical companies can raise the quality and lower the price. But, in our hearts, we think we are unique and thus we desire something unique to us. Except for a few companies like TOG – All Creators TOGether, nobody in the furniture industry has seriously tried to combine the brain and the hand to solve this paradox. At TOG, mass customization – mass production of customizable products – was embraced as a win for the customer, a win for the company, and a win for the craftsman. A virtuous circle: win, win, win for creativity and self-esteem. TOG was the only company that showed clearly that the only acceptable next trend is freedom of choice and freedom to be different. At TOG, there was no style but freedom. For example, an old lady could make her own creation, online or with a company, or
7.1 Maria Maria was a handmade wood and natural fiber woven chair with a backrest that can be customized.
84
07 Starck.indd 84
2018-09-08 11:42 PM
she could have a nice time finishing her own product in a shop, or she could look at a list of customizers and choose a young guy from an Amazonian tribe in the middle of the rainforest to create together the masterpiece of their lives. If she was happy with the product, she could organize a tea party with other old ladies, who could perhaps order some pieces also from this young guy, or from another lady in Africa, and give them the opportunity to start a business. Finally, this developed an interaction between the designer, the producer, and the customer. Customers didn’t just pick a product; they became part of TOG’s global community and helped shape a design (figures 7.1–7.3). With TOG, designers, artists, artisans, industrialists, and end users were ALL CREATORS TOGETHER. The Best Customizer Award we created with TOG, which was won in the first year by a young Italian couple, is a good example of a way to create a new industry.
7.2 The Boss Boss was an office chair that can be also customized.
DEMOCRATIC ECOLOGY In order to save our planet, change our societies and make them more inclusive, we need initiatives, major actions. My concern to develop durable creations has been inscribed in my work right from the beginning, in a responsible, ethical approach. Accessibility, comfort, security, adaptability to needs, timeless design, integrating a genuine industrial process to guarantee a durable quality, and of course respect for the environment and ecological standards. If we want to think about things in the long term, we have to be sure the style we create today will still be acceptable in 50, 100, even 200 years’ time. More than the style, we have to think, imagine, and build with the right materials, at the most reasonable price possible, with an irreproachable quality, and by using the very best technology. It is all of these principles combined that will take us toward a “good future.” I am committed to the future of mankind via a democratic ecology that will help us live in harmony with our natural surroundings. After the process of asking oneself about the legitimacy of a product to exist, we have to focus on all the ways to bring flexibility and customization to the end users: if they order online, or go to the store, if they can participate in the creation of the product, or if they can work with a creator from any society, anywhere in the world. Customization can be about anything: shapes, colors, dimensions, accessories … The beauty is that customization is as infinite as human fantasy. If you decide to work in another field of customization – which is giving your design as open source – this is a philosophical and economic decision through
85
07 Starck.indd 85
2018-09-08 11:42 PM
which your creativity and your ego will be part of a common piece of creativity. It is another time, different needs, a new economy, another way of thinking. Some people will want bespoke products and others want ready-made designs. For some products, everybody is able to customize them, for others, it requires a certain know-how that can be shared online with a community. The world is open and diverse. We have to forget the old idea that one designer creates one piece that is then transformed by others; we have to think about it as if it is one entity with soft and tenuous frontiers. The only acceptable trend is freedom and respect of someone else’s bad taste.
CONCLUSION I aim to offer future generations the possibility of writing their future on a fresh page, so that they can invent another story and a new romanticism. Increasingly aware, we can all take the destiny of the human race in hand instead of drifting into the mechanisms of the market. This consists of giving up the insane cycle of fashion for sustainable, durable objects. Through mass customization, each of us can create our own style – a style that will not go out of fashion. We live in a connected world, with an uberization of all professions. We should be happy to live in a society where each individual can be the top model of his/her own trend.
7.3 Misa Joy is a highly customizable chair.
86
07 Starck.indd 86
2018-09-08 11:42 PM
8 LEARNING AS IT GROWS: THE HUMANIZATION OF OBJECTS
ASSA ASHUACH 87
08 Ashuach.indd 87
2018-10-18 9:50 PM
The Industrial Revolution introduced mass production and millions of affordable and identical products, which turned us into mass consumers of low-value, short shelf-life products. We should begin now to move away from this rigid model. In the past several years my focus has been on developing digital tools that enable a new kind of symbiosis between designers as originators of things and users as consumers of things. Whether this direct connectivity is between a designer and a user or in the traversing of the virtual to the actual world, this is a natural progress that offers a dramatic change and a radical shift in making, distributing, and industrial thinking. I see design as a strategy for questioning and modifying traditional notions of both the aesthetics and the functions of products. Design is a technological catalyst and can be used as an enabler and a problem solver. A considerable part of my practice today is dedicated to re-examining and developing new design methodologies, as a means to achieve new forms of connections between the object and its future owner. I have introduced a technology (UCODO™) that enables designers to add a new layer of 3D user design experience on top of the digitally designed object. I am using 3D coding as a means to devise and develop new tools and solutions and to continue investigating and evolving forms and contemporary aesthetics. My interest in enabling user input has always been an important part of my work. If it is a mechanical action or a virtual opportunity, the user’s selection is an essential element that must be incorporated somehow into the final designed object, and not only in the act of buying it. This vision, in tandem with technological advances, enables a shift in the order of use, whereby a user is informing the functions and aesthetics of the future object. In general, my work and research over the years have been focused on this notion of the “humanization of objects” based on the assumption that if we design “healthy” virtual objects, they could then enable a good co-design and personalization experience or evolve and grow autonomously to become better for their users over time. I am developing both technologies for “co-design within safe boundaries”
(the Digital Forming® technology) and design methodologies to enable semi-autonomous generative object growth (such as STEM 45°, for example, described later in this chapter). The two are linked – if we cannot design well, we cannot provide good co-design and product adaptation. CO-DESIGNING The notion of objects “open within boundaries” was introduced for the first time during my research fellowship at London Metropolitan University in 2005. During the fellowship I developed the Digital Forming® concept and workflow, implemented as 3D software that allows product personalization and reconfiguration within a safe user experience online, connecting the user, designer, and manufacturer. This novel workflow was created based on the assumption that 3D objects are fundamentally a line of code, a code that can be stored, embedded, and rendered as a “virtual open product.” Here, “open” means that objects and the interactive experience of designing them are achieved in an “unlocked” state by the original designer, while the fundamental functional parts are “locked” and protected to be used only by co-designers, ensuring full functionality of the end product. Online and offline, objects can be produced then, using additive layer manufacturing, preferably Selective Laser Sintering (SLS) technology, whereby complex 3D forms can be created, one layer at a time, from laser-fused polymer or metallic powder. This new form of a 3D open object comes with its own interface and set of in-built rules and behaviors. It has a two-way communication system: the designer side and the user side. The designer is designing the object together with the user interaction experience; the user is then codesigning within the boundaries set in place by the original designer. These restrictions relate to the areas of the object that can be manipulated, the ways in which they may be adapted, and the degree of adaptation that is allowed. A key consideration here is in designing objects that are
88
08 Ashuach.indd 88
2018-10-18 9:50 PM
8.1 The Digital Forming Studio Pro 3D software for online customization and shape modification of designed objects.
sufficiently extensible, without jeopardizing the product’s functional soundness and its properties for industrial production. This unique communication method offers a user the opportunity to co-design, personalize, and modify the shape of a virtual 3D object, one that resides within the computer as data and is displayed as graphic-based geometry until it is selected for production. It is a process pipeline devised to introduce a new form of communication and an innovative industrial solution within the triangle of the designer, user, and manufacturer, in what was previously unexplored territory. To implement this vision, I founded Digital Forming® and UCODO™, together with a team from the fields of engineering, business, and online marketing. These two London-based companies were set to democratize the personalization of everyday products by introducing the notion of ODO – Original Designed Object – and CODO – CO Designed Object – using a set of innovative 3D software solutions for online customization and shape modification of designed objects (figure 8.1).
Together with a team of dedicated developers, we introduced two new file formats, ODO and CODO, offering users and designers/brands the platform as a solution to “opening” their products for personalization through shape modification and reconfiguration. These file formats contain the data that defines within the 3D virtual world the anatomy of the objects created by the designer and co-designer (ODO and CODO respectively), within the platforms’ GUI (Graphical User Interface). They consist of binary data, describing geometric forms defined by point co-ordinates, and can therefore be translated into a production format and reconstructed in the physical domain as products via digital manufacturing. The translation process is necessary to convert the object created by the co-designer via the interface into a mesh-based format used in additive layer manufacturing (STL files). They are file formats that act as a bridge between the virtual experience and machine-ready production files. With this platform, it was possible to offer a real-time, online, 3D, co-design experience for
89
08 Ashuach.indd 89
2018-10-18 9:50 PM
8.2 The Loop Light lamp can be co-designed online using a simple interface.
8.3 Co-designed Loop Light lamps.
the first time. There is no longer a middleman and designers can connect to home users directly from their desktops. Brands can connect directly to their user communities. This introduces the notion of caring and connecting in addition to co-designing. I believe that by staying in touch, brands can convert their users into partners. This is a new industrial reality whereby the consumer becomes a user, and a user becomes a partner. This new workflow is offering to democratize the design process of everyday products through the synthesis of innovative 3D software solutions, which allow for online personalization and shape modification (figures 8.2–8.5). It uses a vision of
“postcode” production, whereby products can be produced locally via the intelligent assignment of the user’s submitted designs to the production facilities nearest to them. This architecture also provides producers with the means to capitalize on their redundant production capacities by ensuring that their otherwise idle machine space is being profitably used in the production of users’ CODO objects. In addition, this benefits users with improved delivery times as a result of an optimized production and supply chain. This in turn reduces carbon footprints dramatically; through on-demand production, conventional storage and logistical issues for massproduced items are removed.
90
08 Ashuach.indd 90
2018-10-18 9:50 PM
8.4 The co-design interface for the Helix bracelet.
8.5 Co-designed Helix bracelets.
THE DIGITAL FORMING: OPEN WITHIN BOUNDARIES I see the 3D virtual object as a point cloud with an embedded DNA code. It contains all the required structural and visual characteristics and has a three-dimensional point grid, open to be tagged and assigned with physical attributes at selected points. Here, any 3D object may have a basic topology, but more than this, it may have pre-programmed instructions that dictate its behavior under certain conditions. Put otherwise, it can be called an object with no boundaries or a virtual code that is without limits.
Within the context of the ODO file format, I, or anyone else, as a designer can specify a design within the ODO interface that will enable certain operations, at certain points on the ODO object, that may be manipulated by a user via a CODO interface. Here, the notions of locking and unlocking come into place. For example, you can resize the handle (to some specified degree) on a CODO cup design; however, the surrounding geometry (where the handle meets the main body of the vessel) can be locked so that manipulation is not possible. This then avoids any negative impact and provides a safe co-designing environment. It also ensures the sensitive production process constraints enforced by additive layer
91
08 Ashuach.indd 91
2018-10-18 9:50 PM
8.6 Lemon squeezers: users can modify the shape, assemble, select the color or texture and choose the object’s material.
8.7 Titanium lemon squeezers produced using Direct Metal Laser Sintering (DMLS) machines.
manufacturing are respected; these include issues such as maintaining a minimum wall thickness as well as a safe distance between adjacent but separate moveable elements, to ensure a product of adequate strength and to prevent powder fusion (as in the case with closely positioned parts such as gears) during the rapid prototyping (RP) process, respectively. This creative freedom with varying degrees of preprogrammed control is the concept behind “openness within boundaries.” Users can modify the shape, assemble, select the color or texture, and choose the object’s material (figures 8.6), however, they do this only within the safe boundaries set by a designer to ensure that the design remains functional and able to be produced. This degree of real-time online cocreativity had never before been available prior to the introduction of Digital Forming in 2005.
This new reality empowers the user and offers the opportunity to create objects with greater embedded value. With a simple and intuitive CODO interface, people with no modeling or design experience are able to express their creativity without any concern for the underlying technical issues, and complexity associated with traditional 3D design. In terms of production and delivery, when a user order is received, the request is fed into an intelligent machine allocation system. As previously described, based upon the postcode of the user, an automated enquiry is sent to the nearest corresponding bureau within the network. The bureau responds with data about the availability of their SLS machines. Should a suitable machine be available, the production file is transferred, and the designed object is produced within a previously agreed timeframe (figure 8.7). If
92
08 Ashuach.indd 92
2018-10-18 9:50 PM
8.8 OPENPEN designs.
8.9 A close-up view of an OPENPEN produced using the DMLS process.
there is no availability at the bureau closest to the user’s address, the next one is queried, and so on. This model works for both the end users and producers. Users are given quicker product delivery turnaround times, and production bureaus are able to maximize their machine costs since the operational costs of SLS machines are greater than those of material consumption for part-manufacture. Within this framework the original designs (ODO) chosen are products of a relatively small scale that can fit within an acceptable volume (figures 8.8, 8.9). This is because when a user decides to have their design produced, the system needs to locate a local manufacturing bureau that has sufficient capacity to house it, and therefore, the probability of this increases when the object is small. In addition, as the size and thickness of a part increase, so does
its cost. Therefore, in order to make products financially accessible, this strategic decision was made from the outset. With an ethos of keeping objects virtual and “alive” until they are wanted, huge benefits in terms of waste reduction can be made. As long as the files are not produced, there is zero energy, no material consumption, no warehouse storage, and no shipping or other logistical costs of mass-produced products and packaging, and therefore zero waste while keeping your favorite objects on your virtual shelf. Through my work and research to date, these core concepts have been developed into actual objects and technologies. What follows is a small selection of my work that has been informed by these underlying drivers.
93
08 Ashuach.indd 93
2018-10-18 9:50 PM
8.10a–c OMI.mgx light was one of the first products to be produced entirely as a single nylon part straight from the SLS machine tank.
OMI.mgx Light The OMI.mgx light for materialize mgx (2003, figures 8.10a–8.10c) was one of the first products to be produced entirely as a single nylon part and distributed to the customers straight from the SLS machine tank.1 Its shape, together with the natural flexibility of the polyamide material, creates the joint-less form of a biological mechanism. The object was designed to be transformable through the direct manipulation of the user, by the bending or twisting of the structure, allowing for the sculpting and re-sculpting of the form to create new and personalized sculptural elements. This permeative but yet powerful form of interaction offered the user access to the design process. Letting the user change some of the object’s sculptural elements enables the embedding of personal values – and thus increases the object’s value to its user/owner.
Osteon Chair The Osteon Chair (figures 8.11a, 8.11b) is the first object to have an internal optimized support structure, which is referred to now as “infill” (figure 8.12). Designed in 2004 and produced in 2005, this project has historical significance as it was the first time an object’s internal porosity was designed and optimized to achieve a material usage of a third of the anticipated amount.2 The Osteon Chair is the first chair to be designed using a combination of 3D tools and artificial intelligence. Produced by EOS laser sintering, the chair consists of a cosmetic skin and intelligent internal structure. Like the biological structure and mechanism of bones, the artificial intelligence software knows where to create sufficient support. This is an intelligent product that is growing in free space with an artificial intelligence “DNA” code. This code contains all the information required to ensure that the object will transform perfectly from a virtual design into a 3D object that achieves the optimum strength while maintaining the desired visual aesthetic.
94
08 Ashuach.indd 94
2018-10-18 9:50 PM
8.11a and b The Osteon Chair.
8.12 The Osteon Chair features an internal optimized support structure.
USER-INFORMED OBJECTS: ADDING THE ELEMENT OF TIME My current work is the natural evolution of some of the key concepts introduced in this chapter. Users’ personal data are being collected in real time and harnessed into a feed of essential information. Personal sensor data are streamed into a pre-designed, 3D virtual, production-ready object. These virtual objects are “living” and evolving in a “virtual object incubator” based upon the live stream of personal data, creating an essential user information envelope. This live stream of personal user data drives the object’s evolution, adaptation, and optimization. The object is connected to its essential stream of user sensor data, learning its behavior patterns and ergonomic characteristics. With time, this object can adapt and develop better performance with a better fit to its owner, i.e. the future user. A user can monitor the evolution of the object and vote on its progress; the object is ready for production by the end of the learning and training cycles.
95
08 Ashuach.indd 95
2018-10-18 9:50 PM
8.13 The AI Light can morph its structure and generate new behavior as it is “trained.”
Using artificial intelligence and the design of intelligent agents, learning and optimization are at the center of this system. With time, the connected object is being informed and trained to be better. The stream of the user personal data and characteristics is gradually informing and forming the object’s ergonomics, functionality, and aesthetics. There are two possible models of training: •
•
Active training: Whereby a product collects input from user/users via a series of sensors and either locally or remotely delivers this input (i.e. the user behavior patterns) to a processing unit, which, based upon pre-programmed conditions, processes these data and, as a result, alters its behavior in an intelligent and beneficial way. Passive training: A passive feed of personal user data is collected and processed to shape and form the virtual product. By the end of each training cycle, the object can be materialized using additive layer manufacturing techniques. This is solely the choice of the user who is the current owner of the intelligent product data.
This introduces a shift from the notion of controlling products to training, making products better with time through autonomous learning. This has significant ramifications and introduces the opportunity to deliver a wide spectrum of benefits to
the user, such as ergonomic shape adaptation and conformation of geometrical functional elements to the user’s body parts. It is introducing the notion of growth and evolution of objects through time, as illustrated by the following projects.
AI Light The AI Light (figure 8.13) has an artificial intelligence brain that senses space through five different sensors. Using a biologically inspired mechanism, this structure morphs and generates new behaviors according to your personal space. When you first invite it into your home, you have to let it get accustomed to its new environment. Once it is relaxed, the training can begin. It has five senses that track changes in its environment; it slowly develops a set of behaviors that indicate a new character to each light. The user can also interact with the light by playing with it through sounds, light, and movements. This smart structure may behave in unpredictable ways if moved to an unfamiliar space. One of the revolutionary aspects in this project is introducing the notion of “training” over “controlling” products, as described previously. This product has no buttons and no remote control. The product will develop its behaviors by personal training and a natural interface.
96
08 Ashuach.indd 96
2018-10-18 9:50 PM
8.14 The Femur Stool.
The Femur Stool For the Femur Stool3 (figure 8.14) we used a new algorithm that removes any redundant material according to stress zones on the object surface. An optimization of the exterior and material use was made to achieve a light and economical form (figure 8.15). By using laser sintering, we managed to achieve a faster production time and save energy by reducing the laser mileage and distance to the final object layer. Like the human femur, the stool’s shape follows the internal human bone structure in providing support that is optimized for a load of 120 kg; if the sitting load is changed, the object’s form will change too to accommodate and adjust performance. This object shaping is driven by the mathematical intelligence of the human bone formation.
8.15 The Femur Stool was designed by removing redundant material through optimization.
97
08 Ashuach.indd 97
2018-10-18 9:50 PM
8.16 The bamboo microstructure is manipulated into a new structural topology that has enhanced continuity, flow, and larger cell porosity.
8.17 The Venturi Stool, inspired by nano-fiber.
STEM 45°
I have focused on a very fast-growing breed called phyllostachys that, depending on a variety of parameters, can grow extremely fast and can sense its environment to correct and reinforce itself while growing. The bamboo is “learning as it is growing.” This means that its internal structural 3D morphology is constantly changing and adapting to new environmental conditions, growing differently from section to section based upon inherent intelligence and sensory systems. In collaboration with macromolecular and bio-material scientists at D-Lab at the Kyoto Institute of Technology (KIT) in Japan, we have translated the microscopic bamboo’s internal structure into producible 3D structures (figure 8.16).5 Through scaling-up by 3000 percent we can now study the natural geometrical growth patterns of the bamboo, both in terms of its natural structural porosity and its geometrical growth intelligence. Taking the original geometry
When the Osteon Chair (see figures 8.11a, 8.11b) was introduced in 2004 as an object with an “intelligent” bone-like internal support structure, we used a 3D AI algorithm that generated an optimized structural unit within the object’s 3D voxel grid skeleton. I believe that was the first time an object with optimized internal support structure was designed in that way. Today, a feature called “infill’ is widely available within any FDM4-slicing application. While the “infill” is a primitive repetition of the same unit within the object’s interior boundaries, and although very accessible and fast to compute, it is still very wasteful, without any consideration of the object’s load conditions and 3D organic boundaries. In recent research, I have been looking into the internal 3D geometrical growth of bamboo.
98
08 Ashuach.indd 98
2018-10-18 9:50 PM
8.18 The STEM 45° slim chair.
8.19 The STEM 45° stool.
into our studio’s 3D design workflows, we can now manipulate the original data and study bamboo’s natural growth patterns, adding continuity and controlled porosity. We can turn this virtual material into a line of industrial processes that can be repeated and re-used within industrial design and architecture. The new STEM object collection (figures 8.17–8.19) was designed using the actual bamboo 3D microstructure geometry together with my own personal aesthetics preferences and 3D automated scripts. My aim is to start a discussion on future industrial design and architecture at both large and small scales, whereby automated processes will be fed by a combination of human and biological intelligence, designing a new type of toolpath for the robots to follow.
99
08 Ashuach.indd 99
2018-10-18 9:50 PM
CONCLUSION
ACKNOWLEDGMENTS
From the early ventures in primitive user interaction with the OMI.mgx light and the formation of the Digital Forming® technology and UCODOTM co-creation platform, I have been working on the development of new ways and methods to embed personal user data and values into daily designed objects. In Digital Forming®, ODO software allows designers to build an interaction experience that is delivered to the user using the CODO user interface. This new design method allows designers to design virtual, production-ready products with a user interaction to experience a setup that enables the users to personalize and co-design the products better for themselves. My belief is that this ethos and design method should be adopted by the industry and further developed into a new industrial reality. I think that brands and manufacturers need to stay “in touch” with their user communities, demonstrate greater care and responsibility, and be better connected with their user communities (formerly called “consumers”). It is a move from a consumer to a user to a partner. Consumers should be perceived as partners who participate in an exchange of information that, with time, will result in better products, with less waste and consumption. The exchange of information can be active or passive. This new industrial reality will provide a live stream of bidirectional information and a better way to connect directly with the users. In the end, it is all about owning less but with more value and enhanced performance.
Parts of this chapter have previously been published in Piroozfar, Poorang A.E., and Piller, Frank T. (eds.), Mass Customization and Personalisation in Architecture and Construction, London and New York: Routledge, 2013. NOTES 1 Selective laser sintering (SLS) technology provides the manufacture of a single skeletal-like structure, a feat unachievable by any other manufacturing process at that time. The fundamental aspects of “no assembly,” together with the instant digital materialization of 3D geometry, reinforced my practice and were used as a foundation for further innovation and software development. 2 With thanks to my business partner Sia Mahadavi, formally Complex Matters, today Within Autodesk, and to our sponsors EOS. The Osteon Chair is part of the Design Museum Barcelona’s permanent collection. 3 With thanks to our sponsors: 3T RPD, our manufacturing partners, for the SLS digital manufacturing and post processing, and Altair Inspire for the mathematics and FEA simulation. The Femur Stool is part of the Design Museum London’s permanent collection. 4 Fused Deposition Manufacturing. 5 With thanks to KIT and D-Lab, Kyoto, Japan. The bio-material, macromolecular science, and project development was done in collaboration with KIT professors Julia Cassim, Dr Yoko Okahisa (bamboo microstructure and bio-material science), Dr Yukihiro Nishikawa (microscopic scanning and macromolecular science), Dr Kazunari Masutani (PLA bio-material science), and Tomohiro Inoue (AM and 3D printing).
100
08 Ashuach.indd 100
2018-10-18 9:50 PM
9 CUSTOMIZING PROCESS, DEMOCRATIZING DESIGN
FABIO GRAMAZIO AND MATTHIAS KOHLER 101
09 Gramazio Kohler.indd 101
2018-10-18 9:54 PM
9.1 The mTable, designed using a mobile phone and digitally fabricated.
In the last 25 years, we have digitalized almost every domain of our lives. The beginnings of this radically new development, which has changed the way we experience our world and is right now accelerating through our ability to handle big data by using artificial intelligence techniques, coincides with our architectural education. In the early 1990s, computers had yet to become consumer products but were already found in universities and research institutions such as ETH Zurich. As architecture students, we had access to the computer labs and could experiment with these expensive and powerful machines, which not only visualized architecture in three dimensions and real time but – and this was the really exciting property – could be programmed. Belonging to the generation that learned programming in the mid-1980s in a playful way, with toys such as the Commodore 64, we believed that these machines were heralds of a paradigm shift that was soon to happen. We felt that the power of computation for architecture would go beyond the mere drafting and modeling of geometry; although the concept of computational design
was not yet established, the possibility of moving beyond representation, toward the generation of form, was extremely seductive. While Greg Lynn was demonstrating the use of sophisticated techniques such as animation and kinematics in architecture, we felt that writing computer code, as an intellectual activity, was very close to designing architecture. Both implied the creative deployment of rules and logic, order and relationships, hierarchies and sequences. While we were excited by this relationship and were convinced that this new technique could be deployed in a truly creative manner and would allow us to do radically new things, we observed that these two worlds we were interested in, physical and digital architecture, were dramatically drifting apart. DIGITAL MATERIALITY While mainstream architectural practice was ignoring the challenges of digitalization, a new generation, to which we belonged, was enthusiastically embracing the new possibilities
102
09 Gramazio Kohler.indd 102
2018-10-18 9:54 PM
9.2 Dimensioning an mTable using a mobile phone.
9.3 Creating the deformation points and holes in the table’s surface.
offered by the incorporation of the digital into architectural practice. This almost ideological divide between the material production of physical architecture and the manipulation of data through code, the two worlds we equally loved, felt artificial and unnecessary to us. In our minds, the question should not be addressed in terms of “either/or” but ought to be “as well as.” We were striving for the establishment of a direct link between the digital and the physical realms in order to create what a few years later we would define as “Digital Materiality.”1 By the end of the millennium, when we founded Gramazio Kohler Architects, CNC machines, mechanical devices controlled by digital information, were able to establish this direct link. They bridged the gap between the hitherto separated worlds of data and material, not just in conceptual but very practical terms and thus allowed the reach of the digital to extend into the physical dimension of architecture. Suddenly we were no longer limited to the manipulation of geometry in the computer but could eventually interact directly with the real world and focus our efforts on the materialization of
the digital. Moreover, we believed that the potential of this process would definitely lead to the discovery of new and unsuspected architectural qualities. In fact, the synthesis of digital and material properties would affect architecture in the same radical manner as the new physical building materials, such as steel, glass, and reinforced concrete, had done in the nineteenth century.
mTable In these terms, one of the first projects we realized in our young practice was a speculative investigation of the potentials and challenges of parametric design and mass customization. mTable is a table (figure 9.1) that can be designed by the customer on a mobile phone (figure 9.2). On a custom design interface, the user defines points on the bottom surface of a table board and, with the pressure of his or her thumb on the joystick, deforms the surface by applying force to them (figure 9.3). If the size of deformation leads to the intersection of two surfaces (top and bottom), a hole with a very thin and elegant 103
09 Gramazio Kohler.indd 103
2018-10-18 9:54 PM
9.4 The CNC milling machine produces the table “landscape” directly from the data transmitted from a mobile phone.
edge appears in the table. Next, after having defined the dimensions, materials and color, the user sends the parameters of his or her design to our server, where our design algorithm generates a double-curved surface corresponding to the geometry of the table. Finally, a CAM software extracts the fabrication information from the geometry and sends the machine code to a CNC router that mills the surface in a massive wooden board (figure 9.4). Around this time (the early 2000s) many people and companies were experimenting with online configurators which would allow their clients to customize their products according to their personal wishes. By short-cutting the process from the design at the lowest possible resolution of a 176 x 208 pixels mobile phone display directly to the highest thinkable definition of physical material and by eliminating the need – and possibility – of any form of human intervention along this workflow, we radicalized the question of mass customization and design
democratization. We opted for a mobile phone as the design interface because we did not want the user to perceive it as a computer but as a personal accessory. In 2002, the Nokia 7650 was in fact the first mobile phone allowing third party developers to run custom applications on it and, in an irony of history, Nokia never understood the potential and implications of this invention until Apple released the iPhone in 2007 and disrupted the market with the concept of the App Store. However, what made mTable radically different, ambiguous, and ironic at the same time was the fact that it was not limiting the customer’s freedom to the definition of dimensions and material but extended it to the domain of form. By doing this, we deliberately immolated the holy grail of designers: the definition of form. Moreover, by allowing the customers to create holes in the surface of their table, we delegated the responsibility for the basic functionality of this simple piece of furniture to design nonprofessionals.
104
09 Gramazio Kohler.indd 104
2018-10-18 9:54 PM
9.5 Many different mTable designs can be produced effortlessly.
The eight mTables that we produced and sold, as well as the hundreds that potential customers designed (figure 9.5) but never ordered, are all very sophisticated and formally complex objects. Some of them are elegant and beautiful, others rather irritating and clearly dysfunctional. They are all different but still surprisingly similar because they share the same design genome. In fact, although the customer cannot emancipate himself or herself from the distinct rules set by the designer of the algorithm, the customer becomes a co-author in his or her own right. mTable is a statement about the complex relationship
of freedom and control in design. In a playful and ironic way, it explores the potentials, limits, and implications of customization, both for the customer and the designer. While the former needs clear boundaries, the latter should not fear but embrace parameterization as an opportunity. As mTable demonstrates, parameterization does not necessarily imply “loss of control.” If implemented in a proactive and self-aware manner, parameterization can move design beyond the static definition of form toward the authoring of processes and rules that generate controlled but still flexible formal results.
105
09 Gramazio Kohler.indd 105
2018-10-18 9:54 PM
9.6 Floor plan of the house in Riedikon, Switzerland.
106
09 Gramazio Kohler.indd 106
2018-10-18 9:54 PM
9.7 Site plan.
PARAMETRIC DESIGN AND COMPUTING When we try to illustrate the role and significance of parametric thinking for our own architectural practice, we normally use a small private residence we designed and built in 2009 in Riedikon, Switzerland, as it offers two aspects where parametric processes play an interesting and critical role. The pointed geometry of the volume (figure 9.6), which results from two cutting operations on a simple rectangle, is at first glance irritating and needs some explanation. As the client actually is building the new house in its own garden, the first cut in the rectangle preserves the beautiful view on the nearby lake to his former house (figure 9.7). The angle of this cut negotiates between the generosity of the panorama and the size of the new house. The second cut allows parking behind the new house while keeping the garden facing the lake
as generous as possible. Its angle directly depends on the minimal turning circle of a car accessing the property. This design strategy, which follows clear contextual parameters, could easily be formalized in an algorithm and the meaningful solutions calculated by a computer. However, we would never do so because the design space is simple enough to iterate the problem manually, in our head or on paper, fast enough to identify the ideal solution in an efficient manner. This is what we traditionally call “the design process.” While the starting point of the second aspect, the design of the façade, is almost identical to the first one, its consequences for architectural practice are radically different. Here, 315 timber planks are mounted perpendicular to the façade at a distance of 20 cm from each other, enveloping the whole circumference of the house (figure 9.8) and making it appear, in perspective, like a windowless timber barn (figure 9.9). From the interior, the contrast
9.8 Unfolded elevation of the house, featuring 315 different timber planks.
107
09 Gramazio Kohler.indd 107
2018-10-18 9:54 PM
9.9 Exterior view.
9.10 Interior panoramic view through the planks.
between the front side of the thin planks in the foreground and the bright background causes the human eye to focus on the dominating panorama and to cancel out the façade (figure 9.10). However, modulating the planks’ section relative to the glazing dramatically affects the width of the visual field. A simple set of parameters, such as depth, length, position, and inclination of the planks’ cutout (figure 9.11), governs the subtle visual effects of this sophisticated “perception engine” and allows the architect to control the nature and intensity of the visual relationship between the interior spaces and the surrounding environment. Yet, while the parametric logic is per se simple, its application to such a large number of elements, the geometry of which is different but mutually dependent, makes the development of a coherent design virtually impossible. Although manually modeling the geometry of the individual planks is theoretically possible, it would represent a Sisyphean undertaking. As designing implies
iterating through multiple possible solutions, the only way to overcome this impracticability is to write a computer program, which permits fast computation and the visualization of specific solutions. However, being able to compute the design information for the planks would still be irrelevant without a clear and robust strategy for their fabrication. To manually cut the 315 different profiles, or to manually program a machine to do so, would be inefficient and probably too expensive, at least in the context of a modest house. On the other hand, if we use an appropriate CNC machine, we can directly use the previously generated design information to guide a cutting head along each single plank by simultaneously varying its height and inclination. Because such an automated workflow does not need any manual intervention other than feeding the parts into the machine in the right order, the additional costs are, at least in relation to the architectural benefit, marginal.
108
09 Gramazio Kohler.indd 108
2018-10-18 9:54 PM
9.11 Parametric modulation of the planks’ section.
Designing always involves parametric thinking. However, this becomes evident and computationally relevant only under very specific conditions. Recognizing this relationship is key to a meaningful discussion of the potential of computational design for architectural design that moves beyond its common association with spectacular form and iconic architecture. As the façade of this simple house demonstrates, the conceptual combination of parametric design with digital fabrication radically extends the traditional design tools and enriches architecture by fostering the emergence of a new kind of materiality. HUMANS AND MACHINES After having illustrated our general understanding of computational design and digital fabrication by means of these two rather different projects, we will introduce the industrial robot, which we selected as the hardware of choice to pursue our research on “Digital Materiality” in architecture at ETH Zurich.2 This iconic machine, which in the second half of the twentieth century has dramatically shaped car manufacturing and industrial production logic in general, possesses a number of characteristics that make it interesting in architectural applications. First, the industrial robot is large, robust and, as a mass-produced device, comparably affordable. Second, although in the past its field of application was mainly serial production, the industrial robot is fully programmable and thus not dependent on repetition. Third, and this is the most interesting property, the industrial robots’ nature is generic. When delivered, it comes without a tool in its hand, expressing the possibility as well as the need for process customization. This last property makes the industrial robot the ideal companion of the architect, as its openness extends the design to the very definition and customization of the fabrication process. The first experiment we did back in 2005 was set up as a proof of the concept, whereby we equipped the industrial robot with the simplest
109
09 Gramazio Kohler.indd 109
2018-10-18 9:54 PM
9.12 The robot producing one of the “programmed walls,” brick by brick.
tool possible, a hand, allowing it to grip the most basic architectural building element ever: a brick. This simple setup, which we can describe as a “pick-and-place” operation, becomes very powerful when the “placing” can be different for each brick (figure 9.12). Suddenly what holds true for the human mason, namely, the direct dependence between the complexity of the bond and the efficiency of the building process, fails to make sense. The industrial robot does not need to measure but simply “knows” its position in space at every given moment in time and as such, as in a low-resolution 3D printing process, complexity no longer has an impact on the execution time. Before addressing the architectural potential of this paradigm shift, we would like to discuss its impact on the relationship between human and machine. While in pre-industrial times the artisan not only mastered but also was continuously
refining the use of tools, which he – and it was a he – normally personally owned, the Industrial Revolution of the eighteenth century radically disrupted this relationship. Tools became machines, complex mechanical devices that got bigger and so expensive that they were only affordable to capitalists. At the same time, following the principles of Fordism, the division of the production process into several simple steps alienated the human from the process of making and eventually turned the craftsman into a factory worker. This separation of labor and capital, which starts with the industrialization of weaving during the eighteenth century, created a deep divide between humans and machines, whereby the latter expose humans to the permanent risk of becoming the victim of further automation steps. While in such a historic perspective this deep distrust is comprehensible, we argue that the advent of the
110
09 Gramazio Kohler.indd 110
2018-10-18 9:54 PM
9.13 Human-machine relationship.
numerically controlled machine necessitates a reassessment of the human-machine relationship. In fact, this new type of machine needs humans as much as humans need it. In the case of our proof of concept, the machine needs instructions from a human while the human needs the machine’s indifference to complexity in order to be able to materialize the design. At once, historic antagonists turn into strategic partners in a process to which each one contributes with their unique strengths, which fortunately are the other’s weakness (figure 9.13). In terms of programming, the fabrication of a brick wall can be described by two “nested loops” in which simple rules are repeated a specific number of times. This very basic algorithm does no more than what the mason would execute if asked to build a brick wall with a running bond. He would just take a brick, place it next to the
previously placed one and then repeat this action until the wall reaches the desired length. Then he would move to the next row, shift the first brick by half in order to assure proper bonding and start over again until the wall reaches the desired height. However, while a mason can easily perform this simple and repetitive process, as soon as we add some simple calculations to the algorithm, which modify the position or rotation of each brick, the execution complexity explodes. What previously was just the repetitive execution of a simple rule turns into the sequential placing of bricks according to an endless list of complex spatial information. However, what is difficult for the mason is easy for the robot, whereby this division of labor is nothing else than perfect team play that can directly connect the human act of designing with the mechanical process of building.
111
09 Gramazio Kohler.indd 111
2018-10-18 9:54 PM
9.14 Robotically produced brick façade elements of the Gantenbein Vinery in Fläsch, Switzerland.
20,000 BRICKS This early proof of concept prototypes directly led to an architectural commission of a 400 m2 brick façade for a vinery in Fläsch, Switzerland (figure 9.14). Bearth & Deplazes Architects designed the project, and it was already under construction when they invited us to design and build its façade. The initial design proposed a simple concrete skeleton filled with bricks: The masonry acts as a temperature buffer, as well as filtering the sunlight for the fermentation room behind it. The bricks are offset so that daylight penetrates the hall through the gaps between the bricks. Direct sunlight, which would have a detrimental effect on the fermentation, is, however, excluded. Polycarbonate panels are mounted inside to protect against wind. On the upper floor, the bricks form the balustrade of the roof terrace. While we had to solve many technical questions to manufacture 72 façade elements, which we
transported to the construction site by lorry and installed using a crane, the main challenge was indeed the development of a suitable design strategy. We were practically able to design and construct each wall to possess the desired light and air permeability, while creating a pattern that covers the entire building’s façades. According to the angle at which they are set, the individual bricks each reflect light differently and thus take on different degrees of lightness. Like pixels on a computer screen, they add up to a distinctive image and thus communicate the identity of the vineyard. In contrast to a twodimensional screen, however, there is a dramatic play between plasticity, depth, and color, depending on the viewer’s position and the angle of the sun. As technology is neutral, the challenging question is how to make use of the unconventional degree of freedom and generate the information needed for the placement of the bricks in space. Instead of resorting to a pictorial strategy, which
112
09 Gramazio Kohler.indd 112
2018-10-18 9:54 PM
9.15 A 100-meter-long, robotically fabricated brick “ribbon” for the Swiss pavilion at the 11th Venice Architectural Biennale.
could have depicted the company’s logo on the façades, we decided to design a generative process. We interpreted the concrete frame construction by Bearth & Deplazes as a virtual basket and, with digitally simulated gravity, filled it with abstract, oversized “grapes” of varying diameters. Then we looked at the result from all four sides and transferred the digital image data to the rotation of the individual bricks. On the built façades, the visitor discerns gigantic, synthetic grapes, which were virtually inside the building as we developed our design. On closer view – in contrast to its pictorial effect at a distance – the sensual, textile softness of the walls dissolves into the materiality of the stonework. The observer is surprised that the soft, round forms are actually composed of individual, hard bricks. The façade appears as a solidified dynamic form, in whose three-dimensional depth the viewer’s eye is invited to wander. In the interior, the daylight that penetrates creates a mild, yet luminous
atmosphere. Looking toward the light, the design becomes manifest in its modulation through the open gaps. However, the architectural implications of this brick façade are more elaborate and diverse than those of a two-dimensional image. To the human eye, able to detect even the finest difference in color and lightness, the subtle deflection of the bricks creates an appearance and plasticity that are constantly changing along with the movement of the observer and of the sun over the course of the day. THE LINE OF NEGOTIATION Two years later, we radicalized this experience by designing an installation for the Swiss contribution to the 11th Venice Architectural Biennale. We proposed a 100-meter-long, robotically fabricated brick wall to run as a continuous ribbon through the Swiss pavilion (figure 9.15). With its looped form, the wall defines an involute central space 113
09 Gramazio Kohler.indd 113
2018-10-18 9:54 PM
9.16 In-situ robotic fabrication.
and an interstitial space between the brick wall and the existing structure of the pavilion. The first conceptual innovation was to transport the robot instead of the individual prefabricated parts (figure 9.16). By reactivating the concept of the field factory and producing in situ, we gained a huge degree of formal freedom as the transportation costs no longer related to the geometry of the elements. The radical difference from the winery façade was in the chosen design strategy. In order to avoid any additional elements, we activated the form of the elements by inscribing the structural logic directly in the parametric engine generating the design. We gave the wall a “structural behavior” which would guarantee its feasibility, and we defined which additional properties would be non-negotiable. The course of a single, continuous curve carried all the generative information
necessary to determine the design. This curve functioned as a conceptual interface, which enabled the curator of the exhibition to negotiate between the individual spatial requirements of the exhibited groups. As the curve (figure 9.17), which we named “the line of negotiation” was, for whatever reason, modified, the three-dimensional, undulating wall could be automatically regenerated. Its complex shape was determined by the constructive requirement that each single, 4-meter-long segment should stand firmly on its own. Where the course of the generative curve was almost straight, meaning that the wall elements could possibly be tipped over by the visitors, the wall’s footprint began to swing, thus increasing its stability (figure 9.18). Each curvature in the lower layers was balanced by a counter-curvature in the upper layers, thus giving the wall its architectural expression. In addition,
114
09 Gramazio Kohler.indd 114
2018-10-18 9:54 PM
9.17 “The line of negotiation.”
9.18 Floor plan.
the 15,000 bricks were rotated according to the curvature—the greater the concavity of the curve, the more the bricks were rotated. The wall thus adapted its shape according to its course, widening and narrowing, producing tension-rich architectural spaces. We created a parametric design system, which involved the user, in this case, the curator of the exhibition, in the design process. In a strong analogy to the mTable
strategy, we separated the design into hard and soft parameters, where the hard ones guaranteed the quality of the design, both in functional as well as in aesthetic terms, while a co-author would be free to customize the soft ones until shortly before the production started. This freedom is of particular value in an exhibition context, where the curator has to ideally stay open and flexible to changes in the spatial layout until very late in the process. 115
09 Gramazio Kohler.indd 115
2018-10-18 9:54 PM
CONCLUSION In summary, the question of where authorship manifests itself in contemporary digital design practice and how this relates to the alleged openness and democratization of design is key to many of our architectural and research projects. The profound structural change ahead of us not only will overcome the nonlinear relation between complexity and costs that has governed the logic of twentieth-century architecture, but also will facilitate the production of difference, which is a prerequisite to a successful industrialization of the building trade. We believe that the architect should spearhead the development of a novel digital building culture, yet this movement will have to involve all stakeholders in the building industry and eventually reshape the relationship between humans and their tools with the aim of making the machines our best allies. NOTES 1 Gramazio, Fabio and Kohler, Matthias, Digital Materiality in Architecture, Zurich: Lars Müller Publishers, 2008. 2 Gramazio, Fabio, Kohler, Matthias, and Willmann, Jan, The Robotic Touch, Zurich: Park Books, 2014.
116
09 Gramazio Kohler.indd 116
2018-10-18 9:54 PM
10 METADESIGNING CUSTOMIZABLE HOUSES
BRANKO KOLAREVIC 117
10 Kolarevic.indd 117
2018-09-08 11:45 PM
The relative cultural homogeneity that pervades North American cities is in large part conditioned and perpetuated as such through the sameness of the physical environments in which the lives of middle-class city dwellers unfold, from the suburban homes in which they live (figure 10.1), the glassclad, high-rise office buildings in which they work, and the endless low-rise strip malls where they shop. Karl Ove Knausgaard, a contemporary Norwegian writer, vividly captured in a recent essay1 this synergetic relationship between the physical fabric of a typical North American city and its culture: The identical cars are followed by identical gas stations, identical restaurants, identical motels and, as an extension of these, by identical TV screens … broadcasting identical entertainment and identical dreams. Not even the Soviet Union at the height of its power had succeeded in creating such a unified, collective identity. Knausgaard is not the first to attribute such sameness – or more precisely, lack of difference – to a modern system of mass production. Largescale repetition was recognized early on in the twentieth century as necessary for economic, low-cost production, whether of cars, houses or 25-story glass towers. All of that, however, was supposed to change with the emergence of mass customization in the 1990s as a post-Fordist paradigm for the economy of the twenty-first century. Indeed, mass customization, which Joseph Pine defined as mass production of individually customized goods and services,2 did fulfill its promise of increased variety and customization in several segments of the economy, from services to consumer products to industrial production, without a substantial increase in costs.3 As a consequence, for example, today’s consumers can create their own unique, customized shoes, jackets, furniture, bicycles, cars, etc. that cost the same or marginally more than the mass-produced equivalents; they can choose colors and finishes that fit their
esthetic aspirations. Such cosmetic, material or surface customization by customers is now a standard option in a range of industries, including commercial housing; dimensional or geometric customization, however, is rather rare. CUSTOMIZING HOUSES As discussed in Chapter 1, there are websites by architects and/or builders that enable anyone to choose the house design they like, explore available options, and then customize it within some carefully imposed limits. In most cases, customization is limited to the selection of materials, finishes, or colors in the interior or on the house’s exterior, in what could be referred to as surface or material customization, as previously discussed. In many cases, customers are given a chance to choose the appliances they want or could opt to add some relatively minor features – such as fireplaces, bay windows, etc. – that don’t require a change in the overall geometry of the house, in what could be called the feature customization. Some cases involve modular customization, where modules could be arranged in a variety of predetermined configurations, resulting in houses that do have very different layouts.4 These different, currently available, levels of initial customization illustrate a progression from no changes to the shape or form in the case of surface customization, to Lego-like customization of the geometry based on predefined modules. The market has yet to see the emergence of a website-based, interactive house design in which the overall geometry of the selected house – its overall shape, interior layout, and the associated dimensions, including its rooms and various elements, such as doors and windows – could be manipulated. The full dimensional, geometric customization – while technologically possible – is not yet a market reality. Why that remains the case is an interesting issue worth looking into in detail. Lack of geometric customization could in part be attributed to the well-established norms of economy in the industry. Buildings in general are mostly one-off, highly customized “products”
118
10 Kolarevic.indd 118
2018-09-08 11:45 PM
10.1 An aerial view of a housing development near Markham, Ontario, Canada.
and “customization” comes at a (high) price. As discussed earlier, one of the ways of controlling the costs of building is to use repetitive, identical components as much as possible. But that is changing, as uniqueness at the component level no longer commands the same price premium as it did in the past, thanks to the widespread adoption of the digital technologies of design and production. As discussed in Chapter 1, dimensional, geometric customization is now an accepted paradigm in some sectors of the building industry; the forms,
shapes, and internal composition of elements of building assemblies can be customized on a massive scale, in what can be referred to as massive rather than mass customization.5 From a purely technological perspective – given the relative ubiquity of parametric design, digital fabrication, and interactive websites – dimensional customization seems to be a perfectly suitable mass customization paradigm for the housing sector of the building industry. Instead of “cookie-cutter” homes that populate most of the commercial suburban
119
10 Kolarevic.indd 119
2018-09-08 11:45 PM
developments throughout North America (see figure 10.1), the developers of such housing could actually offer truly “custom” homes, that are oneoff, highly customized products – with a unique geometry, i.e. shape and form – and make them available to a broader segment of society, and not only to its more affluent members. As discussed earlier, the technologies to economically deliver the highly customized houses currently exist: parametric design, digital fabrication, interactive websites for design, visualization, evaluation, and estimating (and automatic generation of production and assembly data). The challenges, however, are not technological, but social and cultural. We need a population with a modicum of design literacy – and interest in design – and an appropriate business and market model that would make architectural expertise in customization accessible to society at large. What makes geometric mass customization of house designs based on parametric variation particularly compelling is that it could deliver homogeneous heterogeneity at the scale of the neighborhood. Typologically – and topologically – the houses would be very similar, if not identical, yet their layout and geometry could vary considerably. Such typological (and topological) sameness, with differences in the overall layout, shapes, and forms, is what characterizes many traditional, historic neighborhoods. Adding variable geometry – via parametrics – would add elements of difference that would not compromise the overall sense of stylistic or formal unity that might be an overarching design goal at the scale of the neighborhood. After the initial customization of the house’s overall geometry, internal layout, location and size of the openings, such as doors and windows, the customers are likely to be interested in adaptive customization later on, as needs and the number of people living in the house change.6 Continuous or adaptive customization, however, could be considerably less demanding than the initial customization, depending on the level of customization that the system would permit. So,
any system that aims to provide the capacity to interactively change the geometry of the house should address not only the needs associated with the initial customization, but also the changes that are likely to be needed after the house is built. Even in the highly repetitive modes of housing production that do not offer any initial customization of the geometry, the houses undergo changes over time – additions get built, spaces repartitioned, etc. In other words, adaptive customization is a given, regardless of whether initial customization is an option. METADESIGNING At the conceptual level, embracing geometric customization of any product, whether an armchair, a shelving system, or a house, requires a fundamentally different approach to design than is traditionally the case. First, it requires a definition of metatype by the designer and metadesign, i.e. the design process that can produce a number of variations, individually tailored designs that differ from each other, yet remain within the type defined by the designer. This is not as unprecedented or as radical as it sounds. One could argue that Andrea Palladio in his I quattro libri dell’architettura [The Four Books of Architecture]7 is essentially providing a definition of different villa metatypes (such as having the cruciform layout), with the metadesign of the different types described in detail.8 One could go back even further in history to classical orders as true metatypes, with clearly articulated metadesign rules that specify the proportions and relationship of various elements. (They were parametric too.) Instead of designing a single house, or creating several slightly different designs, the architect would in this scenario (like Andrea Palladio) become a metadesigner who creates a design system that can produce hundreds or thousands of different designs. A parametric definition of a house’s geometry could be then made accessible via an interactive website to the masses, who could then design their own, unique versions of the house.
120
10 Kolarevic.indd 120
2018-09-08 11:45 PM
10.2 Homeomorphic (topologically equivalent) figures.
In metadesigning a house, the parametric definition of its geometry requires careful consideration not only of its parameters, but first and foremost of the house’s topology, i.e. essential relationships that stem from its typological definition, and then the resulting parametric hierarchy, with parameters articulated at different levels, with a clear definition of interdependencies. While the notion of parametric variation is now commonly understood and broadly used in practice to vary and dynamically explore the geometry and spatial articulation of building designs, the actual organization or structuring of dependencies and relationships in designs is in a state of free-for-all where no particular established methodologies exist beyond the already familiar grids and modules. This is in large part due to a relative absence of a proper discourse that addresses the role of topology versus geometry in architectural design. TOPOLOGY VERSUS GEOMETRY After all, nothing is more fundamental in design than formation and discovery of relationships among parts of a composition. — William Mitchell and Malcolm McCullough9 According to its mathematical definition, topology is the study of intrinsic, qualitative properties of geometric forms that are not normally affected by changes in size or shape, i.e. which remain invariant through continuous one-to-one transformations or
elastic deformations, such as stretching or twisting. A circle and an ellipse, for example, or a square and a rectangle, can be considered to be topologically equivalent, as both a circle and a square could be deformed by stretching them into an ellipsoid or rectangle, respectively. A square and a rectangle have the same number of edges and the same number of vertices, and are, therefore, topologically identical, or homeomorphic. This quality of homeomorphism is particularly interesting, as focus is on the relational structure of an object and not on its geometry – the same topological structure could be geometrically manifested in an infinite number of forms (figure 10.2). The notion of topology has particular importance in metadesign, as the focus shifts to relations that exist between and within elements of the geometry. These interdependencies then become the structuring, organizing principle for the generation and transformation of form. What makes topology particularly appealing in architecture is the primacy over form of the structures of relations, interconnections or inherent qualities that exist internally and externally within the context of a project. But what defines the topology of an architectural project? Is there some kind of universal set of relationships and interdependencies that defines forms and spaces? How should those relationships and interdependencies be structured – as hierarchies or networks? How should they be visually manifested and manipulated?
121
10 Kolarevic.indd 121
2018-09-08 11:45 PM
10.3 An interpretation and incremental assembly of Peter Eisenman’s House II based on the concept of “construction” lines and their relationships. Geometric shapes and relations are abstracted and translated into a relational drawing.
For example, using geometric relations, a designer can enforce the desired spatial configurations of building components and spaces – the project’s topology.10 The established relations constrain the design possibilities; they structure possible manipulations (figure 10.3). The choice of relationships applied in a composition may result in dramatically different designs even though a small set of possible relations and a few transformations are available (figures 10.4, 10.5). How the composition is assembled, structured, or restructured, determines its developmental potential. As William Mitchell observes: [T]he choice of modeling conventions and organizational devices that will structure the internal symbolic model ... will determine how the model can be manipulated, and what can be done with it.11 Once the topological structure of the house is defined, then there is the issue of topological operations, which affect the internal relational structure first and, as a consequence, the resulting forms. As a rather simple example, a rectangle could be transformed into a triangle with a single topological operation of deleting one of its vertices. Syntactically and semantically, this is one of the simplest operations that could be carried out. Obviously, the repertoire of topological operations will in large part be determined by what constitutes the topology of the project, i.e. what connects elements of the geometry to each other. The structuring of dependencies – the topology of the parametric model – determines how it transforms as the parameters are varied. A designer must understand the underlying topological (i.e. organizational) structure of the model to operate successfully upon it. This understanding is required on a basic, pragmatic level: if an interconnected element is moved, which other elements will move too? However, if the topology of the model, i.e. its composition or the underlying organizational structure, is too complex, applying a transformation to it might be difficult to control and envision.12
Constructing a topology of a house isn’t as complicated as it sounds. Early work in graph theory that was carried out in the 1970s by Lionel March, Philip Steadman, and others, captured in William Mitchell’s seminal book, Computer Aided Architectural Design,13 contains the seeds of an appropriate definition of the house’s topology. In The Logic of Architecture, William Mitchell offered examples of shape grammars as a way of defining and structuring the topology of houses.14 In well-defined design worlds,15 shape grammars could indeed provide an appropriate basis for the topological definition of the designs.16
122
10 Kolarevic.indd 122
2018-09-08 11:45 PM
10.4 A possible transformation of Peter Eisenman’s House II, based on a relational interpretation illustrated in figure 10.3.
10.5 Another possible transformation of Peter Eisenman’s House II.
STRUCTURING PARAMETRICS
(points, lines, shapes, etc.) are mutually linked. That way, interdependencies between objects can be established, and objects’ behavior under transformations defined.17 Metadesign, as a process of designing the design, requires that designers construct a parametric, computational system of formal production based on a well-understood, topological (i.e. relational) definition of the constituent geometry, create an interface with on-screen controls that affect its outcomes, and then select forms that emerge from the operation of the system. In this digital morphosis, a system
Parametric design is now commonly understood as an enabling digital technology for an infinite variation of shapes and forms, either through the embedded, inherent ways in which geometry is represented within the chosen drawing and modeling software or via visual programming aids or scripting. By assigning different values to the parameters, different objects or configurations can be created. Equations can be used to describe the relationships between objects, thus defining an associative geometry in which geometric elements
123
10 Kolarevic.indd 123
2018-09-08 11:45 PM
of generative rules, relations, and/or constraints is defined first (in-formation), and its interactive controls specified; the resulting structure of interdependencies is often given some generic form (formation), which is then subjected to the processes of trans-formation, driven by those very same relations and rules embedded within the system itself. A variety of systems have been developed over the past two decades to facilitate the sophisticated structuring of geometric models and their parametric variations. The early experiments in relations- or constraints-based modeling led to robust commercially available systems that provide for powerful conceptions of form and spatial organization. In addition to myriad tool palettes and option boxes that provide for interactive creation and manipulation of geometry, many offer ways to extend the system’s capacity via some kind of programming or scripting language. Some provide means of visually programming the underlying structure of the geometry, i.e. in some programs, the coding is no longer required. Developing a parametric model is in fact programming: entities are defined that form larger assemblies, relationships are established, values are assigned to parameters. Some designers do this through scripting (i.e. programming) and some use sophisticated modeling software that features visual programming. Regardless of the way in which it is done, effective parametric design requires abstractions, definitions of relationships, i.e. more than the simple knowledge of syntax of some programming or scripting language or the features of some modeling software. Once the project’s topology is articulated, the geometry can be either procedurally created (via programming scripts) or interactively modeled using some visual context for establishing associations between constituent elements of the geometry. Making a parametric model of the geometry accessible to the masses requires that careful consideration be given to which parameters should be presented for direct manipulation and more importantly, how many at any given
time. Presenting too many could “intimidate” the user and could lead to mental confusion. Only the parameters that are most relevant to manipulate the geometry at the given level should be shown. For example, it doesn’t make sense to make accessible the parameters that drive the sizes of the windows when the overall dimensions of the house are the primary concern. Thus, certain parametric minimalism becomes necessary. This also requires a careful definition of the parametric hierarchies in the definition of the house’s geometry, starting with the “global” parameters that affect the overall geometry of the house, down to the “local” parameters that define, for example, the number and the spacing of window mullions. A simple interface, coupled with the parametric minimalism discussed above, should focus customers’ minds on the most essential design features. While each house in its design family could have a unique geometry, they would all share an identical, “standard” underlying parametric model, i.e. a metadesign. That single parametric definition of the geometry can generate an infinite number of “nonstandard” shapes or forms, all of which belong to the same design family, i.e. the same design space. From a metadesign perspective, a key characteristic of this process is that emphasis shifts away from designing a particular form with a discrete set of dimensions to a parametric system that can produce a range of designs. This requires that designers of the parametric system define dimensional ranges – minima and maxima – instead of discrete dimensions and create design rules that would limit the generation of “bad” designs. ASSESSING PERFORMANCE, BUILDABILITY, ESTHETICS There are also other interesting conceptual challenges in the design of software that facilitates customization related to the functional and esthetic quality of the houses designed by non-designers. For example, the parametric design “engine” should ensure that each dimensionally customized house would perform well, both structurally and
124
10 Kolarevic.indd 124
2018-09-08 11:45 PM
environmentally. Such performance evaluations, for aspects of design that are easily quantifiable, could be performed on the fly so that “bad” designs could be immediately noted as such. If the system relies on modular panelization, with specific rules that drive allowable sizes and combinations, each customized design could be similarly assessed for compliance in real time as the geometry is being manipulated. In other words, it is insufficient to provide only for the interactive parametric manipulation of the geometry; any change in the geometry would have to be assessed in real time for its performativity and buildability. Thus, in an attempt to ensure that altered designs are viable, i.e. that they could perform and be buildable, the designer of the system would have to limit which aspects of the geometry could be manipulated. Restricting the variability of the geometry could lead to a rather small design space that provides a limited range of options that may not elicit much interest among the customers. Allowing too much variability, on the other hand, could lead to the problem of having too much choice that has already been identified as an issue in mass customization.18 The most difficult issue to address, however, is the esthetics. Arguably, the designs should be esthetically acceptable, requiring that purely qualitative and highly subjective aspects of design be somehow measured, i.e. quantitatively assessed within the software. While the esthetic judgment is seemingly transferred to the customer, a designer retains principal control over key elements of the design, deciding which decisions the customer can make. Similar to the previous discussion related to performativity and buildability, in an attempt to control the esthetics of the outcomes, the designer could end up offering too little choice; conversely, too much choice could potentially lead to esthetically questionable results, which, while acceptable to the customers, could be regarded by the designer – or the builder or manufacturer – of the house as having a negative impact on the market perceptions, and thus acceptance, of the designed product.
CO-DESIGNING AS THE FUTURE The extent of responsibility that a customer is able and willing to assume in making certain design decisions is perhaps the key obstacle to a broader acceptance of mass customization as a business and marketing strategy. Selecting colors and materials is within the comfort zone of most customers – and that is the principal reason why such cosmetic customization is the rule at present. Even with the apparent ease with which such customization could be done, still a relatively small percentage of the customers choose to do it. Anecdotally, very few in our community of professional designers have designed our own shoes or jackets by accessing interactive websites – in spite of the fact that many of them have been around for almost a decade.19 The large majority of us still opt to walk into a shoe or fashion store and pick the standard offerings from the shelf or the rack. While cosmetic customization seems to be easy to provide and deliver – and is widely acceptable to customers – customization of geometry, on the other, is anything but. While technologically possible, on both the design and production side, its successful implementation is far from being trivial. The biggest challenge is on the customer side: manipulating geometry requires expertise or, at a minimum, a certain degree of comfort and confidence. Figuring out how to empower customers to become designers (or co-designers) is key to the successful deployment of any production system that aims to provide geometric customization. A potentially promising strategy is to make design expertise accessible to the customer by having a designer work alongside the customer; this can be done physically, i.e. in the same space looking at the same screen, or virtually, which is a more likely and less expensive scenario. This implied process of co-designing, in which both the customer and a professional designer work together, could start virtually, as an interaction via a website, and then move into a physical space as the customer becomes more interested and invested into the design process. This projected dynamic is not that different from what already takes place
125
10 Kolarevic.indd 125
2018-09-08 11:45 PM
in the industry; the challenge is to make that interaction seamless, relatively effortless, and most importantly, enjoyable for the customer. Finally, there is a question of authorship, i.e. who ultimately is the author of the final design – the designer of the parametric system, or the customer (i.e. the co-designer) who chose the parameter values for the design? The legal ramifications of co-designing a house will almost certainly require involvement of a trained professional, at least in the early years of mass customization. As the ways in which customers go about acquiring a house change, and as mass customizable designs begin to proliferate and create a new culture of co-designing, we could end up with a society in which the degree of design literacy – and design confidence – change for the better over time. The principal question for the success of mass customization will remain: how many of us are prepared to become co-designers instead of mere customers? The promise of mass customization was a dramatic revolution in how we shop and consume. The reality is that it is more of a relatively slow evolution, as cultural and social habits take time to change. The mass customization challenges are not technological – they are cultural and social. The number of mass customizable products will grow, but that growth is likely to be slow and difficult for some producers. Eventually, some will get it right – and open the consumers’ minds to the liberating effects of being different. CONCLUSION When parametric design was introduced into architectural discourse in the late 1980s and the early 1990s, for the first time in the twentieth century architects were designing not only the specific shape of the building but also a set of principles encoded digitally as a sequence of parametric interdependencies, by which specific instances of the design can be generated by simply varying the values of parameters. In a radical departure from the traditions and norms
of architectural design, instead of just modeling an external form, designers first and foremost articulate an internal generative logic (i.e., the metadesign), which then produces, often in an automatic fashion, a range of design possibilities – a design space – from which various alternatives can be chosen for further exploration. While interactive parametric design is broadly understood and widely accepted in the architectural community, the real estate industry has yet to effectively leverage the implied promise of making customizable design accessible to a broader segment of society. This potential for the democratization of the design process has interesting implications for the most commoditized sector of the industry – the commercial provision of suburban housing, as discussed earlier. We may soon witness the emergence of interactive websites where customers can fully customize the overall spatial layout and appearance of the chosen house design, altering, for example, the sizes of various rooms, locations and sizes of the doors and windows, etc. down to the number of mullions in the windows, let alone the materials and finishes on the houses (i.e. engage in cosmetic customization that is already on offer). Such customer-designed homes could then be verified structurally or otherwise, with the geometry of various components generated directly for automated production using digital fabrication and robotic assembly. These technologies already exist, i.e. there aren’t any technological impediments to implementing such design and production systems. The challenges are largely cultural, i.e. it is unlikely that most customers would be willing to assume responsibility for the design of their homes. These scenarios also redefine the central task that architects of mass-customized homes would have to undertake: instead of designing with discrete dimensions, they would be designing with dimensional ranges in mind; as already discussed, this also requires a careful definition of parametric hierarchies in the definition of the house’s geometry. And then there is the most important challenge – to ensure that designs that emerge
126
10 Kolarevic.indd 126
2018-09-08 11:45 PM
out of the operation of the parametric system are not only viable, but also “good.” The esthetic challenge will thus remain: we, as a society and a culture, do not have a capacity to weed out bad designs in the world of mass-produced suburban housing, let alone the mass-customized one. ACKNOWLEDGMENTS Parts of this chapter were previously published by the author in two chapters: “Parametric Evolution,” in Brady Peters and Terri Peters (eds.) Inside Smart Geometry, London: Wiley, 2013, and “From Mass-Customization to Design Democratization,” in Tom Verebes (ed.), Mass Customized Cities, London: Wiley Academy Press, 2015. NOTES
12 Kolarevic, op. cit. 13 Mitchell, William J., Computer-Aided Architectural Design, New York: Petrocelli Charter, 1977. 14 Mitchell, William J., The Logic of Architecture: Design, Computation and Cognition, Cambridge, MA: MIT Press, 1990. 15 Ibid. 16 For more information, see Chapter 11 by José Pinto Duarte in this volume. 17 Kolarevic, Branko, “Lines, Relations, Drawing and Design,” in Harfmann, Anton and Fraser, Michael (eds.), Reconnecting, Proceedings of the Association for Computer Aided Design in Architecture (ACADIA) 1994 Conference, Washington University, St. Louis, 1994. 18 See Chapter 3 by Frank Piller in this volume. 19 In the lectures I had given on the subject of mass customization and design democratization, I would often ask the audience how many of them have used interactive websites to design shoes or jackets; typically, very few hands would go up when the question was posed.
1 Knausgaard, Karl Ove, “My Saga, Part 1,” New York Times, Sunday Magazine, March 1, 2015, p. MM34. 2 Pine, Joseph B., Mass Customization: The New Frontier in Business Competition, Boston: Harvard Business School Press, 1993. 3 Piller, Frank, and Tseng, Mitchell M. (eds.), Handbook of Research in Mass Customization and Personalization, Singapore: World Scientific Publishing Company, 2009. 4 See Chapter 13 by Joseph Tanney in this volume. 5 Marc Fornes introduced the term “massive customization” in one of his public lectures several years ago. For more information, see Chapter 5 by Marc Fornes in this volume. 6 See Chapter 15 by John Brown in this volume. 7 Palladio, Andrea, The Four Books of Architecture, New York: Dover Publications, 1965. 8 For the discussion of parametric variation of different Palladian villa designs, see Hersey, George, Possible Palladian Villas, Cambridge, MA: MIT Press, 1992. 9 Mitchell, William, and McCullough, Malcolm, Digital Design Media, New York: Van Nostrand Reinhold, 1991. 10 Kolarevic, Branko, “Geometric Relations as a Framework for Design Conceptualization,” doctoral thesis, Harvard University Graduate School of Design, 1993. 11 Mitchell, William J., “Architecture and the Second Industrial Revolution,” Harvard Architectural Review, vol. 7 (1989).
127
10 Kolarevic.indd 127
2018-09-08 11:45 PM
128
10 Kolarevic.indd 128
2018-09-08 11:45 PM
11 CUSTOMIZING MASS HOUSING: TOWARD A FORMALIZED APPROACH
JOSÉ PINTO DUARTE 129
11 Duarte.indd 129
2018-11-04 5:08 PM
11.1 Hierarchical structure of the design world.
This chapter describes a framework for developing integrated systems for mass customization, using housing as an illustrative example. The aim of mass customization is to lower the costs through recourse to large-scale production methods, while satisfying the unique requirements of each individual household to guarantee customer satisfaction. The integrated system includes a design system that encodes the rules for generating customized designs and a digital production system that makes it possible to construct from such designs. Integration is achieved through a computer tool that enables the easy exploration and visualization of solutions, and automatically generates the information required for production. Discussion is focused on the design system. TYPE AND MODULE In a seminal book, Herbert Simon proposed the notion of The Sciences of the Artificial.1 The role of these sciences would be to unveil the laws underlying the systems of objects designed by humans, as the role of the natural sciences was to discover the laws governing the natural world. From this analogy between the design and the natural words naturally
follows the idea that there must be a code that specifies how to generate designed objects, similar to how DNA encodes the rules for generating living creatures. Also, as DNA is manipulated over time through a process of natural selection to generate living beings that are the fittest for the environments where they live, it should be possible to manipulate the design code to generate designs that are the most appropriate for their respective design contexts. By design context is meant the features associated with the user and the site, as well as the associated social and technological aspects. This is the basic idea of customization. We are then left with two problems: (1) to find the design code; and (2) to discover how it can be manipulated to generate customized designs. To accomplish this complex task, we must delve into the nature of design. This is a rather complex and philosophical endeavor, but I propose a shortcut. In Type and Module, the design world was described as a hierarchical structure where one can find on any given level various “types.”2 These types result from the combination of lowerlevel types and are the modules for composing higher-level types. For instance, on a certain level there can be windows, doors, floors, walls,
130
11 Duarte.indd 130
2018-11-04 5:08 PM
11.2 Basic conceptual model for mass customization.
and ceilings of different types. These can then be combined to form different rooms, for instance, bedrooms, living rooms, kitchens, and so on. These room types can then be combined to form different building types, such as single detached houses, townhouses, high-rise, and so on, which in turn can be used to form various city blocks, and so on. This hierarchical combinatorial process can extend from the very small to the very large (figure 11.1). According to Habraken, a house type is a result of three systems: (1) a spatial/functional system that specifies how rooms are organized in relation to one another; (2) a structural/building system that determines how a house is materialized using an existing building material and technology; and (3) a stylistic/decorative system that defines the rules for embellishing and decorating the house elements (walls, ceilings, windows, etc.).3 Type is not a model to be repeated exhaustively, but rather a very flexible concept that captures and reflects the rules of living and social conviviality of a society in a certain moment in time. It has the flexibility to adjust to different family profiles and to specific social and site contexts, while retaining the values and features that are common to all of them. Type can thus be instantiated in many ways,
depending on the design context. It is, therefore, a very useful concept in a mass customization paradigm. The problem becomes one of defining an appropriate type and finding how to instantiate it for a given context. This equates to developing a design system encoding the rules of the type and applying them in a way to instantiate the type in an appropriate way. A CONCEPTUAL MODEL FOR MASS CUSTOMIZATION: DESIGN AND PRODUCTION The aim is to devise a “system” that enables one to generate customized products. This system (figure 11.2) has two main parts: (1) a design system that takes as input the design context and outputs a design; and (2) a production system that takes the design as input and materializes it into a product. If we give a human the task of designing customized houses for many clients, he or she will take too long to process contextual data and generate differentiated solutions and even longer to construct them. These limitations can be overcome with recourse to computer systems, taking advantage of the computer’s high-processing power to manipulate information.
131
11 Duarte.indd 131
2018-11-04 5:08 PM
11.3 Conceptual diagram of the design system.
11.4 Conceptual diagram of the production system.
The design system is composed of four subsystems (figure 11.3): (a1) a formulation subsystem that can read and interpret the design context (the user profile and the surrounding physical and social context) and generate design specifications, which means mapping features of the context to features of the design; (a2) a generative subsystem that can yield design solutions (shape configurations and material characteristics) according to a predefined system of rules that define a space of design solutions; (a3) an evaluation subsystem that can compare the performance and adequacy of different design alternatives; and (a4) a search subsystem that can browse the space of design solutions and find the best solutions, that is, the ones that most closely satisfy the client’s requirements and have better performance. The production system can take different forms, from handcrafted to automated production (figure 11.4). The latter includes digital fabrication
of parts and their assembly, and it has some obvious advantages, including speed, cost, and quality control. It may also include the multi-material, additive manufacturing of the entire dwelling. To come up with the design system, we have to address four problems:
•
•
Formulating: This problem has two parts. The first is to elicit the user’s needs, that is helping users to understand and communicate what their needs are. The second part is to find a way of linking features of the design context to design specifications; this means defining the mechanism behind the formulation subsystem. Generating: Solving this problem requires one to extract design rules, that is, rules of spatial and formal composition. The most successful systems so far use if-then rules such as shape grammars4 to outline the algorithm and then convert them into parametric design models for
132
11 Duarte.indd 132
2018-11-04 5:08 PM
11.5 Conceptual diagram of the processes for developing and applying design systems.
•
the ease of implementation. Rule inference has been done manually, mainly because automating the process runs into complex shape recognition problems. However, rule inference is a rather tedious and cumbersome problem and theoretically it should be amenable to automation, but more research needs to be developed on this issue. There have been attempts to use other artificial intelligence paradigms, such as constraint propagation and neural networks, but these have not been as successful. On the other hand, “rule” inference in these cases might be easier because it runs into less complex shape recognition problems. Evaluating: This problem means having a way of rating and ranking the performance of designs. This is the realm of simulation and analysis and there is software to do this for specific, individual viewpoints
•
(structural, environmental, etc.), but other viewpoints (energy production, ecological impact, etc.) can and need to be added to have a more thorough customization process and designs with improved performance. Searching: To solve this problem we need to possess a way of browsing the space of design solutions, that is, a search mechanism that is adequate to the particularities of the design space.
Figure 11.5 presents a conceptual diagram depicting the various possibilities for developing and applying generative design systems. We will use this diagram to guide the discussion that follows and map how the problems above might be addressed, particularly, the development of the generative design system. 133
11 Duarte.indd 133
2018-11-04 5:08 PM
11.6 The corpus used to infer the specific grammar for Siza’s Malagueira houses.
INFERRING A SPECIFIC DESIGN SYSTEM The first possibility is to develop a specific design system by analyzing a series of existing designs (figure 11.5, step 1), the corpus in shape grammar terms (figure 11.6) and extract the design rules. The literature contains a wide range of analytical grammars developed after this process, such as Palladian Villas,5 Frank Lloyd Wright’s Prairie Houses,6 and Malagueira Houses,7 to mention just a few. This process is justifiable only if the number of new designs that can be generated by the grammar compensates for the investment in developing the grammar. A specific grammar may capture the rules of a given designer like the Malagueira grammar or those of a collective, historical type like the Queen Anne grammar.8 GENERATING SPECIFIC DESIGNS Once we have a specific grammar, we can use it to generate new specific designs (figure 11.5, step 5; figure 11.7). The formulation subsystem helps us to convert data on the design context into design specifications; the generative system outputs candidate solutions; the evaluation system
determines how fit they are, that is, to what extent they match the design specifications, and what their performance is from different viewpoints, such as energy consumption, structural stability, and so on; finally, the search system helps us to navigate the space of design solutions and find the fittest. To go from the initial shape to the fittest design, there are two possible basic scenarios. In the first, information from the evaluation system is used to constrain rule application to designs with improved performance. In this scenario, the rules are compositional rules, as they are used to progressively compose the design from a very simple initial shape. This was the strategy used, for instance, to generate customized Malagueira houses.9 In the second scenario, the design is first produced using compositional rules and then progressively transformed using transformational rules toward designs with improved performance. The evaluation system is also used in this scenario to guide this progressive transformation. An example of this strategy is the work of Shea and Woodbury.10 Different search algorithms may be and have been used, from heuristic to stochastic search, including best-first search,11 simulated
134
11 Duarte.indd 134
2018-11-04 5:08 PM
11.7 The first stage in the Malagueira specific grammar is the division of a rectangular plot into a basic pattern consisting of four functional zones (living, sleeping, service, yard). In the development of the grammar, four situations were identified: patterns designed by Siza (bold), patterns he did not design but could have designed (gray), patterns he would not have designed but admitted others could, and patterns he did not admit that could be designed.
annealing,12 and genetic algorithms,13 to name a few. The choice of search algorithm may depend on several factors, such as how much is known about the search space, the complexity of the problem measured by the number and type of variables, and time constraints. In general, stochastic algorithms take longer to converge to an optimal solution, and heuristic methods run faster but may not guarantee that the global optimum is reached. In mass customization problems, time might be a limiting factor, particularly if the configurator is to be used by the end user, in which case, heuristic methods are more advisable to avoid having the user staring at the computer for a long time while a solution is being generated.
DESIGNING A SPECIFIC DESIGN SYSTEM The process of generating a solution from existing designs described above has the disadvantage of requiring the existence of previous designs, binding the developer of the grammar to work with an existing style. What if the goal is to develop a new style? What if the developer of the grammar is the designer who wants to develop his or her own design system in his or her own style? The initial idea to solve this problem came from the analysis of Siza’s designs at the Malagueira and from the process of developing the corresponding grammar. Figure 11.8 shows the derivation 135
11 Duarte.indd 135
2018-11-04 5:08 PM
11.8 Derivation tree of the Malagueira specific grammar.
tree for all the designs used to infer the grammar (designs A–E). Design Aa was the first designed by Siza at Malagueira. All the other designs are variations of this initial design that differ from it in the way rules are applied, by manipulating the symmetry properties of shapes in the rules and/or changing the dimensional and functional parameter values associated with them. As such, when Siza created this initial design, he implicitly developed
the underlying grammar. The grammar would be made explicit later in a similar fashion, first, by intrinsically analyzing the initial design, and then comparing it with subsequent designs to extract the rules and understand the limits of design variation. I have used the same procedure in design studios where students are asked to produce a design system for the mass customization of housing in two stages. First, students are asked
136
11 Duarte.indd 136
2018-11-04 5:08 PM
11.9 Grammatical transformations vs. transformation grammars.
to use a housing design from previous studios or to design one following a traditional, intuitive approach. Then, they are asked to analyze their design, understand its formal and functional structure, meaning its parts, their shape, and their function, and write down the rules, in whatever format best suits them. Ultimately these rules need to be organized in an algorithmic fashion so that they can be followed by anyone in the generation of new designs. Students’ initial designs are the immediate precedents for the development of the design system. However, the primordial precedents are the designs they used, aware or unaware, as a reference to create their designs. This observation leads to a more formalized approach to the conception of the design system. In this formalized approach, students are asked first to collect housing designs that are relevant to the current design context. This can be a survey of local house types that may include vernacular housing, signature designs by local architects, or simply a random selection of house models by local developers or builders. It can also include other house designs, for instance, by world-famous architects whose work they admire, and which may have some connection to the current context, such as the building type (singlefamily, high-rise, etc.), the socio-economic level, and so on. Students are then asked to infer the grammar underlying the chosen house type. Then they are asked to transform the grammar to fit their interpretation of the current design context to guarantee the adequacy of the system (figure
11.5, step 9). Finally, they can use this new specific system to generate houses for particular clients on a particular site (figure 11.5, step 4). The design of systems by transforming a specific grammar into a new specific grammar has its roots in Knight’s Transformations of Design.14 In this work, Knight describes stylistic evolution through the transformation of the grammar underlying an earlier style into the grammar underlying the later style. These grammatical transformations (figure 11.9) occur by changing the rule set of one grammar into the rule set of another grammar by deleting rules, adding new rules, or simply changing rules by fiddling with the constraints on parameters’ values, while keeping some rules unchanged. DESIGNING A GENERIC DESIGN SYSTEM There is an alternative way to obtain a new specific system by transforming an existing one. This alternative pathway requires one to develop a generic grammar (Figure 11.5, step 2). A generic grammar encompasses more than one specific grammar. However, the generic grammar is more than the union of the rule sets of specific grammars. First, the different specific grammars used to infer the generic grammar need to be developed in the same format, that is, as additive, subtractive, or grid grammars. Then, rules of the same kind need to be generalized into generic rules. The literature already includes some examples of generic grammars developed according to this procedure. For instance, 137
11 Duarte.indd 137
2018-11-04 5:08 PM
Benrós et al. developed a generic grammar for Palladian Villas, Malagueira Houses, and Prairie Houses,15 Costa et al. proposed a generic grammar for a given ceramic company’s tableware,16 and Garcia and Romão presented a generic grammar for multipurpose chairs.17 An important issue is the degree of “genericness” of grammars. Beirão and Duarte (2018) put forward the idea of encoding the design knowledge of a certain domain into a generic grammar and illustrated the idea by developing a very generic grammar for urban design.18 Similarly, one could develop generic grammars for housing, tableware, and chair design. All these generic grammars encode the rules for generating potentially all the designs of a generic type in a given domain. GENERATING A SPECIFIC SYSTEM FROM A GENERIC DESIGN SYSTEM Having a generic grammar to start with is very useful, as this grammar does not bind the designer to work within a given style but allows him or her to develop a personal style. This requires the designer to restrict the generic grammar rule set by deleting some rules or restricting other rules, limiting their parameter values (figure 11.5, step 3). This process can be rather idiosyncratic, but in doing so, the designer can take into account information on the context. In any case, the designer is defining the rules of the game by determining the features of designs that can be generated downstream, either by himself or herself, other designers or the end users. Generating a specific grammar from a generic grammar through rule set restriction is different from generating it from another specific grammar through rule set transformation. In the latter case, the designer’s work is potentially less limited as he or she can add new rules at any moment. Developing a generic grammar in the first place allows for an increased control of the process and the outcome, but may limit the universe
of design solutions, thereby imposing limits on design diversity and restricting the opportunity for customization. Of course, one can keep the generic grammar open, allowing one to add new rules at any time, thereby gradually increasing its genericness. GENERATING A SPECIFIC DESIGN DIRECTLY FROM A GENERIC GRAMMAR Nothing prevents a designer from generating a specific design directly from a generic grammar (figure 11.5, step 10; figures 11.10 and 11.11), bypassing the intermediate step in which a specific grammar is generated. This process theoretically corresponds to what happens when a less-experienced designer (one who has designed fewer houses) designs a customized home for a client in traditional handcrafted processes. Implicitly, the designer may be defining a personal style, that is, designing a specific grammar, but this is not the main concern as the designer is focused on the design of a house for the client, not on the design of a set of houses for potentially many clients. Steps 6, 7, and 8 in figure 11.5 are also possible but they do not need to be singled out, as they are similar to other steps described above. Step 6 is analogous to step 3, using a generic grammar to generate a specific grammar. Step 7 is like step 2, using a specific grammar in the generation of a generic grammar. And step 8 corresponds to step 1, inferring a specific grammar from an existing design. TRANSFORMATION GRAMMAR VS. GRAMMATICAL TRANSFORMATIONS Above we mentioned how to use grammatical transformations to generate a new grammar from an existing one. These transformations occur directly at the grammar level by manipulating the rules of an existing grammar to obtain a new one. It is also possible to make transformations at the design level (figure 11.9). Such a
138
11 Duarte.indd 138
2018-11-04 5:08 PM
11.10 Existing and new, hybrid designs generated after a generic grammar developed from specific grammars for Palladian Villas, Malagueira Houses, and Prairie Houses.
139
11 Duarte.indd 139
2018-11-04 5:08 PM
11.11 Derivation tree showing the step by step derivation of the layouts of a Palladian Villa, a Malagueira House, and a Prairie House, as well as hybrid house designs between these three styles from the same generic grammar.
transformation of one design into another is useful, for instance, when the aim is to customize existing artifacts. The process can be generalizable so that the transformation grammar transforms any design in a grammar into a design in another grammar. Eloy and Duarte developed a grammar of this kind to rehabilitate Rabo-de-Bacalhau dwellings,
designed from 1945 to 1965, to comply with contemporary living requirements (figures 11.12a and b).19 The grammar was inferred after a small set of experiments in which designers were asked to rehabilitate specific dwellings for specific clients, but it can be used to customize any instance of the Rabo-de-Bacalhau type. More recently, Guerritore
140
11 Duarte.indd 140
2018-11-04 5:08 PM
11.12a and b Conversion of an original Rabo-de-Bacalhau dwelling to satisfy contemporary needs after a transformation grammar: original layout (a) and transformed layout (b).
and Duarte proposed a more generic transformation grammar to convert high-rise office buildings into housing.20 This grammar was designed first by identifying the housing building type into which the office building was more amenable to be converted, and then systematically mapping all the possible transformations of office into dwelling floorplans. At the end, the grammar can be used not just to convert office buildings into housing buildings, but also to design customized dwellings in the process. These layout transformation grammars are akin to the kind of grammar used by Shea et al., mentioned above, for truss design. As such, they can be used in similar optimization processes, where the search starts from an existing design that is progressively transformed into a customized solution, rather than from a very basic shape that is incrementally developed as in the Malagueira grammar. ROLES IN MASS CUSTOMIZATION: GRAMMARIANS, DESIGNERS, AND END USERS In the discussion above, two main tasks can be identified: (1) to create grammars (more or less specific or generic); and (2) to use them. Creating a grammar instead of a design means defining a space of design solutions instead of a single solution. This means that the developer of the grammar needs to anticipate all the possible admissible outcomes. According to the process described above, this task can be facilitated by inferring the grammar from the analysis of existing designs and then transforming it
to fit the current design context. This process helps to ground design decisions, and, to a certain extent, it is like precedent-based design. The process of creating a grammar is mainly one of abstracting; the creator of the grammar first identifies the variables and then frees them. Conversely, the process of applying a grammar, which I would call concretizing, for lack of a better word, consists of the opposite. In this process, the user of the grammar starts from a larger range of possibilities and then restricts the variables down to very specific values. But who among the stakeholders of the mass customization process are the creators and users of grammars? Experienced designers can certainly perform both, but it might be beneficial to assign different tasks to different specialists. To researchinclined designers could be assigned the task of inferring specific and generic grammars. Developing a specific grammar from a generic grammar would certainly be the task of design-oriented designers. The task of generating a specific design from a specific shape grammar could be performed by end users too. However, as the degree of genericness of a grammar can vary, so can the division of tasks. In general, developing or manipulating more generic grammars, I would argue, requires more experienced designers. Theoretically, to increase the likelihood of customization, one should aim at a system that can generate a wide variety of designs. But this has two setbacks: one is the increased difficulty in guaranteeing that appropriate designs can be generated, and the other is the risk of overwhelming end users with the burden of choice. 141
11 Duarte.indd 141
2018-11-04 5:08 PM
NOTES 1 Simon, Herbert, The Sciences of the Artificial, Cambridge, MA: MIT Press, 1968. 2 Duarte, José Pinto, Tipo e Módulo: Uma Abordagem ao Processo de Produção de Habitação, Lisbon: LNEC, 1995. 3 Habraken, John, “Type as a Social Agreement,” paper presented at the third Asian Congress of Architects, Seoul, Korea, November 1988. 4 Stiny, George, “Introduction to Shape and Shape Grammars,” Environment and Planning B: Planning and Design, vol. 7, no. 3 (1980), pp. 343–351. 5 Stiny, George, and Mitchell, William J., “The Palladian Grammar,” Environment and Planning B: Planning and Design, vol. 5, no. 1 (1978), pp. 5–18. 6 Koning, H. and Eizenberg, Julie, “The Language of the Prairie: Frank Lloyd Wright’s Prairie Houses,” Environment and Planning B: Planning and Design, vol. 8, no. 3 (1981), pp. 295–323. 7 Duarte, José Pinto, “Towards the Mass Customization of Housing: The Grammar of Siza’s Houses at Malagueira,” Environment and Planning B: Planning and Design, vol. 32, no. 3 (2005), pp. 347–380. 8 Flemming, Ulrich, “More Than the Sum of Parts: The Grammar of Queen Anne Houses,” Environment and Planning B: Planning and Design, vol. 14, no. 3 (1987), pp. 323–350. 9 Duarte, José Pinto, “A Discursive Grammar for Customizing Mass Housing: The Case of Siza’s Houses at Malagueira,” Automation in Construction, vol. 14, no. 2 (2005), pp. 265–275. 10 Shea, Kristina, Cagan, Jonathan, and Fenves, Stephen Joseph, “A Shape Annealing Approach to Optimal Truss Design with Dynamic Grouping of Members,” Journal of Mechanical Design, vol. 119, no. 3 (1997), pp. 388–394. 11 Duarte, “A Discursive Grammar” op. cit. 12 Shea et al., op. cit. 13 Granadeiro, Vasco, Duarte, José Pinto, Correia, João, and Leal, Vítor, “Building Envelope Shape
Design in Early Stages of the Design Process: Integrating Architectural Design Systems and Energy Simulation,” Automation in Construction, vol. 32 (2013), pp. 196–209. 14 Knight, Terry, Transformations of Design: A Formal Approach to Stylistic Change and Innovation in the Visual Arts, Cambridge: Cambridge University Press, 1994. 15 Benrós, Deborah, Hanna, Sean, and Duarte, José Pinto, “A Generic Shape Grammar for the Palladian Villa, Malagueira House, and Prairie House,” in John Gero (ed.), Design Computing and Cognition ’12, Heidelberg: Springer, 2014, pp. 321–340. 16 Castro e Costa, Eduardo, and Duarte, José Pinto, “Mass Customization of Ceramic Tableware through Digital Technology,” in Helena Bártolo, Paulo J. Bártolo, Nuno Alves, et al. (eds.), Green Design, Materials, and Manufacturing Processes, London: CRC Press, 2013, pp. 467–471. 17 Garcia, Sara, and Romão, Luís, “A Design Tool for Multipurpose Chair Design,” in Gabriela Celani, David Moreno Sperling, and Juarez Moara Santos Franco (eds.), Computer-Aided Architectural Design: The Next City – New Technologies and the Future of the Built Environment, Heidelberg: Springer, 2015, pp. 599–619. 18 Beirão, José, and Duarte, José Pinto, “Generic Grammars for Design Domains,” Artificial Intelligence for Engineering Design, Analysis and Manufacturing, vol. 32 (2018), pp. 225–239. 19 Eloy, Sara, and Duarte, José Pinto, “Transformation Grammar for Housing Rehabilitation,” Nexus Network Journal, vol. 13, no. 1 (2011), pp. 49–71. 20 Guerritore, Camilla, and Duarte, José Pinto, “Rule-Based Systems in Adaptation Processes: A Methodological Framework for the Adaptation of Office Buildings into Housing”, in John Gero (ed.), Design Computing and Cognition ’18, Heidelberg: Springer, 2018, pp. 459–478.
142
11 Duarte.indd 142
2018-11-04 5:08 PM
12 INTERPLAY OF DESIGN, TECHNOLOGY, MANUFACTURING AND BUSINESS
KARL DAUBMANN 143
12 Daubmann.indd 143
2018-10-18 9:59 PM
12.1 Plan of BLU Homes Balance house.
12.2 The Balance house rendering for proposed location in southern California.
Imagine walking into a store to buy a new piece of furniture. Once you pick out the item, the store tells you that they will begin work building the item after you pay a deposit. Once it is complete, the store will let you know the cost to fabricate, and at that point you must pay for it. In addition, they are not completely sure how long it will take to build because they have a number of material suppliers who are not always reliable. Imagine a second scenario where you purchase a new car. Rather than picking it up at the showroom, a crate arrives in your driveway. Initially, you appreciate not having to go to collect your new car but after a few days, a couple of people arrive in a pickup truck and begin work. To your surprise, in the crate is not your car but the parts for your car. Unexpectedly, the truck departs because the team was missing a critical tool. A couple of days go by and a different crew arrives with more parts
and begins assembly. They let you know that it will likely rain so they will be working on a different customer’s car tomorrow because that customer has an enclosed garage. Both of these scenarios highlight the discrepancy between client experience and customer expectations for the purchase of a manufactured product. Both scenarios are not unlike the experience that many first-time residential clients face when designing a home. Most residential clients who have been through the process of commissioning a project have experienced unforeseen costs and/ or delays in the project timeline. Almost everyone will have experienced one or the other, and often both. Much of the client dissatisfaction is the direct result of intentional professional separation originally put in place to protect the client and to protect the profession. While manufactured, modular, and prefab houses have all fascinated designers
144
12 Daubmann.indd 144
2018-10-18 9:59 PM
12.3 The Balance house in Spring Green, Wisconsin.
from an efficiency and aesthetic perspective, an integrated design/construction/business model has the potential also to positively impact the client experience. A typical architectural professional process separates design from construction. Costs vary based on the market and labor. The costs are controlled by those who build, and as a result, this system creates an exclusion of knowledge from those who are involved early in the process. To thoughtfully take on client experience, improve quality, increase quantity and advance professional knowledge, architects should view contractual boundaries as innovative design moments between design, construction, and business. There are opportunities to be gained by allowing design to move upstream.1 In 2010, I began working with BLU Homes as Creative Director and the initial appeal for me was
to design houses as products (figures 12.1–12.3). In 2012, my responsibilities expanded as the Vice President of Design and for two years I worked with a multidisciplinary team taking on the problem of the single-family green home in the US. I would describe BLU as a technology company because that change in mindset from a “home builder” to a technology company empowered many within the team to embrace disruption and reimagine and expand their disciplinary roles. This experience exposed me to numerous instances where design, manufacturing, and business were intertwined in a way that is not easily visible from the outside. In this role I was able to understand the agency and impact that design might have if we look beyond a narrow design scope. It was often through specific manufacturing logistical concerns or temporal constraints that design was able to offer impactful solutions.
145
12 Daubmann.indd 145
2018-10-18 9:59 PM
VARIED METHODS OF MANUFACTURING HOUSES Architects have looked longingly toward manufacturing for generations. There is a perceived efficiency, impact, precision, modernity, and ambition that separate construction from manufacturing. One aspect central to the distance between the two approaches is the difference between client and customer. Residential architects support the client in creating a unique solution for a family, site, and circumstances. The tradeoff for the custom solution is that designers could be losing benefits that might come with the design process for an industrial product, such as market research, incremental learning, application to future products, and a dedicated team to take the project from inception to construction completion. In a typical contractual relationship, architects provide a crafted solution to clients, resulting in a mass-crafting paradigm. Mass customization does not exist industry-wide where mass refers to the masses. In order to attain the goal of mass customization in housing, the collaboration and support of manufacturing and an integrated business model must be considered as part of the design. The history of the automotive industry influences manufacturing, economics, and expectations, and as a result, this industry has had an impact on design. The iconic photo of one day of production at the Ford Highland Park Plant shows a vast exterior space filled with Model T chassis. At its pinnacle in 1926, this plant produced 9,000 cars a day and every single Model T was black; an affordable automobile for the masses (or at least a growing middle class) where no choices were available. Today at the River Rouge Plant the visitors can watch Ford F-150 pickup trucks rolling down the assembly line. Humans work collaboratively alongside robots and different models of F-150s in different colors with different options are produced in a bright and clean, climate-controlled space. Technology has transformed the design, making, and performance of modern automobiles through safety, efficiency, and luxury. In contrast to the performance gains in
automotive between 1926 and present, I compare my own house from 1845 to homes being built in the US today. While the components have evolved, the construction process and methods have been quite slow to change and often for good reason – buildings are expected to last longer than manufactured goods. My house from 1845 has a foundation, a wood structure, cladding, windows, insulation, and a roof, and was built on site by similar trades, using many of the tools recognizable to contractors today. Kieran Timberlake highlighted conceptual advances in manufacturing as a means to transform the construction industry. In Refabricating Architecture, they describe fabrication, prefabrication, the use of “chunks” and simultaneous fabrication in manufacturing as examples of how the process could evolve to incorporate off-site construction techniques.2 This continues to be an important and polemical text while we witness many of these changes happening in our industry, but sadly, these changes are rarely driven by designers. These advances are in the realm of construction. Contractually, most architects are not in a position to dictate the means and methods deployed on construction sites. As a result, it is hard to imagine these techniques as part of design understanding until designers start questioning contractual boundaries as fertile territory for disruption. While architects are not typically in a position to dictate the methods on site, architects work with or specify building products when designing. Historically, contractors may have created their own lumber on site directly from trees while today lumber is an industrial product with predictable dimensions and performance characteristics. Windows, doors, plumbing, and electrical fixtures are now products and fabricated off-site (figure 12.4) to deliver consistency on site.3 This aggregation of parts and products might scale up to include entire rooms like bathrooms, or in the case of residential construction, this product specification could scale up to modular construction where assemblies of large chunks of homes are built in factories and then delivered to a site.
146
12 Daubmann.indd 146
2018-10-18 9:59 PM
12.4 View of the PivoTek prefab bathroom module built in their factory outside of Cincinnati.
On-site construction epitomizes the slow-toinnovate nature of the residential housing market in the US. Building a house on site means that the construction process must confront changing conditions where the site or home is open and exposed for most of the process. The workers are often subjected to extreme heat or cold, depending on the region, and the workers must travel to a new location for their job depending on the location of the project. This mobile crew often lacks managerial oversight and the process and workers pose such a considerable intrusion into a neighborhood that the construction of a new home can even alienate neighbors long before the new homeowners take possession. The fragmentation of the residential market is partially to blame for this slow evolution as it continues in the way it has before. In contrast, off-site construction allows for a consistent interior environment to build that is not subject to weather. It allows for a consistent pool of labor where training and oversight are possible. The cost of distributed, skilled, on-site workers with travel, insurance and oversight is about five times that of prefab off-site construction that occurs in a factory. In a factory, materials and tools are readily available to minimize the travel to and from the lumberyard and hardware store. If a standard suburban house takes between nine months to a year to build on-site, off-site construction could minimize the on-site time to one month. The construction of
the home can occur in a factory at the same time the foundation and site work is being done on the site. A spectrum of off-site construction approaches currently exists between panelized construction and modular construction. The two ends of this spectrum are important to understand as each has benefits and liabilities. Panelized construction builds an entire home in a factory out of prefabricated panels. The panels are assembled in the factory and then taken apart and shipped flat or flat-packed. This approach minimizes shipping volume and shipping costs but these panels must be reassembled on-site. The assembly process can often be slowed by trying to find the correct panel that is needed at that point in the process. The house gets enclosed faster than traditional construction, making it weathertight, but the expensive systems, such as plumbing, electrical and mechanical systems, must be installed on-site by tradespeople. The other end of the spectrum is modular construction. This approach maximizes the volume able to be placed on a truck for shipping. Anyone who has driven through Pennsylvania on Route 80 has probably passed a modular home on a truck with cars ahead and behind signaling the wide load. Modular construction is very fast on-site, but a large crane is necessary to place the modules, and access to the site can be difficult with the large house chunks being brought in. The shipping and cost of the crane must be overcome by efficiencies and cost savings in the factory.4
147
12 Daubmann.indd 147
2018-10-18 9:59 PM
12.5 Plan of BLU Homes Lofthouse showing the condensed core.
12.6 Fritz’s folding A-Frame house as a precursor to folding prefab.
NEITHER MODULAR NOR PANELIZED BLU Homes finds a “sweet spot” between panelized and modular methods by using aspects of both approaches, the best of both worlds. Because modular construction ships the oversized loads on the highway, the usual maximum distance modular companies operate is about a 250-mile radius. When the modular homes are shipped, the majority of the shipping volume is empty space of the interior rooms. BLU Homes sought to find a solution that would provide a national business model while providing consistency and speed. Some of the rooms in a house are denser and more expensive to build like the kitchen, bathrooms, and mechanical spaces. These spaces were aligned on one side of the house into a linear core (figure 12.5). Other spaces like bedrooms, living rooms, and dining rooms, where it is predominantly space with furniture, occupied the other portion of the house module (mod for short).
The spaces that were mostly empty were built with a panelized approach to minimize the shipping of empty volume, but rather than stack these panels up, the panels were hinged and folded against the core. This folding is not without precedent in the industry. A 1967 patent by William Fritz shows a relocatable A-Frame house that unfolds like the wings of a bird (figure 12.6).5 More commonly, modular construction folds down the portions of a gable roof, with hinges for shipping to fit under the overpasses of the highway. When the house is set on site, the roof panels are rotated up into the correct orientation and slope and fixed in the final position. This hybrid approach between panelized and folding allows BLU Homes to deliver prefabricated building modules nationally because the shipping volume of 8’-6” wide does not require the special permits and expensive trucks like that of the larger modular approaches. The 8’-6”-wide module of
148
12 Daubmann.indd 148
2018-10-18 9:59 PM
12.7 Folding steel frame from 8’-6” wide to 22’-0” wide.
BLU Homes unfolds quickly on-site to 22’-0” wide (figure 12.7) and can be watertight or dried-in in one day.6 The shipping width permits a radically different type of business model that allows for a centralized national manufacturing footprint while having sales staff distributed to meet with clients. This distributed sales model also caters for clients who might reside in one state but want to build a second home or retirement home in a different state. Members of the BLU Homes team could meet conveniently in either or both locations. Mass customization is dependent on customers and, as a result, sales – and marketing – are an essential part of this desire for the mass of customers. Designing a house can be difficult, but when adding in the constraints of shipping, this task can become even more complex. One aspect of this complexity is laying out the program and systems and parsing them, based on the layout of the building modules. In the case of BLU Homes, the
folding of building panels adds yet another level of complexity. The width of a building module is determined by shipping height because the floor (when folded) moves into a vertical position. The modules are craned into their final position so the house must be also engineered to be lifted. Given the folding, shipping, and craning, the structure of the BLU houses was made of rigid steel. Steel is not a common material for a house, but given these additional constraints, it makes sense. The steel then translated into a sales and marketing approach that sought out situations where traditional wooden-framed homes might have challenges, such as areas subject to seismic events, high wind loads, and high snow loads. A regional business model might not encounter many of these structural loading characteristics, but once you begin operating at a national scale, more markets and more situations arise. 149
12 Daubmann.indd 149
2018-10-18 9:59 PM
12.8 Steel frame in the factory with insulated panel roof in place (temporary bracing).
WHEN TRADITIONAL RULES OF THUMB ARE NO LONGER APPLICABLE The rule of thumb for steel structures is that lateral bracing with moment frames is the most expensive approach and should be avoided. In a First World economy, labor is always more expensive than material, and the old adage that “time is money” makes this concept well known. In the case of BLU Homes, faster build and delivery meant looking at every option where one could understand the financial implications of a design decision on permitting, engineering, manufacturing, or delivery. The resulting structural approach at BLU Homes was to develop a steel moment frame for widespread deployment so that consistent steel sections could be purchased in bulk and welded repetitively throughout. Had a custom structural approach been used that responded specifically to site loading conditions, the engineering would need to be updated each time. This consistent moment frame minimized design time, engineering time, and permitting. The steel frame was analogous to that of an automotive chassis where different body styles, engines, or options can be interchanged. Opaque walls, walls with windows or glass doors could be swapped in the knowledge that there would be no structural consequences (figure 12.8). The AEC industry is incredibly fragmented which poses a large threat to the desire for a national business model to design and deliver singlefamily homes. By comparison to the automotive industry, where cars receive their permits at a federal level based on the performance of the car models, residential construction is permitted at the state and city levels. If a house is repetitive and built in a factory, then it can be permitted at the state level and inspected in the factory, while anything that is built on-site, like foundations, decks, garages, and building connectors, all receive their permits at the local municipal level and are inspected on the building site. Building codes vary by state and are augmented by municipal zoning ordinances, and then individual building sites might be affected by homeowners associations or
idiosyncrasies like wildfire zones or wetlands and other natural resources. These factors result in the need to have a robust design staff able to do early code research to support the sales team. We quickly realized that every site was unique, where adjacent sites in the same neighborhood had different permitting issues and different issues of shipping access (figure 12.9). As a result, each potential customer receives a conceptual design analysis of codes, customizations, shipping access, crane location, and other idiosyncrasies that impact the process or cost. Colorado became an emerging market for BLU Homes because of the speed of construction. In mountainous areas the construction window is quite small, where some projects might span several years because project sites are closed down for the winter due to lack of access and extreme weather. With simultaneous construction of foundations and the house in the factory (figure 12.10) and a quick set and dry-in period, the prefab approach is economically viable because of the element of time. Remote locations such as islands in the Pacific Northwest were also economically viable for the prefab approach because of a lack of competitive trades in the remote locations. Often only one contractor might exist in remote locations, making the schedule elongated or budget inflated.
150
12 Daubmann.indd 150
2018-10-18 9:59 PM
12.9 Map of BLU Homes project locations (2012).
IMPACT OF DESIGN TECHNOLOGIES 12.10 Placing the prefab module over the cast-on-site foundations.
Because time was the differentiating factor between other home builders, the role of design technology became a critical focus. Building Information Models (BIM) were essential for the efficiency of the design team, especially in the context of having to produce specific drawings quickly for state permits, different drawings for local permits, bills of materials (BOMs) for purchasing, and shop drawings for the BLU factory to construct the house. With the exception of the BOM, each deliverable was required to be hard-copy drawings given to the specific audiences. While designers long for a time when they can submit a 3D model to a review board, 2D drawings still persist. This is an instance where technology outpaces an ingrained process that limits opportunity for potential innovation. Manufacturing costs also impact the approach and tools of the design team. Under one of the directors of manufacturing, the digital tools were exploited and tied directly to the manufacturing process. Light gauge steel members were used as the infill in the structural steel frame. These light gauge members were custom-rolled in the factory, cut to length, and punched for all penetrations, based on the data provided from a CATIA design file. 151
12 Daubmann.indd 151
2018-10-18 9:59 PM
12.11 and 12.12 Metal studs versus wood infill studs.
This process produced zero waste as the light gauge members came off a roll and this process realized the positive potential of file-to-factory collaboration. This approach required skilled machine operators and a high level of quality assurance of the digital files at a time when the company volume was not producing the required manufacturing savings. A new director of manufacturing with a background in the modular industry considered the high-tech approach too expensive compared to using wood studs as infill (figures 12.11, 12.12) that could be cut by a circular saw and adjusted as needed. His job was to bring the manufacturing costs down. The skilled labor was replaced with less expensive laborers and this had design implications. While CATIA was perfect for producing BOMs and going from file to factory, it was not the right tool for the new version of the factory. In the context of having to produce various sets of 2D drawings or simple instructions for the new factory crew, CATIA was replaced with Revit. From a technology perspective, this was viewed by the design team as a backward step but was embraced for the good of the product and company. The shift highlights the circumstantial nature of design – where the
permitting process and labor accessibility dictated an appropriate response. Technology and design process innovations needed to be understood in the context of design, manufacturing, and business. While some functions were lost with the switch from CATIA to Revit (figure 12.13), new functionality also emerged as more was learned in the manufacturing process. The house would move from station to station along the factory floor where each station had the tools, materials, and building elements relevant to the stage of work. One example of the impact of design and purchasing on the process is that of a simple under-counter mounted microwave oven. At one point the house arrived at the finish station where the kitchen cabinets and appliances were installed. The microwave oven was unboxed but it was discovered that the clips to mount the microwave under the cabinets were not included with the purchase. Remember, the assembly line must keep moving and the missing clips could be ordered but would take a week to arrive. At this point the microwave was put back in the box and shipped with the house. The clips were shipped to the site and the site crew were to install the microwave. The cost of
152
12 Daubmann.indd 152
2018-10-18 9:59 PM
12.13 The BIM Revit model, with designer parts tied to the database.
labor of the site crew is five times that of the factory crew. A manufacturing post-mortem led to ways to rethink the design process so that issues like this could be controlled and automated. The design team began working with the purchasing and materials handling teams to develop a database of components and all associated parts along with installation manuals. As a designer placed an object into a digital model, the smart-part would carry links to all the associated data. As BOMs were created, all the associated parts could be ordered and any special installation notes would carry over to manufacturing. This internal development allowed the design team to work on design quickly without having to manage the complexity of the data associated with all of the building elements. The shop drawings and order forms for kitchen cabinets became formatted output as opposed to the painstaking creation of formatted drawings. As appliances vary by product model or customer choice, for example, the location of a plumbing line for an icemaker in a refrigerator is indicated on the framing and plumbing plans. The resulting process in Revit saw better drawings according to manufacturing and fewer purchasing
mistakes. The process included a technical librarian and programmer within the design team and partners from manufacturing and purchasing as well as external vendors from the cabinet shops and appliance companies. The house moved from the structural steel station to carpentry. From there it would move to the plumbing station where another incident triggered a broader discussion. A toilet was being set and the plumber drilled a hole through the subfloor and hit a floor joist. The floor framing had to be adjusted and it meant pulling a carpenter off that station and fixing the problem while the mod sat at the plumbing station. While that particular situation was fixed quickly, it did highlight another moment where the design team was able to automate and imbed intelligence into the BIM models. A design rule was articulated that floor joists would avoid plumbing drain lines. Designers would place toilets in the BIM model, at which point the floor framing would be automated in its layout and fully dimensioned, based on the reference points given by the carpentry team. The fully dimensioned floor framing shop drawing was the direct result of the location of the plumbing
153
12 Daubmann.indd 153
2018-10-18 9:59 PM
fixtures and any other floor penetrations. This approach allowed designers to focus on the layout of the home based on the clients’ input and the technical drawings would be produced in real time in the background. But that rule of avoiding framing members does not apply to everything. Another rule was created to have wall studs located adjacent to electrical fixtures, outlets, and switches. High-end clients want to feel that the things they can see have been carefully considered, including outlets, switches, light fixtures, and HVAC registers. These elements are completely contingent on the framing members that they attach to through the drywall. Once again, as the houses move down the line, framing members are set, the drywall is installed, and the house arrives at the electrical station. If anything needs to be modified based on the electrician’s observations, it is a costly change and a halt to the assembly line. This highlights the importance, specificity, and precision of the BIM model, as almost no improvisation is possible, given the structure of the system. The librarian, programmer, and designers quickly created a new set of scripts that would locate and dimension all wall and roof framing members based on the location of outlets, switches, and fixtures. Again, the designers worked on what was visible to the client and the required technical shop drawings for the carpentry station were automatically created in the background. Most residential architects would not dimension framing members, as skilled trades are able to negotiate many of these clashes on site in real time. The manufacturing constraints require forethought and precision and allow for the time and financial investment in automating these design tools specifically for BLU Homes. An additional design-related offshoot was the online “configurator” that added to a positive client experience. This has been the subject of other academic publications but comes about through many of the opportunities of designing products. A complete digital model existed for each product, with the rules for various optioning strategies such
as different kitchen cabinet configurations, finishes, and appliance packages. These digital assets were linked to a database of pricing and information. The “configurator” was developed and maintained so that the customer could walk through the house virtually in 3D and make different design choices (figure 12.14). As the customer made these choices, a cost sheet would update and give the project cost of the house. Originally, the configurator was a downloadable application available on the website with most recent versions available as an app download to mobile devices. The concept of the configurator is similar to that of the experience when buying a car where a customer can navigate the various choices. The configurator operated to educate clients before they used valuable designers’ time. Such approaches are not feasible when designing one-off custom projects but with designed products, the investment can be made to create tangential tools to support the client experience. LESSONS LEARNED If design is treated as a custom project, with a different team of consultants and contractors for each project, only limited knowledge can be transferred. In contrast, when design is taken on as a product within a strict manufacturing approach, it permits knowledge transfer every time a new house rolls off the assembly line. As soon as the first-of-a-kind product came off the assembly line (and often before), the design team would get valuable feedback from the factory team, the site team, sales people, and the clients. If improvements were identified as process or design elements, these could be folded back into the product model, often updating the BIM model as the new projects were being sent to the shop drawing phase. Feedback could also be updated subsequent to product update cycles. By embracing the interplay between design, manufacturing, and business, various approaches could be explored to address customer input. The first approach used at BLU Homes allowed
154
12 Daubmann.indd 154
2018-10-18 9:59 PM
12.14 BLU Homes’ online “configurator.”
architects to work directly with clients and the resultant custom home often deviated from the norms of the pre-designed products. The result was understandable as the designer wanted to respond to the site and solve the clients’ specific concerns. In response, a second approach tried to only allow sales consultants to work with clients. Rather than resulting in more standardized houses, this approach also resulted in custom homes to guarantee sales. The sales team was tasked with selling houses and when confronted with a situation that seemed to deviate from the pre-design products, they would also attempt to find unique solutions. A third approach was to create a team of a seasoned project architect together with a salesperson where each could temper the leanings of the other. More recent approaches after my departure from the company included a minor design fee to cover the time for any customizations and a project schedule to mitigate the client choices. Where “off-the-shelf” houses could be put on the assembly line almost immediately, houses with customizations would need to go through a more complicated permitting process outside the control of the manufacturer.
Many of the issues facing the BLU Homes initiative were large and beyond the scope of the endeavor. The topic of building permits highlighted the external pressures put on the design technology. Building codes vary by state and it is difficult to understand all the nuances unless you use that specific code frequently. The building permitting process seems also to favor the local architect and the local builder as the person working the counter in a local planning and zoning office knows the local people. The fragmentation of the residential construction industry will be a factor that will continue to limit true innovation and disruption; that is the reason there are almost no national operations attempting this scale of prefab. Since my departure from the company, BLU Homes has scaled back its national ambition and is focused on houses in California. Clients have also become accustomed to the idea of a square foot cost. The difficulty is in comparing what contractors typically include in a square foot cost – no appliances, allowances for plumbing and electrical fixtures,
155
12 Daubmann.indd 155
2018-10-18 10:00 PM
and no accounting for volume as the most efficient square foot costs – with minimal ceiling heights. Many clients want to understand the square foot cost without being able to compare a prefab to a custom approach. Often, internally this insistence on square foot costing would migrate to the analogous mindset of buying wine by quantity or a car on a number of tires basis. While producers of manufactured goods are able to charge more for increased quality, reputation, or experience, homeowners continue to rely on the square foot cost approach. Another big issue (or a potentially wicked problem) is the expectations of homeowners in the US where the notion of the manufactured home continues to have a stigma of low quality. Compare this with the norm of the automotive industry where a manufactured product with limited options has become the expectation. The cost of the home is also responsible for the often mismatched relationship between approach and expectation. Given the focus on the locations of the BLU Homes markets (see figure 12.9), the property values often dictated the architectural response. While it was possible to deliver a $100,000 house, the property values and real estate comparable housing often dictated much more space and program. One does not put a $100,000 home on a million-dollar piece of property. While working with a client on a million-dollar house, the expectations also increase in both features and level of service. Quickly, the client’s mindset migrates to the expectation of a custom home with unique design options. While BLU Homes did not achieve mass customization for the masses in housing, the strong relationship between design, business, and manufacturing existed and points to more potential to co-leverage these discrete bodies of knowledge. Additionally, successes or failures of any particular model or approach point to the dependencies of broader social and cultural expectations and constraints. As technologies advance and broader social norms shift, the possibility of masscustomized, single-family housing may become a reality.
NOTES 1 Brown, Tim, and Katz, Barry, Change by Design: How Design Thinking Transforms Organizations and Inspires Innovation, New York: Harper Business, 2009. A key concept of design thinking is the desire to place design in the hands of non-designers and move design techniques upstream. The notion of upstream is a key aspect when thinking about the manufacturing process. 2 Kieran, Stephen, and Timberlake, James, Refabricating Architecture: How Manufacturing Methodologies Are Poised to Transform Building Construction, New York: McGraw-Hill, 2004. Chapter 5 focuses on architectural applications of these manufacturing techniques. 3 Garrison, James, and Tweedie, Aaron, Modular Architecture Manual, Newark, NJ: Kullman Building Corporation, 2008, p. 10, see figure 1.1.1. 4 Ibid. See chart 1.7.1 for prefab versus on-site construction schedule. 5 For a precursor to a folding building, see patent no. US3460297 A by William Fritz. 6 For clear descriptions of the unfolding process, see patent no. US20140059947 A1 or search YouTube for videos showing a Blu Home being set and unfolding.
156
12 Daubmann.indd 156
2018-10-18 10:00 PM
13 THE MODERN MODULAR: THE MASS CUSTOMIZATION OF THE SINGLE FAMILY HOME
JOSEPH TANNEY 157
13 Tanney.indd 157
2018-10-18 10:03 PM
13.1 Modules of Use.
13.2 Typology matrix.
158
13 Tanney.indd 158
2018-10-18 10:03 PM
Although architects have historically been preoccupied with the single-family home, most people do not live in a space designed by an architect. As a small architectural practice in New York City, Resolution: 4 Architecture’s (RES4) work began with a series of kitchen and bathroom interventions and evolved into apartment, loft, and townhouse renovations. In completing numerous urban interior projects, RES4 has developed many highly efficient, cost-effective, and idea-driven spaces. Combining these strategies for efficient domestic planning with our interest in off-theshelf products and a preoccupation with factorybased construction processes, we have developed a strategy for the “mass customization” of the single-family home.
The Modern Modular is a systematic design methodology that attempts to leverage existing methods of prefabrication to produce custom modern homes, specific to each client and site. It is based on conceptual building blocks we call Modules of Use (figure 13.1), spatial configurations that organize the essential components of contemporary domestic life within the dimensional limits of the modular industry. Based on our explorations, we developed a series of freestanding domestic typologies that embody this essence of utility (figure 13.2). Our built projects are a series of experiments testing and refining this concept (see figures 13.6–13.19). More than a product to be purchased, the Modern Modular is a method of design – a theory really – with its roots reaching beyond the history of just our office.
159
13 Tanney.indd 159
2018-10-18 10:03 PM
OBSERVATIONS In the late 1990s, the US experienced a rise in its collective design consciousness, particularly in modern design. In 2000, Dwell was published, becoming the first magazine completely devoted to modern domestic space. It shone a bright light on the lifestyles of “normal” people living in modern, designed spaces. Simultaneously, stores such as Design Within Reach, Target, and IKEA were making a range of modern-designed goods more available and more affordable than ever before. A decade later, the pervasiveness of the Internet allowed for an expanding global awareness of modernism, which gave the perception that modern domestic space and individual choice were more accessible. In reality, while we could buy modern goods to fill our homes, the proverbial domestic space has remained stagnant while becoming more expensive. Most domestic structures are conceived by developers as products built for profit, belonging more to a fast food society than a culture of cultivation. In many ways, the efficiency of mass production has contributed to the self-same tombstones in the graveyards of complacency otherwise known as the American suburbs. The oversized and inefficient nature of most suburban homes is nevertheless a reality. Americans are living in the past, relying on an invented nostalgia of the American Dream, as they occupy the country’s bloated stock of single-family houses that represent a preconceived image of what “home” should look like. Although our lives are being “modernized” through increased access to better technology, consumer products, and furniture, the actual spaces in which we live have been slow to evolve – both in their design and their execution. The statistic that 96 percent of domestic spaces in this country has been built without direct communication between the end user and the designer represents a chasm. This void provides an opportunity for architects to rethink how we participate in creating the physical fabric in which we live. After all, designer consumer goods are mass-produced and relatively low cost, but architect-designed domestic spaces
are still individually produced and expensive. So, our speculation was that if we, as architects, could design relatively affordable domestic spaces by leveraging mass production, we might be able not only to tap into the country’s broadening design consciousness – a market we felt was hungry for custom modern homes – but also to effect lasting change in the American residential landscape. THE HOLY GRAIL OF MODERNISM Of course, we weren’t the first architects to explore mass production as a solution to making a well-designed, higher-quality, single-family home more accessible. Over the past century, many architects before us have pursued this “Holy Grail” of modernism. Le Corbusier, Walter Gropius, Buckminster Fuller, and Frank Lloyd Wright were just a few who attempted to design a relatively affordable modern home that could be mass-produced. Several commercial institutions, such as Lustron and Sears Roebuck and Company, also participated in the pursuit. Both the academic and commercial paths ended with varying degrees of success. Numerous academic attempts were generally conceived as mass-produced products with a high level of design, but architects often burdened themselves by inventing complicated manufacturing processes. Conversely, commercial attempts were more successful in feasible, efficient manufacturing, but they were limited in design flexibility – seemingly strangled by the process of production. Neither achieved a means to cost effectively allow for a highquality, customizable product. PREFABRICATION METHODS | DIMENSIONAL CONSTRAINTS With this historical context, our research focused not on reinventing the manufacturing process, but on how we could work within the existing commercial methods of residential prefabrication. Employed by successful factories everyday, off-site construction processes fall into three basic tiers, or delivery methods. At the low end, in both cost and quality, the manufactured
160
13 Tanney.indd 160
2018-10-18 10:03 PM
13.3 Components of Domesticity.
home falls under the Department of Housing and Urban Development (HUD) guidelines and offers very little design flexibility. At the high end, the panelized or kit-of-parts delivery method offers the greatest design flexibility, yet often costs as much as stick building on-site. The modular industry falls in the middle. Born out of the manufactured home industry, modular construction adheres to all local codes and offers opportunities for design flexibility – if one designs within the industry limits. The majority of residential modular manufacturers in the United States use a standard module width of 12, 14, or 16 ft, based on Department of Transportation highway shipping regulations. Although module lengths vary by factory, the most common length is 60 ft, which is determined by the physical limitations of the factory. Within these dimensional parameters, we found a kindred spirit. Having completed many renovations of long linear loft spaces in New York City, thinking inside the box was a natural extension of our practice. CONCEPTUAL BUILDING BLOCKS | MODULES OF USE We began to translate our experience working within the strict constraints of the urban environment to work within the existing dimensional constraints of the modular industry. By reorganizing the essential elements of domestic space, such as furniture, fixtures, cabinets, and appliances, based on functional dimensional requirements and efficient circulation, we developed compositional strategies to arrange the Components of Domesticity within the maximum modular width of 16 ft – the Modern Modular’s optimum dimension for domesticity (figure 13.3). The Components are arranged in contiguous relations relative to use – communal or private – creating linear 16-ft-wide volumes that we termed Modules of Use (see figure 13.1). These Modules become our conceptual, not physical, building blocks, for, when combined, they can form a range of customizable typologies (see figure 13.2).
161
13 Tanney.indd 161
2018-10-18 10:03 PM
13.4 Massing variations.
TYPOLOGIES After a series of investigations, plan configurations began to evolve into seven basic forms, or types. The most basic and efficient is the single-bar, single-wide typology, with all Components of Domesticity contained within one 16 x 60 ft module. From there, as the client’s program expands, modules can be stacked for multiple stories, or they can be aligned in plan to form double- or triplewide typologies that allow for larger interior spaces. Further, modules can be arranged perpendicularly to form T, L, Z or courtyard typologies, which define exterior living spaces. These seven basic types can then be manipulated in an endless number of ways by a simple series of design moves, such as pushing, pulling, and slipping boxes (figures 13.4, 13.5). For instance, slipping two stacked bars can open areas for porches or parking space below the overhang of the top box, and the exposed roof surface of the lower level can be used for outdoor living space. Or, pulling apart lowerlevel modules and bridging the second-story modules over the top creates a framed view or entry (see figures 13.13, 13.19). These simple moves permit each home to specifically address the unique site, solar orientation, views, program, and other project-specific factors.
162
13 Tanney.indd 162
2018-10-18 10:03 PM
13.5 Plan variations.
163
13 Tanney.indd 163
2018-10-18 10:03 PM
13.6a–c The Dwell Home (L-series, two-bar bridge), Pittsboro, NC (2004).
THE DWELL HOME We posted our research as a series of typology diagrams and specific case studies on our website in 2002 – not intended as models to be purchased, but as examples of what could be created with our system, as a portfolio of capabilities. The intention was to demonstrate a process of design operating within the modular industry limits, in order to achieve a greater predictability of construction time and costs and a higher level of custom design. Potential clients quickly found our website and began inquiring. Academic institutions and publications also expressed an interest. In 2003, Dwell invited us to submit an entry to the 2003 Dwell Home International Competition for a modern prefab home. We submitted a proposal specific to the site, the client, and the $200,000 budget (figures 13.6a–13.6c). It was the first, and to date the only, international design competition for a home with actual clients on an actual site, with an actual budget – that was actually built. In addition to our proposal, we included our concept of the Modern Modular to show how the strategy can be used to create unlimited variations in response to specific site conditions and requirements. It was important to us that it was understood: our site-specific response, the Dwell Home, was not intended to be a mass-produced product as the competition suggested. It was an example – our first test case – of how an integrated design and manufacturing strategy could be employed to create custom, modern, modular homes – in theory, using any modular factory in the country. Since then, we have designed and built dozens of site- and clientspecific homes from Maine to Hawaii, using modular, panelized, and hybrid delivery methods. Learning more with each project, we have developed a systematic process for how to approach each step of the project, from design through completion. PROCESS To understand our process, it is critical to understand the typical factory processes. Of the 200 residential modular factories in the country, the majority are located in the North-east, predominantly in
164
13 Tanney.indd 164
2018-10-18 10:03 PM
13.7 Factory production, suburban villa (two-story, double-wide), Bethesda, MD (2013).
Pennsylvania. Farther south and west, the factories decrease in both quantity and quality due to lower labor rates and more amiable weather. Most residential modular manufacturers use exclusively wood structural framing and assembly line production – the modules move from station to station while the assemblers remain in place (figure 13.7). Typically, each assembly line has about 15–18 stations, commonly called line spaces. To start, walls, floors, and roof panels are assembled on the flat surface, then are picked up by hoists and moved into place to create a three-dimensional volume. From there, electrical, plumbing, mechanical, sheetrock, and finishes are installed at their respective stations. Typically, each module makes two line movements a day, so in seven to ten days, boxes are going out the factory door and into the yard, where they are prepared for shipping. Surprisingly, it is all very low-tech. There are no robots. Skilled labor is the minority, as most workers on the factory floor are assemblers, not craftsmen. Although precision is critical, they use various off-the-shelf pieces and literally just put them into place. Given how fast everything moves, the factory needs to know exactly what they are building – every detail – and have all materials at the ready before any modules start fabrication. This means all design and construction decisions need to be made up front, as there is no time for changes once the modules hit the assembly line. This is not how most architects and contractors work, so we had to develop a new way to approach design, documentation, and construction. Our typical process is broken into four phases, each four months long.
Phase I | Design | Documentation The first phase is design. Throughout this phase, we spend a lot of time with the clients to make all the decisions. We begin with zoning and site studies to understand our constraints. We develop a series of plan sketches and massing models, based on programmatic requirements. As the scheme develops, the 3D model becomes our greatest tool; we model every single element as we select them with the clients, down to the towel bars. This allows us to walk
the clients through each space, so they understand everything they are getting. Later, this model is immensely helpful as a communication tool with the factory and the contractor as well. By the end of the first phase, we have a preliminary set of drawings for the factory to get the engineers on board.
Phase II | Engineering | Coordination | State Approvals | Bid In the second phase, we focus on documentation and factory coordination for engineering and state approvals. Based on our drawings, the factory produces a set of detailed shop drawings that show how every piece comes together. Simultaneously, we work on drawing sets for local approvals and for bidding to local general contractors (GCs), who are responsible for purchasing the modules and for all the site work. As we work through design and detailing we are always considering shipping limitations and the factory’s capabilities, leveraging them for what they do best while respecting their limitations as assemblers. This means that some elements that are too complex, or cannot be shipped, may not be “factory-friendly,” in which case, we plan for the contractor to complete this work on site. Similar to how we have developed a series of standards for interior layouts, we have developed a series of typical details for trim and millwork done in the factory, along with site-built elements such as trellises, railings, and outdoor showers. These standards not only help streamline the process for the factory and GCs, but also help make our design and documentation process more efficient. By the end
165
13 Tanney.indd 165
2018-10-18 10:03 PM
13.8 Module set, High Peak Meadow House (L-series, two-bar bridge), Maplecrest, NY (2009).
of this phase, we have all our approvals in hand, a contractor signed on, and the factory ready to go. So, before any construction begins, we spend at least eight months working through design, documentation, and coordination.
on a steel frame on concrete grade beams and piles. All of this is considered throughout design of the modules and coordinated with the engineer, so by this time the contractor is focused on execution.
Phase IV | On-Site Completion The fourth phase starts once the modules are set (figure 13.8). All the modules are often shipped out to the site in a day, and the next day a certified set crew and crane are ready. Typically, all modules are set in a single day, usually from sunrise to sundown. From this point, the GC is responsible for all the on-site work. Although approximately 80 percent of the house is finished in the factory, there is still quite a bit to do on site. Mechanical, electrical, and plumbing hook-ups and connections need to be made. To achieve a seamless appearance across multiple modules, exterior siding and flooring are installed on site, and any landscape or site elements, like a pool and decks, are done after the set. PROJECTS
Phase III | Fabrication | Site Preparation With the third phase, factory fabrication takes place. About three of these four months are spent just procuring materials, because they need every material on hand to move as quickly as they do through production. Our houses typically take two to four weeks to fabricate, depending on size and complexity. It is critical for the factory’s operations that the assembly line keeps moving at a consistent pace, but our projects typically have a lot more detail and finishes than the factory’s norm. To accommodate this, after the typical two weeks on the assembly line, our houses are often moved out to the yard, where the installers have more time to do finish work, like custom cabinets, tiles, and countertops, without slowing down production. Simultaneously, the general contractor (GC) preps the site with the foundation, utilities, and site work. The foundations often vary by site and program. At a minimum, a crawl space is needed below the modules to allow room to make all the structural and plumbing connections. When working with a flat site with good soil conditions, a full basement is often the highest value proposition. If we are on a steeply sloped site, we may expose part of the concrete foundation to accommodate a walk-out basement. If there is a lot of rock, we may build up a concrete plinth to sit the modules on. When in the floodplain, the modules likely sit
Over time, we have worked to push the limitations of what the factory can offer. Some projects grow in terms of scale and complexity, with our largest project in recent years totaling 14 modules. In other ways, we have brought new materials and finishes to the factory. For instance, when we started, we used only stock cabinetry, and now almost all our projects have custom millwork installed in the factory. Additionally, all our homes are designed to meet or exceed LEED1 standards. We start with a tight building envelope with the flash-and-batt insulation system we developed with the factory – 2 inches of spray foam insulation followed by 4 inches of R-15 batt insulation2 within our 2 x 6 exterior walls.3 We also typically incorporate low-flow plumbing fixtures, LED lighting, and highefficiency boilers. Increasingly, when budgets allow, we specify radiant floor heating and geothermal systems, and if we are not using roof surfaces for outdoor living spaces, we often use them for solar panels and green roofs. As we have broadened the possibilities available within the existing modular industry, the Modern Modular’s capacity to flex has grown, in order to serve the clients’ needs and budget and to address each unique site. The following five projects exemplify the Modern Modular capacity to grow from very simple to more complex.
166
13 Tanney.indd 166
2018-10-18 10:03 PM
13.9 Exterior view of the Connecticut Pool House (single bar), Sharon, CT (2011).
13.10 Axonometric view of the Connecticut Pool House.
Connecticut Pool House Sharon, CT (2011) | Single-Bar Typology | One Module | 832 sq ft The Connecticut Pool House represents the most basic typology of the Modern Modular system, as a single bar on a conventional concrete foundation (figures 13.9, 13.10). Designed as an oasis for an urban couple to share with their two adult children, this simple pavilion is ultimately a freestanding loft, its simplicity driven by the client’s budget. The 16’ x 52’ module contains a sleeping zone at one end, and a kitchen/ dining/living zone at the other. Without the use of doors to provide separation, a black steel-clad volume containing the bathroom, laundry, storage, and fireplace demarcates private and public zones.
Although this “floating” core does not touch a perimeter wall, it is filled with natural light from an oversized skylight and frosted glass pocket doors. In the warmer months, the pool provides evaporative cooling. Breezes move air over the pool, which is cooled as it comes through sliding doors all along the south side, into the living space, and out the operable clerestory windows running along the north wall. Meanwhile, a cedar brise-soleil, built on site, shades the southern exposure. Although primarily designed for use in the summer, this pool house is sustainably equipped for all seasons. The roof is designed to accommodate solar hot water panels, to heat not only the pool, but also the house itself via radiant heat tubing beneath bamboo flooring. The heat-generating fireplace also provides supplemental heat in this small space. 167
13 Tanney.indd 167
2018-10-18 10:03 PM
13.11 Exterior view of the Bronx Box.
13.12 Axonometric view of the Bronx Box (two-story, singlebar), Bronx, NY (2008).
Bronx Box The Bronx, NY (2008) | Two-Story, Single-Bar Typology | Two Modules + Panelized Saddlebag and Stair Bulkhead | 1815 sq ft The Bronx Box shares a similar footprint to the Connecticut Pool House, but here zoning regulations rather than budget determined the size of the home. It grows vertically with a second-story, private-use module stacked on top of a communal-use module; both are 16’ x 54’. A “saddlebag” of 2-ft-wide accessory modules is added along the side for storage, housing built-in cabinets to run the length of the house (figure 13.11 and 13.12). The urban infill site is located at the foot of the Throgs Neck Bridge on Eastchester Bay in the Bronx, and was an exciting welcome-home present for the client, an Iraq War veteran. Before her deployment, it was deemed more cost-effective to replace the existing neglected bungalow than to remodel. Given the narrow lot, the prefabricated design was able to celebrate the constraints of its zoning envelope. Setbacks, height limitations, and floodplain requirements yield a compact home that allows for off-street parking, a small patch of grass, and an expansive roof deck with stunning views of the bay and bridge. Unlike the pool house, there are neighboring houses along either side, so the interior space is configured more like an urban loft, opening up to the exterior on the ends. The open living, dining, and kitchen area on the first level flows directly onto an elevated deck. Site-built exterior stairs run the full width of the house to serve as an extension of the communal space and lead down to a pier that
juts into the bay. The second-floor master suite features a fireplace, balcony, and skylight that lets natural light into the bathroom. The metal-clad roof stair bulkhead was carefully sculpted within zoning limits to maintain 360-degree views of the water and surrounding neighborhood. The materials respond to the surrounding built fabric, which has helped the house to be accepted as another of the many unique personalities in the eclectic neighborhood. The Bronx Box demonstrates the efficiency and adaptability of the Modern Modular system within a limited and narrow urban lot.
168
13 Tanney.indd 168
2018-10-18 10:03 PM
13.13 Exterior view of the House on Sunset Ridge (L-series, two-bar bridge), Norfolk, CT (2008).
13.14 Axonometric view of the House on Sunset Ridge.
House on Sunset Ridge Norfolk, CT (2008) | Two-Bar, Bridge L-Series Typology | Six Modules | 3035 sq ft Reached via a gently rising spiral drive, this LEEDcertified, modern prefab home sits perched on a small hill just outside a town filled with traditional New England architecture (figure 13.13). Designed as a country house for a Brooklyn family of four, the house is positioned on an axis with a local historical landmark, the fire lookout tower one and a half miles due north on Haystack Mountain. The lower-level modules, forming the “L,” contain the public spaces. The kitchen occupies the overlap of the intersecting bars, with the living and dining to one side and the screened porch to the
other. Private spaces, like bedrooms and a study, are located in the perpendicular upper bar. The guest suite is one exception. By pulling apart the lower-level modules, it is set apart on the first floor with its own framed meadow views and private outdoor space defined by the overlapping module above (figure 13.14). The large screened-in porch off the kitchen becomes the preferred living and dining space during warmer months. The main living area, oriented to align with the distant tower, features a fireplace that is clad in recycled cement 169
13 Tanney.indd 169
2018-10-18 10:03 PM
13.15 Exterior view of the Dune Road Beach House (two-story, doublewide), East Quogue, NY (2012).
13.16 Axonometric view of the Dune Road Beach House.
170
13 Tanney.indd 170
2018-10-18 10:03 PM
13.17 Interior view of the Dune Road Beach House.
board panel. The living space is wrapped in glass, maintaining a strong connection to the outdoors even when weather prohibits outdoor living. Above the living space on the second floor, an exterior deck is located adjacent to the study. This outdoor area features a green roof and fireplace, providing year-round living space with stunning views of the surrounding landscape. Like the traditional homes that surround it, this modern home embodies tradition in the making. Embracing an esthetic of efficiency reflective of today’s domesticity, it can be seen from the lookout tower, becoming its own landmark in the historical landscape.
Dune Road Beach House East Quogue, NY (2012) | Two-Story, Double-Wide Typology | Four Modules | 2200 sq ft Located on famous Dune Road, the lot is one of the last remaining of its kind. The site is on a long thin barrier island, connected to the mainland by bridges, which offers ocean views immediately to the south, and bay and marshland views to the north. The house replaced a dilapidated beach bungalow that was falling apart due to crashing waves, which allowed it, by zoning, to be placed past the typical allowable building line. Riding the crest of a dune, the house sits at the end of a long drive, reached by a long ramp that bridges over the duned
landscape (figure 13.15). The foundation is a series of 45-ft-long wooden piers driven deep into the sand, upon which the two lower floor boxes are set. Zoning and square footage limitations determined not only the home’s placement, but also its length, width, and height. The eastern end of Long Island has a rich history of architect-designed beach houses. Today, it is not uncommon for homes in this area to reach 10,000 or even 20,000 sq ft; the neighbor’s house, for instance, features a tennis court and a helicopter pad. In contrast, this compact house at 2200 sq ft is comprised of just four modules – two 54-ftlong boxes stacked on two 54-ft-long boxes (figure 13.16). Just like the long, narrow site, the home’s interior is organized as a linear composition, with service spaces, including bathrooms, mechanical rooms, the kitchen and vertical circulation to the north, and open living and sleeping areas to the south. The client is a young family who live in a large apartment in downtown Manhattan. Unlike most second homes owned by city dwellers, this house is actually smaller than their apartment. Vacations here are more akin to sleeping on the beach than escaping to a rural fortress. The floor-to-ceiling glass wall in the Dune Road Beach House’s communal space gives the effect of sitting on the beach while retaining all the comforts of home (figure 13.17). The second level has four bedrooms, including a kids’ bunkroom, and interior stairs lead to a roof deck with a fireplace, hot tub, a space for morning yoga, and commanding views up and down the beach, high above the surrounding houses.
Fishers Island House Fishers Island, NY (2012) | Two-Story, DoubleWide/Triple-Wide, Hybrid Typology | Eight Modules | 4470 sq ft Located off the coast of Connecticut and Rhode Island, yet occupying a body of water in New York State, Fishers Island is accessible only by ferry. While during the summer the island’s population swells to over 3000, year-round residents total only about 300. Delivery of supplies and materials is therefore limited. Contractors live on the mainland, creating short workdays due to the commute. 171
13 Tanney.indd 171
2018-10-18 10:03 PM
13.18 Axonometric view of the Fishers Island House (twostory, double-wide/triple-wide hybrid), Fishers Island, NY (2012).
172
13 Tanney.indd 172
2018-10-18 10:03 PM
13.19 Exterior view of the Fishers Island House.
Thereby, construction of a new home takes longer and costs more. Leveraging off-site construction with prefabricated modules – complete with plumbing, electrical, and finishes – makes for a much more cost-effective construction method. The eight boxes were designed not only to suit the needs of the client and site, but also to fit on the standard ferry. The ferry itself can accommodate 16-ft-wide boxes, but the gangplank was only 14 ft wide. So, the house is composed of 12- and 14-ft-wide boxes. Similarly, we used a prefabricated, panelized concrete foundation to minimize the amount of ferry trips required to shuttle materials and equipment. Negotiating the limited access and strict zoning regulations on the island, along with the large scale and hybrid typology configuration, made this our most complex project yet (figure 13.18). The client, a family of four including two recent college graduates, has extended family who also live on the island. This house is designed to accommodate and entertain them all for generations to come. Often
called “the ark” by locals, the modules are arranged to create four levels of living (figure 13.19). Integrated into the sloping landscape, the long and linear vessel is organized by public and private uses. The main level contains communal living spaces that span three parallel boxes, with private spaces for the immediate family on the second floor. The east wing contains guest bedrooms and a stair down to a large bunkroom below for future grandkids. Using all the horizontal surfaces, the upper roof houses solar panels and solar hot water pipes used for radiant floor heating during colder months. The lower roof surface becomes the main gathering area, essentially another level of living space. Partially covered, the roof deck is an outdoor room, with green areas, a kitchen, seating and an outdoor fireplace. An evening of entertaining starts with drinks on the roof deck, then continues downstairs for dinner on the main level. An exterior stair from the rear parking court makes it easy for guests to join the party upon arrival. 173
13 Tanney.indd 173
2018-10-18 10:03 PM
OPPORTUNITIES AND CHALLENGES
NOTES
Since winning the competition and building the Dwell Home, public interest in bespoke modern homes fabricated using mass-production techniques, as a means to make them more accessible, has increased. We have designed and built dozens of prefab homes throughout the United States using the Modern Modular methodology, responding to “economies of one.” With each, we attempt to make small advances in the details, products, and technologies we employ. Of course, energy efficiency and sustainability are central to every project; ultimately, we aspire for each house to produce more energy than it consumes. As these supplemental energy technologies become more feasible at the scale of the single-family home, we have been consistently impressed with our clients’ enthusiasm for creating efficient and sustainable living environments. Although historically slow to change, the modular industry has made significant improvements in the quality of their built product over the last 10–20 years. Yet, as interest from potential modern homeowners continues to build, we have sometimes found the modular industry struggling to keep up with the higher expectations from the new clientele. Ultimately, the Modern Modular is a design and execution methodology focused on the efficiency of use, implementation, and performance. It attempts to leverage existing methods of residential prefabrication in creating custom, sitespecific, modern homes. By becoming intimately involved not only with the design, but also with the production of homes, architects can redefine their professional role in creating domestic space, thus fostering an opportunity to improve the quality of our most precious of all built environments, the home.
1 Leadership in Energy and Environmental Design (LEED) is a rating system developed by the United States Green Building Council (USGBC) to evaluate the environmental performance of a building. 2 R-value is a measure of resistance to heat flow through a given thickness of material; the higher the R-value, the greater the resistance. 3 2 x 6 construction refers to the framing material (1.5” x 5.5” studs, normally referred to as 2 x 6, hence the name) used to build the wooden frame of a home.
174
13 Tanney.indd 174
2018-10-18 10:03 PM
14 MASS PREFABRICATION: INVESTIGATING THE RELATIONSHIP BETWEEN PREFABRICATION AND MASS CUSTOMIZATION IN ARCHITECTURE RYAN E. SMITH 175
14 Smith.indd 175
2018-11-04 3:00 PM
14.1 Peculiarities (challenges) of architectural production: site, program and labor are theoretically overcome with prefabrication.
This chapter investigates the relationship between prefabrication – off-site, modular, and industrialized building – and the subject of this book, “mass customization” in order to realize greater housing design democratization. The terms “prefabrication” and “mass production” or “mass customization” are often conflated in architecture. This ignores the complexity and reality of the expansive proliferation of off-site factory-fabricated components and the processes involved in delivering these products and sub-assemblies – standard, custom, or inbetween. There is also a tendency to isolate mass customization and prefabrication to a formalist parametric design ideology at the expense of the business and production platform strategies that are foundational to manufacturing knowledge in other production industries. The default formalist perspective of prefabrication seemingly follows the dictum: standardization is bad, customization is good. While variation in the human experience of space and environments, living quarters, and working offices, is certainly desirable, the method of producing said environments is not necessarily variable in principle. That is, the level of standardization or customization in the modes of making architecture need not be defined in terms of polarity, but rather moves along a sliding scale of opportunity. When factory production or job site labor is used, whether it is automated or analog, or if the product is more or less customized, should not be as much of a concern as the appropriateness of the production to the question of scope: the architecture and the user program. In one instance, the means of producing architecture is just that, a means and not an end, but ignoring the modes of production risks missing an opportunity to leverage manufacturing principles in order to realize architecture that is not only more enriching, responsive, and experiential, but also more equitable, affordable, and accessible – in short, more democratic.
PREFAB ARCHITECTURE The premise herein is that architecture is part of the building production industry (an integral part, a necessary contributor) and is therefore not discretely valuable on its own without building production. As such, architecture is inherently about producing: designing and making buildings and cities to serve those who inhabit and communicate within them. Architecture is therefore both the process and product; it is production.1 Architecture is peculiar, however, when compared with other production industries (i.e. IT, automobile, film) in three unique ways (figure 14.1): (1) the location of production, the site, is unique every production cycle and the product is ideally specific to this location – climatically, materially, historically, culturally, aesthetically, etc.; (2) the functional program and resulting production scope are unique at each iteration – architecture is fit for the function of its use; and (3) the labor is mutable and inconsistent from product to product.2 Prefabrication conceptually solves these peculiarities. While the site is unique for each housing project, taking the majority of operations out of the job site alleviates schedule overruns and quality challenges associated with weather delays. The sun always shines in the factory. Although the functional program or scope of each project is different, standardized or mass customized product platforms can produce elements (panels, modules, etc.) that can service many different housing configurations and forms. The labor for construction jobsites is mutable, but prefabrication allows for a higher skilled trained workforce to consistently produce quality products. In this way, prefabrication leverages the benefits of manufacturing economies of scale while realizing economies of scope differentiation by virtue of varying levels of mass customization.3 The question of whether standard or custom in prefabrication is not new. The factory-made
176
14 Smith.indd 176
2018-11-04 3:00 PM
14.2 Productivity increases with the fewer number of client inputs.
house has eluded architects for over 150 years. In the past, Le Corbusier, Gropius, Lloyd Wright, and Safdie tried their hand at prefabrication, and more recently, Marmol Radziner, Michelle Kaufmann, and Blu Homes have had fits and starts. These struggles are evidence of the difficulty in establishing and maintaining a viable architect-led prefab operation. The theory is seemingly sound: design a mass-customized system, take construction operations out of the job site, relocate them in a factory environment to control material flow, labor, and risk, and deliver early at a predictable cost. Despite this formula, presently, there are virtually no commercially viable architect-led systems on the market.4 What are architects missing? Architect-led prefab has approached prefabrication as a service business and therefore made it supply-driven. Prefab architecture, before it is architecture, is manufacturing, a product business, which suggests that it is demand-driven. On the sliding scale of opportunity, mass production to mass customization is, first, a business theory and, second, a production theory. Architecture historically is first an artistic theory, then a production theory, and finally a business theory. Architecture as a discipline would do well to focus on the production and business in order to realize greater impact on design democratization. Business theory states that there are two types of commercial operations: services and products. Service industries have a high degree of client interaction, relying heavily on customer input at various intervals. Production theory
says that productivity increases with the fewer number of client inputs (figure 14.2). The client decoupling point is the juncture at which customer inputs are not considered in the manufacture of the product. As appropriate, architects should work with fabricators and builders to ensure that prefabrication is leveraged to reduce unnecessary client feedback and maximize necessary client feedback. This represents a sliding scale of opportunity as opposed to hard-and-fast definitions of production modality and client feedback.5 As interactions and inputs increase, inefficiencies in outputs do as well. Product industries, on the other hand, have less direct customer interaction and are more focused on reducing time invested for output generated. Architecture, the design and building of environments, is conceptually both a service and product industry, involving service processes and nonservice activities. Prefabrication is essentially product-oriented. However, architecture and architect-led prefab are most often service-focused. Architects associate their value in terms of meeting client needs and generating design ideas, while manufacturers typically value moving the service process toward non-service-oriented practices of production, realizing productivity in the process – increased output per unit of input.6 Supply-driven enterprise links capability with application, real or perceived. The importance of the capability of the service supplier is primary; what those skills service is secondary. Supply-driven products run the potential risk that the problem does not actually exist, that there is no real need 177
14 Smith.indd 177
2018-11-04 3:00 PM
14.3 The post-recession multifamily housing boom is increasing demand in the prefabrication sector (Sources: Freddie Mac, U.S. Census Bureau, Moody’s Analytics).
for the service provider’s knowledge and innovation. Service industries, such as architecture and many architect-led prefab companies, are supply-driven and service-minded. Supply risk is managed through inventory control and careful orchestration of process and players. Conversely, demand-driven enterprise connects a market problem with a solution. Identifying the market need drives business. Product development is fueled by finding a solution to that need. The risk is shouldered by the product provider to actually find an appropriate solution to the problem that the market will continue to “pull,” and that the business can accurately assess this demand. Demand risk is managed by good market analysis and awareness of the greater economic and contextual environment. As a service industry and applying supplydriven capability, architect-led prefab has taken their design acumen and technical prowess in meeting clients’ needs and created an innovative prototypical design to deliver on the problem of democratic housing. The risk for architects in this way, as a viable business model, is that access to mid-market affordable housing may not actually exist.7 But today that risk is relatively low. The construction sector capacity is challenged to meet the current post-recession housing demand (figure 14.3). Across the industry, from designer and builders to commissioning agents and facility managers, the construction workforce is lacking in numbers and skills.8 Much of the talent that left during the recession found work in other sectors, leaving a gaping hole in the post-recession demand cycle.9 While prefabrication architecture might
have been perceived as risky prior to the recession, today it is the means by which developers can realize greater growth and general contractors can deliver higher quantities of work without increasing their own staff significantly. Outsourcing 60–70 percent of the scope of work on new multi-family builds allows construction projects to be realized at a competitive cost at faster speeds. It is not that prefabrication is able to provide a low-end disruption, rather, all the elements of construction have increased in cost, including materials and labor because of availability and therefore off-site construction is proving competitive. Additionally, as baby boomers and millennials move back to the city, the densification of metropolitan regions is causing land prices to inflate. As land becomes more expensive, the cost of the built work (sticks and bricks or panels and modules) on the site accounts for far less than the total real estate value of the investment. Demand for schedule and predictability over first costs is where prefabrication can deliver and incrementally innovate.10 Therefore, demand for prefabrication architecture is not the problem. Architects are the problem, insofar as they continue to operate from a service mind-set. There are two terms that describe this situation: autonomy and agency.11 On the surface, the terms appear to be synonymous. However, in the context of prefabrication architecture, they are discrete ideas. Autonomy in architecture is related to authorship, the freedom from external control (i.e. independence). Leveraging digital tools in parametric design, architects have taken mass customization as a design strategy to realize greater individual artistic expression. Agency is the state of acting
178
14 Smith.indd 178
2018-11-04 3:00 PM
14.4 Progression of prefabrication theories since the Industrial Revolution from standardization to mass customization.
and exerting influence, choice, freedom, and power. While autonomy and authorship are one form of agency, they have arguably limited the influence of architects in democratic housing. Embracing the principles of manufacturing – standardization to mass customization – has the potential to empower architects to be able to respond to the current demand and have greater impact on delivering democratic housing design for the masses. STANDARDIZATION TO MASS CUSTOMIZATION The concepts in manufacturing from industrialization in the mid-1800s to today’s mass customization by virtue of CAD/CAM technologies are not exclusive but represent a chronological evolution of production theory (figure 14.4).12 Industrialization, as related to the Industrial Revolution of 1848, marked a change in economic and societal thinking, by virtue of advanced machinery, that is still pervasive today. Standardization is a result of the industrialized society when products became codified. This was most prevalent in developing standards related to military production. Mechanization is an attempt to move standardization to greater economies of scale, by introducing additional mechanized processes that were developed during the war years but furthered by virtue of more advanced mechanical machinery, thus reducing human labor. Mass production thrives on economies of scale – this concept is to produce as much of the same thing in order to bring down the cost of a single item. It has grown concurrently with consumer demand. Automation occurred as a result of digitally informed manufacturing machinery via computer numerical control and CAD/CAM software. Finally, mass customization brings together mass production and automation to deliver an economy of
means. Mass customization maximizes the benefits of mechanization and automation production methods, reducing labor costs, but works to preserve the benefits of variability and in the output. Fordist mass production relies on the economies of scale – as repetition increases, the cost per unit decreases. Likewise, as variation increases, the cost per unit grows exponentially. Conversely, mass customization suggests that variability is possible within an acceptable margin of cost increase.13 Prefab architecture may be standardized or custom. However, these terms do not capture the complexity of the manufacturing and fabrication industry. Fabrication techniques vary with each project. The chief concerns in prefabrication for the fabricator are cost, lead times, and flexibility surrounding custom products. Four terms have emerged in the manufacturing industry to describe the levels of prefabrication completion and associated effort that will be expended in manufacturing. These terms and definitions help architects understand the scope of the project that is being discussed and developed. These are: (1) Made-to-Stock (MTS): (2) Assembled-toStock (ATS); (3) Made-to-Order (MTO); and (4) Engineered-to-Order (ETO).14
•
Made-to-Stock (MTS): MTS products are best handled through inventory replenishment strategies. In order to keep inventory replenished, manufacturers have used standardizing, or reducing complexity and increasing repetition. Supplier-managed inventory has proven successful for some companies and projects, where suppliers take on the job of determining requirements, and maintaining and distributing materials. 179
14 Smith.indd 179
2018-11-04 3:00 PM
14.5 ATS products on a standardized product platform line with customized interior and exterior lining at Sekisui Heim factory in Japan.
•
•
Examples of MTS products include warehoused building goods, such as lumber, wood, steel, and aluminum sections, ceiling tiles, and panel material, such as gypsum board or plywood. Assembled-to-Stock (ATS): ATS products have set designs and established standards. Many of the attributes of MTS are found in ATS, but customization is introduced. The principles of assembly line production and mass customization are often associated with ATS, where customers request variation within a set system of form and relationship of elements to one another. Outside of the building industry, computer companies and shoe companies are now offering customizable options for their standardized products. The most prevalent examples of ATS fabrication in housing architecture include light wood frame (LWF) and light gauge steel (LGS) panels that have varying levels of enhancement (completion in the factory from frame only – called open panel – to fully insulated, pre-wired and lined internally and externally – called closed panel) and LWF, LGS and hot rolled structural steel permanent modular boxes that are finished up to 90 percent complete in the factory and assembled on site. Modular construction is becoming more common for multi-family and hospitality architecture that is repetitive (figure 14.5). Made-to-Order (MTO): MTO products are pulled forward through their supply process to arrive on site just in time. These products are not sitting on shelves in MTS or have a set geometry as in ATS, but have determined the design and engineering options within a product. MTO are
•
not made until the last responsible moment but do require more lead time than ATS products due to their increased variability from product to product sold. Examples include custom windows, doors, and other elements that have a myriad of options and are custom-made for a project within a product line. Many custom prefab housing systems on the market today represent MTO fabricated from panels or modules. Engineered-to-Order (ETO): ETO might also be called designed-to-order. These products represent the most complex and demanding products available. This is, by far, the largest category of building creativity and development in architecture. It also represents the greatest challenge for manufacturers and fabricators trying to determine how to deliver entirely custom products at competitive pricing. ETO products generally have the greatest lead times and the highest price points. Examples of ETO products for building include precast elements, façades, and other per specification design and construction elements.
Architects specify MTS, ATS, MTO, and ETO elements for buildings (figure 14.6). Looking at design through the lens of these manufacturing principles, prefabrication can be a tool by which the design team can control cost. If a product truly does not need to be custom, perhaps a more simplified method of delivery is possible. ETO producers operate shops that manufacture components, panels, and modules that are designed and engineered before production. Some prefabricators maintain large in-house engineering departments as a holistic
180
14 Smith.indd 180
2018-11-04 3:00 PM
14.6 The terms Made-to-Stock, Assembled-to-Stock, Made-toOrder, and Engineered-to-Order are used in manufacturing to define the extent to which a product is customized. This is generally considered proportional to the cost and lead time necessary for production. Prefabrication of architectural elements is within the scope of ATS, MTO, and ETO. Sometimes MTS (off-the-shelf) and MTO (flexible) are used exclusively to describe standardized versus customized products.
delivery of their services, while others outsource engineering and detailing. In addition, some use installers to place their ETO products in buildings. MTS, ATS, MTO, and ETO are not entirely exclusive. ETO uses MTS, ATS, and MTO in order to manufacture their products. In addition, a prefabricator that is primarily dealing with MTO can offer, on a limited basis, ETO products as well. Many prefabricators have their bread and butter and find a market niche in one specialty product. Finally, although the terms MTS, ATS, MTO, and ETO are used to describe the various levels of manufacture with their respective levels of cost, lead time, and flexibility, often MTS and MTO are terms used to generally distinguish between standardized (massproduced) and variable (mass-customized) products.
DESIGN FOR MANUFACTURE AND ASSEMBLY The AIA B141 agreement for architectural services points out that architects are not to take responsibility for the means and methods of construction. Architects are not to suggest modes of building production. While considering prefabrication and mass customization, architects have embraced the digital question head on, leading in software solutions for form and organization of on-site construction data in some cases. However, architects have not, save in few cases, taken the risk associated with designing for manufacturing and assembly (DfMA) – for the very physical work of making in the factory and erecting on the job site.15 Until the late 1700s, manufacturing was a craft-based activity in which one person was responsible for all aspects of manufacturing,
including procurement of materials. This method of manufacture had disadvantages. First, products were supply- rather than demand-driven, making the capacity to meet an increase in demand impossible. Also, new products or new technologies were inefficient because there was no common building block. Finally, manufacturing methods were inefficient due to a lack of repetitions involved in the work.16 The Industrial Revolution introduced more effective sources of power and advances in material manipulation. From drilling and milling to lathing and deforming presses, the primary manufacturing technologies have not changed in the history of industrialized manufacturing, only the tools and materials have been improved. These improvements include:
•
•
Interchangeability: realization of the concept of interchangeability of parts for a given product was developed. This allowed random pieces to be selected and assembled to form a single product. Increase in production rate: separation between primary manufacturing and assembly. Fitting is the process of improvement to allow for product functionality while assembly is a secondary process whereby one manipulates the finished parts into a meaningful spatial relationship.17
Fitting is the making of parts that, when assembled, can be a meaningful whole. In the manufacturing of automobiles, an assembly may be parts that make up a single door, or the door may be a fitted part that when assembled with the rest of the parts constitutes an assembly – the automobile. During the early Industrial Revolution, the act of fitting constituted a great deal of time 181
14 Smith.indd 181
2018-11-04 3:00 PM
and energy. Today, fitting is relatively negligible in manufacturing industry. In the building industry, however, it is quite the opposite. Fitting parts together on site is standard practice. Much in the way early Fords were run on a production line adding one part at a time to an overall assembly, on-site building production relies on the craft of individuals to piece together buildings into an assemblage. Prefabrication implements the concepts of interchangeability and increased production rate discovered in manufacturing and applies them in building construction. The manufacturing terms of parts, subassemblies and assembly refer to three levels of manufacture and fabrication adopted for architecture:
•
•
•
Parts: Parts are fitted products that may be stand-alone materials or may be components for construction. In prefabrication, parts are not erected on site, but rather are joined together in a sub-assembly in the factory. These are MTS elements. Sub-assemblies: This refers to components, panels, or modules that are pieced together with parts to create elements to be assembled on site. These are MTO products. Assembly: This is the act of setting subassemblies together on site in their final location and stitching.
The fundamental strategy then in productivity in production theory is to reduce the number of parts in a sub-assembly and reduce the number of sub-assemblies in an assembly. Architecture and building production can learn much from this DfMA approach. In the 1990s, the concept of mass customization was seen as the business strategy of the future, offering a streamlined approach to delivering infinite variability while reducing cost.18 Although mass customization is beginning to have an impact on architecture, there is still relatively little connection between design software environments and manufacturing output.
Therefore, most products today are still designed with a standardized mentality – shop drawings are submitted, and design and manufacture are rarely integrated. Schodek et al. call manufacturers that have CNC tools “islands of automation” that offer the potential for mass customization, but require architects to engage in a meaningful collaboration with manufacturing in order to realize these benefits.19 The prime example of mass customization having reached increased variety with reduced cost is window manufacturers. No two windows are made the same, and to do so would not offer any reduced cost that was significant enough to warrant standardization over tightly fitting custom windows. Mass customization is much more common in industrial design than in architecture, where variation is not an entirely unique building that in no way resembles the one before it, but rather a product that has many similar products with preferential adaptations – sometimes called personalization.20 In short, the differences between architecture and industrial design production are found in volume and repeatability. Although a fully integrated mass customization model is not entirely possible under the current methods of project development and delivery of architecture, a few models do exist in industrial design that can be transferred to the built environment (figure 14.7):
•
•
•
Component-sharing modularity:21 the same fundamental components with appearance variability within each discrete product (changing cladding options initially from project to project on a panel or module). Component-swapping modularity: the same configuration of appearance with ability to swap component function (changing cladding options post-occupancy on panels or modules). Cut-to-fit modularity: varying the length, width, or height of a product by cutting to size based on a fixed module (increasing or reducing structural panels or modules based on a standardized costly product, such as cladding).
182
14 Smith.indd 182
2018-11-04 3:00 PM
14.7 Mass customization concepts in industrial design. Top row from left to right: component-sharing modularity, componentswapping modularity, and cut-to-fit modularity; and bottom row from left to right: mix modularity, bus modularity, and sectional modularity.
• •
•
Mix modularity: variation is achieved by mixing products (panels or modules that can be added or taken away in fabrication and assembly). Bus modularity: a base structure that supports a number of attachments, sometimes called “platform design” (base frame to which numerous modular boxes can be attached). Sectional modularity: parts are all different but share a common connection method (panels or modules may vary, but the connection to one another is always the same like LEGO or even the standard three-prong electrical outlet).
CONCLUSION Mass customization, enabled by technology, is a business and product platform approach used in prefabrication architecture. While mass customization in architecture has primarily been a formal study, leveraging digital tools, this chapter has presented design for manufacturing and assembly (DfMA) principles and concepts to aid architects in realizing greater agency in the development of democratic housing solutions. Architecture is fundamentally concerned with producing elegant physical environments for people. However, compared to other production industries such as automobile manufacturing, it is peculiar in its unique site, scope, and mutable labor. Prefabrication architecture works to solve these problems not as ubiquitous solutions, but rather, by leveraging appropriate levels of customization in order to increase housing affordability and accessibility while not sacrificing the spatial and environmental experience for its inhabitants. This will aid architects in transitioning from autonomous
authors to embrace greater agency and influence in democratizing design. Mass customization is a result of digital technology and manufacturing tool development.22 However, the objective can only be realized in its full potential if more architectural operations are integrated into the development eco-system of housing. In this way, both product and process are realized. This may be accomplished through different business platforms. Some prefab companies, such as Project Frog and Blu Homes, deliver building products by employing architects on their staff. Architects may become product developers themselves, as in the case of Kasita, who flattens the entire design-to-delivery process. Or, architects may engage with product developers, such as KieranTimberlake and other architects with Living Homes, or Resolution 4: Architecture, offering designs that are a mass-customized system to be adapted to user needs with manufacturer Simplex Homes. The most recent model of endto-end delivery mirrors that has emerged in the prefabrication market in Japan (Sekisui Heim) and Sweden (Lindbacks) over the past 40 years, is that of the vertically integrated prefabrication company. Recent start-ups of Factory OS in Vallejo, California, and Katerra in Phoenix and Spokane have strong partnerships that integrate architects in both design and production planning. And in some cases, as with Michael Green Architects, entire architecture firms have been acquired by the parent company. It is unclear what the future of prefabrication business and product platforms is, however, what is clear is that mass customization is just beginning to see its impact as a theory and practice in democratizing architectural design.
183
14 Smith.indd 183
2018-11-04 3:00 PM
NOTES 1 Smith, R.E., “Production Technics,” Journal of Architectural Education, vol. 71, no. 1 (2017), pp. 2–4. 2 Ballard, G. and Howell, G.A., “What Kind of Production Is Construction?” Proceedings of the 6th Annual Lean Construction Conference (IGLC–6), Guarujá, Brazil, 1998; and Ballard, G., “Construction: One Type of Project-Based Production System,” in Proceedings of the SCRI Forum Event Lean Construction: The Next Generation, Salford, UK: SCRI, 2015, p. 14. 3 Buntrock, D., Japanese Architecture as a Collaborative Process: Opportunities in a Flexible Construction Culture, London: Spon Press, 2002, pp. 105–106. 4 Smith, R.E., “ Connect Homes: Modular Housing Patent and Architect Led Business Model,” Journal of Architectural Education, vol. 70, no. 1 (2016), pp. 168–171. 5 Lidelöw, H., Stehn, L., Lessing, J., and Engstrom, D., Industriellt hysbyggande, Lund, Sweden: Studentlitterature, 2012 (in Swedish). 6 Sampson, S.E., “The Unified Service Theory: A Paradigm for Service Science,” in Maglio, Paul P., Kieliszewski, Cheryl A., and Spohrer, James C. (eds.), Handbook of Service Science, New York: Springer, 2010, pp. 107–131. 7 Smith, “Connect Homes,” op. cit. 8 Henderson, T., “U.S. Construction Is on the Rebound after the Great Recession,” Public Broadcasting Service, August 30, 2016; available at: www.pbs.org/ newshour/economy/u-s-construction-rebound-greatrecession 9 CBS News, “A New Blueprint for America’s Construction Trades,” October 1, 2017; available at: www.cbsnews.com/news/labor-shortage-a-newblueprint-for-americas-construction-trades/ 10 Numerous studies have been published that demonstrate the productivity benefits of utilizing prefabrication. Many of them are cataloged in Smith, R.E. and Quale, J.D., Offsite Architecture: Constructing the Future, London: Routledge, 2017.
11 Luck, M. and d’Inverno, M., “A Formal Framework for Agency and Autonomy,” Proceedings of the First International Conference on Multiagent Systems, 1995; available at: www.aaai.org. 12 Smith, R.E., Prefab Architecture: A Guide to Modular Design and Construction, Hoboken, NJ: John Wiley & Sons, Inc., 2011, p. 68. 13 Ibid., p. 69. 14 Ibid., pp. 122–124; and Smith, R.E., “Prefabrication,” Journal of Architectural Education, Online Artifacts/ Reviews, April 27, 2016; available at: www.jaeonline.org/ articles/exhibit-reviews/prefabrication#/page1/ 15 Boothroyd, G., Dewhurst, P., and Knight, W., Product Design for Manufacture and Assembly (2nd edn, rev. and expanded), Boca Raton, FL: CRC Press, 2002. 16 Redford, A. and Chal, J., Design for Assembly: Principles and Practice, Maidenhead: McGraw-Hill, 1994, pp. 3–4. 17 Smith, “Prefabrication,” op. cit. 18 Pine, J.B., Mass Customization: The New Frontier in Business Competition, Boston: Harvard Business Press, 1993. 19 Schodek, D., Bechthold, M., Griggs, J.K., Kao, K., and Steinberg, M., Digital Design and Manufacturing: CAD/ CAM Applications in Architecture and Design, Hoboken, NJ: John Wiley & Sons, Inc., 2004, pp. 156–157. 20 Smith, Prefab Architecture, op. cit., pp. 183–184. 21 The term modularity in these industrial design principles is not to be confused with “modular” construction that is being developed to produce housing and hospitality projects across North America. However, modular construction, a form of prefabrication, is often used to achieve different types of modularity described in this section, including mix, bus and sectional modularity approaches. More on the term modular and its relation to prefabrication and mass customization is reviewed in Smith, “Prefabrication”, op. cit., and Russell, A.L., “Modularity: An Interdisciplinary History of an Ordering Concept,” Information and Culture: A Journal of History, vol. 47, no. 3 (2012), pp. 257–287. 22 Smith, Prefab Architecture, op. cit., p. 183.
184
14 Smith.indd 184
2018-11-04 3:00 PM
15 FUTURE ADAPTIVE BUILDING: MASS-CUSTOMIZED HOUSING FOR AN AGING POPULATION
JOHN L. BROWN 185
15 Brown.indd 185
2018-11-04 2:36 PM
15.1 The relationship with our environment changes as we age.
A good house can help us to live better – in both good times and bad. Like a pair of properly fitted shoes, a well-designed home can help make life’s road feel a little less bumpy and our stride a bit more confident. For most of us, the houses we live in through childhood and middle age probably suit our needs rather well. The same cannot usually be said about the home in which we grow old and frail, the one we want to live in for as long as possible. Too many seniors live out their last years in a poorly fitted house that is inconvenient, inhospitable, and makes the daily chores of life more difficult and perhaps even dangerous. Even more are forced into institutionalized care before it is medically necessary because they can no longer live independently in their homes. The scale of this problem is expanding as the world’s population rapidly ages. The World Health Organization projects that by 2050, 22 percent of the global population will be 60 or older, up from 11 percent in 2006.1 By the middle of the twentyfirst century, and for the first time in human history, seniors will outnumber children. The situation will become particularly acute in Canada, the United States, Australia, and New Zealand where, over the next 30 years, the population bulge of post-war baby boomers passes through old age. In Canada, for example, the senior population is expected to jump from 4.9 million in 2011 to 10.3 million by 2036.2 We need to rethink the nature of the houses that will help see us through this last stage of life. This extends far beyond things like adding wellplaced grab bars and accessible showers to include a fundamental reconsideration of the way our houses are designed, built, and used.
It’s time to reconcile the design of our homes to the reality of the many changes that are part of the aging process. Change is the only constant in life, and, as we enter old age, the degree of this change becomes increasingly profound, the rate more rapid, and the direction less predictable. Mass customization and design democratization are two ways for architects, and the clients they serve, to help mediate some of the issues that arise from this increasing level of change. FUTURE ADAPTIVE BUILDING Future Adaptive Building (FAB) is a new housing option for twenty-first-century seniors.3 It is an interior system of design, construction, and inhabitation that can adapt to meet changes in lifestyle, physical health, and cognitive ability. FAB supports the dynamic realities of long-term agingin-place across the full spectrum of housing types. FAB is based on the Ecological Theory of Aging that focuses on the relationship between an older individual and his or her environment, rather than on just the person or their environment in isolation.4 During childhood, the demands of our environment clearly exceed our individual physical and cognitive capacity. As we reach early adulthood, most of us achieve a level of competence whereby we are able to completely look after ourselves and live independently. This ability continues until we experience some kind of physical or emotional issue that reduces our capacity to cope with the demands of our environment (figure 15.1). Barring a mid-life injury or illness, this most often occurs during old age. The severity and nature of these disabilities
186
15 Brown.indd 186
2018-11-04 2:36 PM
15.2 FABmodular is a system of prefabricated cabinetry components that replace the fixed interior site-built walls.
15.3 The components include kitchen and bath cabinetry as well as freestanding closets, bookshelves, and display units.
are highly individual and can include mobility issues, cognitive impairment, chronic disease, illness, or any combination of the above. The rate of decline is also highly variable and individual. At some point in this downward trajectory, our personal capacity may fall below the fixed demands of the house and we are no longer able to live independently in this particular situation. Conventional home modifications can sometimes extend the length of this period, but the inherent rigidity of a conventional house, combined with the prohibitive cost of undertaking more extensive interior changes, usually makes this a short-term fix at best. Living in a FAB house can significantly delay, and sometimes even eliminate, this tipping point between individual capacity and environmental challenge. The mass customization and design democratization strategies underlying the FAB system allow each house to be dynamically adjusted to the unique and ever-evolving needs, and abilities, of its resident. FAB helps to increase an individual’s resilience to aging by continually modifying his or her relationship to the surrounding environment according to evolving capacity.
Future Adaptive Building is built on three strategies that foster three types of resilience: 1. The mass customization of FABmodular provides functional resilience. 2. The co-design system in FABstudio promotes design democratization that increases emotional resilience. 3. The medical specialization of the FAB+ components provides physical resilience. Together they form a design-based resilience support system for older individuals that helps maintain well-being throughout the various challenges of the aging process.
FABmodular FABmodular is a system of prefabricated cabinetry components that replace the fixed interior site-built walls typically used in residential construction to define the spaces within a home (figures 15.2, 15.3). The components include kitchen and bath cabinetry as well as freestanding closets, bookshelves, and display units. FABmodular uses mass customization techniques to create interior floor plan layouts
187
15 Brown.indd 187
2018-11-04 2:36 PM
15.4 FAB+ is a series of specialized medical modules and safety systems that can be unobtrusively integrated into the FABmodular interior.
realize within the practical realities of the North American residential construction industry.
FABstudio
that can be custom-configured to meet the specific functional requirements of the resident, without additional cost. Most importantly, as the resident’s needs change over time, the components can be easily rearranged to change all or part of the floor plan. Although our functional needs evolve continuously during our lives, FABmodular is particularly well suited to the homes of older individuals because the rate of change at this stage of life is often more rapid and unexpected. The adaptability of the FABmodular interior means that lifestyle changes, including, for example, the resizing or relocating of the primary bedroom, adding quarters for a live-in caregiver, or providing wheelchair accessibility can be completed quickly and for very little cost. It uses a mass customization strategy of design and production that makes this level of functional resilience affordable and easy to
FABstudio is a web-based information and communication platform that helps older individuals maintain a sense of agency and control over the design and operation of their homes. FABstudio is centered on a DESIGN app that enables residents to customize the FABmodular layout of their home. A COMMUNITY app empowers residents to share best practices with the residents of other FAB houses. FABstudio also includes a MANAGE app for monitoring and controlling all aspects of the home’s operation. A HEALTH app provides access to personal medical vital signs data, a library of health information, tracking of medication and therapy regimes, and distance communication with health care professionals. A LIBRARY app provides access to design and age-in-place resources including instructional videos on the use of the DESIGN app and the operation of the house. The FABstudio interface is designed to be used by seniors and can be adjusted for varying levels of visual acuity, technical ability, and cognitive capacity. Maintaining a sense of agency and control is an important part of well-being in old age. FABstudio enables the resident to maintain a level of control that is dynamically attuned to his or her evolving level of ability. Most seniors experience a loss of control as decisions they used to make on their own begin to be made by family members and caregivers. FABstudio gives older individuals the capacity to continue participating in decisions about the detailed functioning of their home and how they want to live. It encourages older residents to participate in the design process in order to foster a deeper, more substantive relationship with his or her home. Advanced home automation strategies embedded in the house and controlled through the FABstudio interface, also help to create a nuanced level of capability and control that matches an individual’s cognitive ability.
188
15 Brown.indd 188
2018-11-04 2:36 PM
15.5 The grab-bar is integrated into the design of the cabinetry and visually reads as a towel bar rather than a medical device.
FAB+ FAB+ is a series of specialized medical modules and safety systems that can be unobtrusively integrated into the FABmodular interior to provide physical and cognitive support as well as individually tailored levels of home care, including modular cabinet units for advanced bedside medical care (figure 15.4) and an integral safety rail system. Sections of grab-bar can be easily locked into a continuous support hidden within the cabinetry to provide stability and walking assistance in functional areas such as the kitchen and bathroom (figure 15.5). The grab-bar is integrated into the design of the cabinetry and visually reads as a towel bar rather than a medical device. Depending on the size and spatial layout of the house, the grabbar can also be deployed in hallways and other circulation zones to provide continuous runs of support. The FAB+ grab-bar system is specified in the FABstudio DESIGN app. At a meta-level, an occupational therapist can specify a global pre-set prior to the resident starting the design of his or her interior. Selecting this feature at the meta-level ensures that all design configurations created by the homeowner include the grab-bar detail by default. As mobility deteriorates, older individuals tend to reduce or discontinue previous outside activities and spend more time at home. Within the FABstudio environment, the occupational therapist can specify, at a meta-level, which exercise module is most
appropriate for the resident. The resident then codesigns the placement of the module in the location of their choice. Algorithms in the system restrict placement of the modules to only those areas in the floor plan that have sufficient floor area for the type of exercise program enabled by the module. The ability to reach up or bend down to access storage is the most common physical limitation, after mobility, that negatively affects the way that older individuals use their home. When combined with decreasing range of motion in our shoulders and elbows, as well as a loss of upper body strength, the comfortable zone of reach for above-counter storage becomes much smaller as we age. In FABstudio, the health care team can activate a meta-level setting to ensure that only reach-appropriate cabinet types and locations can be selected during the co-design process with a resident who has functional disability issues. Should reach issues emerge at any time after the initial design has been deployed, the modular cabinets in the kitchen, bathroom, and bedrooms can easily be replaced or adjusted in height to suit the evolving ergonomic requirements of the user. All FAB+ units can be quickly and easily added into any FABmodular layout on an “as-andwhere-needed” basis, in collaboration with the resident’s health care team. FAB+ accommodates mobility and frailty issues and facilitates the effective delivery of high quality medical care without undue disruption to the essential domestic character of the home. 189
15 Brown.indd 189
2018-11-04 2:36 PM
15.6 The strategic location of service points is essential to the FABframe’s design.
FAB IN PRACTICE The following speculative case study explores the potential for the Future Adaptive Building system to transform a typical 650 sq ft production-built, one-bedroom-plus-den unit located in a recently completed high-rise housing development. The result is a responsive age-in-place residence that can accommodate many of the changes that seniors face without departing too far outside the normative housing market. The case study highlights four of over a dozen potential floor plan layouts, including the developer’s original design. Any of the details in these examples could be deployed at the time of first construction as an initial customization. They could also be created as part of a continuous process of adaptability at any time during occupancy. Depending on the extent of the changes, additional cabinet modules may need to be purchased and there would also be a nominal cost for a contractor to make the alterations over the course of a one- or two-day period. These costs would be minimal compared to a conventional renovation and would involve very little disruption or damage. The result is a home that is able to meet the resident’s current and future functional needs in terms of both lifestyle and health.
The resident uses the FABstudio DESIGN app to configure the initial deployment of the FABmodular system. As their needs evolve, they would take up the app again and work with their family and health care providers to design a series of adjustments to their floor plan configuration that meets their evolving needs. This includes the type of spatial/functional changes outlined in the first three layouts as well as the introduction of FAB+ elements illustrated in the last option. When the resident’s tenure in the home is complete, the interior can quickly and easily be reset to the needs of the next resident.
FAB Frame The FABmodular system is based on the bus modularity model of mass customization. “Bus” is a technical term that comes from the electronics industry and is the name given to the armature in a computer through which information is transferred, and into which various components are plugged. To make the concept easier to understand think of a school bus, which is really nothing more than an armature of seats in which an infinite variety of children (modular components) can travel to and from school each day.
190
15 Brown.indd 190
2018-11-04 2:36 PM
15.7 Layout One: FABmodular cabinets are deployed to recreate the original developer’s plan.
Bus modularity systems naturally lend themselves to continuous adaptability because the modular components retain their independence after installation. As long as the connections to the bus are properly detailed, the modules can be moved around, exchanged, substituted, or replaced with no damage and for very little cost. FABmodular uses the bus modularity typology of mass customization to meet both the initial customization and continuous adaptability performance objectives for age-in-place functional resilience. In a FAB house, the bus is the base building comprised of the exterior walls, the fixed service components, such as bathrooms and mechanical service chases, and any permanent interior partitions. The modular components are the mass-produced cabinets that define the rest of the interior layout and provide all the different types of storage required by the resident. In the FAB system, the bus armature is called the FABframe, after Bernard Leupen’s recent theoretical work on adaptable architecture.5 The ability of the FABmodular system to adapt to future change rests as much on the ability of the frame, or base building, to accommodate all the potential changes as it does on the moveable cabinet modules, digital interface, and medical components.
The FABframe is not just an empty shell with a rough-in of building services, as happens with open building commercial construction. In the language of modular mass customization, the bus armature is a critical part of the system and must be designed to easily accept the plug-and-play modules in order to ensure that the various floor plan adaptations can continue to be made over the life of the house. Essential to the FABframe’s design is the strategic location of service points (figure 15.6) that can fit a variety of different kitchen and bathroom configurations, and ensuring that all potential locations for the modular cabinetry are kept free and clear of obstructions such as HVAC ducts, structural supports, ceiling drops, and electrical receptacles.
Layout One In the first plan layout, the FABmodular cabinets are deployed to recreate the original developer’s plan for the case study unit (figure 15.7). A series of 12” deep x 30” wide floor-to-ceiling cabinet modules creates the spatial separation for the bedroom. The bathroom is defined by a series of 24” deep x 30” wide wardrobe cabinets that face into both the bedroom and the hallway; 12” deep bookcase modules define the study and the bedroom. 191
15 Brown.indd 191
2018-11-04 2:36 PM
15.8 Layout Two: The kitchen is relocated to the end of the unit and the living and dining areas are combined into one large space.
Layout Two In the second option, the kitchen is relocated to the end of the unit so that the living and dining areas are combined into one large space (figure 15.8). In this scenario, the dining area can be located beside the window wall for an older individual who spends a great deal of time at a table working on a hobby. Replacing some of the bookcase modules with wardrobe cabinets can provide additional clothes storage. The bathroom doorway is relocated to the bedroom to create a true en suite, with additional general storage added in the hallway.
Layout Three The third layout accommodates a resident requiring the use of a wheelchair (figure 15.9). To make the bathroom of the case study unit wheelchair accessible, the bathroom door is located off the circulation zone rather than the bedroom to maximize the amount of maneuvering space in front of the door. The size of the bathroom cabinetry units has been reduced to create a larger floor area in the bathroom for turning the chair. The modular sink cabinet can easily be switched with either a lower height unit designed for wheelchair use, or a mechanically controlled sit-stand countertop, whose height can be adjusted to accommodate
the needs of different members of the household. Removing the glass shower panel between the toilet and the shower makes both the toilet and the shower accessible by wheelchair. In the bedroom, the wardrobe cabinets adjacent to the living area have been replaced with narrow bookcase modules to increase the floor area of the bedroom and to make it easier to maneuver a wheelchair beside the bed. In the kitchen, the counter adjacent to the hallway has been removed to create easy wheelchair access to the kitchen and open up the floor area by the front entry. If this layout was required at the time of initial layout, a global pre-set by the health care team would restrict the cabinetry options to those specifically intended for wheelchair use. If the need for a wheelchair were to arise after the initial deployment, the traditional cabinet units could easily and quickly be switched.
Layout Four The final floor plan deployment contemplates a situation in which the resident is largely bedridden and requires medical equipment to support increasingly complex regimes of nursing care (figure 15.10). The locations of the bedroom and living room have been swapped so that the best, and largest, space in the unit is devoted to the care bed.
192
15 Brown.indd 192
2018-11-04 2:36 PM
15.9 Layout Three: Accommodating a resident requiring the use of a wheelchair.
15.10 Layout Four: Accommodating a largely bed-ridden resident.
193
15 Brown.indd 193
2018-11-04 2:36 PM
The kitchen is located in the middle of the plan in order to allow for a secondary bedroom space at the back of the unit for a spouse or live-in caregiver. The bathroom is wheelchair accessible. The large bedroom space helps health care workers to more effectively do their jobs. The open connection to the small living area redefines the privacy boundaries of a traditional bedroom and makes it easier and more acceptable for family members and visitors to spend time with the resident in a setting that is both living area and bedroom. A series of FAB+ medical cabinets are located on either side of the bed. A headboard panel above the bed provides a service space for any required tubing or equipment. These units would be specified by the resident’s health care team, based on the therapies and treatment that are required. The cabinets would have the same architectural character as the other FABmodular components so that the medical technology blends as seamlessly as possible into the domestic space of the house. Depending on the health needs of the resident, these could include one or more of the following: • •
• •
A respiratory module containing oxygen supplementation. Non-invasive positive pressure ventilation (CPAP and BIPAP), and nebulized medications. A renal module containing equipment for home hemodialysis or peritoneal dialysis. A feeding module containing an IV support and ancillary equipment required for a nocturnal enteral feeding system via a nasogastric tube.
The FAB+ system also includes two cabinet modules for storing general health care supplies and equipment. If necessary, all the medical cabinets can be locked for the exclusive use of visiting health care professionals. The FABstudio system provides an opportunity for the homeowner, his or her family, and the resident’s health care team to open up a
conversation about how best to integrate these increased regimes of care into the home. The familiarity of the platform grants a level of control to older residents and their families that can reduce stress and help them through the difficult decision-making process of introducing more intrusive levels of medical care into their home. CONCLUSION FABmodular increases functional resilience to the stresses of growing old. It uses mass customization to empower residents to customize the initial functional layout of their home and overcome the longstanding dichotomy between one-off customdesigned homes and mass-produced tract-built houses. It builds on the bus modularity model of mass customization exemplified by the smartphone to enable residents to use a small number of component assets to continuously adapt the spatial layout of their homes to meet the changing needs that develop over the course of growing old. The FABmodular system uses advanced builtto-order manufacturing strategies that can deliver this high level of flexibility and customization for almost the same cost as a conventionally built tract project. It effectively integrates into the typical design, sales, and construction practice of the residential construction industry and can be applied to almost any size, price point, or type of housing project, from single-family houses to high-rise apartments. The initial customization and continuous adaptability offered by FABmodular benefit older individuals because it helps them cope with most of the unexpected changes in functional needs that arise from changes in lifestyle circumstance and physical capacity. This includes changes that may be dramatic in nature or rapid to develop. The customization and adaptability are also of benefit to project developers and homebuilders who may be concerned about limiting the size of their market by focusing too exclusively on the needs of an older demographic. As demonstrated in the case study example, this age-in-place
194
15 Brown.indd 194
2018-11-04 2:36 PM
residence can easily be configured into a normative floor plan organization that appeals to a variety of market segments, including healthy young professional first-time buyers, the staple of most large-scale housing developments in North America. This flexibility represents an added benefit to the older homeowner who, when ready to move, can re-configure their age-in-place residence into a unit with much broader market appeal. It also means that older residents are not forced to live in a “seniors only” project because the specific requirements they need for functional, as well as emotional and physical, resilience can be added, at any point, to any of the units in a FAB-enabled project. FABstudio capitalizes on the principles of codesign and design democratization to empower older residents to manipulate the FABmodular system and help create a home that is continuously customized to their needs. This involvement can happen at the level of design engagement in the purpose-driven activity of co-designing your own home, as well as through the deep participation of pursuing design as a form of serious leisure. Both help foster emotional resilience through feelings of control and a sense of agency and help increase life’s meaning and purpose. For example, one of the most difficult stages of growing old is having to move out of the family home and into a more age-friendly residence. Leaving behind familiar spaces and long-term memories for something new and different is typically very stressful, even traumatic, for older individuals. The FABstudio process can help make this transition easier to manage because the individual is moving into a home that they had an active role in creating. Instead of seeing the move as a loss, or a narrowing of life’s choices, the opportunity and excitement of being involved in the co-design of their new home can be a positive step toward creating a new future. For a smaller group of older individuals, the benefits of this process may blossom into a serious leisure activity where they become super-users, offering peer-to-peer design advice and technical support about the operation
of the FABstudio interface and the FABmodular components. FAB+ helps to increase an individual’s physical resilience to the stresses of growing old as a complement to the functional resilience provided by FABmodular and the emotional resilience enabled by FABstudio. With the FAB+ system, each resident, or their health care team, can deploy a tailored selection of these support services to seamlessly accommodate their evolving health needs and reduce the friction that normally exists between the objects and procedures of health care and the domestic qualities of home. Within the broader concept of health as overall well-being and not just the absence of disease, the integration of mass customization and design democratization in the FAB system helps ensure that older residents maintain not only their independence but a sense of agency, dignity, and self-worth in the face of an increasing need for ever more intrusive home health care. Enabling an individual’s resilience to growing old through architecture is not just about making the house safe when we become frail or enabling health technologies to look after us as we become sick. The nursing researcher Joan Liaschenko argues that the goal of home care is not merely in helping patients to stay alive or even healthy but in helping them to have a life. To have a life is to have a sense of agency, to occupy social and political space, to live a temporally structured existence, and to die. Protecting and fostering patient agency not only are part of this, but they are central to all the others.6 Growing old is an architectural, as well as a medical, issue that urgently needs to be addressed. Through systems like Future Adaptive Building, good design, as both process and product, enabled by contemporary advances in mass customization delivery and the design democratization in codesigning strategies, can significantly improve the quality of life for our rapidly expanding senior population.
195
15 Brown.indd 195
2018-11-04 2:36 PM
NOTES 1 Global Age Friendly Cities: A Guide, Geneva: World Health Organization, 2007, p. 3. 2 Housing for Older Canadians: The Definitive Guide to the Over-55 Market, Ottawa, Canada: Canada Mortgage and Housing Corporation, 2012, p. 7. 3 The concepts presented in this chapter were developed as part of the author’s unpublished PhD dissertation, “Going Home: Future Adaptive Building for Aging-inPlace,” RMIT University, 2016. 4 Powell, Lawton, M., “An Ecological Theory of Aging Applied to Elderly Housing,” Journal of Architectural Education, vol. 31, no. 1 (1977), pp. 8–10. 5 Leupen, Bernard, Frame and Generic Space: A Study into the Changeable Dwelling Proceeding from the Permanent, Rotterdam: 010 Publishers, 2006, p. 20. 6 Liaschenko, Joan, “The Moral Geography of Home Care,” Advances in Nursing Science, vol. 17, no. 2 (1994), p. 24.
196
15 Brown.indd 196
2018-11-04 2:36 PM
16 DEMOCRATIZING CREATIVITY
CHRISTOPHER SHARPLES 197
16 Sharples.indd 197
2018-10-18 10:27 PM
The way we work as architects is deficient, and soon, perhaps, it will be defunct, replaced by something new, something that reflects more closely our ambitions, our best values, and the possibilities and responsibilities of the restless time in which we live. The historic divisions between creativity and production, the alienation of labor identified by John Ruskin in The Stones of Venice,1 have only deepened in the 165 years since its publication. We mostly still rely on systems of organization and habits of thought developed in and for other, very different eras. Our methodologies – from school to office to job site – are inefficient, not conducive to healthy collaboration, incapable, as things stand now, of leveraging the full potential of the information we all create and share to shape and guide our collective work to meet the challenges we see before us. Architects and the entire suite of associated professions and trades, all of us need to do better, and fast. Why, for example, is something as simple as the generation and conservation of data across parties and throughout the process of design and construction (not to mention, critically, in postoccupancy, and even beyond, as a seed for the next campaign) not by now an everyday occurrence, the definition of mundane, a given, on every project? Why should championing and implementing progressive methods of design and construction even be considered, this late in the twenty-first century, an innovation of any sort? Architecture is ripe for disruption. This is a fact every one of us knows in our hearts but may not acknowledge. Few architects were surprised, for instance, when a 2017 report by McKinsey & Company, Reinventing Construction: A Route to Higher Productivity,2 identified building in the United States and Europe as “among the least digitized sectors in the world.” Compared to every other area studied – retail, transportation, even a pursuit as seemingly brutal as mining – construction lagged behind, recording negative rates of productivity growth in an increasingly streamlined world. Why? Some seek to find a sort of exceptionalism in design and construction, a quality in its nature and degree of complexity that
leads to this arena of human activity, practically alone now if the McKinsey authors are to be believed, being resistant to the revolutions in communication and instrumentality that new technologies have long made possible elsewhere. Certainly, the necessary interdependence of the associated industries responsible for building (the compound of architecture, engineering, and construction known as “AEC”) is a major factor in this inertia, creating points of friction internal to each field, between them, and with the clientand user-side structures of finance and culture that enable and place demands on its production. That the AEC industries exist in turn within a nexus of civic, legal, and governmental forces, each differentially resistant to change, must also be credited as a delaying factor, and those barriers to change will need to be identified and confronted frankly and directly in order to achieve success in the professional revolutions to come. But it is the force not yet mentioned, labor, that may yet offer the greatest challenge to change and – if we follow the innate potential for humanistic horizontality in our digital tech, not only its powers to amplify productivity and extraction – it is labor, construction workers, all the trades, that could become the beneficiaries of the greatest rewards: increased autonomy, increased creative participation, increased job site safety, and, crucially, decreased alienation from their work. The choice is ours, and we will make it in the decades to come, willfully or by default, through how we steer our tools and shape our methods in this moment of critical transition. Our feeling at SHoP Architects, what we have discovered after two decades of experimentation with nontraditional methods of design and construction, is that the net effect of all of them – the digital modeling softwares, the data management platforms once known as BIM, the dramatic forms of visualization, the suite of methods now known as Virtual Design and Construction – can be used to bring people together in their work in novel ways, to give voice in the creative process to those who may typically be thought of only as
198
16 Sharples.indd 198
2018-10-18 10:27 PM
16.1 Camera Obscura pavilion at Mitchell Park, Greenport, New York (2003).
suppliers or implementers, to make the idea of a fully democratized process of architecture seem possible. This is a disruptive dream very much worth pursuing, if one that, still, faces very significant challenges on every front. How do we make this process – taking buildings from an idea, to physical form, to lasting contributors in the everyday quality of our lives – more organic, efficient, meaningful, responsive, and participatory? How do we make it fairer? How do we begin to achieve the possibilities of an architecture open to all? We need to start by listening to what our tools and techniques are trying to tell us. TOWARD AN OPEN ARCHITECTURE Every technology has an in-built bias. Machine tools suggest orders, sequences, timings. Their logic is the logic of meshed gears, belts driving pulleys at fixed ratios. Their native terrain, the outcome of their predispositions, is the assembly line – purposebuilt to task, fixed in space, and ordering time as a sequence of discrete steps; it is worth remembering that the River Rouge plant produced not only Fords, but also Fordism. Our technologies today, of
course, suggest something completely different, a decentralized mode, or, rather, one centered on the speed and fluidity of data as a unit of communication, and that has very different implications for time and space as they might relate to physical production. At SHoP, we began to discover this first-hand many years ago when the Cooper Hewitt Museum asked us to create a model of our Camera Obscura pavilion (figure 16.1) at Mitchell Park (Greenport, New York, 2003). That project was a breakthrough for us, and I believe an historic first, in being modeled digitally in its entirety, down to the fasteners (figure 16.2), and built from pre-cut CNC components via a set of exported instructions that existed purely as assembly diagrams without a single dimension needed; the structure was put up with no on-site measurements or cuts. For the physical model destined for the museum gallery, we merely dialed down the scale and “hit print” again, a power that then seemed miraculous. Now, of course, the printing has become literal and, in an object-organized platform such as Dassault’s 3DExperience,3 each magnitude of scale represents an interlinked galaxy of possibility in which one can refine and universally coordinate the
199
16 Sharples.indd 199
2018-10-18 10:27 PM
16.2 The digital model of the Camera Obscura pavilion.
200
16 Sharples.indd 200
2018-10-18 10:27 PM
16.3 Dunescape, a temporary event pavilion built in the forecourt of MoMA/ PS1 in Queens, New York (2000).
largest consequences of the most minute detail. For the Camera Obscura, not two decades past, we did all that modeling and re-modeling by brute force, each change necessitating hand-crafted fixes throughout. The tool has become more powerful, but also more accessible, more open, better networked, and so more conducive to supporting broader collaborations. If we are to grow as a profession toward a meaningful transformation of the industry, as such technologies progress by leaps and bounds, our understanding of the implications of these means of production must keep pace. The new spatial scalability, we would soon see as our tools advanced, was paired with an even more powerful scalability in time: digital processes could be made massively parallel, opening the platforms, the models, to multiple hands and minds working and contributing at once, capable of expanding the moment of creation across multiple authors both internal and external to our own studio, coordinating all into a single, purposeful workflow, and then collapsing that collectively generated and organized information into a single process, component, or construction. That capability, that innate direction or implied desire of the digital modeling tools we all use now at only a fraction of their potential, is what, to me, offers the greatest hope for change in the future that must come: a design and construction technology understood as a true communications medium, at once demystifying architecture and opening it, with the entire process of building, to direct creative input by all those who participate in its conception, production, and, eventually, end use.
For SHoP, ironically, we approached and gained a mastery of this manner of nonlinear work through a strictly sequential investigation over many years, progressing through a series of project-based researches we think of now, only in retrospect, as generative – the necessary work of compiling our own DNA. It is important, in keeping with the theme of democratization, to note that even within our own firm these projects only came to be through the contributions of many dozens of individuals over many years, some still with us, others taking their inspiration and expertise to explore similar projects at new enterprises. SHoP is best thought of as an operating system, or a platform; many strains and tendencies and movements coexist among our diverse staff, all organized as openly and horizontally as possible to ensure an equal hearing in the process of discovering the best ideas in their most suitable applications. It is only with hindsight that the early sequence of collaborations began to be identified as generative – in the moment it was project-specific research: a slow building of lesson-upon-lesson, fueled by need as much as curiosity, pressured by budgets and deadlines, that was leading us toward an answer in each moment to what we consider the most important architectural question of our age: how can we build for today in a way that reflects and supports our best values? The first of these DNA projects was also one of our earliest: a temporary structure built in the forecourt of MoMA/PS1, the folly and event pavilion we called Dunescape (Queens, New York, 2000, figure 16.3). The method of construction (field-cut cedar two-by-twos) and the manner of communication 201
16 Sharples.indd 201
2018-10-18 10:27 PM
16.4 A template drawing for the construction of the Dunescape pavilion.
(lengths measured directly off of very large oneto-one templates) represented a move away from drawings as representation and introduced us to an instruction-set approach in which the end-product of design is not a delineation of intentions and extents to be defined by others but directions for an assembly of pre-specified parts (figure 16.4). That our workforce at PS1 was largely composed of friends with varying degrees of literacy in the arcana of traditional architectural representation is not insignificant: we learned instantly through that experiment the democratizing power of using “plain language” in construction. When the Camera Obscura went up a few years later, the professional builders there were working with a kit of parts and an instruction set that looked like it might have come from Lego or IKEA. They loved working that way, too. Looking back, all of it can seem quite obvious, and indeed the breakthrough these projects represented for us at the time can be hard to explain to younger colleagues or visitors to the office, where a very large model of Dunescape greets them in our lobby gallery as a sort of totem. The story is only
complete when they begin to connect it with the images of the pavilion’s construction and use that we have polemically hung as reminders in our two print rooms. The dated fashion and Kodak low-res of those images help everyone to see it for what it is, as history, from a time before it was clear what our new digital tools were really for: fast, accurate, open sharing; rapid and responsive making; the means to a newly inclusive creativity. Designing and building Dunescape and the Camera Obscura brought us into direct contact with the power and possibility of our new tools, but, with one located in the front yard of a museum and the other in a village park, we had not yet field-tested the ideas, or seen what we might learn, by applying them in “the deep end of the pool”: the unforgiving world of Manhattan real estate. Our work on The Porter House (New York City, 2003, figure 16.5), gave us that chance. A residential renovation and addition to an historic warehouse at the gateway to the then-undeveloped Meat Market neighborhood, the maximizing impulses of real estate formed a primary pressure on the project. We won it by proposing an innovative approach to consolidating the air rights
202
16 Sharples.indd 202
2018-10-18 10:27 PM
16.5 The Porter House in New York City, 2003.
16.6 The alphanumeric assembly identifiers were laser-cut into the corner of each panel on The Porter House.
of adjacent properties, paired with an engineering solution that made the resulting cantilever possible. We ensured its architectural integrity, in the end, by two novel forms of collaboration. The first was with the client who, after questioning our decisions in the normal way, became much more trusting after we assumed a direct financial stake in the project. Invested in that manner, with, as the developers say, “skin in the game,” we were also then even more fully responsible for the result. That led us to form far closer, more collaborative, and more equitable – more democratic – relationships not only with our client but also with the overseas material suppliers and the local panel fabricators. It also encouraged an approach that was significantly more parsimonious with the material, a beautiful French zinc, than we,
or any architect, might otherwise have adopted. We learned how to coordinate expressive desires with the unforgiving limits of sheet-size, and how to use our partner’s nesting software to eliminate waste. The façade of The Porter House, now presiding over quite a different neighborhood than the one we first found, stands as a quiet monument to the collective problem-solving made possible by our digital tools as a component of an inclusive, open-minded strategy of cooperative construction. If you look closely, you can see the alphanumeric assembly identifiers laser-cut into the corner of each panel (figure 16.6). We left them there on purpose, to tell this human, technical story, to honor our collaborators, the way that, as Ruskin noted, the hand of a given stone carver can still be discerned on the walls of Venice. 203
16 Sharples.indd 203
2018-10-18 10:27 PM
16.7 Barclays Center arena in Brooklyn, New York, 2012.
The idea of the master builder, the Renaissance professional who could demonstrate and apply expertise over the work from the first cut in the quarry to the last polish of the ornament, who could direct through that expertise an especially informed leadership of every craft and trade, had long held a romantic fascination for us. Now, however, through the course of the on-the-job R&D our tools and techniques were leading us into, we were beginning to see that it might be possible to revive that holistic notion, that productive generalism. It was also becoming apparent that the architectural richness that in past eras was made possible by abundant labor – examples of which run from medieval cathedrals to the early-twentieth-century
skyscraper in which SHoP has its studios today – could now be realized through technology. The rigor and repetition that were both the product and the preferred expression of the last century’s industrial modernism could be replaced. Our technologies could, without delaying the ribbon cutting or breaking the bank, now go beyond the architecture of Fordism and give us – all of us working on a project – a new outlet for a rich, variable expression of the kind that reflects and accommodates an innate humanity. Mass customization, in the sense we first achieved it at the Barclays Center arena (Brooklyn, 2012, figure 16.7), is not a stylistic exercise. It is a human exercise – about change, choice,
204
16 Sharples.indd 204
2018-10-18 10:27 PM
16.8 12,000 individual steel panels for the Barclays Center were tracked from plant to installation using bar-code scanning via an iPhone app.
desire, accommodating difference, reflecting the consensus of a collective will. The ability to create field conditions that are at once intensely specific and situationally responsive – one effective use of mass customization – is both the result and an expression of the effective use of an open creative architecture. At Barclays, for the first time, everyone was “in the model” simultaneously. We could adjust form in response to production limitations in real time. Cost and logistics could be visualized on the same models that we were using to determine the contextual effects of massing. This maximized not only the performative but also the creative interaction of all parties. We could start to see in practice how, as per Ruskin but empowered
by technology tools, the workers themselves could begin to see their creativity incorporated into the form of a new architecture. And it was all made possible, under the intense deadline pressure of that project (it was a little more than three years from the commission to the tip-off of the Brooklyn Nets’ first home game) by the ability, developed on the run during the process itself, to auto-generate dimensional fold and cut information, and directly export it to our fabrication partners’ machines, for over 12,000 individual steel panels (figure 16.8), process them through an accelerated weathering system we assisted in designing, then visually track their progress from plant to installation using barcode scanning via an in-house iPhone app.
205
16 Sharples.indd 205
2018-10-18 10:27 PM
16.9 Botswana Innovation Hub, under construction now in Gaborone.
This accomplishment marked the end of our ironically linear sequence of non-linear discovery. Since Barclays, SHoP’s investigations have been aided by the old-fashioned method of business expansion – more resources to deploy, more hands on deck, more minds on a problem, more ideas resulting, more democracy at work. This has led to a branching of the tree, to new strains of applied research. But all of them are united, still, by the same idea: that the powerful tools and techniques we are fortunate to have access to today can be used by architects in an open process to make our environments and the relationships they shape and reflect more human, not more alienating. We see this at our Botswana Innovation Hub (figure 16.9), under construction now, which has become another benchmark in the development of an open architecture. Unwilling to follow the usual aloof habits of international practice, after winning an anonymized competition, we sent a team to live in Gaborone in Botswana and work on-site with local contractors and engineers to realize the 310,000 sq ft project – a combination of laboratories, creative office spaces, and shared amenities strategically interlinked to promote the exchange of ideas. The trailer where we set up our server at the end of a single long work table performed the exact same function; working there
with our new colleagues over the many years of the project’s realization, we came to a much greater understanding of the material and customary expectations and needs of those who would use our building, and we were able to incorporate these, in real time, into our design as it grew and changed. We were also able to share some of what we have learned over the years about a suite of tools and a manner of working that were not yet in common use in that region, and that had the effect – the democratizing effect – of opening the process more fully to the voices of our collaborators. As the Hub enters its final years of construction, our team has come home and our Botswana partners are managing the job in seamless partnership and with instant, integrated information flow with our office on the other end of the fiber optic cables. It seems almost incidental now that “the job” in question includes an automated instruction set and direct-fabrication effort for the environmentallyperformative self-shading façade that, both in raw element count and formal complexity, dwarfs what we were able to achieve at Barclays in our own backyard. SHoP has also found itself returning to its roots in a series of pavilions. Just as with Dunescape and the Camera Obscura, these small-scale, tight-timeline projects are ideal as
206
16 Sharples.indd 206
2018-10-18 10:27 PM
16.10 Flotsam & Jetsam at the 2016 Design Miami fair.
a working platform for research. The first of this new generation, the entry environment to the 2016 Design Miami fair, a project we call Flotsam & Jetsam (figure 16.10), was designed, fabricated, and assembled over a period of only three months. It is fortunate that we conceived this work from the beginning as an opportunity to implement a truly unified collaborative approach between all parties, coordinated in all aspects on the latest Dassault platform, as such factors as ongoing programmatic changes and the discovery of technical limitations in manufacturing (in this case, how much a materialprinter armed Kuka robot, running continuously,
could print in a given period of weeks) necessitated ongoing and fruitful revision to the designs. The collaboration between our partners at Branch Technologies (printing the main components of two shell structures using 47 miles of a new carbon fiber-reinforced ABS in a material-conserving spatial matrix technique they have pioneered) and the Oak Ridge National Laboratory (fabricating the balance of the festive environment from a recently developed fully biodegradable bamboo-based print medium) represented a new level of integrating diverse expertise, and a new high-water mark toward achieving a democratized workflow. 207
16 Sharples.indd 207
2018-10-18 10:27 PM
16.11 Wave/Cave installation at the 2017 Salone in Milan.
Meanwhile, not long after the experimentation with very contemporary materials in Miami, we were invited to contribute an installation to the 2017 Salone in Milan. This project, known as Wave/Cave (figure 16.11), is especially close to my heart as it demonstrated the possibilities of an architecture for today – an architecture unleashed from the constraints of hierarchy and linear methodologies, conceived in an open process – that nonetheless retains its material roots in a very ancient way of making. Wave/Cave was a 60 m2 topographic composition formed by 1670
unglazed terracotta blocks. Stacked in three tiers, the interior faces of the blocks reveal a surprising ornamental richness as they describe a smoothly curved surface within the 3.6-meter high perimeter wall. Developed in collaboration with NBK Keramik, with a concealed fastening system designed by Metalsigma, the unique extrusions feature regularly fluted exterior faces and a webbed cross-section that exposes its geometrical complexity when cut at various inclinations. To achieve this, we opened a direct line of communications on shared modeling and
208
16 Sharples.indd 208
2018-10-18 10:27 PM
16.12 Stagg Street, a fabrication, prototyping and testing facility for SHoP Architects.
management platforms between our designers and NBK craftsmen and technicians, resulting in a block form that balances structural integrity, manufacturing efficiency, and expressive possibility. The blocks were extruded through a custom die, then cut with CNC equipment into 797 individual profiles. Design and manufacturing, again, were fully integrated in this project, with digital instruction sets for the profile cuts auto-generated and exported directly to the production team. The result was a Ruskinian level of craftgenerated delight, achieved under contemporary economic constraints not by automating the worker out of the equation, but by incorporating their creative expertise into an automationaccelerated process in the context of a truly open architecture. Albeit in microcosm – not unlike the example of Dunescape and the Camera Obscura preceding The Porter House and Barclays, we hope to be able to bring what we have learned in Milan and Miami, as soon as possible, out of the hothouse environment of the design fairs and onto the streets. Aiding us now in that critical long-term project is a facility we have developed over the last several
years not far from Newtown Creek in Brooklyn. Referred to in-house simply as Stagg Street (figure 16.12), where it is located, it is a simple steel-and-sheet-metal Butler building4 in which SHoP has established a fully equipped test-bed for process innovation. Individual projects undertaken at Stagg have ranged from the construction of full-sized performance mockups, to large-scale model-making and materials experimentation, to the automated precision fabrication of actual building components for an innovation and maker lab addition to the Benchmark School outside of Philadelphia (Media, Pennsylvania, 2018). Many of our staff have gained very valuable hands-on experience there already, directing the industrial robot or just working on a lathe. But its true use to us, this small factory where we can iterate under controlled conditions through the whole process of design and construction, is as a laboratory for discovering how design and manufacturing labor, united with the latest technology tools, can best combine to support the democratization of creativity that our industry so badly needs if it is going to meet the challenges of this still-young century. 209
16 Sharples.indd 209
2018-10-18 10:27 PM
AT THE MOMENT OF CHANGE So, architects are almost magicians now – we can create a digital twin of an intended future construction, integrate the expertise of all parties into the most robust formulation, respond nimbly to every constraint, communicate real-time data with craftspeople in such a way that machines may even capture some of the soul of their work. So what? That work is still only as meaningful as how it is applied, which problem it is applied to, which crisis it may contribute to averting, what necessary social progress it may allow. How can we use our new suite of tools and techniques to address the endemic disconnection in the flow of architectural creativity brought about by industrialization, the alienation of the worker from the work, the thinker from a hand in the physical creation of a thought? How can we fully and finally close the gap between action and intention in the built world in a way that captures the consensus wisdom of a true, direct democracy? And how can we, through that new, open architectural integration, begin to solve the problems of our age? As we have throughout our history, as detailed in part above, at SHoP, our instinct is to always approach solution and resolution by the direct application of curiosity through real work, what we think of as everyday R&D. And we do our best work in response to real, pressing problems – perhaps even, as at Barclays, with its very tight timelines, or in Botswana, with its significant logistical challenges, what might be considered under duress. Located as we are in New York City (but seeing as well how conditions here reflect a pattern of crisis that is present now nearly everywhere), our thoughts have gone immediately to housing; in recent years, our city government has stated and restated the need for no fewer than 300,000 new affordable units to be in place by 2026. At the time of this writing, that is only eight years away. This enormous challenge, this provocation, must also serve a catalytic function in every corner of the combined AEC industries, radically and immediately recentering our modes of ideation and production; indeed, at the pace implied by that eight-year
deadline and, critically, the simple human urgency of the demand, we must revolutionize the way we work immediately. The time for the long-delayed disruption in architecture is now. In the context of current ground-conditions and the scope of the challenges we all face together, there is simply no other way to move forward. Let’s imagine now we have reached that future, and remember briefly the old modes of building. Remember the pollution and the waste. Remember the warring, territorial trades, the conflict and inefficiency of the job site. Remember – and one day we will do this, shaking our heads – how the hard-won data that might have been used to drive a project forward through all phases was in fact most often dumped midway through the process, as responsibility changed hands between designer and contractor. These things, which one day soon will be our shared bad memories, are the fruits of the linear process, a twentieth- and even nineteenthcentury way of working tragically misapplied to the resolution of twenty-first-century need, a process that must be phased out as rapidly now in our industry as it has in those industries that, recalling for a moment that damning McKinsey Report, have found a way to remain healthy and productive by welcoming change. Regarding the special problem of housing, and considering the societal factors as diverse as climate change, unemployment, and the effects of heavy machinery on neighborhood quality of life, the solution has to begin by allying our new, integrated methods of technology-enhanced collaboration with off-site construction. I prefer that term because of its inclusiveness, comprising as it does all forms of integrated component production, prefabrication, modular, and other hybrid methods and approaches, existing or yet-to-be-developed, that allow for the controlled and precise delivery of parts to efficiently form coherent wholes. A day may come when machines and systems are capable of real-time production in the field – coordinated, model-guided, AI-equipped printed outputs of the type SHoP once proposed in a short fiction several years ago when asked by the editors of Architectural Design (AD) to imagine the state of architecture in 20505 – but for now, to meet the challenges of today, work must
210
16 Sharples.indd 210
2018-10-18 10:27 PM
16.13 461 Dean Street in Brooklyn (2017), the tallest modular building in the world.
return to the workshop. I use that term because, built ground-up on a digital platform, these sites of industry can be conceived as fundamentally different than a traditional factory: robust, massively parallel, and nimble in all the ways we know our tools and techniques would prefer to have us work with them, if only we would listen. And the benefits of off-site construction, whatever the specific method deployed, are incomparable. The environmental and safety advantages are well documented; the ability to achieve economies of
scale is a given when those scales can at last be achieved. But there are other reasons to turn to offsite now. Chief among them, regarding the problem of building enormous amounts of new housing in already crowded cities, is the significant reduction of site disruption, pollution, traffic, noise, and waste. We saw this first-hand at the project we call B2, also known as 461 Dean Street (Brooklyn, 2017, figure 16.13), the tallest modular building in the world at the time of its completion. Located immediately behind the Barclays Center, it rises 211
16 Sharples.indd 211
2018-10-18 10:28 PM
from a very compact, wedge-shaped site at the edge of a very dense, and very self-protective, neighborhood. It was a marvel to see how clean and clear that site was, how little public ground it had to claim beyond its bounds, with so much of the traditional on-site work taking place in the factory not much more than a mile away at the Brooklyn Navy Yard. The goal at B2, driven by our client and by market forces, was to demonstrate that modular, combined with an advanced digital approach drawing from all our previous experience, could achieve the variety of spaces and external, expressive effects commonly achieved in traditional construction. It follows that there were a large number of module shapes, some triangular to fit the site, most often, in the varied plan, ganged to make generous units (up to three bedroom) with a proportion and a spatial quality that were novel to the type. The façade system, itself a first in its all-factory installation and its ability to support an expressive diversity, also brought the whole into conformity with expectations bred from traditional construction. Future modular buildings may not need to mask their nature to this degree – or, rather, may more frankly express the possibilities of space-planning that are available when a modular approach is married to the practices of mass customization – but B2 succeeds in demonstrating, at the dawn of its great age of need, that modular, with all its other advantages, can exhibit the same situational adaptability to context (including to the market) as conventional construction. And it can do so while, in its optimal application, serving as an arena for the expansion of a democratized way of building. Brought together in the factory, poring over the same model, the same virtual twins, developing them together, iteratively and responsively, what we think of now as independent trades could begin to merge, with the architects and engineers, into a single democratically conceived corps of construction. Those are the types of jobs, skilled and far less alienated, that could begin to fill New York City’s many underused high-bay spaces if and when the
city chooses to pursue a comprehensive modular plan to reverse its vast deficit in housing. The environmental advantages of modular are also enormously significant and begin with its relative lightness. Our engineer for B2, Arup’s David Farnsworth, calculated that its weight was only one-quarter what an equivalent traditionally-constructed building might weigh. And, of course, that lightness translates down the line to much-reduced carbon emissions in everything from raw materials extraction, to component manufacturing, to transportation. Even greater benefits can be achieved – not low atmospheric carbon addition but actual sequestration – in the burgeoning field of mass timber construction; a ten-story conventional building might release 600 tons of carbon compared to 1500 removed from circulation in a comparable timber structure. And there, too, though not cellular in the sense of modular delivery, the elements of construction are milled to such tolerances off-site that buildings are essentially erected, efficiently and with minimal site disruption, as if from a kit. As such, mass timber, like modular, lends itself supremely well to being the end result of an open, integrated, model-based process of design, fabrication, and assembly. And where modular, when planned, as it should, to limit delivery distances, lends itself very well to supporting employment in cities, mass timber brings with it a host of possibilities for economic opportunity in more rural areas, from forestry to milling. As many environmentalists are fond of pointing out, we have all the tools we need right now to solve all the problems we face. Because of the logic of its application to contemporary need, the multivalent nature of the solutions it offers, interest in off-site is exploding – among both private and public entities, both on the demand and supply sides, particularly in housing. The number of inquiries we have received in the last several years to explore modular projects alone, some at scales and scopes that would have once been
212
16 Sharples.indd 212
2018-10-18 10:28 PM
unimaginable – scales, not incidentally, that need to be achieved in order to benefit from the economies of the method – is a testament to its growing popular acceptance by developers and residents alike, and the success of many early efforts across the field. Our current work in this area includes a partnership with a legacy auto manufacturer to develop an efficient new approach to modular housing. This long-term project involves not only exploring the adoption of methods from that adjacent field of making, but collaborating in the creation of the new processes and work flows that will support the ultimate democratization that such a nimble process, the one we must all adopt, will make possible across the industry. CONCLUSION In The Stones of Venice, we find Ruskin writing at full throttle, criticizing how workers were, in his time, dehumanized. Once a master carpenter, perhaps, now a cog in a factory with a repetitive task, putting a head on a nail. In the past, he notes, that same worker might have enjoyed a great deal of creative autonomy: glancing at a shared model of what was to be made, climbing up the scaffolding, then reinterpreting that intention as a component of the beautiful edifice they were a direct part of collectively creating. It is a method of working that is not only much more generous with its access to pleasure in the work, but one as well that is open to benefitting from combined wisdom. This is also a method of working, Ruskin’s idealized view of labor in the early Renaissance, that may never have existed as described but is within our grasp, now, to achieve. So, what do we need to do, now, immediately, to conduct a top-to-bottom revision of one of the most historically inertial arenas of human endeavor? We may only need more examples, from everyone in the allied professions, and then the political will, after understanding them, to remove certain specific legal and regulatory barriers, such as the acceptance of the use
of digital models in governmental approval processes, or the adjustments to the fire code that will allow tall timber structures to play their promising role in mitigating climate disaster. We have the tools to build smarter and faster, in a way that can contribute to building a stronger society. We have, in spades, the driving need to do so. Much of the work to be done is in fact away from the keyboard and the Kuka robot – redesigning a host of human relationships. Between designer and builder, architect and client, ultimately between worker and product, buyer and home. None of our goals can be met – and the cost of failure is no less than dystopia – if the agents in this effort, broadly considered, remain in their silos, protected by defensive legal restrictions and rapidly elapsing social customs. What comes next must be organic and adaptable, networked and horizontal, strong and fast. The processes are all there, waiting to be developed and merged. Shared digital models, as at Barclays, to align collective intentions across the job from suppliers to installers. Automation deployed, as in our Milan project, not to eliminate but to elevate the craft worker. Natural, intuitive instruction sets, as at the Camera Obscura, can be used to lower barriers to entry for employment in construction work. An open architecture, like the one we are so close to achieving in Botswana, can play a role in ensuring that everyone has a hand in shaping their own future. In the act of scripting, coding a model, or simply using it as the universal communications tool it was born to be, you find yourself in a very creative state, a very human state. Let the machines put the heads on a nail. Our human co-workers should be busy mass customizing.
213
16 Sharples.indd 213
2018-10-18 10:28 PM
NOTES 1 Ruskin, John, The Stones of Venice, edited and abridged by J.G. Links, New York: Penguin Books, 2001. 2 See www.mckinsey.com/industries/capital-projectsand-infrastructure/our-insights/reinventing-constructionthrough-a-productivity-revolution. 3 See www.3ds.com/about-3ds/3dexperience-platform/. 4 A metal, pre-engineered building by Butler Manufacturing (see www.butlermfg.com/). They design and manufacture metal building systems for a variety of uses. 5 Nobel, Philip, “After Architects: A Vision of the Near-Future from SHoP,” in Chris Luebkeman (ed.), 2050: Designing Our Tomorrow (Architectural Design), London: John Wiley & Sons, Ltd, 2015, pp. 46–53.
214
16 Sharples.indd 214
2018-10-18 10:28 PM
17 SENTIENCE AND THE SPECIFICITIES OF CITIES
TOM VEREBES 215
17 Verebes.indd 215
2018-10-18 8:49 PM
Throughout the twentieth century, the regime of mass production had contributed substantively to the persistent legacy, and the accelerating proliferation of the ubiquitous uniformity of cities.1 In turn, the current context of urbanization, especially in East Asia, is arguably the most inexhaustible as well as the most globalized era of city building ever to occur. The pervasive narratives and instruments of industrial modern planning are not waning, rather, many of the mistakes of the twentieth century are being repeated in the twenty-first century. New paradigms, tools, and capacities with which to confront the challenges of the unparalleled speed and extent of urbanization are urgently needed to diverge from the default modes of producing standardized, repetitious architecture and urban design.2 At the frontier of the discourses and methods of mass customization lie the challenges of the expanding size of cities, and the problematic of their global proliferation, and the obsolescence of the methods and tools of urban planning.3 Despite the ways in which the design of cities tends toward uniformity in an increasingly globalized world, cities nevertheless persist to articulate their individuated character and identity. At the core of this chapter’s thesis on the current transition in industrial paradigm toward greater democratization and mass customization in design, is a set of speculations on the contemporary city, understood as the product of emerging technologies associated with current models of industrial production. Customization and design democratization, at the scale of the city, exceed discrete architecture, interiors, furniture and products, and speak to the consequences of computational design, manufacturing, and assembly for urbanism. This chapter elaborates these notions through three projects by OCEAN Consultancy Network (CN). The projects are within the disciplinary categories of urban design, landscape urbanism, and master planning. Projecting some of the historical and theoretical implications of information-based design and production tools, the design work of OCEAN CN will serve to explicate the ways in which spaces in cities can be designed to be unique, rather than over-familiar and repetitious. If the challenges of creating unique urbanities
within the transition to a new urban paradigm of distinctive urbanism turn out to be insurmountable, the risk will be to continue to build cities entirely void of specificity, identity, and character. Through unorthodox and unexpected applications of new technologies, the projects presented at the Mass Customization and Design Democratization symposium, and published in this book, are mobilizing this important transition to surpass the industrial paradigm of mass production toward increasingly unique systems, spaces, and experiences, deployed globally through mass customization, the aim of which is to create local specificities.4 REPERCUSSIONS OF THE CURRENT TRANSITION IN INDUSTRIAL PARADIGM FOR URBANISM The mass-customized cities issue of Architectural Design (AD), published in 2015, aimed to explicate the urban consequences of emerging technologies. It was indicative of a significant paradigm shift, beyond the scale of discrete one-off buildings, which continue to revolutionize the ways in which architecture can be conceived, practiced, and experienced.5 The term mass customization resonates with a hint of commercial crassness, but nevertheless, when applied in reference to cities, mass customization, in conjunction with its industrial, political, social, and cultural implications, challenges the legacy of spatial uniformity that has been the hallmark of standardized and mechanized production. At the core of this transition from Fordist to post-Fordist cities lie questions as to the ways in which cities and their qualities can be amplified and differentiated to become identifiable rather than indistinguishable.6 Although the current era of accelerated urbanization is clearly not the first era of mass urbanization to occur, evidenced by instant cities of the first Industrial Revolution, such as Manchester, Chicago, and more recently Shenzhen during its industrialization and rapid urban growth, urbanization is now occurring on a planetary scale, at unforeseen speeds. As articulated by Manuel Castells in The Rise of the Network Society: “The
216
17 Verebes.indd 216
2018-10-18 8:49 PM
information age is ushering in a new urban form, the informational city. Yet, as the industrial city was not a worldwide replica of Manchester, the emerging informational city will not copy Silicon Valley.”7 MASS-MANUFACTURED CITIES AND THE PROBLEM OF THE SAME Confronting the problem of the sameness of cities, the remedial proposition of a mass-customized city directly provokes the ideation of choice and identity for a mass urban populace.8 In recent decades, advances in computational design and fabrication technologies were propelled largely by the distributed democratization of the tools of nonstandard design. Well-rehearsed methods have taken root, and the product has been countless dazzling examples of highly specific architectural schemes, in practice, research, and teaching. The experimental culture of non-standard design is increasingly becoming mainstream, from its pioneering origins of the avant-garde. These new opportunities, experiments, and achievements in industrial machining, manufacturing, and assembly, have been limited to the scale of discrete buildings, façades, pavilions, interiors, furniture, and products but remain largely untested at the scale of urbanism.9 In the mass-customized cities issue of Architectural Design (AD), a series of questions were raised, “at the intersection of the ubiquitous uniformity of cities with the causal legacy of the 20th-century industrial paradigm of standardized mass production,” in an interdisciplinary survey of diverse approaches to the problematic of urbanism today.10 What are the opportunities for computational design and fabrication technologies at the urban scale? If today’s innovative production processes are largely limited to small-scale design projects, what are the repercussions for the vast, unprecedented scale of twenty-first-century cities? Emerging design and production technologies are facilitating this important shift from mass production toward increasingly customized systems in design arenas of architecture, interiors, furniture, and products.11
If modernism, and the modernist city, were the architectural and urban expression of the industrial paradigm of mass production, what are the consequences of non-standard customizing production methods for cities in this century? As noted by Karl Marx in Capital, “Modern industry had therefore itself to take in hand the machine, its characteristic instrument of production, and to construct machines by machines.”12 Again, with the advances of automation, and also the legacy of mass production, machines do construct machines. Today’s machinery is characterized by more lifelike qualities than the mechanical machines of the Industrial Revolution. Perception and processing, through machinery with capacities for sensing, feedback, and learning, are embedded in our personal computers.13 The emerging methods of contemporary nonstandard production have impacted OCEAN CN’s projects discussed in this chapter, which serve to elucidate some of the design consequences of this new paradigm for urbanism. The research agendas of OCEAN CN’s projects confront the difficult issues of the expression of identity in late capitalism.14 Given the diverse cultures surrounding computational design and production technologies, its champions holler about heterogeneity, and its critics insist homogeneous effects are to be blamed. In the past half century, a series of discursive ventures have valorized heterogeneity over the presumed homogeneity of modern abstract space, negotiating oppositions such as globalization and locality, positivism and nostalgia, and difference and sameness. The city has been the pre-eminent host and target for these debates. In the case of Hong Kong, it is the quintessential mass-manufactured city, whose pragmatic architecture is maximized through the spatialization of profit-seeking business plans, extruded as real-life three-dimensional Excel sheets rising from the plots of the city’s precious land. Despite the ruthless reliance upon standardized, monotonous architecture, Hong Kong’s urbanism is nevertheless distinctive, instantly recognizable and unique.15 Perhaps not as characterless as low-
217
17 Verebes.indd 217
2018-10-18 8:49 PM
17.1 User interface for the Parametric Pearl River Delta project.
density suburbia, the high-density urbanism of Hong Kong “raises a paradox of the uniqueness of the city versus the blandness of its architecture,” and the urgent need to find alternatives to repetitious standardization.16 In each of the projects explicated in this chapter, the notion of an interactive, intelligent urban model is developed as the basis of embedding intelligence into the processes of design in urban contexts at various scales. The consequence of feedback and learning is the capacity to produce deeply structured variance of highly specific design outcomes. The conceptual and technical basis of these professional and academic projects evolved from an exhibition project several years ago, Parametric Pearl River Delta, for the Hong Kong-Shenzhen Biennale in 2009–2010 (figure 17.1). OCEAN CN, in collaboration with Gao Yan (Crystal Design, Beijing) and Luis Fraguada (BAD, Barcelona), had designed the front-end interface for users of
the exhibition to interact with the back-end model, in Rhinoceros and Grasshopper, for visitors to the exhibition to control the state space of the urban model, via sliders adjusting the programs, densities, heights, and footprints, associated with ten cities in the Pearl River Delta, with the aim of users designing their own city, on screen. The resulting urban models are the result of scripted relationships to existing infrastructural and morphological patterns of existing urban typologies and densities. In a further, more formal elaboration of these methods, four seed buildings generated a series of massing configurations on a 500 x 500 m urban plot in Shenzhen, China (figure 17.2). The results of this interactive mode of producing variance from the reading and input of existing conditions arise through the variation of accumulation, proximity, and differential density. These nascent methods, initially proposed in the Parametric Pearl River Delta project, had been developed over a series of design proposals by OCEAN CN.
218
17 Verebes.indd 218
2018-10-18 8:49 PM
17.2 Different massing configurations for an urban plot in Shenzhen, China.
MASTERPLANNING, ADAPTIVITY AND SPECIFICITY In the book Masterplanning the Adaptive City,17 the masterplan, when conceived and represented as a single spatial image of a fixed and final future instance, formed the basis of a critique on the divorce between the disciplines and practices of urban planning and architecture. Cities are now comprehended as the spatial and temporal compilation of a plethora of unpredictable and indeterminable processes and interactions. Conventional masterplanning, conceptualized and documented in two dimensions at a fixed point in time, lacks the intelligence to respond to input, or create feedback, and is therefore limited in the capacity to embed intelligent capacities to learn, and facilitate greater sentience and responsiveness to contextual and user specificities.
When compared to the tools and representational media of a static masterplan, the intelligent urban model, and its serially differentiated outputs, are more flexible and adaptable to the contingent complex forces shaping cities. Most masterplans portray final representations, often limited to two-dimensional plans on a singular, or even ideal, outcome. Masterplanning, as a process, is still credible, yet static two-dimensional images are rarely built. Cities are never-ending. As observed by Mark Burry: “Cities always have cranes sticking out of them. There is no such thing as a completed city.”18 In the context of accelerated urbanization, the intelligence of alternative methods of conceiving, comprehending and designing urban order presents an opportunity to create evolving specificities, to confront the spatial and temporal limitations for the Fordist city. The methods of these new
219
17 Verebes.indd 219
2018-10-18 8:49 PM
practices embrace emergence, incompleteness, unpredictability, and indeterminacy, as ways of working through emergent patterns rather than through planning and prescription in a more evolutionary and adaptive model of urban development. This chapter addresses the legacies of architecture and urbanism’s infatuation with the future. The Futurists’ notions of mobility and plasticity were essential to their vision of the city of tomorrow.19 In The City of Tomorrow and Its Planning, translated from the French Urbanisme, Le Corbusier unveiled in 1925 his prototype for a city of 3 million inhabitants, La Ville Radieuse, which remains an emblematic moment of the expression and articulation of standardization and mechanization, and the erasure of the city’s historic fabric.20 In this case, Le Corbusier’s authoritative hand hovered over his scheme for Paris, characterized by a repetitious array of architectural building blocks of his prototype of twentieth-century urbanism. Architecture is the culprit in Le Corbusier’s urbanism. Inherited from early modernism, the core of this problem is the legacy of sameness and repetition, and the hot pursuit, in the past half-century of postmodernism, of urban heterogeneity and customization as a paradigm of our generation. In Rem Koolhaas’ scathing critique of Le Corbusier’s comparison of his Ville Radieuse with Manhattan, Le Corbusier brands Manhattan “an architectural accident,” and offers analogies of human deformities as evidence of his anti-Manhattan tirade.21 Le Corbusier remains to the current generation a pre-eminent kingpin of Cartesian rationalism and a consistent and reliable straw man for contemporary urbanists. The questions posed by Koolhaas’ generation remain regarding the ways in which the qualities of cities can be brought forward toward greater heterogeneity, to confront the proliferation of the ubiquity and sameness of cities.22 According to Antoine Picon, Koolhaas and his generation envisage “an ever-intensifying present, instead of a radically different future.”23 The systemic methodological variation and differentiation of today’s design approaches are essential tools in these transitions in industrial model.
FORDISM, TAYLORISM, INDUSTRIAL MODERNISM, AND THE PATH TOWARD INTELLIGENT URBANISM The longstanding association of Taylorism and Fordism to the standardization of the building industry during modernization was spearheaded by industrialization and the buildings of the industrial era, long before modernism. Taylorism, or the scientific management theories of the late nineteenth century, was assumed to be free of ideological conflict and class struggles, in favor of alternatives offered from familiar economic and political scenarios.24 The complexity of the range of work roles and tasks, and hence also the product of industrial production, were reduced.25 What became known as Fordism, was named after Henry Ford’s application of the assembly line in the production factories of the Model T car, which was appropriated from the automation of food production in the nineteenth century, toward the linear segmentation of tasks of the mechanized factory.26 Fordism accomplished the transition from the batch production workshop of the nineteenth century to the industrial model of the late nineteenth- and early twentieth-century assembly line, with new capacities to produce repeated components and products, synergetic with the legacies of industrial urban modernism.27 The shift from the mechanized Fordist factory, which had taken command with the assembly line, has progressed toward our contemporary conception of automation. The automobile industry has been, with some irony, one of the early adopters of the third Industrial Revolution. Different from the relatively limited variation of the automobile industry offers, how much variety is required in urban environments to cross a threshold to a substantive difference? Still, a hangover throbs from the legacies of standardization, evidenced from Levittown, on the one hand, as the emblematic prototype for cookie-cutter reproduction of sameness in sprawling suburbia, or in another extreme of urban density, the unrelenting repetitious forms and façades of the mass housing of Hong Kong, where Fordism lives on. Despite “flexible specialization” having coexisted alongside mass production since the nineteenth century, the continued dependency of modernist cities upon Fordist repetitious production
220
17 Verebes.indd 220
2018-10-18 8:49 PM
is challenged today by the accelerating culture spurring innovation, through experimentation with emerging production technologies.28 During the transition in industrial paradigm on the urban scale, any consideration of mass customization as a research paradigm begins with the precondition of the limitations of urban masterplanning to unleash algorithmically differentiated spaces. Ultimately, the potential to install greater intelligence and sentience into how cities are organized and how they behave has three goals: (1) the capacity to harness complexity with our new data-driven tools; (2) the capacity to yield organizational specificity and heterogeneity; and (3) the capacity for adaptivity and evolution in time. Given the backdrop of generic modernism and its legacies, a future urbanism is arising out of artificial intelligence and machine learning. For the generation of students and emerging architects of the 1980s, the demi-mondes of postmodernism and deconstruction were tendencies that were part and parcel of a turn towards postFordism. Early postmodernism represented an industrial, economic, and cultural move toward fragmentation and disaggregation.29 The first expression of post-Fordist architecture had vehemently resisted neo-traditionalism. Meanwhile, the post-Fordist cultural theorists of the city in the 1980s and 1990s made advances in theorizing the cultural effects of heterotopia, yet the discipline and profession of urban planning were unable to manage the challenges of the failures of its own doing. How has this shared valorization of heterogeneity changed over the past decade of the continued postmodern era? Architecture, when proliferated as indistinguishable and featureless facsimiles, represents a longstanding quandary in urbanism.30 Cities without substantial design criteria were aptly qualified in Archizoom’s No Stop City project in 1968, expressing “a city without qualities for a man finally without qualities.”31 Mass-produced generic urbanism, devoid of specificities, was expressed by Archizoom with a mechanical typewriter. Contemporary design culture is in transition from standardized modes of producing architecture and urbanism, in which the uniform, universalized
and repetitious spaces are byproducts. Convergence, or the “accidental” homogenization of the contemporary city through the generic spaces of airports and malls, strips and renders meaningless identity and character.32 The “supermodernity” of “non-places” turns the exotic and the familiarity of the local into spectacle, just as it does to history.33 Architecture, along with its consumable aesthetics, has been used to output countless cities void of history.34 As noted by Saskia Sassen in The Global City, “Being in a City becomes synonymous with being in an extremely intense and dense information loop.”35 The consequences of numerically controlled prototyping practices at the scale and complexity of urbanism conflict with the ubiquity of the uniformity that has been inherited from history. Customization to local forces, agents, and actors resonates with Kenneth Frampton’s formulation of critical regionalism, which shunned the “universalizing” forces of technology.36 Critical regionalism was closely associated with, and thus bolstered, other postmodern agendas, including neo-traditionalism, vernacularism, and other nostalgic forays expressing anxieties over the loss of local identity and continuity at the hands of globalization. When scaled up to urbanism, the parametrics inherent in Frampton’s position about creating regional identity, in which forces such as local topographical and climatic conditions, material techniques and systems, and other inherently specific constraints, channel design formulations toward the distinctiveness of each particular city. Shunning the “deliberate regression” into the deepening of history, Fredric Jameson repudiated critical regionalism of its marginality.37 More recently, Frampton has asserted that in the practices of landscape urbanism infrastructural systems are organized to behave as associating to ecological and social systems.38 The design approaches of mass customization, when applied to the city, challenge the authority of the designer, and bias local inputs in bottom-up democratic, participatory processes. In an example of the scaling-up of architecture to the city, E-Grow, a next-generation fabricator based in Shanghai, has a hallmark set of material production methods, combining low-tech, manual 221
17 Verebes.indd 221
2018-10-18 8:49 PM
17.3 An aerial view of the competition proposal.
craft, with advanced fabrication technology, to produce some of the most compelling buildings being built in China. Indicative of the shift from mass production to more articulated and unique architectural outcomes, China had become known as the world’s factory. This migration from mechanical to informational production is characterized by both gradual change as well as sudden leaps.39 The prevailing context of rapid deployment of mass-produced cities and buildings is happening at an unprecedented and alarming extent and speed. An opportunity to differentiate China’s new cities depends upon the enabling potential of intelligent design to infiltrate the hegemony of urban planning upon the large brushstrokes of infrastructural planning, land subdivision and supply, and zoning.
Umekita Second Development Area Osaka, Japan (2013–2015) OCEAN CN & Arup Hong Kong In this invited competition which had sought innovative urban design ideas for a 7-hectare site in Osaka, Japan, OCEAN CN developed the rules and protocols of a back-end 3D model environment, with capacities to process input urban information, in order to produce potentially a multitude of possible projects. Each outcome of this process is understood as the result of interactive operations yielding unformed variations, rather than any of the particular variegated outcomes having the status of the form of any singular proposal. For the
competition proposal, OCEAN CN presented a single image of an instance of the process (figure 17.3), in aerial renderings and a plan, to at once demonstrate clarity, rather than the misinterpretation of this method as indecisiveness. The back-end model, as a way of working with complex yet legible patterns as the graphic space of the project, manages variable inputs using sliders and numerical inputs, toward differentiated configurations and expression. These methods help to develop a manifold of possible future scenarios. Each of the six zoning scenarios, shown in figure 17.4, have a different mix of programming and relations between public and private landscapes, and open spaces. Each of the six scenarios play out different ratios of programs (highrise and low-rise housing, hotels, commercial, etc.), typologies, densities, and footprints, customizing a set of divergent configurations for the development of the site. These models are not different schematic options, but instead, they are potentially different ways of interacting with the dense character of the urban site. Scripted models, by their nature of being formed through diverse influences, are generated in ways to address multiple performance criteria. The models are constrained as definitive typologies associated with different programs. On the east side of the site is the first stage of the Umekita development, comprised of very large deep plan and deep section buildings. To the west of the site is a smaller, more granular traditional scale of Osaka. The series of plans (figure 17.5) are cut at the ground level, and the heterogeneous mixes of the program are indicated through color-coding.
222
17 Verebes.indd 222
2018-10-18 8:49 PM
17.4 Six different zoning scenarios for Umekita. 17.5 Six different zoning plans for Umekita.
223
17 Verebes.indd 223
2018-10-18 8:50 PM
17.6 Site diagram.
In further stages of modeling, these six different controlling scenarios were advanced in the computational development of back-end Rhinoceros 3D model environment. These are automated design systems – creating agency through the negotiation of a set of charged parameters (figure 17.6), such as the proximity of access points from roads, the underground metro system and trains; heights and positions of buildings from one side of the site to the other; the pattern of infrastructural routes and orientations, all of which contribute to the results of a series of models issuing from these and other associations. Each of these models defers from proposing conclusive designed outcomes, but alternatively, the design effort is located in the relationships of the massing components to the context through a more aggregate approach to associations within a
model. Speculating on the potential applications of this modeling approach, one possibility is the management of the development process of a large urban project, between diverse actors and agents in time. The heightening of interactivity facilitated by these modeling environments can enable the control systems of the model to become more inclusive of various constituencies. Designers, the client body, planning authorities and other regulatory and statutory bodies, citizens, and other parties, can potentially engage in a more participatory mode, with an expanded field of contingencies and constraints, as a more intelligent planning and design tool. This is an argument for the potential democratization of increasingly automated design processes, through the integration of a more inclusive group of actors and agents into the design process.
224
17 Verebes.indd 224
2018-10-18 8:50 PM
Yan Jiao Hua Run 4D City Masterplan Hebei, China (2011–2012) OCEAN CN & dot.A
17.7 A computable density map, featuring building types, heights, and footprints.
In this commission for a 25 km2 site in Hebei Province, located 50 km south of Beijing, the client requirements were to simply allocate an infrastructural and land subdivision grid of mega-blocks, with dimensions between 160 m and 220 m squared, and to deploy a stipulated set of programs in a representational massing configuration. OCEAN CN’s response was to create a particular set of conditions, generated from the specific existing features of the site, including a river frontage, a canal bisecting the site, and existing roads, which resisted a uniform grid of streets. From these existing disturbances to the uniformity of a megablock grid, OCEAN CN initially generated a series of variations of urban grids. Applying computer programming to architectural and urban programming, the modeling environment was charged such that the distribution of the range of programs enables alternatives to single-use zoning, and simplistic two-dimensional mixes of zoning. An atomized pallet of programmatic composites was scripted, the range of gradients of a color mapping of intensities and mixes of uses, and consolidated as an abstract, yet computable density map, which was read in terms of building types, heights, and footprints (figure 17.7). Three outcomes were expediently produced for the client (figure 17.8). The project was generated through the specific character and distribution of open spaces. In developing the project, we studied the concentration and dispersal of landscape spaces, between zero and up to higher degrees and qualities of consolidation and diffusion, more linear and hard-edged, or more scattered and porous. Within these different conditions of landscape, six schemes were selected to test their variegated consequences in a more detailed cluster of a few of these mega blocks. Each of these massing schemes describes low-, mediumand high-density urbanism on a hypothetical block (figure 17.9), through relatively simple extrusions of aggregate massing volumes. A series of massing configurations (figure 17.10) are generated by simple extrusion techniques to create an architectural expression, which is albeit relatively generic and normative, yet the urbanism arising is highly specific and customized. 225
17 Verebes.indd 225
2018-10-18 8:50 PM
17.8 Three possible outcomes.
226
17 Verebes.indd 226
2018-10-18 8:50 PM
17.9 Massing schemes of low-, medium- and highdensity urbanism on a hypothetical block.
17.10 Possible massing configurations.
227
17 Verebes.indd 227
2018-10-18 8:50 PM
17.11 Xiangmi Park plan.
Xiangmi Park, Shenzhen (2014) OCEAN CN & SED Landscape Architects In a landscape urbanism project proposed in collaboration with SED Landscape Architects, this 45-hectare park in Shenzhen occupies a city block of approximately 1 km north-south, and 450 m eastwest, comprising an existing dilapidated park. OCEAN CN’s proposal is for a string of continuous groundscraper buildings lining the edge of the site along the road frontage, framing the interiority of the site and the systems and spaces of the park (figure 17.11). Sub-divisional patterns are the basis for the generative form-finding of abstract organizational geometries, patterns, and material textures, as the
distribution of varied species of planting, routing systems, and hard landscaping (figure 17.12). The various systems of the project organize the required 50,000 m2 of building gross floor area, and how one walks inside, underneath, and on top of the buildings. An existing hill is the focus of the routing systems to negotiate the topography between the edges of the side into the interior and the hilltop. CONCLUSION The series of projects presented in this chapter have served to explicate the design consequences for urbanism of the transitions in urban paradigm toward greater sentience, leading to enhanced
228
17 Verebes.indd 228
2018-10-18 8:50 PM
17.12 The diagram array.
organizational specificity. Given the diverse cultures and stances surrounding, within, and against technology, mass-customized urbanism continues critical regionalism’s confrontation with the issue of the expression of identity in late capitalism. These discourses negotiate oppositions such as globalization and locality; positivism and nostalgia; and heterogeneity and homogeneity. Substantial advances have occurred in the capacities of design technologies applied to the making of architecture and cities. Since the Industrial Revolution, and especially in recent decades during the so-called information revolution, the unprecedented use and application of information-based technologies are overhauling the ways in which designers of
urban environments, source and apply data as a basis for differentiated spaces. In a call for the specificity of cities, the implication of the new paradigm of nonstandard variable production for the twenty-first-century city has yet to be fully embraced, but in the future, it may be articulated as distinctive urbanism.40 Concerning the democratization of design, the accelerated power of computational modeling environments will be further distributed to citizens and users, commissioners and investors, politicians, and not least, for designers to be the drivers of these particular participatory processes toward more inclusive bottom-up consumer products, architectural spaces and cities. 229
17 Verebes.indd 229
2018-10-18 8:50 PM
NOTES 1 Verebes, Tom, “Introduction,” in Tom Verebes (ed.), Mass-Customised Cities (Architectural Design, AD), London: John Wiley & Sons, Ltd, 2015, p. 8. 2 Verebes, Tom, “Technological Transitions, Industrial Innovations and the Marching Chinese Urban Revolution: An Interview with Jerry Ku and Philip Vernon, of E-Grow, Shanghai,” in Tom Verebes, Mass-Customised Cities, London: John Wiley & Sons, Ltd, 2015, p. 115. 3 Kuhn, Thomas, “Scientific Revolutions,” in Richard Boyd, Philip Gasper, and J.D. Trout (eds.), The Philosophy of Science, Cambridge, MA: MIT Press, 1994, p. 78. 4 Verebes, “Technological Transitions,” op. cit., p. 117. 5 Verebes, “Introduction,” op. cit., p. 9. 6 Ibid. 7 Castells, Manuel, The Rise of the Network Society, New York: Wiley-Blackwell, 1996, p. 429. 8 Verebes, Tom, “Towards a Distinctive Urbanism: An Interview with Kenneth Frampton,” in Tom Verebes (ed.), Mass-Customised Cities, London: John Wiley & Sons, Ltd, p. 27. 9 Verebes, “Introduction,” op. cit., p. 9. 10 Ibid. 11 Verebes, “Introduction,” op. cit., p. 10. 12 Marx, Karl, Capital, vol. I, Moscow: Progress Publishers, 1887, p. 384. 13 De Landa, Manuel, War in the Age of Intelligent Machines, New York: Zone, 1991, p. 228. 14 Jameson, Fredric, Postmodernism, Or the Cultural Logic of Late Capitalism, London: Verso, 1991, p. 2. 15 Verebes, Tom, “Hong Kong: Appearing Dense but Growing Smarter,” in Philip Joo Hwa Bay and Steffen Lehmann (eds.), Growing Compact: Urban Form, Density, Sustainability, New York: Routledge, 2017, pp. 253–270. 16 Ibid, p. 262. 17 Verebes, Tom, Masterplanning the Adaptive City: Computational Urbanism in the Twenty-first Century, New York: Routledge, 2013. 18 Bates, Donald, “Permanence and Change: An Interview with Mark Burry,” in Tom Verebes (ed.), MassCustomised Cities, London: John Wiley & Sons, Ltd, 2015, p. 82. 19 Banham, Reyner, “Italy: Futurist Manifestoes and Projects,” in Reyner Banham, Theory and Design in the First Machine Age, Cambridge, MA: MIT Press, 1980 (first published in 1960), pp. 99–137. 20 Le Corbusier, Urbanisme, Paris: Les éditions G. Crès, 1925. 21 Koolhaas, Rem, “La Ville Radieuse,” in Carlo Palazzolo and Riccardo Vio (eds.), In the Footsteps of Le Corbusier, New York: Rizzoli, 1991, p. 165. 22 Verebes, “Towards a Distinctive Urbanism,” op. cit., p. 27. 23 Picon, Antoine, “Learning from Utopia: Contemporary Architecture and the Quest for Political and Social Relevance,” Journal of Architectural Education (JAE), vol. 67, no. 1 (March 2013), p. 22.
24 McLeod, Mary, “Architecture or Revolution: Taylorism, Technocracy, and Social Change,” Art Journal, vol. 43, no. 2 (Summer 1983), pp. 133–134. 25 Gartman, David, “Introduction,” in David Gartman, From Autos to Architecture: Fordism and Architectural Aesthetics in the Twentieth Century, New York: Princeton, 2009, p. 39. 26 Verebes, “Introduction,” op. cit., p. 12. 27 Jones, Bryn, “Past Production Paradigms: The Workshop, Taylorism and Fordism,” in Bryn Jones, Forcing the Factory of the Future: Cybernation and Societal Institutions, Cambridge: Cambridge University Press, 1997, pp. 23–50. 28 Amin, Ash, “Post-Fordism: Models, Fantasies and Phantoms of Transition,” in Ash Amin (ed.), Post-Fordism: A Reader, Oxford: Blackwell, 1994, p. 14. 29 Harvey, David, and Scott, Allen J., “The Practice of Human Geography: Theory and Empirical Specificity in the Transition from Fordism to Flexible Accumulation,” in W.D. Macmillan (ed.), Remodelling Geography, Oxford: Blackwell, 1988, pp. 217–229. 30 Verebes, “Introduction,” op. cit., p. 10. 31 Branzi, Andrea, Weak and Diffuse Modernity: The World of Projects at the Beginning of the 21st Century, Milan: Skira, 2006, p. 71. 32 Koolhaas, Rem, “The Generic City,” in Rem Koolhaas and Bruce Mau, SMLXL, New York: Monacelli Press, 1995, p. 1248. 33 Augé, Marc, “From Places to Non-Places,” in Non-Places: Introduction to an Anthropology to Supermodernity, London: Verso, 1995, pp. 110–111. 34 Mars, Neville, “Cities without History,” in Neville Mars and Adrian Hornsby (eds.), The Chinese Dream: A Society Under Construction, Rotterdam: 010, 2008, pp. 520–531. 35 Sassen, Saskia, “The Global City: An Introduction to a Concept and its History,” in Rem Koolhaas, OMA, and the Harvard Project on the City, Mutations, Barcelona: Actar, 2000, p. 106. 36 Frampton, Kenneth, “Towards a Critical Regionalism: Six Points for an Architecture of Resistance,” in Hal Foster (ed.), The Anti-Aesthetic: Essays on Postmodern Culture, Seattle: Bay Press, 1983, pp. 16–30. See also Frampton, Kenneth, “Critical Regionalism: Modern Architecture and Cultural Identity,” in Kenneth Frampton, Modern Architecture: A Critical History, London: Thames and Hudson, 1985, 2nd edn, pp. 313–327. 37 Jameson, Fredric, “The Constraints of Postmodernism,” in Neil Leach (ed.), Rethinking Architecture: A Reader in Cultural Theory, New York: Routledge, 1991, p. 248. 38 Waldheim, Charles, “Landscape as Urbanism,” in Charles Waldheim, The Landscape Urbanism Reader, New York: Princeton Architectural Press, 2006, pp. 35–53. 39 Amin, op. cit., p. 3. 40 Verebes, “Towards a Distinctive Urbanism,” op. cit., p. 30.
230
17 Verebes.indd 230
2018-10-18 8:50 PM
18 CITY SCIENCE: TOWARD A NEW PROCESS FOR CREATING HIGHPERFORMANCE ENTREPRENEURIAL COMMUNITIES
KENT LARSON 231
18 Larson.indd 231
2018-10-18 10:32 PM
18.1 Shikumen “lane houses” in Shanghai, China.
18.2 Socialist-style housing blocks, Shanghai, China.
We live in a period of extreme urbanization. While cities account for only 2 percent of the landmass of the planet, they will accommodate 90 percent of future population growth and create most of the wealth and opportunities. Cities also produce more than 70 percent of global CO2 emissions and consume 65 percent of energy production. Cities are where the great societal challenges are concentrated and where the most effective solutions will be deployed. Today, the center of gravity of political and economic power is rapidly shifting to urban areas, as state and federal governments become increasingly paralyzed and dysfunctional. Cities have become the primary incubators of the technological, cultural, social, political, and design innovations that will shape our planet, and mayors are taking the lead in finding solutions to transnational problems, such as climate change, human rights, public health, and employment. But cities are evolving in profoundly different ways. In the United States, talent, venture funding, and high-paying jobs are increasingly concentrated in thriving “innovation cities” on the east and west coasts, as smaller cities and rural areas in the heartland are left behind. These forces are also making the most successful cities unaffordable for
young professionals, who are their lifeblood. In Europe, the Brexit vote in Great Britain and the surge of support for the right-wing party in France reveal a backlash against the cultural and economic capitals of London and Paris. It is clear that new ways must be found not only to make our great cities more equitable and but also to make shrinking cities more innovative and attractive to young people. Societies in China, Africa, Latin America, and India are being transformed by a mass migration from rural to urban areas as people seek job opportunities and better lives, and an estimated 2.5 billion people will be added to urban areas by 2050. The low quality of life found in many of these places indicates that a far more effective process is required to create future cities. We propose that a key solution to the challenges of urbanization, applicable to all cities, whether small or large, rich or poor, will be the building of strong communities that minimize resource consumption and maximize creative interactions and quality of life. Achieving this goal will require a wide range of urban innovations, plus a new and more effective process that goes well beyond the conventional approach to urban planning and public policy.
232
18 Larson.indd 232
2018-10-18 10:32 PM
18.3 China’s rapid urbanization from compact autonomous communities to auto-dependent superblocks as seen from high.
TOWARD HIGH-PERFORMANCE, ENTREPRENEURIAL COMMUNITIES For thousands of years, people lived in compact, autonomous communities where almost everything needed for daily life was available in close proximity to home. Cities evolved into a network of neighborhoods, as can be seen in the 20 arrondissements of Paris, or Manhattan’s Greenwich Village, East Village, etc. In the mid-twentieth century, inexpensive automobiles, massive government highway programs, global supply chains, and Euclidean zoning led to single-purpose segregated zones made accessible only by the automobile. Cities such as Atlanta and Los Angeles in the US experienced low-density suburban sprawl, while Chinese cities such as Beijing and Shanghai saw higher-density tower sprawl. In China, attempts are now being made to reverse this trend by privileging complexity, density, and social interactions, as found in traditional neighborhoods. In the United States, many mayors are exploring ways to create active urban “innovation districts.”
Urban Communities in China The spectacular view from the 118th floor of Shanghai Tower shows the various phases in the development of the city’s complex urban fabric. Below are the traditional Shikumen “lane houses,” with their vibrant laneways, that emerged in the late nineteenth and early twentieth century (figure 18.1). In other districts are the relentless four- to six-story Socialist-style housing districts built during the 1950s to the 1980s (figure 18.2). On the horizon are the copy-and-paste high-rise buildings of endless superblocks, developed over the past 20 years, where many of Shanghai’s new residents now live (figure 18.3). As is the case with most cities, older communities that naturally evolved before the adoption of private automobiles provide far better access to shopping, schools, jobs, healthcare, and other amenities that build strong communities. The newer developments, particularly the vast superblock projects, isolate residents in single-function ghettos connected to services primarily by automobile via increasingly clogged highways. This dysfunctional development model, reinforced by antiquated land use regulations, is finally being challenged in China. 233
18 Larson.indd 233
2018-10-18 10:32 PM
In February 2016, China’s State Council established a new set of urban development guidelines that threw out many of the development conventions used during the two-decade period of extreme urbanization. These urban planning recommendations are intended to privilege walking and public transit over car use, preserve historical and cultural complexity, encourage mixed live-work developments, and promote the efficient use of natural resources. Newer urban planning proposals coming out of China are beginning to reflect these values. In addition, Chinese President Xi Jinping has described his vision of environmentally friendly cities that are “a good place to live” with an economy supported by innovation and “top-notch public services.” In January 2018, he called for the development of big data analytics and artificial intelligence to propel the country’s long-term growth, including an emphasis on the creation of “smart cities.” This indicates support for a new approach to urbanization that would bring together community building empowered by the use of new technology.
Kendall Square in Cambridge, MA, is considered by many to be the world’s best model of an urban innovation district. Anchored by MIT, the area is home to many corporate research labs including Google, Microsoft, Biogen, Genzyme, and countless start-ups. The area is well served by subway, buses, shuttles, shared bicycles, and ride-sharing. But Kendall Square is a poorly functioning community as the area lacks a grocery store, pharmacy, public school, clothing store, commercial day-care facility, diverse restaurants, and many other amenities necessary for daily life. With an extremely low residential density, most people who work in Kendall Square travel from other areas each day, and as a result, there is very little nightlife or weekend activity. The lack of social interaction opportunities and time lost to commuting lower the innovation potential of the district. Our conversations with mayors indicate the need to bring together design, technology, and public policy innovation to create more effective, vibrant, urban innovation communities.
Urban Innovation Districts in the United States
TOWARD A NEW PROCESS FOR CREATING CITIES
The world’s first university research park, near Stanford University, played a key role in the evolution of Silicon Valley. Since then, many “research parks” have been created to increase the concentration of high-tech companies in strategically planned, spatially isolated work environments to promote innovation and develop commercial products. Research parks were typically located in suburban corridors, accessible only by car, and conceived as 9–5 weekday, work-only environments. Over the past ten years, in response to the desire of many educated professionals to live and work in cities, there has been a rapid shift away from research parks toward urban innovation districts where leading-edge anchor institutions and companies cluster and connect with startups, business incubators, and accelerators. Studies have shown that over 90 percent of innovation, as measured by patent applications and R&D investment, now takes place in cities.
Contemporary government interventions, such as tax laws or zoning restrictions, are not sufficiently agile to respond to changing social and economic conditions. Market forces, even with government incentives, will rarely result in the diversity and complexity needed by healthy communities since most property developers naturally focus on the highest and best use of land from a financial rather than a community perspective. “Smart city” technology is typically limited to the optimization of existing systems. We believe that the most successful cities in the future will consist of a network of healthy, high-functioning communities empowered by highly responsive services for residents and new systems to produce energy, food, and clean water near the point of consumption, and minimize the need for mechanized mobility. To achieve this vision, a new integrated process is proposed that brings together agile urban design methodologies, emerging
234
18 Larson.indd 234
2018-10-18 10:32 PM
18.4 Traffic congestion was estimated for every hour and road by combining telecoms data with a small sample of vehicle traffic counts; displayed on Andorra CityScope model.
technologies, public policy, and inclusive decisionmaking. We propose four key components of an evidence-based, data-driven process for cities in the future: insight, transformation, prediction, and consensus.
from Andorra is used to understand behavior at the scale of the country. Hundreds of specialized sensors called TerMITes2 are deployed to understand complex meal-preparation behavior at the scale of the urban apartment kitchen.
INSIGHT (UNDERSTANDING CURRENT CONDITIONS)
Insight Case Study 1: Andorra Telecom Data Analysis (Country Scale)
Cities around the world are embracing a policy of “open data” and making anonymized government datasets accessible online for use by researchers, urban planners, entrepreneurs, and citizens. Most commonly used is GIS (geographic information system) data that allow the manipulation and analysis of a variety of spatial and geographical data. Increasingly, government census, transport, and surveillance data are also used to better understand how cities function. It is difficult, however, to obtain fine-grained information related to complex human behavior. Where do people of various demographic profiles live, work, shop, play, recreate, and interact socially and professionally? When and why do they make mobility mode choices? How do urban design features and the availability of amenities in the city impact human behavior? How do urban interventions such as new bike lanes or ride-share systems alter commuting decisions? Mobile phone and social media data offer the unprecedented ability to obtain answers to these questions. Below, we summarize two projects by the MIT City Science Initiative to explore new ways to understand complex human behavior. Geolocated Radio Network Controller (RNC) telecom data1
One of the goals of MIT Media Lab’s City Science group3 collaboration with Andorra was to develop tools to better understand the mobility patterns in the country, with a particular focus on the experiences of tourists. The inherently irregular and seasonal travel demands of tourists cannot be captured using typical approaches such as household surveys, creating a need to develop a transportation model for Andorra which could capture the temporal variation in travel patterns. Telecom data provide a powerful source of information about individual mobility patterns and preferences on a large scale and naturally capture the time-varying nature of these patterns. However, these data also suffer from some limitations: the spatial resolution is generally not high enough to place individuals at specific roads or buildings and the devices observed represent only a subset of the total population of interest. Vehicle traffic counts, on the other hand, measure traffic volumes at near-perfect accuracies but are generally only available at a small number of junctions in the road network. By using a suitable technique – a Mathematical Program with Equilibrium Constraints (MPEC)4 – the strengths of these two data sources were combined in order to provide an accurate time-varying model of mobility patterns in the whole country (figure 18.4). 235
18 Larson.indd 235
2018-10-18 10:32 PM
18.5 RNC telecoms data were processed to identify those areas where dense clusters of human activity tended to occur. Multivariate linear regression models were used to find the relationship between cluster occurrence and urban features.
236
18 Larson.indd 236
2018-10-18 10:32 PM
18.6 TerMITe sensor boards are designed for extensibility and ease of data sharing. An Augmented Reality tablet displays localized TerMITe readings.
A complementary line of research was concerned with understanding the way citizens and tourists in Andorra make use of public space. Specifically, how do the features of an urban area – including amenities, natural landscape, built environment, etc. – influence the propensity for large, diverse, and persistent clusters of human activity to occur? The occurrence of such clusters can be considered as a proxy for vibrancy or attractiveness of the area, so an understanding of this relationship can help us to design successful urban places in an evidence-based way. In the first step of this analysis, the RNC data were filtered for stay-points – events where an individual stayed in roughly the same area for at least 20 minutes – and then these stay-points were further analyzed to find clusters of activity characterized in terms of their size, persistence, and diversity (figure 18.5). Multivariate linear regression models were then used in order to identify associations between the formation of these clusters and various urban features. Some of the urban features that were found to be important were the presence of service and entertainment amenities, natural water features, and the betweenness centrality of the road network. Others, such as educational and park amenities were shown to have negative impact on vibrancy. Ultimately, this approach suggests a “reversed urbanism” methodology: an evidence-based approach to urban design, planning and decision-making, in which human behavioral patterns are instilled as a foundational design tool. A major part of City Science collaboration agreements is dedicated to sharing and collecting unique datasets such as these. Chinese, European, and
South American telecom companies as well as global tech companies are part of the City Science effort to enhance the understanding of urban environments through the collection, aggregation, and analysis of data.
Insight Case Study 2: TerMITes for Data Analysis at the Home Scale In cases where existing data sources do not capture the range of information required for insight generation, City Science researchers have also developed new sensor technologies that can easily be deployed for site-specific data collection. The latest generation of MIT Environmental Sensors – TerMITes – have been developed for low-cost, rapid installation in locations where permanent sensor infrastructure is unavailable (figure 18.6). The TerMITes provide information about how spaces are used, such as when and where people are spending their time, as well as data that help characterize the ambient environment, such as lighting levels, temperature fluctuations, and relative humidity. The TerMITes were designed to make largescale deployment possible for non-expert users, and consequently require no special infrastructure. Rather, they connect to a standard WiFi network and automatically upload data to Cloud-based storage. Recorded data are immediately available for analysis or visualization using tools and applications developed by the City Science research lab. One such application is RePlace, a 3D data visualization platform that takes the normally invisible activity of sensors and data loggers and displays that information as 3D sprites over an 237
18 Larson.indd 237
2018-10-18 10:32 PM
18.7a and b RePlace visualizations of a residential apartment showing a histogram of sensor activation indicated by size of spheres, and a close-up of a kitchen activity sequence during a 30-minute period of meal preparation.
architecturally accurate model of the building interior. RePlace was used to help a major manufacturer of home furnishings understand how customers interacted with product prototypes in an apartment instrumented with TerMITes. RePlace provided reconstructed sequences of object interactions to allow researchers to visualize behavior as if watching the traces of an invisible person touching items in the home (figures 18.7a, 18.7b). Such a tool can promote high-level discussions about home or
workplace activities without the need for video recording that would compromise privacy. Thousands of researchers in academia and industry are investigating the use of “big data” analytics to better understand how cities currently operate. Complementary to this work, small-scale “pilot” interventions using technologies like the TerMITes can begin to test alternative scenarios with measurable impact on the lives and wellbeing of individuals.
238
18 Larson.indd 238
2018-10-18 10:32 PM
18.8 The PEV autonomously roaming the bike lines near MIT campus.
TRANSFORMATION (URBAN INTERVENTIONS TO IMPROVE CONDITIONS)
Transformation Case Study 1: PEV (Persuasive Electric Vehicle)
Over the next 10–20 years, cities will undergo rapid change as new systems are integrated into the urban fabric, and cities replace heavy infrastructure with more flexible and cost-effective distributed systems. It is becoming increasingly possible to deploy distributed energy production, combined with low-energy buildings and systems, to enable net-zero energy communities. Highperformance food systems, such as aeroponics and hydroponics, are allowing healthy food to be produced year-round near the place of consumption. Shared-use autonomous shuttles and ultra-lightweight autonomous personal vehicles offer the possibility of more convenient and efficient last-mile mobility. Cheap and robust nanofiltration technology will likely make possible the local purification of drinking water. Below, we summarize two projects by the MIT City Science Initiative to explore new approaches to mobility and housing that may provide the benefits of urban density (vibrancy, safety, equity) without the negatives often associated with density (congestion, loss of light, loss of access to nature, etc.).
There has been dramatic disruption within urban mobility over the last few years. First, the availability of shared bikes and safer bike-lane infrastructure has led to significant increase in urban cycling for commuting and pleasure, just as many cities have implemented policies that discourage the use of private automobiles. Second, the popularity of ride-share services such as Uber, Lyft, and Didi China have convinced many urban dwellers that alternatives to private automobiles are convenient and affordable. There is little doubt that autonomous vehicles, particularly robo-taxis and autonomous ondemand shuttles, will create a third disruption. Researchers at the MIT City Science Initiative have prototyped the PEV, a lightweight, autonomous three-wheeled vehicle (figure 18.8) designed to combine the best features of bike-share, ride-share, and autonomy. This three-wheeled, bike-like, ultralightweight, shared-use, single-person vehicle can operate legally in a bike lane. The passenger calls for it with an app in a similar manner to calling for an Uber car, and the PEV comes to the required location autonomously. The passenger then rides the vehicle to the destination, with the human energy input tuned to their personal exercise preferences,
239
18 Larson.indd 239
2018-10-18 10:32 PM
18.9a–c Collision avoidance in action: Pedestrians on a designated bike lane were detected through the PEV’s LIDAR and sensors; routing and pathplanning algorithms then set a new safe course around; the result is a continuous movement even in crowded areas.
which can range from fully electric to fully human-powered, according to the context. At the destination, the vehicle reverts to autonomous mode and pre-positions itself in the city according to predicted demand. The fleet can be sized to maintain acceptable wait times at peak hours and can be available to deliver small packages autonomously during off-peak hours as a “land drone.”
The PEV is designed to expand the demographic profiles of people who use bike lanes. It provides more protection than a two-wheeled bicycle, thereby addressing concerns of individuals for whom safety is a major consideration. Future versions will offer a four-point harness and roll bar, eliminating the need to wear a helmet. It has a canopy, making it more usable in light rain, and heated and cooled seats will provide comfort during extreme weather conditions.
240
18 Larson.indd 240
2018-10-18 10:32 PM
A key benefit of a shared-use autonomous fleet of PEVs is the elimination of both stations and the need for labor-intensive rebalancing. In our preliminary study of a bike-sharing fleet’s performance, substituting the traditional twowheeled, non-electric bikes of Boston’s bikesharing program with PEVs resulted in a 47 percent reduction in fleet size while achieving a passenger wait time of less than five minutes (pick-up) for 75 percent of the trips. With the support of leading private and public sector organizations such as DENSO and Taipei City Government, the team is actively piloting new Human-Machine Interaction5 features in real-world pedestrian environments in Taipei, Taiwan. This work will ultimately enable the PEV to understand, predict and respond to the actions of pedestrians and other road users, and communicate its intentions to humans in a natural and non-threatening way (figures 18.9a–c).
Transformation Case Study 2: CityHome Mayors all over the world have realized that creating affordable, high-quality housing for young people is essential if a city is to remain competitive in an interconnected world. But young people are increasingly priced out of the market in the areas where they want to live and work, and often must endure long commutes from suburban areas where they can find more affordable housing. This is particularly challenging in cities that are the most innovative and entrepreneurial. In additional, attitudes are rapidly changing with respect to urban housing. Young professionals increasingly choose service-rich leasing over ownership. In addition, global communications and the changing nature of work are transforming the home into a center of work, entertainment, health care, communication, and commerce, and the boundary between living and working is effectively blurred. Occupants have diverse needs, values, and activities, and
the current standard and rigid solutions do not serve the population well. Residents are looking for more customized solutions, but generic commodity apartments do not respond to today’s highly personalized society. Market rate housing development is significantly behind other industries in deploying innovation to respond to the changing needs and values of their customers. In response to the demand for affordable housing for young people, many developers are experimenting with “micro-units” that feature about 50 percent of the area of a conventional studio apartment. These tiny living spaces typically include small beds, little storage, barely functional kitchens, and only basic technology. The current approach to micro-units, with their severe compromise on quality of life, will likely prove to be an inadequate, transitional solution. The aim of the MIT CityHome project is to develop a scalable strategy for compact, hyperefficient, technology-enabled urban apartments that can help make living more affordable, efficient, comfortable, productive, and enjoyable for young people in increasingly unaffordable cities. Researchers aspire to develop solutions that make use of innovative technology that can be cost-effective for thousands of market-rate urban apartments. City Science researchers have created a number of CityHome prototypes. CityHome 1.0 is a 200 sq ft apartment designed to have the functionality of an apartment three times larger. The unit features a queen-sized bed, an office with a 6-ft desk, a dining space for six people, a living room space for six people, a fullyfunctional kitchen, and a handicap-assessible bathroom complete with a washer/dryer (figures 18.10a–f, 18.11a–h).6 The MIT CityHome was commercialized as a Media Lab spinoff company, ORI. ORI is now working with developers in North America from Miami to Vancouver to take these ideas to scale via commercial real estate development projects.7
241
18 Larson.indd 241
2018-10-18 10:32 PM
18.10a–f The six floor plans of the 200 sq ft MIT CityHOME 1.0 with robotic walls, tables, bed, and other furniture to dynamically and effortlessly reconfigure small apartment space, allowing it to accommodate a much wider range of activities than with conventional housing.
18.11a–h The 200 sq ft transformable MIT CityHOME 1.0 prototyped at the MIT Media Lab showing (clockwise from upper left) living room for six people, handicap accessible bathroom, bedroom with queensized bed, office, open party space, use of hand gestures to initiate automated transformations, kitchen extension, and dining room for six people.
242
18 Larson.indd 242
2018-10-18 10:32 PM
PREDICTION (MODELING AND SIMULATION TO ESTIMATE THE IMPACT OF URBAN INTERVENTIONS) In the past few years, MIT City Science group has been developing CityScope, a Tangible User Interface for urban modeling. CityScope is an open-source, next-generation, tangible, augmented reality platform, to predict the impact of new urban systems that enable more entrepreneurial, livable, high-performance districts. CityScope platforms are designed for collaborative design processes though rapid prototyping and feedback loops. A generic CityScope setup includes a physical model of the urban area, projection array,
18.12a–f CityScope Kendall model of a 1 sq km area of Kendall Square used as an urban data observatory and simulator showing (clockwise from the upper left): (a) color-coded land use; (b) the various mobility networks including subway, bus, shuttle, shared bikes, and
and sensors to detect human interactions. CityScope has proven to assist decision-making in diverse use cases involving urban design, transportation planning, and tourism analysis as well as refugees’ accommodations.
Prediction Case Study 1: CityScope Kendall Square In 2016, MIT was selected as the developer of a 14acre parcel in Kendall Square, Cambridge, MA. As the last significant opportunity to transform Kendall Square into a model innovation district, City Science researchers built a large CityScope platform of the district (figures 18.12a–f, 18.13) to model the following features:
parking; (c) simulation of solar radiation; (d) locations of Twitter messages updated every 8 minutes as a proxy for activity of young people; (e) simulation of wind flows; and (f) a visualization of pedestrian and vehicular traffic flows.
243
18 Larson.indd 243
2018-10-18 10:32 PM
• •
•
•
Proximity: Prediction of walkable access to parks, housing, jobs, mass transit, schools, amenities, and other parameters. Mobility: Visualization of current traffic movement and prediction of the impact of alternatives to private automobiles, including shuttle-on-demand systems, bike-sharing, ondemand autonomous vehicles, and other mobility strategies that enable an increase in density and streetscape vibrancy without traffic congestion and waste of valuable land for parking. Live/Work: Visualization of current live/work behaviors, and simulation of the impact of housing for young professionals to increase diversity, vibrancy, and affordability in the creative heart of the city for those who would otherwise be priced out of the market where they most want to live and work. Sustainability: Prediction of the impact of land use and mobility proposals on mobility and building-related embodied energy and use energy.
18.13 A radar plot showing the attributes of the district updated in real time, with a calculation of building energy per person, mobility energy per person, social well-being (walkability, access to parks, etc.), and innovation potential (probability of creative collisions). Using linear
CONSENSUS (FACILITATING INFORMED DECISIONS BY DIVERSE STAKEHOLDERS) The CityScope platform can be adapted to facilitate consensus building for complex issues with multiple stakeholders related to master planning, rezoning, expansion of bike lanes, bus rapid transit proposals and other uses requiring community engagement. We have used versions of the CityScope in Andorra, Helsinki, Boston, and Shanghai. Perhaps the most compelling use of the CityScope to date was to help the City of Hamburg accommodate refugees from the war in Syria. In 2015, City Science created several CityScope platforms and a community engagement process set to examine the impacts of the Bus Rapid Transit (BRT) system in greater Boston area. This was the first attempt to explore CityScope as a consensus-building platform outside of the confines of research at MIT. CityScope BRT includes an interactive platform that uses physical models and
regression models based on the CDRs and RNC data, we can predict the district activity density, diversity and connectivity (proximity) for each axes of the radar, and weight the values in response to community values and the performance goals of the city.
244
18 Larson.indd 244
2018-10-18 10:32 PM
18.14a and b CityScope during FindingPlaces project, Hamburg, 2016. The tool could react to the participants’ proposals by displaying urban regulation, constraints and suggestions.
3D projection, enabling community members to engage in neighborhood and street-level decisions, such as alternative bus corridor designs, stationlevel variations, prepay boarding or dedicated lanes.
Consensus Case Study: FindingPlaces, Hamburg, Germany Proven to be instrumental in building consensus for urban challenges, a few months later CityScope was again deployed in the context of consensus-building, now for housing for refugees in Hamburg, Germany. In 2015, approximately 1.2 million asylum applications were filed in Europe, one-third of which were in Germany. This persistent influx of asylum seekers posed major challenges to the German federal states. As a consequence, housing solutions were implemented ad hoc and in many cases, refugees were accommodated in tents, warehouses, or gymnasiums. In June 2015, Hamburg’s Mayor Olaf Scholz and MIT Media Lab signed a research collaboration agreement that promoted the establishment of the City Science Lab (CSL) at HafenCity University Hamburg (HCU). In
February 2016, Mayor Olaf Scholz assigned CSL and MIT to convey a participation process that would enable citizens to engage in finding accommodations for the prediction of ~79,000 refugees. The aim was to incorporate the citizens’ personal experience and local knowledge into the political and administrative evaluation of potential locations. The results and proposals emerging from the participation process were to become recommendations for political decision-making. The mayor’s office allowed only three months for conception and development of this project. In response to the call, CSL and MIT employed CityScope as a decision-making and knowledgesupport tool to facilitate the public participation (figures 18.14a–b, 18.15a–b). A series of public workshops was planned that centered around several interactive CityScope stations displaying task-related data to citizen groups as they worked out decisions. Between May and July 2016, a total of 34 two-hour workshops were held at HCU with nearly 400 participants. In total, 161 underused or vacant locations were suggested by the participants, of which 44 passed legal confirmation by the authorities. With these, accommodation solutions for almost 24,000 refugees were proposed, exceeding the initial targeted goal of 20,000. 245
18 Larson.indd 245
2018-10-18 10:32 PM
18.15a and b CityScope FindingPlaces TUI platform in HafenCity University, Hamburg, 2015. The tangibility of the tools created a common ground for discussion, even for complex and sensitive matters.
FindingPlaces is currently being cited as an exemplary case study for public participation in complex, heavy-weighted urban challenges. In 2017, FindingPlaces won the URBACT award for Good Practice, and was suggested as a scalable consensus-building platform for the entire European Union: FindingPlaces is relevant and transferable to other European cities … the issue of massive refugee influx and their ad hoc accommodation is a challenge shared by many cities ... The key tool – the CityScope – is ready for mobile application in other places too ... In a more general sense, the practice and technology of FindingPlaces and CityScope can be applied to a broad range of similar urban problems, especially the identification of appropriate locations for specific uses. FUTURE WORK We have outlined four components of a new process that may enable more high-performance, entrepreneurial communities. We described
the use of data collection and analysis to achieve a fine-grained understanding of current conditions, identification of the design and technology interventions to positively transform a community, modeling and simulation to predict the impact of proposed interventions, and a consensus-building system to help multiple stakeholders achieve a common vision of the future. Over the next few years, MIT City Science researchers will investigate the addition of a fifth and critical component – governance. In a highvalue urban district, such as Kendall Square in Cambridge, MA, property is owned and controlled by relatively few landowners who make business decisions according to the highest and best use of land with respect to financial returns. Inclusionary zoning is a tool that mandates that market-rate housing developments subsidize a small percentage of “affordable” units for lower-income residents. This crude and inflexible mechanism has little impact on the availability of housing needed for the workforce in the district. We outline below a vision for governance to incentivize pro-social real estate development and pro-social behaviors of residents.
246
18 Larson.indd 246
2018-10-18 10:32 PM
•
•
Algorithmic zoning: We envision the replacement of static zoning regulations with a system of algorithmic rules that would permit the fine-grained delineation of community needs, and their mapping to dynamically adjusted incentives for developers and land owners. Dynamic bonuses could be adjusted continuously for an entire district in response to changing economic conditions and community values, eliminating all variances and streamlining the approval process. Agreements could be autonomous and self-enforcing, using “smart contracts” documented on a blockchain public ledger. Algorithmic zoning ordinances may be developed as an opt-in system, in parallel to conventional rules, eliminating site-specific zoning variances. Such a system could be used to incentivize the production of non-market rate housing (for young professionals, families, workforce, and the elderly), amenities not supported by the market (grocery store, day care, pharmacy, theater, museum, high school), contribution to shared infrastructure (bike lanes, mass transit, parks, playgrounds), and the deployment of renewable energy production. The aim of this system would be to incentivize developers to make pro-social decisions in addition to maximizing financial returns. Local token currency: We are also investigating the deployment of a decentralized token economy, enabled by blockchain and public ledgers that operates in a geo-fenced area with rules and exchange rates to optimize for social, cultural, and environmental benefits. In such a system, residents are both consumers and investors with equity in a community. Using a smartphone, and deployed with a system similar to Apple Pay, the value may increase if a person both lives and works in the district, if one does not own a car, or if shopping is at a locally owned
establishment rather than a franchise. Young professionals may elect to have a portion of their salary paid with pre-tax tokens that could subsidize local housing and food. Peer-to-peer exchange of services could be facilitated with a high-tech barter system. Local businesses and employers may invest in such a system as it could improve the diversity and availability of potential employees in the district, and it could lead to an increased desirability of the district, resulting in higher property values. We envision that the CityScope platform could serve as the “oracle” to help communities develop a data-driven process to establish the values embodied in algorithmic zoning and a dynamic local economy.
City Science Network The MIT City Science Initiative has developed an international network of cooperative City Science Labs at Tongji University (Shanghai), Taipei Tech (Taipei), HafenCity University (Hamburg), Aalto University (Helsinki), ActuaTech (Andorra), and Ryerson University (Toronto). At MIT, we are developing concepts and key technology that can be extended, deployed, and evaluated by our collaborators in unique contexts around the globe. Over the next few years, we plan to develop new and exciting projects in less-affluent, rapidly growing cities in Latin America, Africa, and India, where the impact of a new process to address the challenges of urbanization may be the greatest.
247
18 Larson.indd 247
2018-10-18 10:32 PM
NOTES 1 Call Detail Records (CDR) is the prevailing telecom dataset used for large-scale spatial research. CDRs provide customers’ usage history, cell tower ID and approximate coverage region of each station, together used to estimate location-time of customers. In third generation (3G) Universal Mobile Telecommunications System (UMTS), the governing element is the Radio Network Controller (RNC). One function of the RNC is to keep track of the locations of devices as they move around the coverage area. Each observation includes a unique ID for the subscriber, a timestamp, the coordinates of the device at the time of the update, and the home network of the subscriber. See Ariel Noyman et al., “Reversed Urbanism,” Environment and Planning B, (accepted for publication), 2018. 2 The TerMITes are small (approximately 25 mm x 45 mm x 5 mm) battery-powered devices that can be attached to any object and will register information each time the object is moved. Combined with background recording of environmental data (light, temperature, relative humidity, barometric pressure) at regular intervals, motion-activated sensors make it possible to record environmental changes in response to human activity. 3 See www.media.mit.edu/groups/city-science/overview/ 4 See Luo, Z.-Q., Pang, J.-S., and Ralph, D., Mathematical Programs with Equilibrium Constraints, Cambridge: Cambridge University Press, 1996. 5 Ishii, H. and Ullmer, B., “Tangible Bits: Towards Seamless Interfaces Between People, Bits and Atoms,” Proceedings of the ACM SIGCHI Conference on Human Factors in Computing Systems, 1997, pp. 234–241. 6 See www.youtube.com/watch?v=f8giE7i7CAE. 7 See www.orisystems.com.
248
18 Larson.indd 248
2018-10-18 10:32 PM
19 THE PHYSICAL IMPLICATIONS OF A MASS CUSTOMIZATION ECONOMY
THOMAS FISHER 249
19 Fisher.indd 249
2018-11-04 2:55 PM
19.1 The modern nomad.
Recent years have seen the rapid growth in people’s ability to 3D print the products they use, access the on-demand services they need, and occupy property they don’t own. This reflects a shift away from a mass-production economy, in which companies determined the products and property available to us, toward a mass-customization economy, in which consumers can also become the producers of goods and services themselves. Although still a small part of the overall economy, this trend promises to lower the cost of living and increase the economic opportunities for millions of people, while also overturning a number of established business models and professional assumptions. How will architecture change, for instance, in an economy in which people increasingly prefer access to buildings rather than the ownership of them, and what form will cities take in an era in which people increasingly live, work, and make things in close proximity, even in the same building, rather than commute long distances between home and office? To answer such questions, we must first remind ourselves of how recent architecture and cities really are in human history. People have occupied fixed buildings and permanent settlements for only 5 percent of our time as a species. For the other 95 percent of our past, people occupied lightweight structures and used local materials, they moved frequently and carried little with them, they stewarded the land and shared what they had with others in their tribe, and they left little or no trace of themselves behind (figure 19.1).
Despite the long history of our doing so, we have come to think of the last 5 percent of our past as normal: building and occupying permanent structures affixed to their sites, fabricating materials, and purchasing products from all over the world, and buying more goods than we immediately need and throwing away what we no longer want. We may think of ourselves as the smartest of animals, but such behavior makes us the outlier on the planet and in some ways one of the most primitive animals, since unlike other species – and unlike our own ancient ancestors – we have largely lost the ability to make all that we need from what we have at hand and from the materials immediately around us. Which makes the rise of an on-demand, masscustomization economy so significant. It appeals not only to those of us who want to lower our ecological footprint and reduce our financial burden, but it also promises, in its own high-tech way, to return us to a time in which we, too, fully utilized what we had available to us in our immediate environment and sharing what we have with others seeking access to it. This in turn suggests that we may be coming to the end of a 10,000-year experiment, an experiment with fixed buildings and permanent settlements that the environmental historian, Jared Diamond, has called “the worst mistake in the history of the human race.”1 He made that claim because of the environmental damage and human health hazards that came with large-scale agriculture and heavy industry.
250
19 Fisher.indd 250
2018-11-04 2:55 PM
19.2 DUS Architects 3D printed a canal house in Amsterdam using equipment contained in this structure.
Like all transformations in human existence, this one threatens to leave many people behind economically, which can breed the anger and fear that we have seen play out in recent politics around the world. But by increasing the freedom of many over the control of a few, as Hegel argued, this transformation will ultimately prevail because it puts the means of production and vast amounts of information into the hands of almost everyone and gives people access to goods and services that, formerly, only a few could afford. In many ways, this new way of being combines aspects of both the hunter-gatherer and agricultural/industrial eras that preceded it. It depends upon the most advanced technology for its operation and yet enables people to live much more lightly on the planet in a condition that has come to be called digital nomadism (figure 19.2). OUR PLANETARY PONZI SCHEME Environmental and economic reasons have both driven this change. We know that we cannot sustain the rate at which humanity has used up the planet’s finite resources and we also know that people in too many countries cannot meet their basic needs while those in the wealthiest countries consume and waste far too much. The unsustainability and unfairness of that reflect a Ponzi scheme that the developed world has had with the planet for at least the last 200 years. As we learned from the Ponzi scheme that Bernard Madoff perpetrated in
the financial markets earlier this century, these schemes enrich those at the top of the profit pyramid by exponentially exploiting those at its base, eventually running out of people, causing the scheme to collapse, suddenly and without warning. The developed countries have engaged in a similar Ponzi scheme, exploiting low-cost labor in poorer nations and borrowing heavily from future generations by over-consuming non-renewable resources and exhausting what our progeny will need. (What, I wonder, will we tell our grandchildren when they ask us why we used up all the easily accessible oil?) As we also learned from Madoff’s fraud, when Ponzi schemes get large enough we don’t see them or at least don’t want to see them, which has been the case with humanity’s Ponzi scheme with the planet. By refusing to recognize it, we now find ourselves at the point where it now stands on the brink of collapse, demanding, by some estimates, about one-and-a-half planets to meet humanity’s current annual needs – and five planets to sustain the level of consumption in countries like the US. When Ponzi schemes run off the planet like this, they fall. This suggests that the transition to a much lower level of consumption and much higher level of sharing assets and mass-customizing goods will happen more quickly than many might think. Madoff’s scheme crashed over a weekend and so might humanity’s. And when it does collapse, the billion or so people who have profited enormously from this scheme will have the farthest to fall, have the most to lose, and possibly have the hardest time adjusting to a post-Ponzi existence. But this will force us to do what we should have done a lot sooner: use resources much more efficiently, leverage assets much more fully, and spread opportunities much more broadly. The on-demand, mass-customization economy holds great promise in helping us achieve that, although we have to ensure that it doesn’t become just another pawn in the Ponzi scheme, fueling further consumption, exploiting people even more, and borrowing from the future ever faster. 251
19 Fisher.indd 251
2018-11-04 2:55 PM
19.3 Droog Design have used discarded items to create humorous, low-impact products.
A POST-PONZI FUTURE Knowing that such schemes eventually collapse, we need to start asking what a post-Ponzi existence might look like and how we can steer the new economy in constructive ways toward it. Rather than keep building new space, for example, we need to start using the space that we already have much more fully than what we do now. About the time the Ponzi scheme began, people began to design buildings with specialized rooms and cities with single-use zones, all of which led to an incredible waste of space and over-consumption of materials and fuel. In a post-Ponzi world, we will need to return to a pre-industrial notion of space as widely adaptable to a number of uses over the course of a day or week (figure 19.3). Look at the homeless populations who occupy many American cities. While the collapse of our Ponzi scheme may render many more people homeless, that population in our cities right now sits outside all night, often next to office buildings that, fully serviced, sit empty at the same time. Finding a way for the former to occupy the latter represents one example of a post-Ponzi way of thinking. Space is space, and those who learn how to occupy it in more nomadic-like ways, temporarily and creatively, will thrive in the future, and those who don’t, won’t. That mindset echoes the comment of Muhammad Yunus, who said recently to a reporter at The Guardian,
Human beings are not born to work for anybody else. For millions of years that we were on the planet, we never worked for anybody. We are go-getters. We are farmers. We are hunters. We lived in caves and we found our own food, we didn’t send job applications. So, this is our tradition.2 The on-demand, mass-customization economy draws from that tradition. Human beings have an inherent ability to innovate, the result of our capacity to recombine aspects of what we know and have experienced in new ways, and yet, since the rise of the Ponzi scheme and the need for workers who can follow orders and who won’t question things, our educational system has attempted to drum our native inventiveness out of us. In elementary school, when teachers start giving students an “A” for the correct answer rather than the creative one, this process begins. A post-Ponzi existence has to turn that on its head. We will need to have everyone innovating, if for no other reason than that we will have to re-imagine and re-invent almost everything we have come to expect and depend on from purveyors of the grand pyramid scheme. The mobile, digital environment, which in some sense puts the whole world in each of our hands, will greatly facilitate that innovation, and it remains up to us to use it for positive ends.
252
19 Fisher.indd 252
2018-11-04 2:55 PM
19.4 Urbee 3D printed car.
THE NEW NOMADISM Digital devices give us not only intellectual mobility, but also physical mobility, which we will need in the face of climate disruptions. As humans experience ever more violent and frequent storms, those who live in permanent shelter especially along our coasts become sitting ducks. We have carved up the land into fixed structures and fenced-off pieces of property, inhibiting our ability to disperse across the landscape; we have paved over wetlands, destroying their capacity to absorb floodwater; and we have created vast swaths of impervious surface, heightening flood levels and speeding up water currents. And when we do try to get away from coastal storms, we funnel cars onto limited access highways, whose subsequent congestion slows the evacuation to a crawl. We have a lot to learn from the more sensible adaptations of indigenous people, who typically lived on higher ground, away from coasts, and who let the natural defenses of the land, such as dunes and wetlands, protect them from storm surges. With this in mind, we need to stop being sitting ducks and to live more like real ducks, with buildings and settlements that can float with rising waters. Or, we might think of ourselves like mighty ducks and measure our strength according to how well we move, with an architecture of lightweight, portable enclosures able to be folded up and moved in advance of a coming storm and ready for rapid deployment afterward. The climate change
that has resulted from the air pollution of our Ponzi scheme presents us with an opportunity to explore smarter and more adaptable ways of living. We can continue to make ourselves sitting ducks in the face of future storms, or we can think and act more creatively, more like (mighty) ducks. THE THIRD INDUSTRIAL REVOLUTION Such opportunities to rethink how we occupy the land relate to how we rethink the making of things. Economists like Jeremy Rifkin have argued that, unlike the first Industrial Revolution in the nineteenth century, which began with the mechanization of work and the second Industrial Revolution, which led to the mass production of goods and services, the third Industrial Revolution of this century flips the previous two on their head.3 Rather than overproduce almost everything, as we did in the last century, the new economy takes advantage of this excess capacity: Airbnb has leveraged unused beds and apartments, without owning any rooms, while Uber and Lyft have leveraged empty seats in cars, without owning any vehicles. Complementing the provision of services using existing assets is the idea of enabling people to make goods themselves, when they need them. The previous two Industrial Revolutions began with the vehicles: the steam engine in the nineteenth century and the Model T in the twentieth century. The third Industrial Revolution may mark its beginning with the Urbee, the first 3D printed working car (figure 19.4), 253
19 Fisher.indd 253
2018-11-04 2:55 PM
19.5 Wikihouse, an open-source, downloadable design.
printed at RedEye out in Minnesota by Stratasys.4 This is how industrial revolutions happen. The mass producers of cars cannot compete with someone who can print, assemble, and drive away their own vehicle. 3D printing, CNC fabrication, laser cutting, virtual reality – all are potentially revolutionary tools that bring with them very different assumptions about how we will make things, how we will empower people to make things themselves, and how the design community will work in the future. POST-PONZI PRACTICES As people work in new ways in the future, so will the design community. Architects, for example, will do fewer one-off designs for individual clients and more downloadable designs that an unskilled and semi-skilled labor force can assemble as needed. This will happen not only because the number of people able to afford custom designs will shrink in proportion to the total population, as is already happening, but also because the real need for design services – and design thinking – rests with the vast majority of people who do not now have access to architectural services. In an economy based on access, becoming as accessible as possible will become the way to prosper. An example of this: WikiHouse (figure 19.5), an open-source, downloadable design, created by
designers Alastair Parvin and Nick Ierodiaconou, that lets people customize the plans, cut the parts of plywood using a CNC router, and erect it within a day with unskilled labor.5 In a digitally nomadic future, we will find a growing demand for such easily erected structures that use readily available materials and local labor and that can be rapidly disassembled and reused or re-erected where needed. And in such a future, we may distinguish architectural firms not by their size, but by those that work in the creation of downloadable designs, on one hand, and those that work in local communities by helping them customize those designs for local needs, on the other. The architect becomes, in such a scenario, less a generator of unique, one-off solutions and more of a facilitator and guide, characteristic of architectural practice prior to the Ponzi scheme. THE NEW MOBILITY Digital nomadism will affect transportation as much as it will architecture. Digital technology has prompted the rapid rise of shared autonomous vehicles (figure 19.6), in which well-established technology from the aerospace industry has made its way into the automobile industry to create a mode of transportation that will save millions of deaths and injuries a year from negligent drivers, will provide mobility to large numbers of people
254
19 Fisher.indd 254
2018-11-04 2:55 PM
19.6 Keolis and Navya autonomous shared ride vehicle.
who cannot drive now, and will dramatically reduce the economic and environmental impact of vehicles as part of low-cost, shared systems. This will also free up a lot of land now devoted to the parking of vehicles, as much as 30 percent of the land in the average city by some estimates. Using that land in productive ways, as places for growing food, cleaning water, improving habitat, and constructing affordable shelter, can position communities well for a post-Ponzi existence, one in which people will need to share more and become more self-sufficient at the same time. While shared autonomous vehicles will save money, lives, and land, they will also disrupt existing employment. The most common job in America right now is truck driving, and technology will soon disrupt that line of work as it did the two previous most common jobs in America: farming in the nineteenth century and secretarial work in the twentieth. With a post-Ponzi scheme existence comes the requirement that we do all we can to retrain people to take advantage of the new opportunities the sharing economy allows. Oxford professors Carl Benedikt Frey and Michael A. Osborne have shown how a sizable number of dangerous and repetitive jobs will likely be automated in the coming years.6 When we look at what remains for people to do, the jobs fall into what I call the “5 Cs” – caring, communication, creativity, craft, and construction – all work that
requires human interaction, imagination, and manual skill and that recalls what people did in communities prior to the Industrial Revolutions and before the advent of mass production. NOMADIC URBANISM With changes in the nature of work will come changes in where we work and live. Every major economic transformation brings a change in urban form: the first Industrial Revolution led to the rise of large, industrial cities and factory towns in which people lived near where they worked. Cities became like the machines they increasingly housed: efficient, polluting, and often oppressive. The second Industrial Revolution, with the rise of mass production, demanded a different kind of urban form – suburbia – able to absorb the excess production of goods coming off the assembly lines. Suburbs required that everyone own a car, a lawn mower, a washing machine, and so on, sustaining war-time levels of production and consumption that sent our Ponzi scheme with the planet into overdrive. The third Industrial Revolution has begun to overturn this. Even though the new economy allows us to live and work anywhere, many people are choosing to live in higher density settlements, with less personal space and more communal space, as we see with the rise of micro-unit 255
19 Fisher.indd 255
2018-11-04 2:55 PM
19.7 Mixed-use, public space in Malmö, Sweden.
developments and co-housing communities in cities around the world (figure 19.7). Digital nomadism promises a future in which more people will own less, share more, and have lower costs and greater mobility, with an emphasis on experiences over possessions and the quality of things over their quantity. How urban design responds to this remains to be seen, since it typically follows economic transformations more slowly than other design disciplines, but it suggests a city in which we will see a much greater mix of uses in closer proximity to each other and much more usable public space as private space shrinks.
DIGITAL HIGHER EDUCATION Universities will not escape these disruptions. While much of the thinking about digital nomadism has emerged from universities, these institutions remain among the most conservative in our time and they have largely acted as if the new economy won’t affect them very much. The number of online courses has increased and adjunct faculty have become more common, all characteristic of an economy in which asynchronous learning and “gig” employment have become common. But universities remain burdened by a lot of fixed real estate and
256
19 Fisher.indd 256
2018-11-04 2:55 PM
tenured faculty that make it difficult for these institutions to change very quickly. Universities have also become their own worst enemy. As the authors of the book Platform Revolution7 argue, any organization that operates as a gatekeeper with an asymmetry of knowledge and a demand for that knowledge faces major disruption by digital platforms able to offer access at a much lower cost. When universities brag about their selectivity, with the asymmetry of knowledge between professors and students and with a high demand for that knowledge, it puts these institutions directly in the path of platform-based disruption. This doesn’t mean that universities will disappear, but the old Marshall McLuhan idea that every new technology turns the old one into an art form, requires that we think about everything currently done on campuses in terms of what cannot be done online or download. Faceto-face conversations, serendipitous interactions, creative collaboration – those activities, once thought of as marginal, will become among the primary reasons to attend a university. Likewise, education will have to shift from the transfer of information in a lecture hall, something that the digital world can do faster and often more accurately, to the provocation of ideas in more active learning locations. With this will come an increase in the mass-customization of curriculums, in which students will be able to mixand-match courses in the creation of new, hybrid fields of study, working in interdisciplinary ways with a variety of faculty. Digital nomadism may take many forms in higher education in the future. In a class I have taught, for instance, I assign a student in each class to decide where we will meet on campus, using social media to let me and the others know where to gather. My students rarely choose a classroom in which to meet, more often selecting a place with comfortable furniture, daylight, and other activities going on nearby. This pedagogical experiment shows that people can learn in a far greater range of places than campus planners
typically assume, that a campus can become a memory device that helps students remember the content of a conversation by their connecting it with a particular place, and that the listening in and occasional participation in the class by people sitting nearby provide a primary reason to come to a campus: to learn in unexpected and serendipitous ways. ITINERACY Place matters not just when learning, but also when understanding the relationship among people and things in the world. I have led workshops as part of a local leadership development effort by the American Institute of Architects, in which I have asked participants what kind of firm they would want to lead. These recently graduated or licensed architects frequently envision firms without the word “architect” in the title, and many of them imagine practicing in highly flexible and mobile ways, with offices that work in the job site or that set up shop in a client’s organization in order to understand it better. Such a deep engagement with a client or construction site shows how mobile digital technology, combined with pop-up workspace, can offer insights not possible when working in an isolated, professional office setting. As I have written about with my colleague Jacob Mans,8 such digitally nomadic architectural practices may lead to “decentralized design labs,” in which equipment and space move from one site to the next, leveraging the local human and natural resource assets of a place and evolving, with communities, the most appropriate and equitable solutions possible. This recalls pre-modern medieval forms of practice in which architects – or spatial facilitators and provocateurs, as we might someday call ourselves – moved from one site to the next, living and working there until the completion of the work, and then moving on. And it opens up economic opportunity and environmental responsibility to the populations and ecologies of a place that do not now exist for too many people. 257
19 Fisher.indd 257
2018-11-04 2:55 PM
19.8 Messenger II House by MacKay-Lyons Sweetapple Architects.
PLACE-BASED KNOWLEDGE As architectural practice becomes more mobile and itinerant, it will also become more physically grounded and place-based. Some of today’s most compelling architectural firms have evolved their work in response to particular locations and cultural contexts: MacKay-Lyons Sweetapple Architects in Halifax (figure 19.8), Salmela Architects in Duluth, Patkau Architects in Vancouver, Koning Eizenberg Architecture in Los Angeles, and Lake/Flato in San Antonio, among many others. These firms, tied so closely to the places in which they practice, characterize the new nomadism more generally. Nomadic cultures typically evolve in specific climates and environmental conditions, husbanding the land and all that lives there, knowing that by
doing so, that landscape will be there to support the nomads when they return. Place also has become a compelling way to organize our knowledge about the world. Humanity has done so much damage to the natural environment and to indigenous cultures because we no longer understand the world as it is actually organized. We have dismembered reality into a number of disciplines and specialties that, while they may study particular places and cultures, are themselves distinct cultures unto themselves, disconnected from any specific location. The world is, instead, organized spatially, with the knowledge of disciplines all existing in every place, which suggests that in a digitally nomadic future, we would be better served mapping our knowledge visually and spatially
258
19 Fisher.indd 258
2018-11-04 2:55 PM
in particular places rather than sorting it into disciplinary classifications. Digitally mapping what we know through Geographical Information Systems may evolve to be the most useful and insightful ways to understand the world around us and to act appropriately in response to that understanding. CITY-BASED REGIONS Digital nomadism challenges not just traditional design practices, but also the very ideas of a nation state. Jane Jacobs many decades ago argued that city-based regions would become the primary economic unit in the future, and her prediction has begun to come true. The website Nomad List shows this. This site allows digital nomads to decide in which cities to live, based on quality of life, bringing work with them and finding work where they go. This shows how city-based regions have become the primary competitors in the new economy and the strongest attractors of an increasingly nomadic population. Jacobs also thought that healthy regions would replace the imports from elsewhere with local materials and local labor, suggesting that the future architectural practice – a locally grounded profession – can play a much more central role in anticipating economic development and in helping people overcome their fear of change by visualizing the opportunities that lie ahead. And what are those opportunities? To start living as lightly, opportunistically, and adaptively as possible, like most of the other species with whom we share this planet, and to start designing shelter and settlements that can sustain us without exhausting the resources that our progeny will need. Architects have much to contribute to this new, global transformation, but it won’t be with architecture as we have known it.
NOTES 1 Diamond, Jared, “The Worst Mistake in the History of the Human Race,” Discover, May 1987, pp. 64–66, available at: http://discovermagazine.com/1987/ may/02-the-worst-mistake-in-the-history-of-thehuman-race 2 Cosic, Miriam, “We Are All Entrepreneurs: Muhammad Yunus on Changing the World, One Microloan at a Time,” The Guardian, March 28, 2017, available at: www.theguardian.com/sustainablebusiness/2017/mar/29/we-are-all-entrepreneursmuhammad-yunus-on-changing-the-world-onemicroloan-at-a-time 3 Rifkin, Jeremy, The Third Industrial Revolution: How Lateral Power is Transforming Energy, the Economy, and the World, New York: St. Martin’s Press, 2011. 4 See https://korecologic.com/. 5 See https://wikihouse.cc/. 6 Frey, Carl Benedikt, and Osborne, Michael A., Oxford Martin School, University of Oxford, available online 29 September 2016 at: http://dx.doi. org/10.1016/j.techfore.2016.08.019. 7 Parker, Geoffrey G., Van Alstyne, Michael A., and Choudary, Sangeet Paul, Platform Revolution: How Networked Markets Are Transforming the Economy and How to Make Them Work for You, New York: W.W. Norton, 2016. Available at: https:// platformrevolution.com/. 8 Mans, Jacob and Fisher, Thomas, “The Itinerant Architect: Toward a Land-based Architectural Practice,” Journal of Architectural Education, vol. 71, no. 2. (2017).
259
19 Fisher.indd 259
2018-11-04 2:55 PM
260
19 Fisher.indd 260
2018-11-04 2:55 PM
APPENDIX AUTHOR BIOGRAPHIES PHOTO CREDITS PROJECT CREDITS INDEX
261
20 Back Matter.indd 261
2018-11-04 1:33 PM
BIOGRAPHIES
BRANKO KOLAREVIC University of Calgary Calgary, Canada
JOSÉ PINTO DUARTE The Pennsylvania State University State College, PA, USA
Branko Kolarevic is a Professor at the University of Calgary Faculty of Environmental Design, where he co-founded the Laboratory for Integrative Design (LID) and is a co-director of the multi-disciplinary Computational Media Design (CMD) program. He has taught architecture at several universities in North America and Asia and has lectured worldwide on the use of digital technologies in design and production. He has authored, edited or co-edited several books, including Building Dynamics: Exploring Architecture of Change (with Vera Parlac), Manufacturing Material Effects (with Kevin Klinger), Performative Architecture (with Ali Malkawi), and Architecture in the Digital Age. He is a past president of the Association for Computer Aided Design in Architecture (ACADIA), past president of the Canadian Architectural Certification Board (CACB), and currently is president of the Association of Collegiate Schools of Architecture (ACSA). He is a recipient of the ACADIA Award for Innovative Research in 2007 and the ACADIA Society Award of Excellence in 2015. He holds doctoral and master’s degrees in design from Harvard University and a diploma engineer in architecture degree from the University of Belgrade.
José Pinto Duarte is the Stuckeman Chair in Design Innovation and Director of the Stuckeman Center for Design Computing (SCDC) at Penn State, USA. Duarte holds a professional degree in architecture from the Technical University of Lisbon and after obtaining master’s and doctoral degrees from MIT on the mass customization of housing, he returned to Portugal where he helped launch technology-oriented architecture programs and digital fabrication labs. Duarte served as Dean of the Faculty of Architecture at the University of Lisbon, as well as president of eCAADe, a European association devoted to education and research in computer-aided architectural design. Duarte’s work displays a record of uniting academic research and industry and fostering multi-national partnerships. He created the Design and Computation research group at the University of Lisbon, devoted to interdisciplinary and collaborative research, and helped establish the MIT–Portugal program. His research focuses on the use of computation and AI to support context-sensitive design across different scales. In 2008 he was awarded the TU Lisbon/Santander Prize for research excellence in architecture.
262
20 Back Matter.indd 262
2018-11-04 1:33 PM
ASSA ASHUACH ASSA STUDIO / Digital Forming London, UK
JOHN L. BROWN Housebrand Calgary, Canada
Founded in London in 2003, Assa Ashuach Studio works closely with R&D and business innovation units of companies such as Samsung, Nike, and Panasonic, among others, and public institutions such as the UK Government Technology & Strategy Board, the Science Museum, the Design Museum, and the Victoria & Albert Museum on design technology for social and cultural progress. Ashuach’s work focuses on three main areas: (1) industrial research and product design consultancy in digital design and manufacturing methods, specializing in SLS polyamides and DMLS metal alloys within a wide range of additive manufacturing technologies, offering codesign and user interaction coding of dedicated 3D design and engineering tools as part of the studio’s unique innovative methodologies; (2) design and manufacturing of lifestyle products, such as footwear, mobile phones, shaving devices, lighting and furniture, among others; and (3) self-production of limited edition studio pieces, using new design and production methods. At the center of the studio’s developments is the virtual ‘life’ of an object before it is physically produced. Opened to the user’s input, the digital object can be produced using a wide range of additive manufacturing technologies (widely known as 3D printing) with a mixture of mass production solutions. Aesthetics, adaptation and reconfiguration of forms are considered elements in flux within this new product design workflow. Assa Ashuach is also a founder of the technology company Digital Forming.
John Brown is a registered architect, a founding principal of the architectural firm Housebrand, and Dean of the Faculty of Environmental Design, University of Calgary, Canada. He is a recognized authority on residential practice and new models of architectural practice. In 2003, he received the Royal Architectural Institute of Canada (RAIC) Award of Excellence for Innovation for his development of Housebrand, a vertically integrated practice that combines residential architecture, construction, interior design, and real estate services in a one-stop shop for homebuyers. In 2009, he received a Leadership Award from Residential Architect magazine in recognition of his work to increase public awareness about the value of design. In 2012, Dr. Brown expanded his research interests into home health and began a long-term collaboration with researchers in the Cumming School of Medicine to look at agingin-place design strategies that incorporate mass customization and co-design opportunities. He recently completed a PhD on this topic entitled “Going Home: Future Adaptive Building for Aging-in-Place,” at the Royal Melbourne Institute of Technology, Australia. In 2015, Dr. Brown’s aging-in-place research was recognized with a Mayor’s Urban Design Award in Housing Innovation for the development of the Aging-inPlace Laneway Housing Research Project.
263
20 Back Matter.indd 263
2018-11-04 1:33 PM
KARL DAUBMANN Lawrence Technological University Detroit, MI, USA
THOMAS FISHER University of Minnesota Minneapolis, MN, USA
The values of craft, materiality, and technology were all tangible in the fabrication shop where Daubmann grew up. With over 30 years of CAD/ CAM experience, he learned how to operate a boom crane and use computer-controlled equipment long before he could drive a car. He is a registered architect with a record of distinguished projects inspired and driven by his interests in design technology, manufacturing, and multidisciplinary design. Daubmann began the DAUB research studio as a means to focus on those same preoccupations and to develop work to push the disciplinary limits of those interests. He served as the Creative Director and Vice President of Design for Blu Homes and as principal of PLY Architecture. He is a Fellow of the American Academy in Rome. He is Dean and Professor at Lawrence Technological University in Detroit.
Thomas Fisher, a graduate of Cornell University in architecture and Case Western Reserve University in intellectual history, was previously the Editorial Director of Progressive Architecture magazine. Recognized in 2005 as the fifth most published writer about architecture in the United States, he has written 9 books, over 50 book chapters or introductions, and over 400 articles in professional journals and major publications. Named a top-25 design educator four times by Design Intelligence, he has lectured at 36 universities and over 150 professional and public meetings. He has written extensively about architectural design, practice, and ethics. His latest book is Designing Our Way to a Better World (University of Minnesota Press, 2016) and his new book entitled The Architecture of Ethics will be published by Routledge in 2019.
264
20 Back Matter.indd 264
2018-11-04 1:33 PM
MARC FORNES MARC FORNES / THEVERYMANY New York, NY, USA
Marc Fornes is a registered and practicing architect specializing in computational design and digital fabrication. He leads MARC FORNES / THEVERYMANY, a Brooklyn-based studio which has developed a prototypical strategy to unify surface, structure, and spatial experience into a single tectonic system, namely, through the invention of Structural Stripes. Over the last ten years, the studio has designed and built a number of thin-shell pavilions and installations that push the limits of form, structure, and space. These permanent and temporary projects are situated between the fields of art and architecture, with an emphasis on public-facing, spatial artwork. Some of those prototypical architectures have been acquired and exhibited by institutions, including the Centre Pompidou (Paris), the FRAC Centre (Orleans), Art Basel Miami, and the Guggenheim (New York), and have sold at auction at Phillips De Pury. The studio has been acknowledged by the American Institute of Architects as part of New Practices New York in 2012 and by the Architectural League, winning its annual prize in 2013. Several projects of the studio have gone on to win international awards. Marc has shared his research as a TED Fellow, and in graduate design studios at Columbia GSAPP, Harvard GSD, University of California, University of Michigan and Die Angewandte. As a project architect with Zaha Hadid Architects, he directed research for an experimental mediatheque in Pau, France – which would have been the largest carbon shell structure to date.
FABIO GRAMAZIO AND MATTHIAS KOHLER Gramazio Kohler Architects Zurich, Switzerland Fabio Gramazio and Matthias Kohler are architects with multi-disciplinary interests ranging from computational design and robotic fabrication to material innovation. In 2000, they founded the architecture practice Gramazio Kohler Architects, which has received numerous awards for its realized designs. Current projects include the design of the Empa NEST research platform, a future living and working laboratory for sustainable building construction. As professors, Gramazio and Kohler developed the first architectural robotic laboratory at ETH Zurich. The ensuing research has been highly influential in the field of digital architecture, setting a precedent and initiating a new research field focusing on the integration of industrial robots in architectural design and construction. They have contributed to numerous exhibitions around the world, such as the 2008 Architectural Biennial in Venice, the Storefront Gallery for Art and Architecture in New York in 2009, Flight Assembled Architecture at the FRAC Centre Orleans in 2011, and Rock Print at the 2015 Chicago Architecture Biennial. Their recent research is outlined and theoretically framed in the book, The Robotic Touch: How Robots Change Architecture.
265
20 Back Matter.indd 265
2018-11-04 1:33 PM
KENT LARSON MIT
Cambridge, MA, USA Kent Larson directs the City Science Initiative at the MIT Media Lab, USA. His research focuses on developing urban interventions that enable more entrepreneurial, livable, high-performance districts in cities. To that end, his projects include advanced simulation and augmented reality for urban design, transformable micro-housing for millennials, mobility-on-demand systems that create alternatives to private automobiles, and urban living lab deployments in Hamburg, Andorra, Taipei, and Boston. Larson and researchers from his group received the “10-Year Impact Award” from UbiComp in both 2014 and 2017. This is a “test of time” award for work that, with the benefit of hindsight, has had the greatest impact over the previous decade. Larson practiced architecture for 15 years in New York City before joining the MIT Media Lab. His book, Louis I. Kahn: Unbuilt Masterworks was selected as one of the Ten Best Books in Architecture of the year by the New York Times Review of Books. He is a founder of ORI, a company focused on the commercialization of architectural robotics, and Larson Living Labs, which provides design, technology, and community engagement consulting services for large-scale urban projects.
GREG LYNN GLForm Los Angeles, CA, USA Greg Lynn was an innovator in redefining the medium of design with digital technology as well as pioneering the fabrication and manufacture of complex functional and ergonomic forms using CNC (Computer Numerically Controlled) machinery. His work is in the permanent collections of the most important design and architecture museums in the world, including the CCA, SFMoMA, ICA Chicago, and MoMA. In addition to designing consumer products using new materials and manufacturing technologies with companies like Vitra, Alessi, Nike, and Swarovski, he is also a co-founder and Chief Creative Officer of the Boston-based intelligent lightweight mobility company Piaggio Fast Forward. In 2002, he left his position as Professor of Spatial Conception and Exploration at ETHZ (Swiss Federal Institute of Technology Zurich) and became an Ordentlicher University Professor at the University of Applied Arts in Vienna. He is a Studio Professor at UCLA’s School of Architecture and Urban Design where he is currently spearheading the development of an experimental research robotics lab. He was the Davenport Visiting Professor at Yale University from 1999–2016 and Visiting Professor at Harvard GSD for the 2017–2018 academic year. He graduated from Miami University of Ohio with degrees in both architecture and philosophy and later from Princeton University where he received a graduate degree in architecture. He received the American Academy of Arts & Letters Architecture Award in 2003. In 2008, he won the Golden Lion at the 11th International Venice Biennale of Architecture. In 2010, he was awarded a fellowship from United States Artists.
266
20 Back Matter.indd 266
2018-11-04 1:33 PM
FRANK PILLER RWTH
Aachen, Germany Frank Piller is a Chair Professor of Management at the Technology & Innovation Management Group of RWTH Aachen University, Germany, one of Europe’s leading institutes of technology. He is also a founding faculty member of the MIT Smart Customization Group. Prior to his current position in Aachen, he worked at the MIT Sloan School of Management (2004–2007) and was an Associate Professor of Management at TUM Business School at the Technical University of Munich (1999–2004). His research focuses on value cocreation between businesses and customers/users and the interface between innovation management, operations management, and marketing. He is regarded as one of the leading experts in the fields of mass customization and customer-centric value creation. His 1997 article in the German edition of the Harvard Business Review and his first book on mass customization put this topic onto the management agenda in Germany and other European countries. He is the author of numerous papers and has written or edited six books. His Mass Customization book, published by Gabler Verlag in German, was awarded five prizes. His recent analysis of “Threadless” (co-authored with Susumu Ogawa), an innovative crowdsourcing business model in the fashion industry, was chosen as one of the Top 20 articles in MIT Sloan Management Review.
B. JOSEPH PINE II Strategic Horizons Dellwood, MN, USA B. Joseph Pine II is an internationally acclaimed author, speaker, and management advisor to Fortune 500 companies and entrepreneurial start-ups alike. He is cofounder of Strategic Horizons LLP, a thinking studio dedicated to helping businesses conceive and design new ways of adding value to their economic offerings. In 1999, Pine and his partner James H. Gilmore wrote the best-selling book The Experience Economy: Work Is Theatre & Every Business a Stage, which demonstrates how goods and services are no longer enough; what companies must offer today are experiences – memorable events that engage each customer in an inherently personal way. Published in fifteen languages and named one of the 100 best business books of all time by 800ceoread, in 2011, The Experience Economy came out for the first time in paperback as an updated edition with new ideas, new frameworks, and many, many new exemplars. He also co-wrote Infinite Possibility: Creating Customer Value on the Digital Frontier with Kim C. Korn, Authenticity: What Consumers Really Want with James H. Gilmore, and in 1993 published his first book, the award-winning Mass Customization: The New Frontier in Business Competition. Pine consults with numerous companies around the world, helping them embrace the ideas and frameworks he writes about, in order to develop concepts for creating more economic value, and see those concepts become reality. In his speaking and teaching activities, he has addressed the World Economic Forum, the original TED conference, and the Consumer Electronics Show. Today he is a lecturer in Columbia University’s Technology Management program in the School of Professional Studies. 267
20 Back Matter.indd 267
2018-11-04 1:33 PM
RONALD RAEL Emerging Objects Oakland, CA, USA
VIRGINIA SAN FRATELLO Emerging Objects Oakland, CA, USA
Ronald Rael is an Associate Professor and the Chair of the Graduate Committee in the Department of Architecture at UC Berkeley, CA, USA. He directs the printFARM Laboratory (print Facility for Architecture, Research and Materials), holds a joint appointment in the Department of Architecture, in the College of Environmental Design, and the Department of Art Practice. He is a partner in Rael San Fratello and Emerging Objects with Virginia San Fratello. Rael is the author of Borderwall as Architecture: A Manifesto for the U.S.-Mexico Boundary, which advocates a reconsideration of the barrier dividing the US and Mexico through design proposals that are hyperboles of actual scenarios that have occurred as a consequence of the wall, and Earth Architecture, a history of building with earth in the modern era to exemplify new, creative uses of the oldest building material on the planet.
Virginia San Fratello is an architect, artist, and educator. She is an Associate Professor in the Department of Design at San Jose State University, USA, and Director of the Interior Design Program. San Fratello recently won the International Interior Design Educator of the Year Award and her creative practice, Rael San Fratello (with Ronald Rael), was named an Emerging Voice by the Architectural League of New York. She is also a winner of the Metropolis magazine Next Gen Design Competition. San Fratello, along with her partner Rael, is the author of the book Printing Architecture: Innovative Recipes for 3D Printing. The work of Rael San Fratello has been published widely, including in Interior Design, Architect, the New York Times, Wired, MARK, Domus, Metropolis, and PRAXIS and is recognized by several institutions, including the San Francisco Museum of Modern Art, MoMA, and the Cooper Hewitt, Smithsonian Design Museum.
268
20 Back Matter.indd 268
2018-11-04 1:33 PM
CHRISTOPHER SHARPLES SHoP Architects New York, NY, USA
RYAN E. SMITH Washington State University Pullman, WA, USA
SHoP Architects was founded to harness the power of diverse expertise in the design of buildings and environments that improve the quality of public life. The studio has been recognized with awards such as Design Miami/Art Basel’s “Design Visionary” designation in 2016, Fast Company’s “Most Innovative Architecture Firm in the World” in 2014, and the Cooper Hewitt, Smithsonian Design Museum’s “National Design Award for Architecture” in 2009. As a founding partner of SHoP, Christopher has led many of its most significant projects, including the Barclays Center, the East River Waterfront Esplanade, the Botswana Innovation Hub, SITE Santa Fe, the Fashion Institute of Technology, a new headquarters complex for Uber, and B2, the tallest modular residential building in the world. He has held prestigious teaching positions at Cornell University, Parsons School of Design, Yale University, the City College of New York, Columbia University, and the University of Virginia.
Ryan E. Smith is Director and Professor in the School of Design and Construction at Washington State University. He has researched and consulted on architectural production with an emphasis on design for manufacture and assembly (DfMA) and prefabrication since 2004. Smith is author/ co-author of Prefab Architecture (Wiley, 2010), Building Systems (Routledge, 2012), Leading Collaborative Architectural Practice (Wiley, 2017), Offsite Architecture (Routledge, 2017), and Prefab Housing and the Future of Building (Lund & Humphries, 2018). He is founder and past Chair of the National Institute of Building Sciences, the Offsite Construction Council and a Senior Research Fellow in the Centre for Offsite Construction & Innovative Structures at Edinburgh Napier University in the UK.
269
20 Back Matter.indd 269
2018-11-04 1:33 PM
PHILIPPE STARCK STARCK
Paris, France Philippe Starck is a French designer known for his prolific work in many areas, from products of our daily life (lemon juicer, furniture, electric bike, or individual wind turbines) to architecture (hotels and restaurants) and space and naval engineering (mega yachts). But his global notoriety and his protean inventiveness must not overshadow the essential. He considers as a duty to share his ethical and subversive vision of a fairer planet, and creates unconventional places and objects whose purpose is to be “good” before being beautiful. His latest projects include the renovation of the Gran Caffé Quadri in Venice, the interior design of a habitation module for private space tourism with Axiom, the launch of a new fragrance collection Starck Paris, and Maison Heler Metz, an out-of-scale hotel with phantasmagorical architecture. All his creations, whatever form they take, must improve people’s lives. Phillipe Starck is one of the pioneers and one of the central figures of the concept of “democratic design.” He is now engaged in a new battle for “democratic architecture” and “democratic ecology.”
JOSEPH TANNEY Resolution: 4 Architecture (RES4) New York, NY, USA Joseph Tanney founded Resolution: 4 Architecture (RES4) in 1990 with Robert Luntz. Since its inception, his ten-person New York office has been internationally published and highly acclaimed, completing projects in the residential, commercial, and public realms. RES4’s most recent preoccupation is THE MODERN MODULAR, a systematic methodology of design that leverages existing methods of residential prefabrication. In 2003, RES4 won an international competition for a modern prefabricated home, The Dwell Home. In addition to having been profiled in newspapers nation-wide, Tanney’s work has been featured in popular magazines, such as Time, Newsweek, WIRED, and Wallpaper among others, and architectural publications such as Architecture, Architectural Record, Architectural Review, Domus, etc. Tanney has lectured throughout the US and has been a visiting critic at numerous universities. His work has been exhibited in various museums including the Walker Art Center in Minneapolis, the Hammer Museum in Los Angles, and MOMA in New York. In addition to recently receiving consecutive AIA Design Awards over the last several years, THE MODERN MODULAR has received a 2005 American Architecture Award, a 2006 National AIA Housing Award for Concepts in Innovative Housing, and a 2006 Honor Award for Housing Design Research from the Boston Society of Architects and the AIA New York Chapter.
270
20 Back Matter.indd 270
2018-11-04 1:33 PM
TOM VEREBES OCEAN CN
Hong Kong, China Tom Verebes is the Associate Dean for Academic Affairs and tenured Professor in the School of Architecture & Design at New York Institute of Technology (NYIT). He is also the Director of his practice, OCEAN CN, based in Hong Kong. He has been the director of the AA Visiting School Shanghai (2007–2018), and AA Visiting School in Shenzhen (2018). His former roles include Provost of Turenscape Academy (Beijing; Huangshan, Anhui Province), Associate Dean for Teaching & Learning (2011–2014) and Associate Professor at the University of Hong Kong (2009–2016), Co-Director of the Design Research Lab at the AA in London, where he had taught from 1996 to 2009, and Guest Professor at Akademie der Buildenden Künste (ABK), Stuttgart (2004–2006). Recent visiting roles include Visiting Professor at the University of Pennsylvania, Professor of Practice at RPI, Visiting Critic at Syracuse University, RMIT, Singapore University of Technology & Design, and the University of Tokyo. Among over 150 publications, he has published Masterplanning the Adaptive City: Computational Urbanism in the Twenty-first Century (Routledge, 2013), a guest-edited issue of Architectural Design entitled “Mass Customised Cities” (2015), and Shanghai Ten Folio (ORO, 2017). Verebes’ work has been exhibited in over 50 venues worldwide, and he has lectured extensively in Asia, Europe, North America, Africa, and the Middle East.
271
20 Back Matter.indd 271
2018-11-04 1:33 PM
272
20 Back Matter.indd 272
2018-11-04 1:33 PM
PHOTO CREDITS COVER
Gramazio Kohler Architects
1 Kolarevic & Duarte 1.1 Reebok 1.2 Blue Homes 1.3 Housebrand 1.4 Resolution: 4 Architecture 1.5 Bernard Cache 1.6 Fabio Gramazio and Matthias Kohler 2 Pine 2.1 Stan Davis 2.2–6 Joseph Pine 3 Piller 3.1 Frank Piller 4 4.1 4.2 4.3–9 4.10–15
San Fratello & Rael Emerging Objects Matthew Millman Emerging Objects Matthew Millman
5 Fornes 5.1a–b MARC FORNES / THEVERYMANY 5.2 Brice Pelleschi 5.3–6 MARC FORNES / THEVERYMANY 5.7 Miguel de Guzmán / Imagen Subliminal 5.8 Brice Pelleschi 5.9 NAARO 5.10 Brice Pelleschi 5.11–12 MARC FORNES / THEVERYMANY 6 Lynn 6.1–2 Greg Lynn FORM 6.3 Marvin Rand / Greg Lynn FORM 6.4 Carlo Lavatori / Greg Lynn FORM 6.5 Greg Lynn FORM 6.6 Bronny Daniels / Greg Lynn FORM 6.7 Greg Lynn FORM 7 Starck 7.1–3 Nicole Marnati / TOG / Starck Network 8 Ashuach 8.1–19 Assa Ashuach Studio, except: 8.7 Jon Vickers 9 9.1–8 9.9–10
Gramazio Kohler Gramazio Kohler Architects Roman Keller
9.11 9.12 9.13 9.14 9.15 9.16–18
Gramazio Kohler Architects Gramazio Kohler Research, ETH Zürich Jules Spinatsch Gramazio Kohler Architects Alessandra Bello Gramazio Kohler Research, ETH Zürich
10 Kolarevic 10.1 IDuke, CC BY-SA 2.5 10.2 Nia Garner 10.3–5 Branko Kolarevic 11 Duarte 11.1–9 José Pinto Duarte 11.10–11 Debora Benros 11.12a–b Sara Eloy 12 Daubmann 12.1–3 BLU Homes 12.4 Karl Daubmann 12.5 BLU Homes 12.6 US Patent website 12.7 BLU Homes 12.8 Karl Daubmann 12.8–9 BLU Homes 12.10–12 Karl Daubmann 12.13–14 BLU Homes 13 Tanney 13.1–19 Resolution: 4 Architecture 14 Smith 14.1 Ryan E. Smith 14.2 Helena Lidelöw, Luleå University & Lindbacks 14.3 Freddie Mac 14.4–7 Ryan E. Smith 15 Brown 15.1–10 Housebrand 16 Sharples 16.1 Seong Kwon 16.2–4 SHoP Architects 16.5 Seong Kwon 16.6 SHoP Architects 16.7 Magda Biernat 16.8–9 SHoP Architects 16.10 Robin Hill 16.11 Tom Harris 16.12–13 SHoP Architects
273
20 Back Matter.indd 273
2018-11-04 1:33 PM
17 Verebes 17.1–2 OCEAN CN / Crystal CG / LaN 17.3–6 OCEAN CN & Arup 17.7–10 dotA & OCEAN CN 17.11–12 OCEAN CN & SED Landscape Architects 18 Larson 18.1 Kent Larson 18.2 Verkhovensky, CC BY 2.0 18.3 John J. Kim, Chicago Tribune 18.4–5 Ariel Noyman 18.6–7 Carson Smuts 18.8 Jimmy Day 18.9a–c Michael Lin (scans), Jimmy Day (photos) 18.10–12 Kent Larson 18.13 Luis Alonso 18.14 Ariel Noyman, Walter Schiesswohl 18.15 Walter Schiesswohl 19 Fisher 19.1 Krzysztof Aaron, Flickr CC BY-SA 2.0 19.2 Forgemind ArchiMedia, Flickr CC 19.3 Jordanhill School D&T Dept, Flickr CC 19.4 3dilla, Flickr CC 19.5 Andy Roberts, Flickr CC 19.6 Kristain Baty, Flickr CC BY-SA 2.0 19.7 La Citta Vita, Flickr CC BY-SA 2.0 19.8 MacKay-Lyons Sweetapple Architects (GS+, Flickr CC)
274
20 Back Matter.indd 274
2018-11-04 1:33 PM
PROJECT CREDITS CHAPTER 17 TOM VEREBES
CHAPTER 18 KENT LARSON
Umekita Second Development Area OCEAN CN: Tom Verebes (design director) Mohamad Ghamloush (designer), Nathan Melenbrink (designer, scripting), Andrew Haas (post-production, scripting), Terence Ho; ARUP Hong Kong: Mark Swift, Gabriel Lam, William Loasby, Ray Tang, Ben Luk
CityScope Andorra Luis Alonso, Arnaud Grignard, Ariel Noyman, Yan Zhang, Dalma Foldesi, Jung In Seo, Juanita Devis, Núria Macià (Fundació ActuaTech), Marc Vilella (OBSA), Marc Pons (OBSA), Guillem Francisco Giné (OBSA), Cristina Yañez (UdA), Kent Larson
Yan Jiao Hua Run 4D City Masterplan dotA: Gao Yan, Duo Ning, Chiang Qiang, Wang Xin, Crystal Yiu; OCEAN CN: Tom Verebes, Nathan Melenbrink
CityScope Kendall Ira Winder, Carson Smuts, Mohammad Hadhrawi, J.T. White, Estelle Yoon, Suramya Kedia, Sotirios Kotsopoulos, Manos Saratsis, Kent Larson
Xiangmi Park OCEAN CN: Tom Verebes, Andrew Haas; SED: William Huang, Grace Gu
CityScope Volpe Luis Alonso, Yan Zhang, Arnaud Grignard, Ariel Noyman, Yasushi Sakai, Markus ElKatsha, Ronan Doorley, Kent Larson
Parametric Pearl River Delta OCEAN CN: Tom Verebes, Li Bin; Crystal CG: Gao Yan; LaN: Luis Fraguada
FindingPlaces Ariel Noyman, Tobias Holtz, Johannes Kröger, Katrin Hovy, Nina Halker, Gesa Ziemer, Kent Larson MIT CityHome Hasier Larrea, Daniel Goodman, Oier Ariño, Phillip Ewing, Kent Larson PEV Michael Lin, Abhishek Agarwal, Phil Tinn, Chang-Qi Zhang, Yi-Cheng Jiang, Tai-Yu Chen, Kent Larson RePlace Carson Smuts, Jason Nawyn, Kent Larson Reversed Urbanism Ariel Noyman, Ronan Doorley, Zhekun Xiong, Luis Alonso, Arnaud Grignard, Esteban Moro, Kent Larson TerMITes Carson Smuts, Jason Nawyn, Lucas Cassiano Pereira Silva, Chrisoula Kapelonis, Kent Larson
275
20 Back Matter.indd 275
2018-11-04 1:33 PM
INDEX A adaptability, 188, 190, 191, 194, 212 adaptable architecture, 191 adaptation, 88, 95, 96, 182, 191 adaptive building, 10, 186, 190, 195 customization, 37 adaptivity, 219 additive manufacturing, 38, 42, 47, 48, 88, 89, 92, 96 Adidas, 71 agile manufacturing, 80 AI (see artificial intelligence) Alessi, 70, 74–76 Alberto, 74–76 artificial intelligence (AI), 94, 96, 98, 102, 133, 210 Ashuach, Assa, 10 assembly, 61, 64, 65, 165, 180, 202, 203, 212 associative geometry, 123 assortment matching, 36 authorship, 6, 116, 126 auto(mobile) industry, 78, 183 automation, 8, 9, 110, 179, 182, 188, 209, 213 flexible, 35, 37 autonomous learning, 96
B Bangle, Chris, 78 behavior pattern, 95 bespoke consumer objects, 70 performance, 80 product, 70, 85 structure, 80 BIM (see building information modeling) biomorphic, 44 Blu Homes, 4, 5, 10, 145, 148–150, 154–156, 177, 183 BMW, 30, 35, 58, 78 Brown, John, 9, 10 building information modeling (BIM), 151, 153, 198
C Cache, Bernard, 6, 7 CAM (see computer-aided manufacturing) Castells, Manuel, 216 Chocri, 33 choice, 8 complexity, 35 navigation, 31, 32, 34–36 CNC, 2, 7, 70, 74, 80, 103, 104, 108, 182, 199, 209, 254 co-design, 11, 35–37, 88, 89, 91, 125, 187, 189, 195
experience, 88, 89 platform, 36 system, 187 toolkit, 36 co-designing, 125, 126, 195 code (building), 5, 150, 155 color variation, 2, 8, 11 combinatorial variation, 10 complementarity theory, 37 complexity, 34, 35, 42, 116 computation, 56, 58, 102, 108 computational design, 102, 109, 217 computer-aided design (CAD), 42, 179 manufacturing (CAM), 42, 104, 179 concretization, 11 configuration system, 34, 36 toolkit, 35, 37 configurator, 33, 36, 72, 79, 104, 135, 154 constraints-based modeling, 124 consumer-configured shoes, 79 cosmetic customization, 11, 125, 126 Courouble, Fred, 80 craft, 42, 51, 181, 182, 209, 222 CRM (see customer relationship management) customer, 7 co-design, 36 experience, 33 interaction, 7 preferences, 34, 35 relationship management (CRM), 17, 18 satisfaction, 8, 11, 22, 130 customering, 9, 14, 16, 18–20, 26, 27 customization adaptive, 37, 120 continuous, 120 conventional, 34 cosmetic, 11, 125, 126 dimensional, 118, 119 feature, 118 geometric, 118–120 initial, 118, 120, 191, 194 massive, 3, 8, 9, 56, 58, 59, 68, 119 material, 118 modular, 118 product, 34 rule sets, 59 space, 36 surface, 118 system, 8
276
20 Back Matter.indd 276
2018-11-04 1:33 PM
D data analytics, 38 Daubmann, Karl, 9, 10 Davis, Stan, 9, 14, 15, 21, 24, 25 deformation, 7 delayed product differentiation, 34 democratic craft, 42 design, 42, 84 ecology, 84 democratized design, 70, 183 workflow, 207 design customizable, 126 experience, 88 information, 108 space, 124, 126 system, 8, 11, 120, 130–135, 137 world, 122, 130 differentiation, 34 serial, 73 digital fabrication, 2, 4, 6, 9, 109, 119, 120, 126, 132 dimensional customization, 118, 119 variation, 2, 8, 9, 11 downloadable design, 254
fragmentation, 38 Frampton, Kenneth, 221 Fritz, William, 148 Fuller, Buckminster, 160 functional organization, 8 performance, 9 fused deposition modeling, 42, 98
G G-code, 45, 52 generative design, 133 form-finding, 228 rules, 124 system, 132 genius platform, 26 geographic information system (GIS), 235, 259 geometric customization, 118–120 Gramazio, Fabio, 7, 10 Gramazio Kohler, 7 grammar (see shape grammar) Gropius, Walter, 160, 177
H
Eisenman, Peter, 70, 122, 123 embedded configuration capability, 36 toolkit, 37 esthetics, 125 evaluation, 132, 134 extreme mass production, 69, 78, 81
Habraken, John, 8, 131 handcraft, 9 heterogeneity, 31, 38, 120, 217, 220, 229 homeomorphism, 121 house, 10, 144, 145, 148, 149, 155, 159, 160, 166–173, 177, 186, 191, 194 design, 4, 5, 8 layout, 8, 189 type, 8, 131, 137 Housebrand, 4, 5 housing, 2, 6, 8, 9–11, 119, 120, 126, 130, 137, 138, 155, 156, 178, 183, 195, 210, 241
F
I
fabrication (see also digital fabrication), 42, 53, 56, 61, 64, 146, 165, 166, 179, 206, 209, 212, 217, 222, 254 Factory OS, 183 FDM (see fused deposition modeling) Fiat, 33 Fisher, Thomas, 10, 11 fitting, 181, 182 flexibility, 14, 34, 35, 194, 195 flexible automation, 35, 37 Ford, 146, 182, 199 Fornes, Marc, 9
IBM, 14, 15, 35 Ikea, 202 indeterminacy, 56, 220 individuality, 45 individualization, 24, 38 initial customization, 118, 120, 191, 194 integration, 130, 195 intelligence, 26, 33, 99 (see also artificial intelligence) intelligent agent, 96 interaction, 7, 100, 126 experience, 88, 100 system, 36
E
277
20 Back Matter.indd 277
2018-11-04 1:33 PM
interactive application, 7 controls, 124 design, 7 manipulation, 6, 125 platform, 244 toolkit, 30 website, 2, 4–6, 119, 120, 126 interchangeability, 181, 182 interface, 104, 191
J Jacobs, Jane, 259
K Kasita, 183 Katerra, 183 Kaufmann, Michelle, 177 Kieran Timberlake, 146, 183 Knausgaard, Karl Ove, 118 Knight, Terry, 137 Kohler, Matthias, 7, 10 Koolhaas, Rem, 220 Kusama, Yayoi, 66, 68 Kwinter, Sanford, 70, 78
L Larson, Kent, 10, 11 Le Corbusier, 160, 177, 220 Lego, 20, 118, 183, 202 Leupen, Bernard, 191 Lindbacks, 183 Living Homes, 183 Louis Vuitton, 66 Lynn, Greg, 10, 102
M MakeTime, 80 Mantel, Stone, 26 March, Lionel, 122 market analysis, 178 development, 24, 25 research, 32, 146 method, 32 methodology, 32 share, 38 marketing, 9, 14, 16–18, 20, 27, 35 mass production, 3, 10, 16, 20, 33, 34, 38, 44, 70, 76, 78, 84, 88, 118, 160, 174, 176, 179, 216, 220, 250
massive customization, 3, 8–10, 56, 58, 59, 61, 68, 119 masterplanning, 219, 221 material customization, 118 innovation, 42 variation, 2, 11 McCullough, Malcolm, 121 McLuhan, Marshall, 257 metadesign, 120, 121, 123, 124 metatype, 120 Mitchell, William, 121, 122 modular, 35, 42, 148, 152, 160, 164–167, 174, 176, 187, 190–192, 194, 195, 210, 212 building block, 42 cabinetry, 187, 190–192 components, 191, 194 construction, 147, 148, 161, 180 customization, 118 design, 5 house, 4, 5, 144, 147 panelization, 125 strategy, 10 system, 194 modularity, 20, 21, 35, 182, 183, 191, 194 module, 35, 148, 159, 160, 162, 166, 169, 182, 189, 191, 192, 194, 212 morphology, 60, 98 mymuesli, 30, 37
N Net Promoter Score (NPS), 22 Nike, 59, 70, 71, 79, 80 non-standard, 6, 124 design, 217 products, 7 shapes, 124 Norton, Dave, 26 NPS (see Net Promoter Score)
O Ocean Consultancy Network, 10, 216–218 off-site construction, 146, 147, 160, 173, 210, 211 on-demand economy, 250–252 machining, 80 production, 42, 46, 90 on-site construction, 147 installation, 2 open architecture, 206, 213 optimization, 95–97
278
20 Back Matter.indd 278
2018-11-04 1:33 PM
P Palladio, Andrea, 120 panelized construction, 147 parameter, 45, 58, 59, 79, 108, 121, 123, 124, 136, 138, 160 parametric definition, 120, 121 design, 2, 4, 6–8, 58, 103, 107, 109, 115, 119, 123, 124, 126, 132, 176, 178 engine, 111, 124 hierarchy, 121, 124, 126 interdependencies, 126 logic, 108 manipulation, 125 minimalism, 124 model, 122, 124, 132 system, 8, 115, 124, 126, 127 thinking, 109 variation, 120, 121, 124 parameterization, 105 participation, 245 Peppers, Don, 18 performance, 6, 37, 64, 80, 81, 95, 97, 124, 133, 134, 146, 191 evaluation, 125 functional, 8, 9, 34 measurement system, 38 structural, 64, 65 value chain, 32 permutations, 32 personalization, 38, 70, 88–90, 182 Piaggio, 80 Picon, Antoine, 220 Piller, Frank, 9 Pine, Joseph, 3, 9, 30, 118 prefabricated house, 5, 144, 167 prefabrication, 2, 10, 146, 176–183, 187, 210 procedural modeling, 70 operations, 71, 72 tools, 71, 73 variation, 72 process customization, 109 design, 31–33 modularity, 35 satisfaction, 36 product adaptation, 88 attributes, 31, 32 development, 38, 178
differentiation, 34 personalization, 88 platform, 38 reconfiguration, 88 solution, 36 variety, 34 production theory, 177 Project Frog, 183
R Radziner, Marmol, 177 Rael, Ronald, 9 Rand, Paul, 60 rapid prototyping (RP), 92 recombination, 57, 58, 62 reconfiguration, 88, 89 Reebok, 3 relational structure, 122 relations-based modeling, 124 resilience, 187, 188, 191, 195 Resolution: 4 Architecture (RES4), 5, 159, 183 risk-pooling effect, 34 robot, 109–111, 146, 207, 213 robotic assembly, 126 Rogers, Martha, 18 Roy, Lindy, 70, 78 Ruskin, John, 198, 213
S Safdie, Moshe, 177 San Fratello, Virginia, 9 Sassen, Saskia, 221 scalability, 38, 201 Schnapp, Jeffrey, 80 Searls, Doc, 18 Sekisui Heim, 183 selective laser sintering (SLS), 88, 92–94 sensor, 26, 95, 96 sentience, 228 shape adaptation, 96 grammar, 8, 122, 132, 134, 136–141 modification, 89 reconfiguration, 89, 90 Sharples, Christopher, 10 SHoP Architects, 198, 202, 206, 210 Simon, Herbert, 130 Simplex Homes, 183 simulation, 9, 133, 243 sintering, 76, 88, 94, 97 (see also selective laser sintering)
279
20 Back Matter.indd 279
2018-11-04 1:33 PM
Siza, Alvaro, 134–136 SLS (see selective laser sintering) smart product, 26, 37, 38, 70 Smith, Ryan, 9, 10 solution space, 31–33, 34, 36, 38 development, 31–33, 37 spatial variation, 11 standardization, 2, 176, 179, 182 Starck, Philippe, 10 Steadman, Philip, 122 stereolitography, 70, 72 supply chain, 33, 34, 38, 90 surface customization, 118 Swarovski, 70
part-to-part, 60 spatial, 11 texture, 2, 8 topological, 2, 8, 9, 11 variety, 14, 20, 30, 34, 73, 182 vendor relationship management (VRM), 18 Verebes, Tom, 10, 11 virtual construction, 198 design, 198 object, 95 product, 96, 100 Volvo, 70, 78 VRM (see vendor relationship management)
T
W
Tanney, Joseph, 9, 10, 78 tessellation, 56, 57 texture variation, 2, 8 Toffler, Alvin, 9 TOG, 84 Toll Brothers, 4 topological definition, 122, 123 operations, 122 structure, 121, 122 variation 2, 8, 9, 11 topology, 57–59, 91, 120–122, 124 typology, 5, 120, 121, 161, 162, 164, 167, 191
waste reduction, 93 waste-free manufacturing, 42, 46 website, 2, 4, 5, 8, 34, 119 Wright, Frank Lloyd, 134, 160, 177
Z zoning, 5
0–9 3D knitting, 71 3D printing, 2, 38, 42–53, 70, 72, 74, 76, 77, 80, 110, 254
U urban modeling, 243 urbanization, 216, 232 user interaction, 88, 100 persona, 38 user-configured product, 79
V vacuum-forming, 72 value chain, 32, 33, 35 creation, 34 variability, 35, 47, 125, 182 variation, 8, 45, 123, 135, 182, 183, 220, 222 color, 2, 8, 11 combinatorial, 10 crafting, 45 dimensional, 2, 8, 9, 11 industrialized, 70 material, 2, 11
280
20 Back Matter.indd 280
2018-11-04 1:33 PM