Transforming the Bodleian 9783110289398, 9783110289213

Citation preview

Transforming the Bodleian

Current Topics in Library and Information Practice

Transforming the Bodleian Edited by Michael Heaney Executive Secretary, the Bodleian Libraries and Catríona Cannon Associate Director (Collection Support), the Bodleian Libraries

ISBN 978-3-11-028921-3 e-ISBN 978-3-11-028939-8 ISSN 2192-2742 Library of Congress Cataloging-in-Publication data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.d-nb.de. © 2012 Walter de Gruyter GmbH, Berlin/Boston and The Bodleian Libraries, University of Oxford Typesetting: Medien Profis GmbH, Leipzig Printing: Hubert & Co. GmbH & Co. KG, Göttingen ♾ Printed on acid-free paper Printed in Germany www.degruyter.com

Preface The genesis for this book came at the beginning of 2012 as a time of enormous upheaval and unprecedented change at the Bodleian Libraries came to an end. We were, and are, proud of what the staff of the Libraries achieved and wanted to celebrate it. More importantly, we realized that we had accomplished something unprecedented and wanted to share with our professional colleagues what we had done and what we had learned. We successfully delivered several major projects simultaneously, and we showed how innovative methods and a dedicated problem-solving approach enabled us to code and move millions of books and other items in record time. We did all this without a break in service. We achieved it by the integrated management of complex interlinked projects and by bold financial planning enabling multi-million pound budgets to be met and savings made. The Bodleian Libraries’ collections are extremely large and complex – a fourhundred-year history might be seen as a drag on flexibility and responsiveness. But we showed between 2008 and 2011 how major libraries can meet twenty-firstcentury challenges and provide a model for the management of major library projects. The majority of the chapters in this book have been written by the people who actually played the major role in delivering the projects under the leadership of Bodley’s Librarian Sarah Thomas. The bookmoves were overseen by Bruce Wainwright, who has since moved on from the Libraries on the completion of a remarkable set of bookmoves. Michael Williams was responsible for perhaps the most complex and demanding of the projects, the barcoding, and has been assisted in the description of the achievement by two of his colleagues, Alasdair MacDonald and Cath Walsh. Andrew Bonnie was responsible for the delivery of both the integrated library system and the warehouse management system under the Head of the Systems and e-Resource Service Dave Price. Michael Heaney came relatively late into the staff relocation project, for which much of the meticulous planning and execution was carried out by Denis O’Driscoll. Richard Ovenden had overall responsibility not just for the Weston Library but also for all aspects of the decanting of its contents and the construction of the Book Storage Facility. The reader service provisions were overseen in different aspects and at different stages by several staff, who have collaborated in describing their activities. None of our contributors and project leaders could have reached their objectives without the whole-hearted support and co-operation of virtually the entire staff of the Libraries.

VI 

 Preface

Not only have our authors demonstrated great initiative in delivering their projects at such a challenging pace, they have also delivered the contents of this book to equally tight deadlines, and we are grateful to them for their industry and patience in the face of editorial impositions. Each of the chapters can be read and understood independently of the others, though we have tried to keep repetition to a minimum. At the same time the scale and complexity of the achievement can only be appreciated by reading the whole. At word of explanation about terminology may be helpful. ‘The Bodleian Library’ is the building in the middle of Oxford where the library service began. It is sometimes known as ‘the Old Bodleian Library’. ‘The New Bodleian Library’ is the building constructed a short distance away from the Bodleian Library in the 1930s. In this book the term ‘Bodleian Library’ is used for the original building, or ‘Old Bodleian Library’ where it is being contrasted or juxtaposed with the New. Together with the Radcliffe Camera these constitute ‘the central Bodleian Library’, referred to as such in this book, administered as an operational unit until the integration of library services in 2000. ‘The Bodleian Libraries’ is the integrated library service created in 2000, including the central Bodleian Library and approximately thirty other libraries (the number has varied over the course of the decade). From 2000 until 2010 this integrated library service was known as Oxford University Library Services. In this book the service is referred to throughout as ‘the Bodleian Libraries’ even where this is anachronistic, except where the immediate context (e.g. citation) demands otherwise. Michael Heaney and Catríona Cannon

Contents List of illustrations 

 IX

Sarah E. Thomas Introduction   1 Sarah E. Thomas Chapter 1: Background 

 5

Sarah E. Thomas and Michael Heaney Chapter 2: The Bod Squad   23 Laura How Chapter 3: Finance 

 35

Michael Hughes Chapter 4: The special collections 

 43

Jonathan McAslan Chapter 5: The purchase of electronic journal backfiles  Toby Kirtley  Chapter 6: The building of the Book Storage Facility 

 61

 67

Boyd Rodger Chapter 7: Book Storage Facility ingest and retrievals operations 

 83

Catríona Cannon Chapter 8: ‘And moveth all together, if it move at all’: the Bookmoves Project 2010–2011   95 Michael Williams, Alasdair MacDonald and Cath Walsh Chapter 9: Barcoding – the Inventory Control Project 

 109

Andrew Bonnie Chapter 10: The Book Storage Facility Information System 

 137

Dave Price Chapter 11: Implementation of a new integrated library system 

 149

VIII 

 Content

Michael Heaney Chapter 12: Staff relocation 

 163

Richard Ovenden Chapter 13: The creation of the Weston Library 

 175

James Legg Chapter 14: Underground Bookstore and Old Bodleian Access Project  Donald Mackay, Alena Ptak-Danchak and James Shaw Chapter 15: Reader services during a time of major changes  Catríona Cannon Conclusion   223 Index 

 225

 211

 191

List of illustrations Figure 1.1 1.2 1.3 1.4 1.5

Overcrowding in the New Bodleian bookstack The planned Osney depository: initial design The planned Osney depository: revised design The New Bodleian under construction showing the steel frame The Lankester Room in the Radcliffe Science Library

2.1 2.2 2.3 2.4 2.5 2.6

A Bod Squad dashboard The Bod Squad Log Blog Blog categories and word cloud of tags Blog entry for Facilities Management Bodley’s journey Bodley’s Librarian thanks staff for their achievement

4.1 4.2

Moving the special collections Shipment record

6.1 6.2 6.3 6.4 6.5 6.6 6.7

The Book Storage Facility under construction The Book Storage Facility from the south A narrow aisle Storage tray Planchests High-level order-picker Book Storage Facility plan

7.1 7.2 7.3 7.4

Accessioning A high-level order-picker in action Final design of process bench A filled chamber

8.1 8.2 8.3 8.4 8.5 8.6

Oxford libraries and bookstores The progress plan for the repository at Nuneham Courtenay Taking books from the shelves Sealing the crates before transport Loading onto the van Bookmoves progress spreadsheet: New Bodleian

X 

 List of illustrations

9.1 9.2 9.3 9.4 9.5 9.6

SERS warehouse filled with DeepStore deliveries Inventory Control Project records on SOLO Abbreviated catalogue record Barcoders at a trolley Chart of metres of material barcoded by area Chart of items barcoded by each method

10.1

Shelf, tray and book barcodes

12.1 12.2 12.3 12.4 12.5 12.6

Staff moves spreadsheet Clarendon Basement before the moves Clarendon Basement initial plan Bodleian (Old Schools) Quadrangle before the move Bodleian (Old Schools) Quadrangle after the move Clarendon Basement: final configuration

13.1 13.2 13.3 13.4 13.5

The New Bodleian under construction Illustrative view of the planned Broad Street frontage of the refurbished New Bodleian Libary Illustrative view of the Blackwell Hall Illustrative view of zoning in the redesigned New Bodleian Bodzilla

14.1 14.2 14.3 14.4 14.5 14.6 14.7 14.8

The Gladstone shelving Old and New Bodleian Libraries and Radcliffe Camera Old Bodleian stairwell The tunnel – before The tunnel – after The stairwell to the Radcliffe Camera Informal study area Quiet study area

15.1 15.2 15.3 15.4 15.5

JIT workflow Workload of JIT requests Sources of alternatives Results during downtime The Meebo live chat widget

Sarah E. Thomas

Introduction Before Google, before Amazon, before Wikipedia, and before the British Library, there was the Bodleian Library. This institution has earned a reputation as a destination for scholars and as a compendium of knowledge, as represented through its diverse and rich collections. Officially designated as opening in 1602, the Bodleian was in fact the third incarnation of a library whose predecessors were a chest holding a few precious manuscripts in a chamber of the University Church whose existence was known in 1320, and the first true University library, now known as Duke Humfrey’s Library, a magnificent oak-timbered structure completed in 1488 which occupies the space above Oxford’s medieval Divinity School. Duke Humfrey’s glory faded within the next hundred years, and owing to changing values, the splendid library had been eviscerated by the last decade of the sixteenth century, inspiring the retired diplomat and founder Sir Thomas Bodley to re-establish the institution according to strict regulations and modern standards in 1602. From that time the Bodleian Library has enjoyed renown and respect from scholars around the globe, the ’Republic of the Learned’ for whom Bodley opened the Library, and the portal to the great collections which it has since amassed over the past 400 years. Indeed, it was the weight of the books and journals, manuscripts, maps, music, and other objects that filled the repository of the Bodleian that propelled the transformation of the Libraries in the twenty-first century. As the collection expanded, so, too, did the need for shelving. Twice in the twentieth century Oxford had innovated to resolve pressing issues of space. Early in the 1900s, an underground cavity had been excavated beneath Radcliffe Square between the Radcliffe Camera and the Bodleian Library, and filled with mobile racking, the first of its kind, designed to provide compact shelving for as many as 650,000 volumes – decades of relief from the concern about where to house the growing number of books and journals accumulating in the University. The Underground Bookstore opened in 1912, but already by the 1920s Oxford was debating the possibility of relocating a library store a few miles north to Wolvercote, perhaps following the example of Cambridge, which planned to relocate its university library to the western edge of its lands, with ample room for future expansion. Oxford opted instead for the construction of a massive bookstore, around which were wrapped reading rooms and staff accommodation, in the centre of the city, a few steps away from its historic core research libraries. The New Bodleian Library, begun in 1937, was completed in 1940 and immediately conscripted for the war

2 

 Sarah E. Thomas

effort, housing a variety of bodies including intelligence units until 1945. When King George VI formally opened the Giles Gilbert Scott edifice in 1946, librarians anticipated that its capacity would be copious enough to sustain the Library almost indefinitely. Yet the post-war explosion in publishing, from around 20,000 UK monographs a year in the 1950s to around 70,000 a year in the 1990s, coupled with the Bodleian’s acquisition of large collections of archival materials, left the 20,000 m2 building congested. As a legal deposit library, the Bodleian has the right to claim one copy of every UK and Irish publication, and receipts through this programme, coupled with purchased and donated materials, meant that 350,000 items annually flowed through the New Bodleian’s loading dock. By one informed estimate, over 50 % of the Bodleian’s collection of 9 million printed books had arrived through legal deposit or its predecessor agreement with the Stationers’ Company. At the same time, the professional standards for storage of archival documents had developed, and by the beginning of the twenty-first century it was that clear massive intervention was needed to protect and store the collections. Rehabilitating the existing stacks in the New Bodleian, no longer fit for purpose, would have been prohibitively expensive. Instead, the Bodleian, like many other research libraries suffering from similar space and environmental pressures, developed a plan to relocate a major portion of its closed stack collection to a new book depository, to be constructed at the edge of the city in an industrial park. High-density remote storage had been pioneered by Harvard in the 1980s, and dozens of US and European libraries had developed variations on the theme. In the UK, the British Library was preparing to tender for a high-bay facility with an automated storage and retrieval system (ASRS) in a low-oxygen environment.¹ Following a review of storage facilities from California to Norway, the Bodleian settled on a depository that, like the British Library’s Additional Storage Building, would offer robotic retrieval and a low-oxygen environment for fire suppression. The rapid automated retrieval and the Osney Mead location, approximately 20 minutes away, would ensure that very little degradation in service would occur as a result of transferring a large segment of the collection away from the central site. Much of the material was served from closed stacks. Readers would wait an average of 120 minutes for delivery of books from the main Bodleian book-

1 Dawn Olney, ‘A UK first: an automated, high-density storage solution for the British Library’, [paper presented at] Where shall we put it? Spotlight on collection storage issues, NPO conference, London, 2004, accessed 28 June 2012, http://www.bl.uk/aboutus/stratpolprog/ ccare/pubs/2004/DOlneyNPO %20Conf2004.pdf

Introduction 

 3

stacks housing 3.5 million items, or several hours for items from the depository at Nuneham Courtenay, a building constructed in stages from the 1970s to the 2000s, situated in the countryside about 30 minutes away and holding over 1.2 million volumes. The move to remote storage was therefore less wrenching from a service perspective than similar decisions taken by research libraries transferring collections from browsable shelves to offsite storage. The saga of the depository, its planning and rejection, its metamorphosis into a different breed of remote storage and the concomitant revolution in the Bodleian Libraries’ thinking, is the impetus for this book. It is the physical manifestation of the transformation of a venerable institution on the decline into one which is becoming a twenty-first-century leader. The Bodleian rests on great traditions, but one of those traditions is to reinvent itself and to introduce innovations which take the library and its services to new heights. Such it was when Bodley, finding the university library at its nadir in the 1590s, Duke Humfrey’s manuscripts dispersed and destroyed, and even the slanted desks to which they had been chained sold on, rebuilt the reading room, redesigning it to incorporate modern standing cases for the printed books that had become the staple of knowledge dissemination. The Bodleian had celebrated its four-hundredth anniversary in 2002, ushering in the twenty-first century with much fanfare, extolling its brilliant historic collections, studded with gems of medieval manuscripts; Arabic, Chinese, Hebrew and Persian treasures; great swathes of early printed books, particularly comprehensive in seventeenth- and eighteenth-century English and European publications; maps, music, and political papers. And, following decades of internal Oxford reports, the recommendation to integrate the many separate faculty and departmental libraries of the university into a single administrative structure which also incorporated the central Bodleian Library had resulted in the creation of Oxford University Library Services (OULS; now the Bodleian Libraries) in 2000.² Yet despite the valuable legacy of the heritage collections and the founder’s eponymous library, and the progress toward a more efficient and cohesive administration, the Libraries found themselves beleaguered. In 2004 the incoming Vice-Chancellor of the University, a New Zealand Rhodes Scholar with a background in corporate management as well as recent experience in higher education, introduced a number of reforms at Oxford, including

2 Oxford University Library Services (OULS) were renamed as the Bodleian Libraries in March 2010. For the sake of clarity and consistency, the Libraries are identified as the Bodleian Libraries throughout the narrative of this book, including the period 2000-10 when the official designation was Oxford University Library Services.

4 

 Sarah E. Thomas

more rigorous accounting practices. The Bodleian Libraries, which had operated in deficit mode since the turn of the century, fell under increasing pressure to bring the budget into balance. A draconian establishment review had reduced the number of staff, but had exposed the tension and dissension about the Libraries’ direction in a very public way. The integrated library system was superannuated, long overdue for replacement. The University authorized funds for a new system, but the requirement to customize an off-the-shelf product to adapt to the idiosyncrasies of the multifaceted Oxford libraries was creating choppy waters for the implementation. Duke Humfrey’s Library had been restored magnificently in the late 1990s, but the infrastructure which fed it, the core stacks of the New Bodleian Library, were deemed unfit for purpose. Overcrowded, inadequately resistant to fire or flood, they were on the verge of being condemned by The National Archives as a space unsuitable to house collections of national importance. In sum, Oxford’s libraries encompassed enormous wealth and enjoyed the high esteem accorded an institution with a patina acquired over hundreds of years, but as it moved through the first decade of the twenty-first century, it faced challenges which threatened its ability to meet its mission. Like Richard III’s kingdom at risk for loss for want of a horse on the battlefield, the Bodleian seemed at the point of collapse because of the woes encountered in its quest to build a depository to store its collections. Bad press harangued it, and yet, throughout the University and among its loyal readers, sympathy for the Libraries ran deep. Continuing as it had was not a sustainable course of action, but the way forward was not straightforward. In the end, the Bodleian Libraries needed to undergo a dramatic metamorphosis to become a robust and confident organization able to serve contemporary scholars and students. What follows is the story of how the Bodleian transformed itself. Although many of the elements of the story are particular to Oxford, a unique and complex university, there are universal lessons which are applicable to other situations. This is a story of determination, of hope, and of the power of a team which accomplished more than they or any other person believed was possible. It is a success story.

Sarah E. Thomas

Chapter 1: Background 1 Space: the constant frontier Libraries are perennially short of space, and the Bodleian Libraries were no exception. In 2007, the Bodleian was receiving 5000 new items to be added to the collections weekly, and archival materials being acquired through gift, purchase, and deposit strained the capacity of the principal storage facility of the Libraries, the New Bodleian Library, beyond the breaking point. Its ziggurat of eleven storeys, constructed in the 1930s with the prospect of providing growth for a hundred years,¹ was stuffed to the gills. Space that had been intended to accommodate readers had been given over to shelving, and some floors had piles of books stacked in the aisles. (Figure 1.1) In the 1970s the University had expanded its storage capability through the construction of a building at Nuneham Courtenay, a small village about twenty minutes’ drive from Oxford. The building had been constructed in modular fashion over a twenty-year period, and by 2002 over 1.2 million items were housed there. The Libraries had initially planned to continue the expansion of the Nuneham Courtenay depository to channel the New Bodleian overflow. The building, however, was in a rural area with sensitive planning requirements, and when the Estates representatives explored the option of further expansion, they were advised that further planning permission was not likely to be forthcoming. Thwarted in that direction, the Bodleian actively considered alternatives. A further complication, beyond the need for additional space to house the growing collections, was that, upon examination, the existing New Bodleian stacks proved unfit for refurbishment. The investment to bring them up to standard would have been excessive, and consequently, there was a rethinking of the appropriate programme for this Grade II Listed Building in the heart of Oxford. (Listed Buildings are considered to be important elements in the environment, and may only be altered with permission from the local authority.) The estates development programme for the Bodleian Libraries morphed into three major strands which were reflected in the Libraries’ Vision for 2011 document,

1 Library provision in Oxford: report and recommendations of the commission appointed by the Congregation of the University (Oxford: Clarendon Press, 1931), 58.

6 

 Sarah E. Thomas

Figure 1.1 Overcrowding in the New Bodleian bookstack

approved by the Curators of the University Libraries in January 2006.² In a section entitled ‘Delivering the estates vision’, amidst a number of other objectives, were three major projects representing a bold development of the Library’s estate: – Construct a new high-density depository at Osney Mead. Using a ‘high-bay’ design with an Automated Storage and Retrieval System, the depository

2 ‘Oxford University Library Services: Vision for 2011, approved by the Curators of the University Libraries January 2006’. Accessed 28 June 2012, http://www.bodleian.ox.ac.uk/__data/ assets/pdf_file/0005/22919/vision2011.pdf

Chapter 1: Background 

–

–

 7

would have a capacity of some 8.25 million volumes and would allow the Libraries greatly to increase the proportion of stock held in low-cost, highquality storage (currently 11 %); Develop plans to construct a major new library on the Radcliffe Infirmary site, when it became available for redevelopment. The new library would focus on provision for undergraduate and taught postgraduate study, as well as first level-research, in the Humanities and Area Studies, replacing up to twelve existing libraries. Library hours and services would be tailored to undergraduate and taught postgraduate needs, and collections would be stored on open access shelving; Secure University support for redevelopment of the New Bodleian Library as an integrated Special Collections Library (embracing Western and Oriental special collections, including manuscripts), providing new and enhanced services for Humanities research. The intention was to make progress with the design for the redevelopment such that the project could start in 2008, subject to funds being available.

The depository would be a key enabler for site rationalization and for refurbishment projects requiring major decants of material. Subject to obtaining planning permission, we hoped that construction of the depository would commence in Spring 2006 and be completed in late summer 2007. The estates activities outlined in the Vision for 2011 document were estimated to cost approximately £ 150 million and were to be completed for the most part within a five-year time horizon. This was an aggressive plan designed to address almost simultaneously in a compressed timeframe issues which had been developing over decades. The lynchpin for this complex series of estates endeavours was the construction of a book depository which would relieve overcrowding, provide for growth in collections, and permit the reconfiguration of other facilities to be more compatible with contemporary expectations of library service delivery.

2 The Osney Mead saga The plan was to consolidate in a high-density storage facility materials which were distributed across a fragmented estate: the 1.2 million lower-use items in the Nuneham Courtenay depository, which were delivered to the Bodleian daily; an estimated 600,000 twentieth-century volumes packed densely on 1912 mobile racking in the Underground Bookstore harboured under Radcliffe Square;

8 

 Sarah E. Thomas

and the heart of the legal deposit, general and special collections in the New Bodleian Library comprising around 2.7 million items. Additionally, there were over 1.2 million maps, hundreds of thousands of music scores, and manuscript archives measured at 19 kilometres (12 miles). When the New Bodleian stacks had reached saturation without any practical options for relieving pressure internally within the Libraries’ estate or without any other solutions in Oxford, the Libraries looked to commercial storage as a temporary means of managing the overflow. A contract with DeepStore, a company which rented storage space in underground salt mines in Cheshire, provided a secure and environmentally sound environment with the ability to retrieve material weekly or more frequently, if the demand was sufficient. Eventually, over a period of three years, from 2007 to 2010, almost 2 million items were boxed and transferred to these excavated caverns 150 metres below ground. The construction of a depository to hold these dispersed collections in an efficient and safe manner was therefore urgent. The University had acquired property in an industrial estate known as Osney Mead including library administrative offices and an adjacent warehouse, and the warehouse became the preferred location to build the depository. Its principal advantage lay in its proximity to the Bodleian, about 1.3 km (0.8 miles) away as the crow flies, although in Oxford traffic, deliveries were estimated to require about 20 minutes. This factor was deemed paramount in order to limit the degradation of deliveries of books requested by users. Oxford’s deficiency in access to its collections was a sore spot in its reputation, with unfavourable comparisons to Cambridge University Library, where much of the stock was available for consultation on open shelves, and even material requiring fetching is said to be delivered in as little as 15 minutes.³ Bodleian critics gibed that scholars were so put off by slow deliveries that they found it more convenient to drive three hours to Cambridge than to wait for their requests to be fetched from the bowels of the New Bodleian. With such criticisms ringing in their ears, Bodleian staff sought to compensate for the disadvantage of the closed stack through the selection of a site within the Oxford ring road and by the introduction of robotic retrieval machinery which would significantly reduce the time fetching material over manual labour. The architect Scott Brownrigg was engaged, and plans were drawn up for a depository on Osney Mead that would employ ASRS (Automated Storage and Retrieval System) technology. To maximize efficient storage, such high-density facilities were typically 10 to 15 metres high, and in Oxford, because the footprint of the

3 ‘[Cambridge University Library.] Frequently Asked Questions’, accessed 28 June 2012. http://www.lib.cam.ac.uk/students/faqs.html.

Chapter 1: Background 

 9

Fig 1.2 The planned Osney depository: initial design, picture courtesy of Scott Brownrigg

Fig 1.3 The planned Osney depository: revised design, picture courtesy of Scott Brownrigg

building would be constrained, the architects pushed the height still higher. (Figure 1.2) Initial consultations with city planners and other interested parties provided feedback that the warehouse was too brutal in its large rectangular domination. In an effort to mitigate this observation, the architects redesigned the building, creating an undulating roof that rose to a height of over 18 metres. This feature would result in the application of the so-called’ Carfax rule’, in which buildings constructed in Oxford were not allowed to exceed the height of this thirteenth-century tower in the centre of the city in any significant way.⁴ It also compromised the efficiency of the building, as the stacks rose to irregular heights. (Figure 1.3) The proposed building was dogged by derogatory comments, not only about its bulk, but also about its ability to protect one of Oxford’s most valuable assets,

4 Oxford local plan 2001-2016, adopted 11 November 2005 (Oxford: Oxford City Council, 2005), accessed 28 June 2012, http://www.oxford.gov.uk/PageRender/decP/Oxford_Local_Plan_occw. htm, specifically Section 5.0, ‘Historic environment’, 69, policy HE.9, accessed 28 June 2012, http://www.oxford.gov.uk/Direct/3427105.pdf.

10 

 Sarah E. Thomas

the Bodleian’s books. The intended location proved extremely controversial, as Osney Mead was part of a flood plain, adjacent to the Thames, and virulent objections were raised over the risks associated with placing the Bodleian’s priceless collections where they were vulnerable to flooding. Satirical internet postings mocked the initiative, headlining it as Academe’s first underwater library’.⁵ A continuing saga of woe enveloped the planning, with the matter being debated by Congregation, that is, the University’s governing body, consisting of the 3,000person strong body of academics. Modifications raising the estimated cost of construction by £ 2 million were proposed, and consequently, University gave its approval to proceed in late 2006. The next step entailed obtaining planning permission from the city of Oxford. Tainted by the spectre of waterlogged collections, the depository aroused further negative reaction. Although its surroundings included rusty unkempt sheds and an extremely large blue-clad facility housing the local newspaper, the proposed site for the Bodleian store also lay between conservation lands and the Thames river walk. It would be visible from the elevations rising above Oxford, with fears expressed that it would mar the historic views of the dreaming spires of the ancient university town. A pitched media battle ensued. The date at which planning permission was to be considered slipped from month to month, from February 2007 until July 2007. During the last week of July, torrential rains relentlessly lashed down on the UK, with the consequence of widespread, serious flooding. The road from Oxford to Osney Mead was closed, with water rising to a metre in some areas. Osney residents found the lower floors of their homes submerged and cars parked in the neighbourhood ruined. Although the particular spot designated for the depository remained dry, and in any case, the bunds designed to withstand floods were deemed to provide more protection than Westminster, the intensity of emotion made it a poor moment to seek planning permission. Oxford City Councillors entered the town hall and told dramatic tales of the destruction in their sodden homes. University members queried how books would be delivered when roads were impassable. The University strategically and sensitively requested a delay in consideration until September in order to conduct studies on the impact of the flooding. Over the next few weeks analysts reviewed the data about the vulnerability of the site, and in September 2007, the University presented its plans to construct the depository to Oxford City Council, The City planners recommended approval, councillors spoke for and against, and in a suspense-filled vote, the

5 ‘Welcome to Akme‘s bodleyworld sub-index’, accessed 29 June 2012, http://www.btinternet. com/~akme/bodindex.html

Chapter 1: Background 

 11

plan squeaked through by a single vote. Before the Bodleian could uncork champagne, indeed, before the representatives of the University emerged from the hearing room, there was talk of taking the matter to consideration by the full City Council. This was not a case of a town/gown divide, but rather one in which both the academic community and local residents were split, with some sympathetic to the critical need for expanded storage, but many others suspicious of the choice of the site and antithetical to the building proposed for it on both aesthetic and engineering grounds. By November 2007, when the full Council convened, there was a full head of steam powering the opponents, and the planning application was refused. Dejected, Bodleian staff returned to their offices to consider next steps. The University was keen to appeal, and on the advice of expert counsel, felt it had a strong chance of success. Although an appeal would require a significant commitment of funds, it seemed the most prudent course of action. Much money had already been invested in acquisition of the land, the design, and planning consultants. In a high-level meeting of senior administrators of the University and the Queen’s Counsel (senior barrister) advising on the appeal, the Vice-Chancellor asked Bodley’s Librarian if construction of a facility in close proximity to the central Bodleian would be the best possible scenario for the Libraries, and as phrased, the only possible answer was yes. No other solution seemed to tick all the boxes and to offer the promise of completion in time to forestall what was becoming an economic and reputational crisis. The Bodleian, already running a deep deficit of £ 4 million on a £ 30 million budget, was incurring additional costs through the use of DeepStore, and the substandard conditions in the New Bodleian were placing in jeopardy the collections housed there. Decisive and immediate action was necessary. A thorough legal preparation over the next several months ensued. At the same time, staff and townspeople pressed for a Plan B, implying that the Bodleian leaders were insane not to have an alternative in their back pocket. Inside the Libraries and the University, staff and administrators were adamant that a repository at Osney Mead was the only workable solution. The success of the project depended on having a location inside the ring road in order to preserve timely delivery of requests. Well-meaning alumni suggested various sites sprinkled around Oxford where there was undeveloped land or property becoming vacant, but upon investigation, these parcels failed to meet the criteria for development. They were equally unlikely to achieve planning permission; their owners had other plans for them; their availability was uncertain; or the property was not suited for such a large warehouse needing a large flat area. And of course, there would be an additional outlay of funds to acquire them and almost certainly a redesign of the plans that had been customized for Osney Mead.

12 

 Sarah E. Thomas

Bodley’s Librarian was new to Oxford, having arrived from the USA in February 2007. There were many things she did not know. During her tenure at Cornell University a high-density storage facility had been built, although not one that used ASRS. By comparison, the proposed Oxford warehouse seemed expensive, the fact that there were no plans to barcode the collections being transferred was puzzling. Experienced staff explained that since much of the collection had been organized into categories of size since her predecessor Nicholson’s time in the nineteenth century, the plan was to move books in shelfmark order, retaining their Bodleian organization. It was thus not feasible to separate individual highuse items from low-use ones as it would leave gaps in the shelfmark array. Groups of low-use material would be sent to storage. The use of ASRS was also perplexing to a newcomer, as it normally was employed to compensate for the loss of immediacy when consulting open-access collections, substituting speed of retrieval for direct access. Generally facilities with ASRS were located adjacent to the main library, although in the UK, it was being specified for the British Library’s new Additional Storage Building.⁶ ASRS would accelerate the retrieval from the shelf to the bench for pacing in the delivery tote, but in Oxford this was only a small part of the workflow. Bodley’s Librarian spent much of 2007–08 learning about procedures in the Bodleian and preparing for the appeal which was ultimately scheduled for a hearing in mid July before an inspector assigned by the Secretary of State. Confidence was high that the University would win its appeal, and planning consultants felt strongly that any discussion of alternatives to the Osney Mead depository would provide opponents with ammunition that the Oxford plan did not absolutely need to prevail. Consequently, no effort was made to consider other options. On 9 September 2008 Bodley’s Librarian travelled with other University colleagues to Cambridge to visit a number of university buildings which might serve as models for another planned construction project, the Humanities Building and Library on the Radcliffe Observatory Quarter (ROQ). Returning that afternoon to Oxford, she listened to a garbled voicemail, of which the only distinguishable word was ‘Depository’, and the voice speaking did not sound elated. We soon learned the results of the appeal had just been announced, and the inspector had determined that the depository would indeed mar the view and could not go

6 Dawn Olney, ‘A UK first: an automated, high-density storage solution for the British Library’, [paper presented at] Where shall we put it? Spotlight on collection storage issues, NPO conference, London, 2004, accessed 28 June 2012, http://www.bl.uk/aboutus/stratpolprog/ccare/ pubs/2004/DOlneyNPO %20Conf2004.pdf

Chapter 1: Background 

 13

forward as planned.⁷ When a sympathetic Vice-Chancellor enquired about her response to the news Bodley’s Librarian reassured him that this was, in her view, an opportunity.

3 The new Academic Strategy How might one move millions of volumes outside of Oxford proper and not have disastrous results for readers? From Bodley’s Librarian’s perspective as a librarian whose career had been shaped in the USA over more than three decades, the Bodleian’s services laboured under crippling conditions. Library services were fragmented, with almost 40 different libraries providing the services. More than half the collection did not circulate. As a legal deposit library, the Bodleian has the right to receive free copies of works published or distributed in the UK and Ireland. This has the advantage of creating a superb collection without requiring as much university contribution to its acquisitions budget as international peers, but it also meant that the Bodleian lacked the resources to reallocate from print to electronic. A specially commissioned consultant’s report in 2005 had highlighted the Bodleian’s need to enter the electronic age more decisively, and University funding to acquire electronic journals had improved the Bodleian’s position substantially, but print was still the dominant force.⁸ With a fragmented estate, the cost of remaining open was high and there was a greater need for duplicate copies of texts and reference books. We began to ask new questions. Might it be possible to speed up deliveries by allowing the vans to unload directly at the Old Bodleian? Part of the old-fashioned charm of the central Bodleian was the conveyor that connected the New Bodleian stacks with the Old Bodleian for distribution of books to the Upper and Lower Reading Rooms, Duke Humfrey’s Library, and the Radcliffe Camera. As cases of requested and returned books rattled their way over a series of rollers, it created a rhythmic mechanical sound as comforting as train rolling along a railway. Yet delivery from remote store to the New Bodleian and subsequent transfer to the conveyor added time and complexity to the work. We could see, however, that Bodleian vans penetrating the sanctity of the space defined by some of Oxford’s

7 Appeal Decision: Appeal Ref: APP/G3110/A/08/2063341, 9 September 2008. Bristol: The Planning Inspectorate, 2008. Accessed 28 June 2012, http://www.pcs.planningportal.gov.uk/ pcsportal/fscdav/READONLY?OBJ=COO.2036.300.12.133683&NAME=/DECISION.pdf 8 Mel Collier and Derek Law, ELISO: an electronic library and information service for the University of Oxford (London: Electronic Publishing Services Ltd., 2005).

14 

 Sarah E. Thomas

and England’s finest architectural specimens would be a sacrilege. Over and over we traced the flow of material to readers, trying to eliminate a step and reduce the time it took to process a request and get the item in the hands of the reader. The anticipated number of requests from closed stacks was 450,000 per year-- not so many when set against the total use of materials of the collections, calculated at around 2 million. The eureka moment came when we realized we were trying to solve the wrong problem. The challenge was not how to deliver books faster, but rather, how to connect readers with the information they needed. It was not about 450,000 requests for items from the closed stacks, but about managing the information needs of a clientele that used a diverse set of resources. Two areas which users of US academic libraries rated highly on LibQual+TM surveys, measuring the degree to which libraries met their ideal, were access to electronic resources and the provision of services in which the readers were empowered to help themselves. If users of the Bodleian also valued these elements of library service, one could resolve the conundrum of slow delivery either by placing more high-demand books on open shelves for direct access or by meeting more of the demand through electronic provision. Both of these approaches combined would have the effect of reducing the number of items requested from storage whilst simultaneously offering the user immediate access. If the number of items requested from the depository were about 250,000, that would be in the range of US storage facilities, where a rule of thumb was that they were efficient to operate if demand were about 3 % of stock requested annually. Such a ratio would enable the Bodleian to locate its operation outside the ring road, perhaps as far as an hour away. That thinking opened up a vast array of possibilities for sites in areas where warehouse construction would not be as sensitive as in congested, historic Oxford, and where open land was more plentiful and more affordable. Providing direct access to high-demand materials was dependent upon meeting two conditions. We had to be able to identify or predict which books would be requested frequently and we needed accessible space where users could consult them. The Bodleian had circulation information on items, and it could readily define one broad category of material that accounted for a high proportion of requests: English-language monographs published within the last 3 to 5 years. With that target in mind, the next step was to decide where to put them. At the centre of Oxford was Radcliffe Square, under which had been excavated a two-level underground bookstore that was connected to the Radcliffe Camera and via a tunnel to the Old Bodleian. Opened in 1912, it was a storage innovation in its time for its employment of mobile racking, suggested by William Gladstone, Victorian Prime Minister, as an idea to overcome the scarcity of space to

Chapter 1: Background 

 15

accommodate burgeoning book collections.⁹ In the literature promoting it, the Underground Bookstore was reported to have the capacity of a million volumes, although Bodleian staff calculated the stock at less than two-thirds of that.¹⁰ In the business plan supporting the depository at Osney Mead, the Libraries had planned to transfer those materials to the depository and abandon the space as unusable, in order to reduce space charges. The new plan proposed that the space should not become derelict, but rather open-access shelving for users to consult at their convenience books which they had previously had to request to be fetched. This kernel of an idea remained at the heart of the development of the Gladstone Link, although like most of the initial ideas presented following the loss of the appeal, it was transformed and improved as other members of the staff and the University contributed to its development and implementation. The second strand of mitigation related to the increased access to electronic resources. For many scholars and students, e-resources are a superior way of accessing information, since they allow 24/7 consultation and are not place-limited. The ability to drill down into texts and to search across multiple documents enhances discovery. Even for those who wish to work with the original print publication, they are an electronic filter which can make research more productive. In the context of the rethinking of the depository, the Libraries proposed buying electronic backfiles of a significant number of periodicals, thus enabling the transfer of large swathes of the print collections to storage. Evidence of the decline in the use of materials from the Radcliffe Science Library stacks indicated that e-journal access was a substitute for use of paper copies. This trend at Oxford was supported by studies conducted in the USA of material identified for transfer to remote storage. Items in electronic form supplanted use of paper publications. Another very critical factor in the loss of the appeal was the ticking time bomb in the continued storage of the largest concentration of Oxford’s most valuable assets in the New Bodleian Library. Whilst many people had been agitated about the potential dangers of flooding for collections housed at Osney Mead, they were mostly oblivious of the less visible threats posed by the antiquated structure of the New Bodleian, a forest of steel columns supporting eleven floors of books which, in the intense heat of a fire, might pancake down, creating rubble and ashes of centuries of recorded knowledge. (Figure 1.4) Sludge-filled pipes

9 Edmund Craster, History of the Bodleian Library, 1845-1945 (Oxford: Clarendon Press, 1952), 234–35. 10 The new underground book-store of the Bodleian Library (Oxford: W. Lucy & Co., 1914), [5]; Library provision in Oxford, 20.

16 

 Sarah E. Thomas

Figure 1.4: The New Bodleian under construction showing the steel frame

traversed the stacks; a broken seam might spew foul contents over sixteenthcentury royal bindings or obliterate the ink on letters penned centuries ago. The inevitable delay in completing the construction of the depository and the ambiguity about when that completion would occur left the Bodleian’s collections highly vulnerable to a number of high risks. Decisive action was needed. With the imperative to relocate the rare and unique collections to a safer location as a priority, the Bodleian moved to decouple the protection of its special collections from the plan for a storage facility, at least in the immediate future. It proposed to create an interim space that would be compliant with the national standards for the storage of archival collections (British Standard BS5454), thus addressing the concerns of The National Archives, and removing the most egregious risk the Library faced. Initially, plans focused on Nuneham Courtenay, but as its isolation posed security challenges as well as logistical delivery issues, staff were inspired to consider another option, one made possible by the proposed acquisition of electronic journal backfiles: the basement of the Radcliffe Science Library. Here, at the heart of Oxford, only five minutes away from the central

Chapter 1: Background 

 17

Figure 1.5 The Lankester Room in the Radcliffe Science Library (picture by Greg Smolonski)

libraries of the Bodleian system, was a copious stack which could be upgraded to BS5454.¹¹ And, in the spirit of teamliness and support of larger institutional goals, staff at the Radcliffe Science Library agreed to vacate a large portion of an underground reading room, the Lankester Room, in order to create a space for readers consulting special collections. Underneath the Lankester Room, a reading room dating from the 1970s resembling a space ship or perhaps a cruise ship, (Figure 1.5) a far cry from the medieval Duke Humfrey’s Library, lay a vast store. Housed in these subterranean stacks were hundreds of thousands of scientific periodicals alongside important treasures such as an Audubon elephant folio. Over time, as the Bodleian purchased backfiles of journals, the use of the printed periodicals had declined sharply. We took the decision to transfer these titles to DeepStore, a commercial storage option taking advantage of the naturally constant temperature and humidity found in salt mines in Cheshire. The next question to be considered was access to the collection. Some research libraries, such as the Vatican Library, had been known to close during periods of renovation, even for periods as long as three years.¹² The British Library, faced

11 BS5454:2000 Recommendations for the storage and exhibition of archival documents (London: BSI, 2000). 12 David Willey, ‘Vatican Library closure irks scholars’, accessed 29 June 2012, http://news. bbc.co.uk/1/hi/6901606.stm

18 

 Sarah E. Thomas

with moving large portions of its collections to storage, embargoed materials being relocated to Boston Spa from London for up to six months.¹³ The Bodleian leadership was committed to maintaining access to its collections during the transition, ideally without major degradation of service. An element of the new plan was to allocate funds to support innovative approaches which would mitigate any disruption. The final element in the new plan was much less visible to users of the Libraries, but essential to its success: the modernization of collection management in the Bodleian. Books in the Bodleian closed storage were for the most part organized by shelfmark and by size. Permanent barcodes linked to bibliographic records in the online catalogue were not a feature of Bodleian inventory management, in contrast to other libraries which had adopted high-density storage facilities as an approach to shelving large collections. Yet evidence from these models in place throughout the USA and elsewhere pointed in the direction of automated control as an essential ingredient in effective management. Barcoded collections were standard, and they were multiples faster to scan in when transferring holdings and automated readers of barcodes produced a higher quality, more accurate result than manual handling. Nonetheless, there was reluctance to adopt barcoding in the Bodleian, and costs were predicted to be prohibitive, with calculations of as high as £ 2 per item proving chastening to planners. Although Bodleian leadership was uncertain about how the full collection could be barcoded, it was determined to automate, and we bid for £ 1.5 million in funding from the University to embark on an automated inventory of its collections. These plans were developed and presented to the University as a new Academic Strategy in six short weeks after the final failure of the plans for the Osney Mead repository.¹⁴ The Libraries’ agility in responding to this perceived disaster, and the focus on new ways of meeting the needs of the academic life of the University, won immediate favour and did much to restore their reputation.

13 British Library collection moves (Reader news bulletin, Issue 02 - December 08), accessed 29 June 2012, http://www.bl.uk/collectionmoves/pdf/readernewsdec.pdf. 14 ‘Academic Strategy for Oxford University Library Services’, Oxford University Gazette (26 June 2009), accessed 29 June 2012, http://www.ox.ac.uk/gazette/2008-9/weekly/260609/ notc.htm#5Ref.

Chapter 1: Background 

 19

4 Putting the strategy into effect The material we intended to move from the Radcliffe Science Library to make way for the special collections would ultimately have to be moved a second time to the completed Book Storage Facility, but we deemed that the additional expense of a double move was justified by the urgency of our need to rescue the New Bodleian special collections from risk. The preparation, undertaken hurriedly, provided a number of lessons to the Bodleian that were valuable in improving workflows and processes for future moves. In this initial trial of a large-scale move, staff learned, often the hard way, the importance of good communication with affected parties, the need to have a strong and well identified project manager, the requirement for a well developed risk management strategy, the critical value of having adequate bibliographic description, and the essential need to exercise care in negotiating contracts. Despite many tense moments and a series of problems, the principal goal was achieved. The RSL stack, its scientific holdings transferred to DeepStore, offered its void to be upgraded to BS5454 to contain many of the Bodleian’s most precious assets as well as special collections and rare books that would form the mainstay of the materials used in the temporary special collections reading room over the next four years. The project met its budget and kept to its deadline. It was messy, fraught with conflict and uncertainty, but ultimately a success, and perhaps most significantly, it forged a resolve to learn from harsh lessons to build an improved project management strategy. Concurrently with the preparation of the RSL space for the temporary special collections service, there were activities under way to organize the move of the rare and special collections to the RSL once it was ready to receive them. With the benefit of an experienced hand who had played a major role in the relocation of archives and manuscripts when the British Library moved into its St Pancras location, the Bodleian special collections transfer was well organized, with the security and safety of its priceless heritage documents of paramount concern. And simultaneously, the foundation on which the future of the Bodleian’s collection management would be based, the creation of the Book Storage Facility, was being shaped. The key requirement for the depository had been that it be as close as possible to the centre of Oxford so as not to further compromise deliveries of requested items. Now, with decisive steps being taken to rebalance the load and to reduce the total number of fetches predicted to be required for delivery by fulfilling requests through a combination of electronic delivery and direct access, the Bodleian could be more flexible in its requirements. Working with the University’s Land Agent and the Estates Directorate, the Bodleian began almost immediately, in October 2008, to seek parcels of land that were within one hour’s travel time of the centre of Oxford, that were already in possession of planning

20 

 Sarah E. Thomas

permission for the construction of a large warehouse, and which were geographically configured to permit the erection of a high-bay facility. By December 2008 the field had been narrowed from 64 prospects to two sites that merited a visit by the Vice-Chancellor and Bodley’s Librarian, and by January 2009, representatives of the Curators of the University Libraries had satisfied themselves that the location in South Marston, near Swindon, would be a suitable plot on which to build. Within the various committees of the University of Oxford, the plans were moving forward and being approved at every stage. This venerable institution, known for its deliberate pace – a place where the old joke about how many dons it took to change a light bulb concluded with the punch line: ‘Change?’ – was moving at the relative speed of greased lightning. Funds to acquire the land, sufficient to construct three depositories, allowing room for expansion, were voted by University Council in January 2009, and by March the exchange of the property took place. In parallel, a redesign of the structure of the depository was under way, taking advantage of the larger footprint the Swindon site provided, and moving away from the automated storage and retrieval system: an ASRS did not represent value for money when the number of items being fetched dropped below a certain threshold., making the speed of delivery a lower priority. The ASRS, with a predicted lifespan of twenty years and an annual maintenance cost expected to be £ 800,000, was replaced by the less glamorous, but tried-and-true ‘Harvardstyle’ high-bay shelving system pioneered at Harvard University in the 1980s. As a consequence of removing the ASRS from the specifications, of developing plans for a more straightforward ‘shed’ without the architectural embellishment of the undulating roof introduced to mitigate the criticism levelled in Oxford about the ugliness of the building originally proposed, and of the fact that a bund to protect collections from flooding was no longer necessary in the new location, the cost of construction dropped by 10 %, and the Bodleian was able to return over £ 3 million to the University’s capital fund. The introduction of Harvard-style shelving with barcoded stock required a compatible warehouse management system, so the procurement of such a system became part of the strategy. The warehouse management system had to talk to the integrated library system. Coincidentally – and some might say with some foolhardiness! – the Libraries were also replacing the integrated library system at the same time. This had been planned for some years previously – the current system was no longer supported - but unforeseen circumstances had led to a series of delays and the need to start the process again in 2008. This meant that the warehouse management system would have to talk to both the current and the new library systems. To ensure that the projects did not unwittingly diverge, we appointed a single project manager to oversee the implementation of both the warehouse system and the new library system.

Chapter 1: Background 

 21

The Bodleian’s application for planning permission for the South Marston site was unanimously passed by the Swindon Borough Council in May 2009, and construction commenced in August. A year later, in September 2010, with the project on time for completion and on budget, the Bodleian took possession of its new Book Storage Facility, a building with a capacity of 8.4 million volumes, and began its commissioning phase. Staff to process collections being transferred to the BSF were hired and trained in the months preceding November 2010, when, on 8 November, the anniversary of the opening of the Bodleian Library in 1602, the first book was ingested into the BSF. Clever development of the barcoding workflows, and a flexible approach to the allocation of costs, stretched the £ 1.5 million sum allocated so that it covered twice as many items as had been originally envisaged; and over the lifetime of the initiative to transform the Bodleian, savings in other areas were channelled to barcoding, with the end result that £ 3.6 million supported the barcoding of almost 7 million volumes including the creation of bibliographic records and item-level records. In the fourteen months that followed the commissioning of the Book Storage Facility, over 7 million volumes and over 1.2 million maps were relocated from the New Bodleian Library, Nuneham Courtenay, the Underground Bookstore, DeepStore, and other scattered sites where the Bodleian’s vast and overflowing resources had been crowded onto shelves and into boxes. Their position on the gleaming steel racking installed in the Book Storage Facility was a triumph of team work and planning. These various strands formed the Academic Strategy of the Bodleian Libraries. The Academic Strategy was developed in the Bodleian and communicated to the members of the University through a cascading series of committee meetings, town meetings, and other sessions with the very large community of Bodleian users who were passionate about the libraries and their collections. The rethinking of scenarios for service was considered to be uniquely agile for Oxford, where decisions are taken only after much careful deliberation. The Bodleian benefited from a supportive administration, a team of dedicated managers who went into overdrive to grapple with multiple challenges of enormous proportions, and from the ability of staff to think creatively to develop solutions that were persuasive in testing certain assumptions about the way people thought the library should operate and which, in the end, were pragmatic and achievable, as well as audaciously ambitious.

Sarah E. Thomas and Michael Heaney

Chapter 2: The Bod Squad 1 Multiple projects The Bodleian Libraries had a plan, the Academic Strategy, and buy-in from all parts of the University.¹ Various proposals and business plans passed through stages of review and approvals, with funding for the Book Storage Facility, for renovations to the Radcliffe Science Library to accommodate special collections being transferred from the New Bodleian Library, for revamping the Underground Bookstore to hold direct-access collections, for making improvements to the Old Bodleian which would support disabled access, for the purchase of electronic backfiles of journals that would sharply reduce the call on the print journals which would be transferred to the Book Storage Facility (BSF), for creating an automated collection inventory, and for moving millions of volumes in a complex sequence of over twenty separate moves. The New Bodleian renovation plans were progressing, and with them, the need to relocate almost 200 staff from various locations. Alongside these activities came preparations for the selection and implementation of a new integrated library system. Planning was also under way for the Humanities Library on the Radcliffe Observatory Quarter. There were very high stakes: the University had committed capital funding of over £ 120 million to the various project initiatives and we were determined to offer continuing access to the collections, hoping to minimize disruption to services. Although the projects were in many cases dependent on each other for success, it soon became evident that they were being advanced in silos, without adequate consultation and communication. Worried librarians would send Bodley’s Librarian emails about collections being boxed up without adequate bibliographic control, and managers from one activity would lay the blame for failing to meet deadlines or to stay within budget at the feet of other managers, implying that their portion of the work had been held up because of the inefficiency or incompetence of another. Consultants were hired to assist with the workload, but that generated criticism about their costs, which were often much more than

1 ‘Academic Strategy for Oxford University Library Services’, Oxford University Gazette (26 June 2009), accessed 29 June 2012, http://www.ox.ac.uk/gazette/2008-9/weekly/260609/ notc.htm#5Ref.

24 

 Sarah E. Thomas and Michael Heaney

those of salaried employees. There was much stress, tensions between colleagues and some evidence of increased sick leave. It seemed that although the development and communication of the Academic Strategy had gone surprisingly smoothly, the implementation path was crooked and rocky. Senior project staff did not have enough authority to make decisions and there was no overall coordination. The situation was chaotic and the threats to achieving success were great, despite heroic endeavours on the part of individuals.

2 The creation of the Bod Squad Something needed to be done, or else the potential to fail was huge. There was a danger that people would lose confidence in the organization’s ability to deliver projects on time and on budget. Thinking about a business-like approach that would focus on the most important things that needed to be done and then confirm that they had been accomplished as intended, Bodley’s Librarian established a group which first met on 2 April 2010, representing the various initiatives under way, and set up twice-weekly meetings. Participants, who included the leaders of the principal groups such as bookmoves, the barcoding initiative, and other critical areas, were expected to list three objectives they intended to achieve, and in the second meeting held at the end of the week, to report on what they had accomplished. After a brief period of these twice-weekly meetings, it became obvious that weekly meetings would be more efficient and practical, and they were collapsed into a single Friday morning session lasting about 90 minutes. The meetings were chaired by Bodley’s Librarian, or in her absence by another member of the Libraries’ Executive, and as many of the Executive as possible attended each meeting. The group was dubbed the ‘Bod Squad’. Unlike many groups or committees in the Bodleian, the Bod Squad participants were not selected by their place in the management hierarchy, but rather by the importance of the activity they were overseeing in relationship to the fulfilment of the Academic Strategy. That meant a certain unevenness in terms of experience among the participants and it meant that some managers were excluded from meetings which their subordinates were having with senior managers in the Bodleian. There was some discomfort with this, but it did not prove to be a significant problem.

Chapter 2: The Bod Squad 

 25

3 Bod Squad documents A feature of the meetings was the dashboard, a single-sheet device intended to record the important elements relating to the projects: overall goals, milestones, financial data, key performance indicators, risks, accomplishments and goals for the coming week. (Figure 2.1) Each dashboard included a ‘traffic light’ system of recording the status of each element. In the case of numerical data this was coded by formula, so that, for example, a risk with high impact and high likelihood would automatically be flagged as red for danger, while one with high impact but low likelihood would be flagged as amber. In other cases (such as the planned ‘Next actions’) the traffic-light status was assigned manually. Bodley’s Librarian chaired the sessions, which benefited from strong support from the recorder, who ensured that agendas and dashboards were prepared in advance and notes of the meetings were distributed promptly. It was essential to produce reports of meetings quickly so that problems could be logged and dealt with, and so that we could be certain to review the following week the proposed ‘Next actions’ and verify that they had been accomplished. We recorded notes of actions and problems on a private blog, The Bod Squad Log Blog. (Figure 2.2) Each topic had its own blog entry: we assigned category topics based on the specific project. If a topic affected two projects we could give it two categories. We used tags to describe the status of the topic: – New task: for the first blog post about a specific topic; – Progress: to show that the tasks associated with the topic were under way, and under control; – Hiccups…: to indicate that a minor problem or delay had occurred (for example, a planned meeting had been postponed); – CHALLENGE!: to draw attention to a problem for which there was not an immediate answer, but did not immediately threaten the entire project; – RISK!!!: to flag a problem that posed a significant threat to the completion of the project in good time; – Done!: to show that the topic and its tasks had been completed. The Wordpress blog automatically generated a word cloud of the tags, which gave a sense of progress overall. (Figure 2.3) Blog posts could be filtered by category and/or tags, making it easy to study the progress of a specific project or to review all the unresolved risks. The blog entry dealing with the move of Facilities Management staff was typical. The topic was raised at the Bod Squad meeting on 23 January 2011 and the blog entry was created two days later. (Figure 2.4) Two threads of the move were combined in the entry: the move of engineers and the move of the Facili-

 Sarah E. Thomas and Michael Heaney

Figure 2.1: A Bod Squad dashboard

26 

Figure 2.2: The Bod Squad Log Blog

Chapter 2: The Bod Squad   27

28 

 Sarah E. Thomas and Michael Heaney

Figure 2.3: Blog categories and word cloud of tags

Figure 2.4: Blog entry for Facilities Management

ties Management team leaders. By 1 April we had made some progress , but the topic was tagged as a ‘RISK!!!’ because of the uncertainties about receiving Listed Building Consent for alterations to the destination building in time for the move of the Facilities Management team leaders. A challenge faced by many projects is that the project staff become focused on reaching their own targets and lose sight of the wider picture, and of the way

Chapter 2: The Bod Squad 

 29

in which their actions impinge upon other aspects of the organization’s work. Where there is just one project, that can be annoying, or may interfere with other activities. We had a dozen overlapping projects working simultaneously. The weekly meetings of the Bod Squad allowed all of the project managers to hear about the progress and challenges faced by the other teams. On more than one occasion this saved us from potential failure; or, more positively, one project team could provide an answer to a problem faced by another. At each meeting two of the project managers presented their dashboards in a detailed review. The other project managers simply reported on their recent actions and next intended actions, and highlighted any current problems. The Bod Squad developed a strong ethos of problem solving, with a constant awareness of the three major deadlines we would reach in July 2011: the need to vacate the DeepStore repository by the start of the month, the migration to a new integrated library system in the middle of the month and the need to have all the books, people and furniture out of the New Bodleian by the end of that month. The barcoders in particular had to devise several new methods and workflows to deal with the increasing levels of complexity they found in the relations of the books on the shelves to the catalogue entries. (See chapter 9) Sometimes we found it desirable to draw additional staff into the meetings so that they could share in the atmosphere and appreciate the rapid responses we needed. Although the team managing the implementation of the new integrated library system and the warehouse management system team came to each Bod Squad meeting, the staff responsible for the current integrated library system did not. However, that team was responsible for a number of key tasks: generating the checklists of shelfmarks for barcoding, loading the records of barcoded items back into the integrated library system, and transferring that data to the warehouse management system. We brought a key member of the current systems team into the meetings so that they could understand how even a small delay or problems could have serious knock-on effects. In similar fashion we asked a member of the special collections staff to attend some meetings to advise on the best way to move the collections safely but also within our timetable. It is natural for special collections staff to take great care of all the items in their charge, and to want to find the best method of proceeding in preference to the quick method. In this case we needed the methods that would deliver the result – the items being delivered to the BSF – within our time limits.

30 

 Sarah E. Thomas and Michael Heaney

4 Escalation If the Bod Squad meeting identified a problem which required a higher-level decision, this was passed to an overarching Project Board which also met weekly. The membership of the Project Board was broadly similar to that of the Executive, the Libraries’ senior management team, and in the later stages of the life of the Bod Squad the Project Board discussions were absorbed into the regular weekly meetings of the Executive. Typically, the matters considered by the Bod Squad Project Board fell into two areas: interrelations between different projects, and specific obstacles that had arisen. The Project Board meeting of 12 January 2011 discussed some typical issues. At this stage the Book Storage Facility was open, and the decant of material to the Facility from the DeepStore salt mine had just begun and was proceeding at full speed. Material from DeepStore was delivered daily to the temporary warehouse at Osney for barcoding; it was barcoded and sent on to the BSF for ingest. We found that the bookmoves team was delivering the books to the BSF before the notification of the barcode had passed from the barcoding team to the library catalogue and from there to the warehouse management system. This was an issue affecting four projects: barcoding, bookmoves, the warehouse management system and BSF ingest. The solution was simple – hold the material for 24 hours after barcoding before delivering the books to the BSF – but it required us to have an overview of all of the projects. (An additional factor was to ensure that there was space in the temporary warehouse to hold the stock, as we knew that there was no room to do this at the BSF itself.) At the same meeting of the Bod Squad Project Board on 12 January 2011 we considered various risks and challenges. Among them were: – BSF services: the first utilities bills indicated running costs were significantly higher than anticipated. This was passed to the University’s Estates Directorate for review. We noted that in the first few months the high levels of ingest activity could affect the power consumption; – Barcoding: a key staff member had resigned; – Bookmoves: progress was 544 m behind schedule – not serious in itself, given that the forecast daily move rate in this period was 451 m, but it was necessary to continue to monitor closely; – Underground Bookstore/Old Bodleian: delays caused by the bad weather had removed all remaining flexibility from the timetable; – Underground Bookstore / Old Bodleian: the lift being used to deliver books (before the completion of the new lift as part of the project) was becoming subject to frequent breakdowns. The Bodleian’s Head of Reader Services and Customer Relations was asked to look at alternative delivery methods.

Chapter 2: The Bod Squad 

 31

The two projects most closely interlinked were the barcoding and bookmoves projects. Bookmoves could not take place until the material had been barcoded. Each manager had a choice about which of the sequences of material on the eleven floors of the New Bodleian bookstack should be tackled when, but decisions taken in isolation could lead to difficulties. For example, a floor of the bookstack might have four sequences of material, of which three were straightforward for the barcoders but the fourth contained material difficult to work with. The bookmovers did not want to move onto a floor until it was entirely ready. Just a small number of problem items in a sequence could render an entire sequence off-limits for the bookmovers. During the most intense period of the moves the two managers met several times a week to review and organize their teams so as to maximize their effectiveness. None of this activity replaced the normal project management procedures in place, following guidelines for best practice. The capital projects – the Book Storage Facility, New Bodleian, and Underground Bookstore/Old Bodleian – each had their own Project Sponsor Group. All the non-capital projects reported to their line managers and had their own project meetings. The Bod Squad meetings, notes and blog made it very easy to gather appropriate reports for these meetings. They also made it easy to draw together the main lines of progress and concerns for other University bodies. The University’s Planning and Resource Allocation Committee is the senior financial policy-making body for the University, and was the body to whom we had to apply to release funds for the projects and to make adjustments to budgets. (See chapter 3) The reports emanating from the Bod Squad proved to the University that we had strong control over the projects and allowed us to persuade them of the merits of our case when changes were needed – for example, to allow us to apply savings made in one area to address extra spending in another.

5 Stressful times The period from 2009 to 2011 was undoubtedly a time of tension. The extra work, the deadlines, the changes to normal routines, the disruptions caused by building works all contributed to increased levels of stress for staff. We organized staff development sessions on the management of stress – both for managers and line employees – and ran ‘well-being’ weeks. We made sure that all staff were aware of the regular help that the University could provide. We contracted with an external employee assistance provider to provide an independent service. This covered not just problems directly related with work, but all kinds of support. Often, it is a combination of factors which leads to stress for staff – difficult conditions at work

32 

 Sarah E. Thomas and Michael Heaney

Figure 2.5: Bodley’s journey

combined with a problem in the family or the home, such as a child’s illness or financial worries. The provider offered online, telephone and face-to-face counselling on all these topics, and was able to refer callers to appropriate qualified legal, medical or psychological personnel. The service was available 24 hours a day, seven days a week. Communications were also important, so the Head of Communications attended each Bod Squad meeting. At first the emphasis was on communicating to our readers in the University and worldwide, especially where their study plans might be disrupted by system downtimes, by building work or by books and manuscripts being unavailable. However, we soon realized that staff communications were just as important. We promoted team spirit by introducing a special ‘Bod Squad’ diamond-shaped logo (see figure 2.2) and by distributing notepads and coffee mugs bearing it. We also tried to communicate progress in engaging ways. The barcoding project required us to barcode about 192 km of books – coincidentally the same

Chapter 2: The Bod Squad 

 33

Figure 2.6: Bodley’s Librarian thanks staff for their achievement

distance as the journey from Thomas Bodley’s birthplace on Gandy Street in Exeter to the Bodleian Library in Oxford. We therefore mapped the journey using Google maps to record his metaphorical progress from Exeter to Oxford, marking notable milestones either in books barcoded or significant towns reached en route (Figure 2.5).

6 Celebrating success We also tried to make sure we celebrated the successes. When the two-millionth barcode was recorded, special T-shirts bearing a large barcode (appropriately enough, the barcode for a book about barcodes) and the legend ‘2,000,000 and counting’ were produced and distributed to staff. Celebratory parties marked our achieving the deadlines relating to the New Bodleian closure, in particular a party in the now empty New Bodleian – home for so many years to dozens of staff – in the week before closure, giving them a last chance to look around the building together with the frisson of being able to eat and drink in a building where such

34 

 Sarah E. Thomas and Michael Heaney

activity had long been forbidden. We also invited former staff who had worked in the building, including one member who had started work before it was built over seventy years before. Gifts were handed to dozens of staff who had played a key part in achieving the milestone. (Figure 2.6) Many agency staff were of course coming to the end of their contracts, and it was important to mark their success too. A celebration especially for them was held in the fourteenth-century Divinity School.

Laura How

Chapter 3: Finance 1 Allocation of costs The original Academic Strategy presented to the University in autumn 2008 identified a series of major estates projects as being critical for the delivery of a 21stcentury library service for the University.¹ The Academic Strategy summarized the key features of the Libraries‘ overall strategy as follows; – Expanded provision of electronic resources; – Direct user access to high-demand titles on open shelves; – Enhanced user services, such as delivery to the desktop; – Increased protection of at-risk collections of high value; – Updated approaches to collection management; – A coordinated estates programme. The estates programme comprised: – the renovation of the New Bodleian Library as a special collections library; – the development of a Humanities Library to integrate libraries supporting faculties moving to the Radcliffe Infirmary site; – a high-density storage facility with a capacity of 8 million volumes; – the reconfiguration of selected existing closed stores to support direct access by users; – the transfer of the Nuneham Courtenay depository to the University museums; – the transfer of collections in commercial storage to the new high-density storage facility. In the Academic Strategy, the Bodleian Libraries divided the programme of works supporting that strategy into three categories: construction and refurbishment, enabling works, and staff.

1 ‘Academic Strategy for Oxford University Library Services’, Oxford University Gazette (26 June 2009), accessed 29 June 2012, http://www.ox.ac.uk/gazette/2008-9/weekly/260609/ notc.htm#5Ref.

36 

 Laura How

The construction and refurbishment costs consisted of the costs of redeveloping the New Bodleian Library; of buying new land for the construction of a depository and building it, and of reconfiguring the Underground Bookstore and Radcliffe Science Library. The enabling works included the cost of barcoding materials and the cost of purchasing backfiles of electronic journals. They also included the external storage fees arising from the continuing rental of space at the salt mine operated by DeepStore in Cheshire until such time as other elements of the strategy came to fruition. Staff time and cost comprised the cost of staff to carry out the book moves that enabled all of the above projects; and the cost of staff moves, though no separate budget was initially indentifed for this.

2 Construction and refurbishment costs Work progressed in the Libraries and the University’s Estates Department to quantify the costs of these works. The costs presented in Table 3.1 were the initial ones at different stages of draft and further work on them resulted in changes before the University Council was asked to consider authorizing them later in 2009.

Total £ 000s Construction and refurbishment costs New Bodleian

75,670

Depository

21,540

Radcliffe Science Library Underground Bookstore and Camera

730 3,185

Enabling works Purchase of electronic journals

1,000

Barcoding costs

1,500

DeepStore storage costs

3,500

Staff time and costs Book moves Total Table 3.1: Academic Strategy initial cost estimates

3,808 110,933

Chapter 3: Finance 

 37

In the spring of 2009 planning for the New Bodleian was at an advanced stage, and detailed work by the Project Sponsor Group had resulted in a cost plan totalling £ 75.7 million. This was a provisional figure still under review, in particular for costs such as the movement of staff, opportunities for VAT planning and inflation. Opportunities to reduce VAT on these buildings included taking advantage of the one-year reduction in VAT rates to 15 %. We also applied for some of the Book Storage Facility building to be zero-rated for VAT, based on the proportion of legal deposit material held there. The cost of doing this increased the cost of barcoding but the increased cost was more than offset by the saving from the VAT reduction. We also identified other uses of space in the redeveloped New Bodleian Library that could be eligible for zero-rating. We undertook discussions with a peer institution to understand the basis on which their storage facility was fully zero-rated for VAT, to ensure that full advantage was taken of Oxford‘s status as a legal deposit library.

3 Enabling works £ 1 million was allocated for additional electronic journal purchases, which were duly acquired and have been heavily used since acquisition (see chapter 5). The first £ 500,000 was spent on three major publisher packages that benefited all academic disciplines, and one single title (Nature). The purchase resulted in the flexibility to reallocate 320,000 volumes from central Oxford to remote storage without compromising access to the content, which is available electronically. In total this project provided Oxford readers with access to over 20,000 years of journal publishing from their desktops that was previously only available (if at all) as physical volumes on the shelves of their libraries. The Academic Strategy noted that: A decision taken decades ago not to create individual records for sets or periodicals means that many pieces are treated as a block and that information on use of some materials is not recorded and maintained. Much of the collection is not barcoded. Although the cost of comprehensive barcoding and cataloguing of all individual items is seen as prohibitive, selective barcoding can bring improved control in areas known to be of high use. Going forward, all items added to the collection will also be barcoded. This will also improve the management of materials once the high-density storage facility is constructed. The Library is therefore requesting funds to introduce barcoding to the collections, targeting this to the most highly used items and taking advantage of any exercise to move and sort items that is already taking place.

38 

 Laura How

The cost of barcoding part of the collections was initially estimated at £  1.5 million, whilst recognizing that it would not be sufficient to provide complete coverage of the entire collections. However, it would jumpstart the process and result both in lower one-off costs and in lower ongoing operating costs. The Academic Strategy made clear that selective barcoding of the highest-use items was necessary even if the main projects did not proceed. As already noted, we considered it likely that the VAT savings on the Book Storage Facility would justify a large element of the costs of barcoding. As the project planning developed, further funds were allocated to barcoding from the VAT reductions achieved through the separation of legal deposit and non-legal deposit items, and the barcoding budget was revised to £ 2.6 million. In 2009 the Libraries were paying some £ 500,000 per annum for the costs of storage in DeepStore, a secure storage facility in Cheshire that is managed to The National Archives’ standards. In the Academic Strategy we included total storage costs of £ 3.5 million for around 2 million volumes stored there. This was a vital ingredient in the efficient operation of our complex set of initiatives before the Book Storage Facility became available, both in continuing to hold most of the new intake of around 300,000 items a year, and in holding material displaced during the initial bookmoves.

4 Integrated library system In a separate but parallel strand of activity the Libraries had also received funding of £ 2.589 million to replace their integrated library system. The need was urgent, as maintenance of the existing system was unsustainable. The replacement timetable had commenced in 2008 and we planned to replace the existing system in the summer of 2011. This coincided with the need to complete the barcoding of the New Bodleian collections and accomplish their ingest into the Book Storage Facility. As part of the construction of the Book Storage Facility we had to acquire a warehouse management system, which had to communicate with the integrated library system. Close collaboration between the warehouse and library system projects was essential, so we appointed a single project manager to oversee both projects. It was natural, therefore, to include the integrated library system replacement project in the overall consideration of the budget envelope, though it had not originally been an element in the Academic Strategy.

Chapter 3: Finance 

 39

5 Building projects The building projects included the construction of the Book Storage Facility at Swindon (£ 26.165 million), the redevelopment of the New Bodleian building (£ 78.808 million), the refurbishment of the Radcliffe Science Library to house temporarily the special collections (£ 901,000), and the refurbishment of the Underground Bookstore to transform it into the Gladstone Link, together with the associated works in the Old Bodleian Library (£ 5.336 million). This last was later increased by £ 407,000 to include the separately funded project to raise the Bodleian Quadrangle at the same time. These building projects required a complex series of bookmoves which were provided with an initial costing of £ 3.432 million. This sum was revised to £ 3.808 million in June 2009. As the projects developed and further work to establish the detailed methodology for the projects progressed, we reached a more rigorous understanding of the Inventory Control Project, the Bookmoves Project and the Book Storage Facility ingest programme, covering the methodologies, relationship to other projects, and associated costs. On 4 January 2011 the rate of VAT was increased from 17.5 % to 20  % and this raised the projected budget for the New Bodleian from £ 78.8 million to £ 80.5 million.

6 Budgetary control In November 2011 the Bodleian Libraries submitted a request to the University’s Planning and Resource Allocation Committee to be allowed to redistribute funds within the umbrella of the combined budgetary allocations which had been approved under the heading of the Academic Strategy on 26 October 2010. Normal practice within the University was for project savings to be clawed back to the central University, but for cost overruns to be borne by the department. However, recognizing the unusually tight interrelation and interdependency of the Bodleian’s projects, the Committee approved the request. The approval was based on the requirement that the Libraries’ projects would remain within the total funding envelope, with the exception of the additional VAT charges for the New Bodleian project. The Committee acknowledged that addressing the VAT increase would need further consideration. The University Council approved the funding for the New Bodleian project on 2 May 2011, including the additional VAT element and PSG approved a reallocation of funds at its meeting in June 2011.

40 

 Laura How

The Academic Strategy projects continued to move forward successfully with regular reports made to the Curators of the University Libraries and to the University’s Audit and Scrutiny Committee. The decant of the New Bodleian building, the main phase of implementation of the integrated library system, and the completion of the Underground Bookstore and Old Bodleian Access Project (UBOB), were all accomplished in July 2011. The Bodleian Accounts department sent monthly budget reports to each of the project managers for review. In addition, because of the close interdependence of the projects, we held monthly meetings of the bookmoves, barcoding and Book Storage Facility ingest teams to review the actual figures for the projects and their forecasts. This regular monthly monitoring of the project expenditures identified a further series of re-allocations within the Academic Strategy portfolio. The principal change was found in the Inventory Control Project, where additional costs were incurred in meeting both the complexities of the manipulation of existing data, and the need to process material rapidly in order to meet the demanding schedules imposed by the overall project goals. Fortunately, savings were made by the bookmoving project, tray procurement for the Book Storage Facility, and the decision to withdraw collections from DeepStore on an accelerated timescale. Whenever necessary, we prepared papers for relevant committees, seeking approval to vire funds Original allocation £ 000s

Revised allocation (1) £ 000s

Revised allocation (2) £ 000s

Revised allocation (3) £ 000s

Bookmoves & ingest

3,808

5,269

4,879

4,773

Book Storage Facility

26,166

25,110

24,849

24,721

2,589

2,276

2,225

2,225

78,808

78,808

80,477

80,477

5,743

5,743

5,743

5,743

Integrated library system New Bodleian Underground Bookstore Inventory control

2,289

2,609

3,331

3,600

DeepStore

3,500

3,088

3,068

3,033

122,903

122,903

124,572

124,572

Total:

Table 3.2 Successive budget allocations

Table 3.2 summarizes the revisions made to the individual project budgets over time. Revision (1), in November 2010, increased the bookmoves allocation in order to establish a budget for the ingest into the Book Storage Facility. It also

Chapter 3: Finance 

 41

took into account savings on the purchase of the storage trays following a tender process together with savings on the Book Storage Facility Information System. The reallocation of some of that saving to inventory control enabled the concomitant identification of legal deposit material. Revision (2) in June 2011 took account of the VAT increase for the New Bodleian (so increasing the overall envelope) and further reallocated savings in bookmoves and the integrated library system, so as to support additional inventory control costs. The last revision, in January 2012, also diverted savings to support inventory control. There will be at least one further movement after a review of the final budgets of those projects that have finished; at that point we shall move any remaining funds to support the bookmove project to refill the Weston Library in 2014.

Michael Hughes

Chapter 4: The special collections 1 Background At an early stage in the plans for the refurbishment of the New Bodleian Library, we recognized that the most significant and most used special collections in the Bodleian’s custody, which were stored in the New Bodleian stack, needed to be retained in a suitable central location to ensure continued access to readers without incurring the high risks to collections that would be imposed by the remote storage of these unique materials. Additionally, we accepted that such central storage had to meet national and international standards in order to meet the requirements imposed on the Bodleian Libraries by external authorities for its continued recognition as a suitable repository for the custody of special collections. We quickly established that only one other of the Bodleian Libraries had the capacity required: the Radcliffe Science Library (RSL). This had closed-stack space for some 24 km of printed books; because of their differing formats and storage requirements, this equated to space for approximately 12 km of special collections, or about half of those stored in the New Bodleian. Selection of the RSL stack had three key implications: the need to empty the space of almost all of the RSL’s own stack holdings, refurbishment works required to bring the stack space up to the required standard, and the identification and adaptation of a suitable area for secure reading facilities within the RSL’s existing reading rooms. The need to act quickly to establish central special collections facilities before the closure of the New Bodleian to readers meant that the initial move of RSL collections out of the building had to take place in advance of the completion of the Libraries’ intended storage facility. The only way we could achieve this was by transferring these collections elsewhere on a temporary basis, for eventual transfer to the Book Storage Facility (BSF) once feasible. Commercial storage at DeepStore in Cheshire, already used by the Bodleian for storage of some materials, was selected for this purpose, and the emptying of the stack was completed very rapidly, between March and July 2009. This first phase of the project was followed by a period of refurbishment of the RSL stack and the selected reading area in part of the RSL’s underground reading room, the Lankester Room. The stack refurbishment included installation of environmental controls and automated fire-extinguishing equipment, upgrading of security provision, and adaptation of the stack shelving to accom-

44 

 Michael Hughes

modate special collections material. Reading room works involved segregation of an area to provide self-contained, secure reader facilities with space for openaccess reference materials directly relating to those special collections which we were moving. The refurbishment was undertaken between August and November 2009. The final phase of the operation was the move of the selected special collections to the RSL. This was undertaken between November 2009 and March 2010. The new reader facilities at RSL opened at the beginning of the move period as a skeleton service initially, expanding to a full service as the move progressed and special collections reader facilities in the central Bodleian Library could be transferred to RSL. The new temporary service was fully functional by March 2010, so we could turn our attention to dealing with those collections remaining in the New Bodleian stack and other storage sites being consolidated into the BSF. While many of the features of the move of remaining material to BSF were similar to those for the transfer of other library holdings, aspects of the preparatory work required were peculiar to special collections. The quantity of material involved was augmented by collections stored remotely at DeepStore in Cheshire and at the Libraries’ Nuneham Courtenay repository, which were also to be transferred to the BSF. We undertook the planning of a programme of the processing, boxing and inventory work which we would need to enable these materials to be barcoded and accommodated in the BSF in summer 2010, and the New Bodleian and DeepStore work took place between September 2010 and July 2011 in coordination with other barcoding and move operations. A final stage in the moves to BSF saw the processing of special collections stored at Nuneham Courtenay between September and December 2011. Large-scale moves of special collections material will not be finally complete until the refilling of the newly completed Weston Library is undertaken after the building is handed back to the Libraries in 2014. Planning for this project is already under way in tandem with other work projecting the operation of the reconfigured special collections library, and will draw upon the experience and expertise developed in the earlier stages.

2 The move to RSL 2.1 Project management At the outset of the planning of the move of special collections out of the New Bodleian, we constituted the SC-RSL Relocation Group to manage the transfer of

Chapter 4: The special collections 

 45

reader facilities and key collections to the Radcliffe Science Library. This team met to consider and manage all aspects of the move, including: – the identification of a preferred location of the new Special Collections Reading Room (SCRR); – the fitting out of the new SCRR; – the move of materials out of the stack space to be occupied in RSL; – the conversion of the stack space to conform with relevant standards; – the move of collections into the prepared RSL stack; – communications with staff and readers about the moves. A smaller team formed the SC-RSL Move Sub-Group to manage the move of collections. This met regularly before the move to ensure planning was progressing on schedule, co-ordinate ongoing activities and identify any further actions that were required. Several key personnel were involved, including a special collections move co-ordinator, estates project officers, the conservation manager responsible for moves and packaging, the head of the bookmoving team, and others closely involved in associated operations and services. A project plan was created and reviewed regularly, using Prince2Lite management principles. We planned that during the move we should hold a regular weekly meeting of those most closely involved, to review progress during the week, review staff and other resources required, monitor the overall programme, ensure adequate communications and agree work streams for the coming week. In the event we found that these meetings were necessary only occasionally because regular contact in the course of operations meant that this kind of information was being shared and reviewed regularly on a more informal basis. We identified certain constraints and dependencies of particular significance to the duration and successful completion of the move project: – Lifts: Lifts at both the New Bodleian and RSL limited the rate at which crates could be transferred from picking shelf to destination shelf. We put contingencies in place as far as possible in case of lift breakdowns; – Weight limits: Weight limits in the New Bodleian and RSL car parks and roadways. These limited the number and size of vehicles that could be used for transporting filled crates; – Data capture: The time required to record the contents of the crates limited the move. We monitored these and other risks using a risk register within the move plan.

46 

 Michael Hughes

2.2 Identification of priority materials A key task in the early planning stages was the identification of materials to be relocated to RSL. Each of the five curatorial sections (western manuscripts, rare books, Oriental collections, music collections and maps) initially considered this question separately. The sections then pooled the lists and measurements of the selected materials, and the SC-RSL Relocation Group assessed them to ensure that the quantities involved would fit in the space available. The principal criteria for selection were the significance of the collections, their physical vulnerability and their frequency of consultation by readers, bearing in mind that remaining collections would be stored remotely at the BSF. In the case of rare books, some of the Bodleian’s most important printed collections were not selected for RSL for three additional reasons. Firstly, open-shelf material for post-1701 collections was principally housed in the Old Bodleian’s Upper Reading Room and we felt that it would disadvantage readers if they could not read both together. Secondly, numbers of orders for some high-use collections were such that the temporary reading room would have struggled to accommodate the number of readers involved at busy times. Thirdly, some of the most important collections were generally restricted for conservation reasons and we felt that these were better housed at the BSF to free up space in the RSL for more frequently used items. The overriding rationale for selecting rare books collections for RSL was to isolate the pre-1701 collections to form a homogeneous collection, along with accompanying open- shelf reference materials. Selected collections were mapped to a collections layout showing where the various sequences would be stored within the RSL stack.

2.3 Enabling work on collections A certain amount of preparatory work was necessary before collections could be moved. For rare books, this involved checking and verifying the Libraries’ catalogue and boxing the most vulnerable of the collections to be moved. For other curatorial areas, work focused on ensuring that those uncatalogued collections to be moved were adequately packaged and identified, together with some boxing of catalogued collections.

Chapter 4: The special collections 

 47

2.4 Consultation and communication with stakeholders and advisory bodies Regular consultation with The National Archives was essential during the planning process to ensure that the proposed facilities met requirements for continued approval of the Bodleian Libraries as a place of deposit, a factor underpinning the whole of the rationale for the move. Consultation with the Libraries’ committees with readers on the intended division of collections between central and remote storage ensured the dissemination of plans to the academic readership. In addition, web pages informed prospective users of the project and were regularly updated to facilitate access to collections at the appropriate location. Because of the high value of the collections we were moving, we carried out consultations with various authorities while planning the practicalities of the moves. The national Museums, Libraries and Archives Security Supervisor was consulted about the proposed move procedures and his approval obtained. We informed the University Security Service about the intended move and they gave advice on the detailed procedures to be employed.

2.5 Constitution of move teams The move of the collections involved a number of staff, filling four types of role. The Special Collections Move Co-ordinator was responsible for co-ordinating the managerial and administrative aspects of the move, including issue and collation of paperwork, scheduling the collections to be moved, reviewing progress and day-to-day troubleshooting. Four curatorial supervisors, drawn from serving curatorial staff, were responsible for the audit and security of collections during the filling and emptying of crates at the New Bodleian and RSL stacks. They reported to the Move Co-ordinator on a day-to-day basis. Their responsibilities included oversight of packing and unpacking to ensure materials were handled, audited and recorded appropriately as they were being packed; securing of crates before transport, and checking and confirmation of the receipt of the collections at RSL. Budget constraints meant that two of the curatorial supervisor posts were filled by drawing from a rota of curatorial staff. A curatorial supervisor’s time was exclusively devoted to this role, with a tie-in between move rotas and reading room rotas, and back-fill of curatorial reading room cover as required.

48 

 Michael Hughes

Two couriers drawn from staff of the Conservation and Collection Care department were responsible for the security of collections from the time they were brought to the loading area at the New Bodleian until they were despatched to the stack at RSL. They reported to the Move Co-ordinator on a day-to-day basis and liaised closely with the bookmovers at all times. They oversaw the orderly transfer of crates from the New Bodleian stack to the RSL stack, and travelled with loaded vans to the destination. Twenty-five bookmovers were responsible, under the supervision of the Head of Bookmoving and his deputy, for the physical move of collections from the New Bodleian to RSL. Their duties covered removing material from the shelves at the Bodleian and loading it into crates, transferring the crates to and from the vans, and moving the collections onto the shelves at RSL.

2.6 Move methodology The collections were moved to RSL in sequences reflecting their mapped location in the stack. With two move streams working at the same time, it was important to keep the activities of each segregated as far as possible to avoid interference with each other’s work. Collections were moved in alphanumeric shelfmark order within each sequence: the opportunity was taken to draw together collections (or parts thereof) stored ‘out of sequence’ in the New Bodleian stacks. The way the move was carried out meant that collections were moved in reverse order within each of the sequences. The duration of move depended on the daily rates achieved. The move was scheduled to last twenty weeks and to be completed by 2 April 2010. Work was phased to concentrate on collections consulted in the Special Collections Reading Room in the New Bodleian initially, and those consulted in Duke Humfrey’s Library (the special collections reading room in the Old Bodleian) in the latter stages. This enabled the Special Collections Reading Room at the New Bodleian to be closed at the earliest possible stage, allowing its operations to be fully transferred to the new reading room at RSL. The move commenced in November 2009 and was completed slightly ahead of programme in March 2010. At the height of the move, each stream moved an average of 86 linear metres of material per working day to its final place on the shelves in RSL. The separate stream of large format material was moved at an average rate of 1.6m per day. Besides the special collections moves, some time was spent transferring the RSL’s own collections between sites, to refill that part of the stack retained for its use. The early completion of the special collections move allowed the spare capacity to be utilized in consolidating the storage of

Chapter 4: The special collections 

 49

Figure 4.1: Moving the special collections

remaining special collections within the New Bodleian, which proved of great assistance when preparing them for their move to BSF. The move of open-shelf reference materials from each reading room was undertaken by external contractors and scheduled to take place at times that minimized disruption to readers as far as possible. Special collections were taken from the existing shelving, placed in plastic, sealable, stacking and nesting transport crates with integral lids, with appropriate packing and protection (Figure 4.1). The current shelf-sequence was strictly maintained. Wherever possible, books and boxes were placed in transit crates so that their shelf-marks were visible. Books and boxes ran in sequence within the crates from left to right, as on the shelf. All slips and tags (including reader request slips) that were found inserted into the sequence were moved with neighbouring material and sufficient space for the missing items allowed for when filling shelves at the RSL. Crates were clearly tagged with a permanent colour-coded number on the left-hand side of the crate and a barcode on the right-hand side of the crate. Co-ordinated, staggered start times were used for van departures from the New Bodleian and RSL so that vans did not have to wait to access a loading bay.

50 

 Michael Hughes

Packing of material at the New Bodleian was carried out under the overall supervision of the Head of Bookmoves. For the move, 240 crates on 80 skates were used; after the first few weeks, when we reviewed the move rates we had achieved, this was increased to 264 crates on 88 skates. The crates were equally divided into two streams, blue and yellow, and sets of 30 (later 33) crates were packed at a time. They were clearly tagged with a permanent colour-coded number and barcode. Because of the need to maintain shelf-sequence, crates were filled in reverse order, i.e. the third crate of three on a skate was filled first and the first crate last. A crate measure stick was used to measure accurately the stock on the shelf so that the third crate was filled from the correct point in the sequence. As each crate on the skate was filled, the next empty crate was stacked above it and filled in turn. The curatorial supervisor recorded the contents of each crate during this process (see Audit, below), and placed a copy of the shipment record in the top crate before seeing it closed and padlocking the security cable for the three crates on a skate. As skates were filled, they were lined up in a convenient aisle ready for transfer to the loading bay. Before the arrival of a van at the loading bay, sets of crates were moved from the pick area to a buffer zone close the lift at the New Bodleian. On arrival of the van, empty crates were unloaded and then the full crates were transferred up to the loading bay in co-ordination with the courier, who scanned each set of crates as it passed the scanning point before seeing them loaded onto the van. When all skates were in the van, the courier secured the van. The drivers carried out vehicle inspections prior to each journey, while the courier checked that the mobile phone in the vehicle was fully charged each day. The University Security Service briefed drivers, co-drivers, and couriers during the induction period, describing the routine security actions to be taken and special procedures in the event of an incident en route. En route the courier notified the Deputy Head of Bookmoving of the van’s impending arrival, upon which the streams leader opened up the unloading area in readiness for the arrival of the van. The sets of crates were scanned by the courier as they passed the scanning point in RSL, before being taken via the lift to the RSL stack. At the same time, empty skates were transferred up ready to be loaded onto the van for its return journey. In the stack, the sets of crates were lined up close to the shelves being filled. The curatorial supervisor unlocked each set of crates, retrieved the shipment record from the top crate, and checked the contents of each crate as it was

Chapter 4: The special collections 

 51

unpacked (see section 2.7 Audit and security below). The bookmovers removed the material from the crates and placed it on the shelves, resetting shelf heights as required. Empty crates were moved to a buffer zone ready for return to the New Bodleian. Items that were vulnerable or unusual in format, shape or size (including photographic collections, papyri, rolls etc.) were packed and moved by Conservation and Collection Care staff under the direction of the Conservation Manager. Such moves were audited and recorded with input from the relevant curatorial staff. These moves made use of bespoke lockable cabinets designed to accommodate a wide range of materials. The movement of the cabinets was undertaken with the utmost care, particularly when dealing with vulnerable materials. A specific period was set aside to enable moves of especially vulnerable items using air-ride vans hired for the purpose, rather than the standard vans to be used in the main move.

2.7 Audit and security Each curatorial section was responsible for undertaking any pre-audit of their collections prior to the move taking place. The time available precluded a comprehensive audit, so some sections prioritized collections of particular significance or complexity. The recording procedures employed during the move (see below) enabled a check to be made against existing inventories and handlists should any discrepancies arise. A shipment record was completed for each set of three crates on a dolly. (Figure 4.2) The completed forms for a day’s moves were passed to the Move Coordinator by curatorial supervisors. The form was in duplicate to allow a record to be retained the central Bodley Library for security, in addition to that travelling with the material. The shipment record included the date, time and crate numbers, and details of any anomalies in the sequence of items. At RSL, the contents of each crate were checked against the information in the shipment record and the checklist signed off, with any anomalies noted. Once all three crates on a skate had been checked, the record was initialled by the curator, retained and passed to the Move Co-ordinator daily. The shipment record will be kept as a permanent record of the move. Since the move, staff have had recourse to the shipment records to confirm the transfer of an item to RSL and help locate it there. For the benefit of readers, regular reports were produced of the shelfmark ranges moved to RSL.

52 

 Michael Hughes

Figure 4.2: Shipment record

Chapter 4: The special collections 

 53

3 Moves to the Book Storage Facility 3.1 Preparation and inventory of collections For the move of special collections to the Book Storage Facility, each curatorial section worked on the necessary preparation and inventory work largely independently, but with initial common planning to establish what needed to be achieved, and how this would fit into the wider barcoding and moving operations. The Special Collections Move Co-ordinator continued to monitor progress of all sections to ensure the project remained on target. The preparatory work involved differed between curatorial sections. For manuscript and archive collections, much of the material involved was uncatalogued and in many cases inadequately identified or boxed, meaning that a considerable amount of inventory and boxing work was required to enable material to be recorded by means of skeleton entries on the library catalogue (which was not generally used for cataloguing or description of such material). When time allowed, and especially where the collections concerned were likely to be in demand, brief boxlists were compiled to facilitate the identification of items when stored remotely. The boxing and inventory work was of huge benefit for the general management of the Bodleian’s manuscript and archive holdings, resulting in a comprehensive list of collections held and their precise extent. In addition to preparing the historical collections, the western manuscripts team processed the extensive holdings of the official library records stored in both the New Bodleian and DeepStore. Existing holdings were augmented at all stages of the move by administrative papers and other materials uncovered in various locations by the ongoing emptying of stack areas, or assigned to library records during staff relocations. For rare books collections, pre-processing work focused on large-scale and often complex data checking and manipulation of existing library catalogue entries to achieve the accuracy and consistency required to enable barcoding to be undertaken by agency barcoders. The data in the catalogue was significantly flawed as a stock-control database, with massive duplication of records across some sections. The pre-processing work effectively ‘cleaned’ the database, creating a one-to-one relationship between record and physical item for the first time. Many tens of thousands of multi-volume works which were represented on the system with a single record were expanded likewise to create a one-to-one relationship – the vast majority of this work was done manually to ensure that every single rare book was surveyed. Some processing of uncatalogued material was undertaken as well, but formed a minority relative to the significant proportion of catalogued rare books holdings earmarked for transfer to the BSF. The prepara-

54 

 Michael Hughes

tion of the John Johnson collection of printed ephemera involved a huge amount of boxing and inventory of large quantities of uncatalogued ephemera. Much of the Oriental material was recorded only in card catalogues, meaning that a considerable amount of inventory work was required to create records in the online catalogue. The fact that the material was in non-Roman scripts meant that language specialists had to work very closely with agency barcoders, creating and editing records and dealing with the many queries that arose from the inventory process. As shelving arrangements had been by language and/or topic, much of the inventory work surrounded identifying the rare books and separating them from the rest of the collection so that they could be boxed and recorded as special collections material before their transfer to Swindon.

3.2 Barcoding Barcoding was in many respects similar to the wider barcoding operations being undertaken, with certain special considerations. Paramount among these was the requirement to avoid damaging unique binding materials through the application of barcode stickers to the outside of books and documents. Instead of direct application, we inserted acid-free slips (to which the barcode was applied) into the rare books. This had the disadvantage of potential segregation of book and barcode slip, but we felt that the conservation benefits outweighed this. As many books as possible were boxed in the time available, to enable barcodes to be applied to the boxes instead of using slips. Because many of the manuscript collections comprised modern archival papers, we took a decision at the planning stage to box all such material before barcoding; the same applied to the John Johnson collection. Besides barcodes, we applied collection codes to groups of special collections material. These were intended as a management tool, firstly to distinguish special collections from other material being transferred to the BSF, and secondly to help with the future refill of the Weston Library through the identification of blocks of material for transfer back to the refurbished library, rather than relying exclusively upon barcode data.

3.3 Moves The move methodology was similar to that for other collections being transferred to BSF, described elsewhere. This was itself grounded in the methods used in the SC-RSL move.

Chapter 4: The special collections 

 55

Because of the move rates necessary for the transfer of collections to the BSF within the timescale available, it was not possible to undertake a detailed audit of collections as part of the move methodology. However, the inventory and barcoding work undertaken beforehand meant that detailed and accurate electronic records of the collections being transferred were available, and could be checked against records of material ingested at the BSF. As an additional precaution, some sections additionally scanned barcodes of material immediately prior to its despatch from the building.

3.4 Maps The move of maps to the BSF was undertaken independently of the remainder of the special collections moves, principally as a result of the very different physical format and storage requirements of maps; and also because the maps retrieval operation from the BSF was designed to function outside the Integrated Library System. Barcoding and creating retrievable catalogue records for over a million items was deemed to be impractical given the substantial volume of material due to be moved. A project management committee was set up to prepare and deliver the move, with representatives from the Map Room, the BSF, Bookmoves, and Document Delivery section. Meetings between March and August 2010 devised a workable methodology concentrating on preparing the ground for the decant and setting in place a mechanism for retrieving maps ordered to the Bodleian from Swindon. From March 2010, all out-of-sequence map material was re-sorted, creating an overall maps sequence in shelfmark order. By June, a target decant figure of 50–60 drawers per day had been established, later modified to around 6,000 maps per day [= 42 drawers], scheduled to run for the eight months, and in August 2010 a fleet of 44 bespoke map trolleys was ordered to facilitate the move. A total of 3,247 map drawers required emptying from the New Bodleian, located on three floors of the bookstack. This figure did not take into account the substantial number of maps placed loose on top of the cabinets. Map Room staff were asked to prepare an operating schedule to show how best to order the emptying and subsequent filling of drawers, whilst retaining shelfmark order. Meanwhile, at Swindon, pre-planning of empty drawers for growth space was undertaken. It was imperative to have Map Room staff supervising the ingest, alongside three BSF staff appointed specifically to fill the map drawers. An efficient methodology for unloading the 44 map trolleys was devised, so the drawers could be filled quickly, safely, and most importantly, in shelfmark order.

56 

 Michael Hughes

The move began on 17 November 2010 with 144 drawers of Ordnance Survey mapping transported to Swindon. Three van deliveries throughout the day took twelve trolleys each time with four drawers of maps per trolley. By 14 February 2011, the New Bodleian was finally emptied of maps, three months after the move commenced, and five months ahead of schedule. The rapid decant rate was largely due to excellent work by experienced bookmovers who had honed their skills during the special collections move to the RSL. The final part of the move saw around 140,000 sheet maps from 1,525 drawers, along with some 68 map cabinets transferred from the Nuneham Courtenay repository to the BSF during the summer of 2011. We had pre-allocated empty drawer space pre-allocated at the BSF for most of this material. A small relational database is used to monitor requests for map material from the BSF. An automated request system was developed whereby Map Room staff can to input data which is printed out in Swindon, providing BSF staff with details of the reader, their barcode, delivery destination, map shelfmark, map drawer identifier at the BSF, and general notes. This system has continued to work well.

3.5 Music For printed music, the situation was complicated by the fact that the material was partially (but only partially) represented in the main library catalogue and these categories were intermingled on the shelves. Also, much printed music is thin, flimsy and, for the most part, unbound. Such material is grouped together and stored in conservation boxes so the music area of the New Bodleian bookstack contained a mixture of bound volumes of sheet music, more substantial standalone items and boxed material, some of which was represented in the online catalogue but the vast majority of which was recorded only in the card and slip catalogues in the reading room. Since specialist music staff have always had complete control over the music area of the stack, fetching and replacing music scores for readers themselves, these factors had not hitherto caused particular problems. However, with the imminent prospect of remote storage, a quick solution had to be found which would enable the collections to be recorded in such a way that they could be moved and retrieved for readers by BSF staff when required. Because of the insubstantial and fragile nature of most of the unbound sheet music and the frequently less-than-obvious arrangement of the scores in the boxes to a non-specialist, we considered it undesirable to empty the boxes of sheet music for ingest into the BSF, so we decided that whenever an item in a box was requested the whole box would be returned. This meant that a ‘retrievable unit’ at the BSF could be any of

Chapter 4: The special collections 

– – –

 57

a single item (where substantial enough to stand alone on the shelf); a bound volume; or, a box of unbound items.

In the second and third cases there would not be a one-to-one relationship between a bibliographic record in the online or card/slip catalogues and a ‘retrievable unit’. It was therefore necessary to provide some kind of link between a finding aid for an individual item of printed music and something which represented the physical entity to be fetched. The mixture of catalogues (along with a certain amount of uncatalogued material) was an additional complication. In the light of the tight timeframe and the need to keep costs down, we decided to create a dummy record in the online catalogue for every ‘retrievable unit’, against which stack requests could be placed, whether or not that item was already represented in the catalogue. Now, when a reader initiates a request for an individual item, either online or manually, music staff re-order the item against the appropriate dummy record. A temporary member of staff was appointed for the 12 weeks leading up to Christmas 2010 to work with an existing member of the music team to produce a series of spreadsheets which recorded basic information for every stand-alone item, bound volume or box on the shelves of the Music stacks. In practically all cases, the information recorded was no more than the shelfmark of the item, with additional qualifying information if necessary. A total of 55,060 items was recorded in this way, representing approximately half a million individual music scores. Each line of the spreadsheets was then converted into a basic MARC ‘bibliographic’ record which was then loaded into OLIS, with an item record attached to allow for the tracking of the movement of the items on the system. We subsequently created a further 300+ additional records to record uncatalogued or ‘non-collection’ material from the Music Section which could no longer be accommodated in the staff office after the move.

4 Weston refill methodology Work on the planning of the refill of the refurbished Weston Library, which will be dedicated to the storage of special collections, has already begun. Initial considerations cover the selection of those special collections to be housed in the new stack areas, as there will be a shortfall in the space available compared with the total of all special collections in the custody of the Bodleian Libraries. Besides collections falling within the curatorial areas formerly housed in the New Bodleian,

58 

 Michael Hughes

collections such as those of the Bodleian Library of Commonwealth and African Studies and the Oxford University Archives will be integrated into the Weston Library. In general, collections considered of low use (especially uncatalogued collections) or of low vulnerability will be retained at the BSF. We intend that collections will be stored in shelfmark order, reflecting a strong preference to maintain the discrete identity of many special collections and facilitate several aspects of their management. The transfer of collections back from RSL and certain other locations will involve a relatively straightforward replication of existing storage sequences. In the case of collections temporarily transferred to BSF, a reconstitution process will be necessary, and trials are under way to establish the best method of achieving this and assessing the impact on move rates.

5 Statistics Quantum moved (linear metres) To RSL/new reading room locations Western Manuscripts

6,186

Rare Books

3,597

Oriental collections

1,810

Music collections

164

Maps

34

Open access material

387 (+2525 microfilms)

Total

12,178

To the Book Storage Facility Western manuscripts – from New Bodleian

3,705

– from Deepstore

500

– from Nuneham

90

Library Records – From New Bodleian

590

– From Deepstore

620

– From Nuneham

113

Chapter 4: The special collections 

Rare Books – From New Bodleian

11,502

– From Nuneham

2,725

John Johnson collection

2,605

Oriental collections

535

Oxford University Archives

400

Music collections

1683

Maps

1,2000,000

Total (linear measure only)

24,478

Grand total

36,656

 59

Jonathan McAslan

Chapter 5: The purchase of electronic journal backfiles The reasons for the purchase of electronic journal backfiles were twofold: first, and most pressing, was the need to provide electronic access to the print journals held in the Radcliffe Science Library so that they could be removed from there to remote storage to allow the move of Special Collections from the New Bodleian to the Radcliffe Science Library, and second, to provide, in line with the Bodleian Libraries’ strategy, increased access to electronic resources for users so as to reduce demand for materials from the Book Storage Facility. The Bodleian Libraries had begun expanding their electronic resources through the Electronic Library and Information Service for Oxford(ELISO) initiative in 2005 that significantly increased online content.¹ ELISO injected nearly a million pounds into the purchase of electronic resources in response to a commissioned report that had highlighted how far Oxford lagged behind other major universities in its provision of electronic resources.1 While journal backfiles did not feature heavily in the ELISO purchases, subscriptions to the electronic journal collections of many leading publishers did. In the following years the e-journal backfiles from three major publishers were purchased, but these were costly, and there was considerable demand to increase current content. The need for backfiles, while not necessarily as pressing as for current content, was increasingly recognized as important. As users came across references to articles not available in the current ‘front file’ of a journal, the demand for backfiles grew. The purchase of the Web of Science backfiles indexing back to 1945 (and other bibliographic databases pushing their indexing further back in time) provided users with an increasing number of references to older journal content . Reports around this time highlighted the benefits of journal backfiles and the reasons why a university like Oxford would need them.² We made the case for additional funding for the provision of electronic backfiles to our governing body the Curators of the University Libraries in September

1 Mel Collier and Derek Law, ELISO: an electronic library and information service for the University of Oxford (London: Electronic Publishing Services Ltd., 2005). 2 Journal Backfiles in Scientific Publishing (London: British Library, [2006]), accessed 29 June 2012, http://www.bl.uk/reshelp/atyourdesk/docsupply/productsservices/digitisation/ journalbackfileswhitepaper.pdf

62 

 Jonathan McAslan

2008 as part of a strategy involving the changes to the New Bodleian Library that would involve the temporary move of Special Collections from the New Bodleian to the Radcliffe Science Library. In order to make space available in the Radcliffe Science Library it would be necessary to remove considerable amounts of print stock to remote storage. The purchase of the equivalent electronic versions of print journals would be a quick and efficient way to allow large amounts of print stock to be removed. Not only would users not suffer any disruption to their access to journals, they would also have enhanced access to the same journals and more. Because it was necessary to purchase the backfiles quickly so as not to delay any of the projects, we began quotes and discussions with publishers in early September 2008. Agreement on an investment of £ 500,000 provided by the University was reached shortly afterwards in October 2008 and the first purchase was made on 14 November 2008. The need for e-access to content previously held in print in the Radcliffe Science Library was pressing as the move of the print to storage was critical in the move of special collections to the Radcliffe Science Library. What we needed was a lot of space in a short period of time. The purchase of backfile journal content was a quick way of freeing up shelf space. We concentrated on journals from UK publishers so as identify the largest amount of relevant material quickly. We established early on that selecting UK publishers such as Taylor and Francis (whose content we held on our shelves as far back as 1798), Cambridge University Press and Blackwell Publishing would enable space saving in the Radcliffe Science Library. We also gave consideration to non-UK publishers where large runs existed in print (e.g. Wiley back to 1799). We felt that we should to spread the purchases across all the main subject areas as most previous journal backfile purchases had been heavily weighted towards the physical sciences. Therefore, after the initial purchases to allow the Radcliffe Science Library decant to proceed, some of the money was set aside for humanities and, to a lesser extent, social sciences purchases that did not have to be journal backfiles. All the purchases we made from this first tranche of funding, however, were journal backfiles: over 1,000 titles from major publishers. The University granted us a further £ 500,000 and this was used to buy 600 more journal backfiles as well as major UK newspaper archives, some e-books and a major Chinese reference collection. By the end of the project 74 % of the money had been spent on journal backfiles, 10 % on newspaper archives and the remaining 16 % on a range of reference works and e-books. The first round of negotiations with publishers and suppliers established what was available and the list price of that content. The purpose was to create a list, with pricing, of all that was available that fell within the criteria for purchase. The initial list was for e-journal backfiles from those publishers offering

Chapter 5: The purchase of electronic journal backfiles 

 63

the largest amount of content that matched our print holdings. At this stage it was not necessary to have costs that fitted exactly within the £ 500,000 budget as the list was intended to be comprehensive so that a selection could be made from it. While there were a lot of backfiles available, there remained many that were not yet digitized. Some publishers were still working through their digitization programmes and did not have all, or even close to all, their journals available. In one case, a publisher was willing to offer a pre-publication price, but this was unique as other publishers were only willing to declare a price when the content became available. As one of the publishers who could not provide pre-publication sale was a major publisher of UK journals this meant that not all backfile space savings were possible. Once the list of potential purchases had been drawn up it was submitted to the Bodleian Libraries’ Cabinet, a meeting of its twelve most senior staff, for a decision. The main driver at this stage was space saving so the approved list was composed of over 70 % UK publishers for which the libraries had print legal deposit copies. The remainder was for non-UK publishers for which we had substantial print holdings. The first list was approved unchanged as it had been drawn up with the space saving as its priority; but the approval came with the proviso that it had to fit, as closely as possible, within the £ 500,000. The Cabinet also made it clear that where a publisher offered title-by-title selection that there must be an equitable financial split between the key subject areas. To this end we requested a subject-classified list from publishers and the costs, rather than the number of titles, for each subject area were selected. With the list approved, we began the negotiations with publishers. Staying within budget was slightly complicated by the fact that some of the backfiles were priced in foreign currency and thus liable to fluctuating exchange rates. We therefore felt it necessary to have a bit of a cushion. The first negotiating approach was to reduce costs based on the Bodleian Libraries buying most, if not all, the backfiles from a publisher. This was largely successful with publishers offering discounts as high as 20 %. One publisher originally came with an offer of 15 % on one collection and 20 % on another but agreed to settle for 20 % on both. As we were being cautious about overspending due to currency fluctuation, the first backfile purchases, completed by the end of November 2008, came in at just under £ 496,000. This provided access to 980 journals dating back to 1798. In terms of the number of years (which would approximate volumes) the new backfiles provided access to over 20,000 years of journal content. The space savings were considerable as one journal with a backfile to 1958 took up over 14 metres of shelf space. If we accept this as typical for all the purchased journals, then we freed over 13 km of shelf space.

64 

 Jonathan McAslan

But the purchase had a secondary purpose and one that tied in with the Libraries’ strategy: providing users with access to additional electronic content. We hoped and expected that usage of electronic journals would increase with the added backfiles being made available. The purchase of the backfiles was publicised through an announcement on the Libraries’ web pages on 20 January 2009. In this Bodley’s Librarian highlighted the advantages to users: ‘Our goal is to provide services that anticipate and respond to the needs of our users. Online journals are a fundamental asset for 21st-century scholars. The retrospective journal files we are providing will benefit academics and researchers across the disciplines.’³ Following the purchase of the first £ 500,000 of backfiles and the move of print journals from the Radcliffe Science Library to remote storage, the University released the remaining £ 500,000 for further purchases. For the second £ 500,000 we had more time for consideration as the pressure to free up space in the Radcliffe Science Library was not a factor. We consulted the subject librarians on what they would like and drew up a new list, still prioritizing space savings but also taking into account greater access and ease of use by cutting down on requests for print material that had to be fetched from closed stacks for users. The list was drawn up with a column indicating the reason why the content should be purchased, and Cabinet made the final decision on what to buy. Readers requested newspapers frequently so we had identified them early on as possible backfile purchases. In addition to newspapers more e-journal backfiles were becoming available as more publishers digitized their content. But this time Cabinet, taking into consideration the need to spread the purchases across the subject areas and the requests from the humanities and social sciences, agreed to use some of the backfile money to buy e-books. The purchase of e-books was seen as a way of reducing the number of print copies that the Libraries purchased. The e-books were predominantly copies of books on undergraduate reading lists, or reference works. Negotiations with publishers and suppliers again concentrated on discounts for large purchases. We bought all the newspapers from the same supplier and thus we were able to negotiate a considerable discount on their purchase. It was also possible, in the case of one publisher, to negotiate a reduced cost for a current e-journal deal based on buying the backfiles of the journals. Another publisher was willing to offer a considerable pre-publication discount on their backfile collection.

3 ‘20,000 years of journal publishing added to OULS e-resources’, Bodleian Libraries, 29 January 2009, accessed 29 June 2012, http://www.bodleian.ox.ac.uk/news/2009_jan_20.

Chapter 5: The purchase of electronic journal backfiles 

 65

The purchases of the second phase of the project began in March 2009 and were completed in August 2009. This included the backfiles of 695 journals, bringing the total number of purchased backfiles out of a total of £ 1 million funding to 1,675, with the earliest issues from 1798. The implementation of the backfile purchases commenced as soon as each had been paid. Publishers and suppliers provided lists of the titles in the backfiles and their coverage years. We amended the journal holdings on the Libraries’ A-Z e-journal list from the link resolver system SFX. There was quite a bit of manual entry as some of the publisher backfiles were not specific collections and therefore not bundled as such on SFX. Also, as some of the backfile collections were newly digitized they did not appear as discrete collections on SFX. We also had to check individual titles to ensure that all the backfile was available. As some of the content was in the process of being digitized not all content was available and thus did not match the coverage years listed by the publisher. A particular frustration with the information provided by the publishers was that some (most) publishers did not separate journals on their websites to take account of title changes. The providers of SFX, Ex Libris, created linking in SFX based on the lists provided by publishers and linked to current titles without providing links to previous titles of the same journal. We do not know precisely what effect this has on discoverability but we can assume that it may have some negative impact on usage. Once SFX had been updated we uploaded the records into the library catalogue via Ex Libris’s MARCit service. The addition of the backfile content gave Oxford one of the most extensive backfile collections then in existence. With some publishers we were the first library to buy their complete content. Combined with earlier backfile purchases, Oxford had the complete or nearly complete e-journal backfile collections of several major publishers. The success or otherwise of the backfile purchases could be measured in two ways. First, the successful decant of print stock from the Radcliffe Science Library without causing disruption to user access to journal content. Second, the use of the backfiles would indicate whether users were getting an advantage out of their purchase. The Radcliffe Science Library decant was completed on time with the backfiles in place to minimize disruption. That is not to say that there were not occasional problems where journals believed to be in a backfile collection were sent in error to storage, but these minor problems should not be seen as detracting from the overall success of the decant and backfile purchases. Based on the usage of previously purchased backfiles from Elsevier and Nature we had expected that about 3 % of the usage of a journal collection would be from backfiles. However, usage of the newly purchased backfiles averaged over

66 

 Jonathan McAslan

15 %: two collections had 14 % backfile usage, one 19 %. This may have been in part due to the inclusion of considerable social science and humanities material, where historical content has much greater interest for users. Today, across all our purchased backfiles (including those purchased by the national consortium JISC Collections on behalf of UK universities), usage is over 10 % (or approximately 450,000 articles per year) of the collections containing backfiles.

Toby Kirtley

Chapter 6: The building of the Book Storage Facility 1 Introduction and strategic need To deliver the programme for improving and rationalizing our library estate the Bodleian Libraries first required a high-capacity off-site Book Storage Facility (BSF). We needed sufficient capacity to accommodate new growth in the collections, lower-use stock from libraries being merged or re-purposed, and largescale collections from buildings being emptied for refurbishment, in particular the New Bodleian Library. In order to provide optimal conditions for the storage of documents the BSF would need to comply with British Standard BS5454, especially with regard to fire.¹ It would need to meet the Libraries’ storage requirement for a 20-year period and be high-density and so cost-effective to build and operate. It would have to be sited in an industrial area, given its bulk and height, and yet it would need to be close enough to libraries to allow for the rapid delivery of requested books. These requirements were met through the construction of the University’s largest ever book repository at South Marston on the east side of Swindon, some 45 kilometres (28 miles) from central Oxford. After an extended planning period during which we overcame a number of challenges, construction of the £ 24 million Book Storage Facility began in September 2009. (Figure 6.1) The completed BSF opened in September 2010 with a capacity of 247 linear kilometres of shelving and over one million maps.

2 Background Oxford University’s centrally funded libraries first began to make use of offsite storage in 1975, when the then Libraries Board opened the first of a series of low-cost book storage modules at Nuneham Courtenay, 13 kilometres (8 miles) outside Oxford. By 2003, seven modules had been constructed in all, to a stan-

1 BS5454:2000 Recommendations for the storage and exhibition of archival documents (London: BSI, 2000).

68 

 Toby Kirtley

Figure 6.1: The Book Storage Facility under construction

dard pattern. Each module comprised a single-storey 500 m2 room containing four blocks of conventional mobile shelving or planchests (for maps), arranged around two gangways. On average a module provided some 5,500 m of storage and the whole 3,477 m2, development provided storage for 37,423 m or 1,272,000 nominal volumes. An ancillary wing provided facilities for staff and a small room to serve as a reading room for occasional use by visiting readers. Although some air handling was provided, the development did not comply with the relevant standards for library storage and particularly lacked adequate fire engineering systems. The slow development of the Nuneham Courtenay Book Repository, over nearly thirty years, and the limited size of its component modules, meant that it was unable to provide a strategic solution to Oxford’s remote storage needs. A notable feature of the repository was its location within an eighteenth-century walled kitchen garden and, from the outset, planning permission was severely limited. Although the University brought forward plans in 2002 to more than double the size of the facility to some 65 linear kilometres, its planning application was refused and the University had to look elsewhere to meet its strategic needs. Faced with a significant lead time to develop plans for a new, larger store, libraries were forced to begin making use of commercial off-site stores. By the

Chapter 6: The building of the Book Storage Facility  

 69

time of the BSF’s opening, some 2,000,000 books had been placed into commercial storage with the companies Iron Mountain and DeepStore. The high costs of contracting out storage meant that this could not in itself provide a sustainable strategic solution.

3 Towards a high-density storage solution The move away from the Nuneham Courtenay Book Repository and its limitations brought an opportunity to make use of higher-density storage methods that, per book stored, would reduce both the capital cost of construction and annual running costs. In 2000 the University had acquired warehouse units on the Osney Mead Industrial Estate 1.3 kilometres (0.8 miles) and just under 4 kilometres (2.5 miles) by road from the central Bodleian. Although these buildings were themselves old and of poor quality, their sites, covering approximately 1.5 hectares (3.75 acres), had considerable potential. Transfer times between Osney Mead and the central Bodleian were about 20 minutes, allowing requested items to be rapidly delivered from storage to reading rooms. There were precedents for large-scale industrial storage developments on the estate, including tall structures. Moreover, the sites were immediately adjacent to a library support services building, the Osney One Building, which was to be refurbished as accommodation for the Libraries’ departments of Conservation & Collection Care and Technical Services. In March 2004 a group of Libraries staff undertook a study tour of high-density storage systems on the west coast of the United States. The group saw three different storage solutions. High-level static, narrow-aisle shelving with access floors was seen at the University of California’s Northern Regional Library Facility (NRLF) at Richmond, California. A ‘Harvard Model’ fixed rack with access by High-Level Order-Picker was viewed at Stanford University’s SAL3 facility at Livermore, California, and Automated Storage and Retrieval Systems (ASRS) were seen at Sonoma State University Library at Sonoma, California and the Lied Library at the University of Nevada, Las Vegas. Within the UK, the group visited a number of retail distribution centres using similar technology, and made initial enquiries with the leading providers of logistics systems. Further reference-site visits were made to the US and to the National Library of Norway’s Book Repository at Mo i Rana in 2005. We carried out a number of internal option appraisals in 2004 and 2005, exploring potential layouts on the Osney site and comparing a range of building types and storage solutions. We recognized that a specialist logistics consultant would be needed to undertake a thorough assessment of options and in Febru-

70 

 Toby Kirtley

ary 2005 Total Logistics Ltd were appointed. Their initial study compared four options (Harvard Model, ASRS, conventional mobile shelving in a multi-storey building, and NRLF-type high-bay shelving). Their report recommended ASRS as the lowest cost option in terms of recurrent costs per book.

4 An Automated Storage and Retrieval System Detailed plans were drawn up for an ASRS facility at Osney. This comprised a fourteen-aisle configuration of high-bay racking, with each aisle served by a fully automated crane controlled by a warehouse management system communicating directly with the library management system. One of the fourteen aisles was to hold 620 map planchests with the remaining thirteen housing books stored in large polypropylene trays. The book aisles were served by a cross-link conveyor moving trays between storage racks and operator positions, where individual books could be picked or re-filed. The facility would be fully compliant with BS5454: 2000 and would be protected by a low-oxygen environment to prevent any fire ignition. The site stood partly on the flood plain of the River Thames, at or below the 1-in-100-year flood contour. As a result, the design incorporated a defensive bund wall to protect the building against a 1-in-5,000-year flood, and sub-floor voids to hold flood water and prevent consequential flooding of neighbouring property. The proximity of the site to the city centre meant that the impact on the famous skyline of ‘dreaming spires’ had to be minimized. Our attempts to meet the City Council’s objections led to a highly unusual design for an ASRS. The overall height of the building was limited to 18.2m and a wave-form, or ‘sinusoidal’, roof was adopted to soften the building’s profile. Varied materials and finishes were carefully chosen to break up the mass of the building and help to blend it into its surroundings. The novel roof line meant that height of the rack and cranes inside the high bay had to vary between 7 m and 15 m. The University submitted a planning application in February 2007 and it was passed by the city’s Area and Strategic Planning committees. However, such were the sensitivities around visual impact and flooding that the application was subsequently reviewed in a full meeting of the City Council in November 2007, and rejected.An appeal lodged by the University also failed and it became clear that a new site had to be found.

Chapter 6: The building of the Book Storage Facility  

 71

5 A new site and a new approach After a detailed search of more than 90 sites across the region a highly suitable site was identified at South Marston, on the eastern edge of Swindon. Some 45 kilometres (28 miles) south-west of Oxford, the location allowed a transit time of only 45 minutes. The site was well elevated and, being adjacent to other high-bay industrial units, had no onerous planning restrictions. The hiatus resulting from the change in site allowed us to review the Libraries’ strategy and the storage solution itself. The Osney Mead site had been close enough to city centre libraries to house material regardless of how heavily it was used, since a fast and frequent delivery of material to reading rooms would have been possible. A more remote site would not allow such rapid access and a different strategic approach was called for. The proposals described in other chapters of this book enabled the remote store at Swindon to hold lower-use material. Accordingly, the case for a rapid, automated retrieval system was reduced and a review of storage solutions showed that a Harvard Model facility would be more suitable. In particular, we felt that a more staff-based (though still mechanized) solution would be more flexible, with scope to reduce operating costs should the demand for physical books decline with increasing digital access. Ultimately a Harvard Model repository could be served by a very small staff, while an ASRS repository would require ongoing support for a large fleet of automated cranes. We undertook a further study tour to the USA in May 2008, focusing on the ‘high-density’ repositories at Harvard, Yale and ReCAP. All of these Harvard Model facilities use the same storage approach. Each item is given an individual barcode recorded in simple warehouse inventory software which communicates with the libraries’ management systems. Items are placed in barcoded cardboard trays which are held, several deep, on barcoded steel shelves in banks of high-bay racking separated by narrow aisles. The inventory system records the tray, shelf and aisle location for every item in the store and tracks the picking and re-filing of individual documents as they are withdrawn for reader use and returned. Staff access the high-bay racking from onboard electric High-Level Order-Picker vehicles, riding on access platforms that lift them and their picking trolleys high into the air. In the UK the system is known as Very Narrow Aisle storage. The use of this industrial logistics approach for library storage was pioneered in the US by the late Reese Dill and there are now over sixty high-density storage modules across the USA. Although it is widely used in commercial logistics it had never been applied to library storage in the UK. The Libraries were able to benefit from Reese Dill’s tremendous experience through his consultancy services, though a number of deviations from his established model were required because of differences between US and European regulations, products and suppliers and the specific nature of the Bodleian’s project.

72 

 Toby Kirtley

6 Designing the South Marston Book Storage Facility The building housing the storage facility was designed closely around the operations that were to take place there, and in particular around the immense shelving system used to hold the stock. (Figure 6.2) The shelving in turn was designed closely around the collections that it would hold, and the requirements of the staff and mechanical handling equipment used to access it. Accordingly a great deal of early work was done to understand more fully the size, nature and usage of the Libraries’ collections. A planning base was drawn up to record the quantities of manuscripts, printed books, microfilm, maps and other material identified for transfer to the BSF on completion. We analysed the dimensions of library material, in the various size groupings that have been in use in varying periods of the Libraries’ 400-year history. A range of sample material was weighed to determine both the load that the shelving system would have to carry and the maximum weight of items that staff would need to lift. We analysed reader requests to estimate the overall number of requests that the BSF would have to handle as well as the day-to-day and hour-to-hour patterns of usage. This preparatory scoping work showed that the BSF would need a capacity of some 8.4 million nominal volumes and 1.2 million maps. It would need to accommodate objects up to 20 kg in weight and some 550 mm x 1,000 mm in size. The initial estimate of overall demand for items was that this would be approximately 230,000 each year, with a peak requirement for 1,595 retrievals a day. To achieve a good balance between high-density storage and efficiency of operations the design of the shelving system needed to take a number of considerations into account. To maximize storage density, the blocks of shelving would need to be as tall and long as possible, with access aisles as narrow as possible and shelving as deep and as closely spaced as possible. (Figure 6.3) At the same time, staff would have to be able to access all material quickly and safely, with minimal manual handling. The height of the shelving installation, and of the building overall, was governed by planning restrictions. Allowing for a roof structure and service zone above the shelving resulted in a shelving height of 11.4 m. The High-Level Order-Pickers required to reach the top of this racking required an aisle width of just 1.5 m. Increasing the height to beyond 11.4 m (had the site allowed this) would have required larger, wider vehicles with correspondingly wider aisles and this would have significantly reduced density. The length of the shelving blocks was determined in part by the size and layout of the site itself and partly by the need for staff to be able to travel to all areas of the shelving relatively quickly –excessively long blocks would increase travel time to the shelves.

Chapter 6: The building of the Book Storage Facility  

Figure 6.2: The Book Storage Facility from the south

Figure 6.3: A narrow aisle

 73

74 

 Toby Kirtley

A block length of 71 m would allow efficient storage and access, and would result in a building size that was acceptable from a planning perspective and which would allow for the further development of the site in the future. The depth of the shelving units was determined by the limits of human reach. Staff had to be able to withdraw all items from the shelf from a safe standing position without the need to lean out into the shelving. The length of the shelves was determined by structural considerations. Although the use of wider shelves would improve density horizontally by reducing the number of support uprights, this gain would be lost through a corresponding reduction in density vertically because, to maintain strength, wider shelves would also need to be thicker. A cost-benefit analysis was undertaken to determine whether we should use specially designed shelves sized to our own requirements, or use existing commercially available shelf sizes. We found that adopting existing shelf sizes would be the most cost-effective approach, with minimal impact on density. The optimal dimensions for a standard shelf were found to be 1,283 mm wide x 990 mm deep. The vertical spacing, or pitch, of the shelving was a crucial factor in achieving high density. For structural and health and safety reasons shelf installation and alterations could only be done by the shelving contractor. As a result, we had to determine all shelf pitches in advance. We would require a number of standard pitch settings, enough to accommodate the range of stock heights without incorporating too much free space. We gave careful consideration to the size groupings occurring within the collections. In doing this the Bodleian departed from the usual practice followed at American Harvard Model depositories, where on arrival all incoming material is split up and sorted into new depository-specific height bands according to sizing templates. This practice has largely been necessary because of the amount of stock coming from libraries holding unsized material in Library of Congress classifications. The Bodleian, from which the vast majority of the BSF holdings would come, had already been sorting and storing stock by size in a number of different schemes from the seventeenth century onwards. The BSF could take advantage of these existing size bands. Being able to ingest existing stock into the BSF without sorting by size would also allow a rapid initial load of the facility, so reducing unit costs, allowing savings in commercial storage and allowing the New Bodleian to be vacated promptly ahead of refurbishment. Initially, eight BSF shelf pitches were proposed, but this was reduced to six by combining two of the size bands with adjacent size bands. The storage trays form a key component in the overall storage solution. The requirement was for these to be of sufficient quality to withstand physical handling and protect their contents for a design life of 50 years, all at a relatively low unit cost, given the large numbers required. The overall concept was based on that used in the United States but dimensions and the form of construction

Chapter 6: The building of the Book Storage Facility  

 75

Figure 6.4: Storage tray

were altered for the BSF. Archival-quality corrugated board was used because of its rigidity and light weight, allowing a high-throughput, automated production process and therefore lower costs. Archival corrugated board has not been used much within the UK and a special make was produced for the BSF by Smurfit Kappa in Norwich from Billerude White paper imported from Germany. Flat tray matrices were die-cut by Caps Cases Ltd of Newmarket and delivered on pallets to the BSF, to be made up on site as needed. Nylon grab-handles and backing plates were provided by Item Products Ltd. During fire tests at the Building Research Establishment the original polypropylene handles were found to present a fire risk and fire-resistant nylon was substituted. Five different tray sizes were produced, with the largest in two variant forms to accommodate books and two-part flat archival boxes. At the prototyping stage, sample trays were tested under load to simulate 50 years of usage. The largest items, above 408 mm in height, would be stored laid horizontally, directly on the shelves, not in trays. Two different tray lengths were adopted. Trays A-B measure 488 mm in length and are stored two deep on the 990 mm-deep shelving. Trays C, D, E and XE are for larger items. In order to reduce tray weight they are shorter, at 323 mm, and are stored three deep on the shelves. Trays are arranged in up to seven ‘files’ across the front of a shelf.

76 

 Toby Kirtley

Figure 6.5: Planchests

Not all the material stored in the BSF is stored within this tray-based shelving system. One storage aisle has been built to a different configuration – a five-level multi-tier comprising standard beam pallet racking with in-fill access floors reached by stairs and served with a goods-only hoist. Four of the five levels hold some 600 map planchests with deep shelving above and the fifth is fitted out entirely with deep shelving for rolled maps and other items not suitable for the tray-based system. Trays are loaded onto the shelving and individual items picked and later refiled by staff on High-Level Order-Pickers. Selecting and modifying these was an important part of the design process since they interacted closely with the shelving and had to provide a safe working environment for operators. The orderpicker selected was the Jungheinrich EKS 312, in a configuration where the operator faces forwards over the engine compartment to drive the order picker along the aisles, into position at a shelf, and then turns around to work from a trolley

Chapter 6: The building of the Book Storage Facility  

 77

Figure 6.6: High-Level Order-Picker

carried on forks at the rear of the vehicle. The order-picker has a triplex mast extending to a maximum pick-height (the height of the highest tray handle) of 10.6 m. Items stored on the lowest four shelves cannot be retrieved from onboard an order picker and staff dismount to access them. Two variant forms of the EKS 312 are used. The ZG model is used for loading the shelving and has a pair of moveable forks on the front, allowing heavy trolleys of filled storage trays to be raised and lowered to a comfortable working height. The LG model order-picker is used for picking and re-filing individual books and has non-moving forks that hold lighter book trolleys in a fixed position. Although manually driven by staff, the electrically operated order-pickers are steered by a wire guidance system embedded in the concrete floor while they are in the shelving aisles – any slight deviation from the centre line of an aisle is immediately compensated for. Separate transponders in the floor limit the speed of the order-pickers as they approach the ends of an aisle.

78 

 Toby Kirtley

Two designs of steel trolley were specially made for the BSF. The heavier trayto-shelf (TTS) trolleys have three shelves carrying up to 36 fully loaded trays or 1,224 nominal volumes. The top two shelves fold back when empty, to give access to the shelf below. In operation, the order picker operator brings a shelf of the TTS trolley up level with an empty storage shelf and filled trays are pushed out over the intervening gap into their final storage location. The lighter picking trolley has two fixed shelves below a fixed work surface at chest height. In use, this work surface is brought level with a storage shelf of book trays; the required tray, and any tray in front of it, is pulled out onto the trolley work surface and the requested book selected. The tray and its remaining contents are then pushed back into position on the storage shelving. This handling of trays on the trolleys and at the shelf is the principal manual handling activity in the BSF and the process and equipment were designed with the close involvement of health and safety consultants and University Safety Office staff. At no point in the BSF process do staff have to lift a loaded tray. Material received by the BSF for ingest into storage, and the items being picked and refilled, are all processed in an ancillary hall next to the high-density storage area. This is an open-plan area that can be fitted out and re-configured for a variety of functions, from large-scale ingest projects and normal day-to-day book delivery operations to scanning and box-making. The benching and equipment in use here is therefore designed to be movable. Power and data are delivered from the ceiling as well as from the perimeter. The ancillary hall was sized to accommodate normal operations rather than the large-scale ingest operations that were needed in the first eighteen months of the building’s life. Initially, we thought that we would need a temporary extension for this initial ingest project. However, through careful design, an efficient, linear arrangement of processing benches, the rapid processing of inbound stock at up to six benches and the temporary use of the high-bay for the storage of tray flats, the ancillary hall functioned efficiently without additional space. Ingest benches were designed to minimize manual handling. Hydraulic lift platforms were used to raise incoming transport totes of books to a comfortable working height, bench tops were equipped with roller-ball tops to allow trays to be pushed and pulled along through the process, computer screens, keyboards and barcode scanners were mounted over the benches on adjustable arms and electric stacker trucks were used to lift the heavy tray-to-shelf trolleys to bench-top height and to move them in and out of the high-bay area.

Chapter 6: The building of the Book Storage Facility  

 79

7 A building to house BSF operations The building that houses this ‘storage solution’ comprises two distinct parts. The high-bay store houses the 31 aisles of tray storage and the map multi-tier aisle. In order to reduce risk it is sub-divided into four four-hour fire compartments, each with a capacity of some two million volumes. In day-to-day use, the compartments operate as one space linked by cross-aisles at the front and back of the high-bay. In the event of a fire, however, the connecting doorways close with automatic fire shutters to create four separate protected chambers. The high-bay is protected with an ‘in-rack’ water sprinkler system. Four levels of distribution pipework and a close arrangement of sprinkler heads ensure adequate protection. The system is pre-charged with water and in fire conditions the sprinkler heads activate automatically as the air temperature reaches 68°C (152°F). Because no UK or European standards exist for the installation of sprinkler systems in high-bay shelving (as opposed to racking), we had to conduct full-scale burn tests at the Building Research Establishment. These were highly successful, demonstrating that the sprinkler system design could effectively contain a fire for 90 minutes. Other fire-engineering systems incorporated within the design include air-sampling early fire detection, smoke curtains, smoke extract and, around the external perimeter of the building, an access roadway for fire tenders. The high-bay walls are made of pre-cast, high-thermal mass insulated concrete panels carried on a steel frame and sealed with gaskets. The roof is of a heavily insulated, pitched construction with a single-membrane roof covering. The overall level of insulation and air-tightness is high, resulting in a highly stable internal environment with little or no variation in conditions from day to day. The environment is maintained with separate air-handling plant serving each of the four fire compartments. The concrete floor of the high-bay was cast ‘super-flat’ to ensure a smooth ride for order-pickers when travelling between the rack faces. It is cast with the minimal number of joints and these remained stable as the storage shelving was loaded with book stock. Aisles are well lit, at high level, with luminaires controlled by movement sensors to ensure that stock is held in darkness for as long as possible. One of the fire compartments also houses the ‘pit-stop’ a bay for parking order-pickers and for dropping off and picking up trolleys. Adjacent to this is a battery charging room where batteries removed from the order pickers can be placed on charge. This room is separated from the high-bay storage area by a four-hour fire shutter. Alongside this 9,561 m2 high-bay lies the single-storey ancillary block, providing 1,425 m2 of accommodation for operations, in the form of a reception area and security room, a sub-dividable meeting room, an office, common room, toilets, an IT hub room and a conservation quarantine room. The greater part of the ancillary

80 

 Toby Kirtley

Figure 6.7: Book Storage Facility plan

block is comprised of the ancillary processing hall, an open-plan concrete-floored room of some 708 m2. At the rear of the ancillary block is the loading bay with spaces for two 3.5-tonne box vans side by side, one with a 600 mm-high dock.

8 Specification The Storage Facility is 45 kilometres (28 miles) by road from the New Bodleian Library. – The building is compliant with BS 5454:2000 – Net area 10,986 m2 – Net storage area 9,561 m2 – Four separated four-hour rated fire compartments – sprinkler protection to BS EN 12845:2004

Chapter 6: The building of the Book Storage Facility  

–

– – – – – –

– – – – – – – – – – – –

 81

Fully air-conditioned, temperature and humidity controlled environment. The design temperature is 18 °C ± 1 °C; the design humidity is 50 % RH ±5 %. Air Filtration to Class F7 of BS EN 770. 31 aisles totalling 3,224 bays with 97,132 shelf levels Total area of shelf surfaces 120,666 m2 Height of the shelving is 11.4 m, aisle length is 71 m When full the BSF will hold 745,000 trays for book storage BSF capacity is 247,000 linear metres (153.5 miles) of book stock equating to about 8.4 million volumes One of the four compartments contains a fire-separated Map Store with one Ground Floor and four Mezzanine Levels built to house 600 Map Cabinets holding over 1.2 million maps and large-format items 1,600 linear metres microfilm Maximum tray weight 19 kg Maximum shelf load 242 kg Maximum bay load – 6,118 kg (excluding weight of shelves) Map cabinet load 680 kg Five high-level order-pickers (VNA trucks) for use in regular operations Mechanical handling equipment: Jungheinrich UK Shelving: SSI Schaefer Ltd. BSF Information System: Generation Fifth Applications Inc. Main Contractor: Mace Ltd. Logistics Consultant: Total Logistics Ltd Architect: Scott Brownrigg

Notes A linear metre (lm) contains 34 nominal volumes (based on ‘d’ sized books (9–12 inches in height) in the Bodleian’s Nicholson classification). All floor areas are of Net Internal Floor Area (NIFA) in square metres.

Boyd Rodger

Chapter 7: Book Storage Facility ingest and retrievals operations 1 Introduction When the planning approval was granted to build the Book Storage Facility (BSF) in Swindon, it was clear that the timeline for starting up the operations was going to be challenging. The timeline of key milestones was: – August 2009 – construction on the site begins; – September 2010 – the completed building is handed over to the University; – October 2010 – new staff are appointed, trained, and the building is furnished for operational work. The ingest of material also begins; – November 2010 – the retrieval operations and book delivery service begins; – 30 June 2011 – all material from the DeepStore salt mine is ingested; – 31 July 2011 – the material from the New Bodleian is ingested; – 30 December 2011 – the material from the Nuneham Depository and Iron Mountain stores is ingested. This gave us 2 years and 3 months from the start of construction to the completion of the ingest project. The purpose of the BSF is to store low-use material on behalf of the Bodleian Libraries and retrieve the material on request. To achieve this aim the ingest project needed to process an estimated 6.5 million volumes of material (books, journals, microfilms, etc.) and 1.2 million maps. We understood that the key dates for emptying existing storage locations were the drivers of the ingest projects, determining the sequence in which material needed to move. The dates of the other capital projects emerged as our performance targets. In addition to the target dates, the material had to be separated at the BSF into legal deposit and non legal deposit storage areas, for tax reasons. With the volume of material being ingested within the time targets, and the separation of legal and non legal deposit material, it was clear that the ingest process needed to be slick, efficient, safe, and simple to learn.

84 

 Boyd Rodger

2 Research 2.1 Visits to Library warehouses in the USA As part of researching what the project team could learn from other organizations, we visited library warehouses in the USA and industrial warehouses in Britain. These proved to be very useful in focusing our thinking around the design of the ingest process and the retrieval operations. During the visit to the ReCAP depository at Princeton in 2009, we asked the Executive Director what advice she would give to starting up operations at the BSF. Her advice was: I would bring in the management team ahead of your staff. It was chaos for us on Day One where no one knew more than anyone else. We learnt quickly but it would have been better if the managers started first. The other advice I will give you is barcode, barcode, barcode. Barcode everything!

The library warehouses we visited in the USA were those at Harvard, Princeton, and Yale universities, which operate the Very Narrow Aisle (VNA) system of high density storage. The books are stored according to size (providing the most efficient use of space) in 30-feet-high aisles. They are retrieved by the staff member driving a forklift truck which takes them up to the required shelf height. The book is retrieved from the tray in which it is stored and subsequently despatched to the nominated library. Visiting these US library warehouses made us aware that there are three stages of accessioning material (what we later referred to as the ‘ingest process’). The first stage involves sizing the books to fit predesigned storage trays. The shelving in their warehouses is set to provide just enough height for the trays to sit on and be retrieved. This is the essence of high-density storage where library material is stored according to height. The second stage involves accessioning the library material into the warehouse management system. (Figure 7.1) In all three US library warehouses this software was provided by Generation Fifth Applications (see chapter 10). The software records the barcodes of the items and matches them to the tray barcode. This created a draft data file into the system. The third stage repeats the second stage and acts as verification that the initial draft data file correctly matches items barcodes with their tray barcode. From this point onwards the trays and their contents are ready to be accessioned into the warehouse. An operator in the warehouse takes the trays on specially designed trolleys and attaches the trolleys to their forklift truck. The trucks are driven into the

Chapter 7: Book Storage Facility ingest and retrievals operations 

 85

Figure 7.1: Accessioning

appropriate aisle and elevated to the required height where there is an empty shelf. (Figure 7.2) At the shelf the operator places the trays in location and scans the barcode of the tray and the location barcode of the shelf, by means of a handheld terminal. The data in the hand-held terminal is subsequently downloaded on to the system and the warehouse record is now live, with the material being available to be requested. This three-stage model of accessioning was adopted as the Bodleian Libraries’ preferred method. However, the project team were concerned about the three stages of accessioning being conducted in separate locations as we saw in the USA. To replicate this would involve loading and unloading the books several times in the process. Multiple handling would take longer to achieve and present safety risks to the accessioning teams . The ingest process design team consisted of the Libraries estates project officers, the business applications development manager and an applications manager, together with a consultant from Total Logistics Ltd providing specialist advice. The team decided to design a unified ingest process where the three stages are merged into a single process bench. The design team encountered a problem at an early stage. When the length of the bench was added up, plus the provision of space of lifting equipment at either end of the bench, it measured about 10 metres (over 30 feet) in length. With seven benches being estimated as necessary to process the material we thought there would not be enough room in the ancillary hall for the seven benches. Then the design team experienced a ‘eureka moment’ when one team member presented a drawing with the process bench being U-shaped. This shape could

86 

 Boyd Rodger

Figure 7.2: A High-Level Order-Picker in action

include the three stages of accessioning as well as occupying less space in the ancillary hall. So, the U-shaped ingest process bench emerged from our research. (Figure 7.3)

2.2 Predicting reader demand Another significant area of research was around the pattern of reader demand for material. The aim was to predict peak, average, and low periods of retrieval activity to ensure we had enough staff and forklift trucks to meet the variable levels of demand and yet operate on a cost-efficient basis. Historical data was used from the Automated Stack Request system to analyse patterns of reader activity. Two key lessons were learnt: – Reader demand patterns vary significantly throughout the academic year. There were three periods of peak activity : Michaelmas Term (October to December), Hilary Term (January to March), and Trinity Term (April to June).

Chapter 7: Book Storage Facility ingest and retrievals operations 

 87

Figure 7.3: Final design of process bench

–

Reader demand has peaks throughout each day. It was clear that one-third of requests for the entire day were made prior to 10am. Depending on the type of delivery system used to convey the books from the BSF in Swindon to the Bodleian Libraries in Oxford, it would be possible to deliver a significant proportion of books the same day.

2.3 Working with a key supplier and a consultant To ensure our data was validated for prediction purposes and to confirm how many staff we would need, we worked with Jungheinrich, the supplier of the forklift trucks, and Total Logistics Ltd. Jungheinrich processed our demand pattern data through their software. The calculations factored in forklift truck travelling time from the pit-stop to the aisles to retrieve the books, as well as time to load and unload the truck batteries each day. We presumed it would take one minute to retrieve each book. This one-minute-

88 

 Boyd Rodger

per-book target was based on the experience in the USA and role-play exercises conducted by the design team. Jungheinrich confirmed that we could meet the peak requests with four forklift trucks dedicated to retrievals and refiles operations, and one truck dedicated to putting the new trays of material in the warehouse. For the purpose of the initial ingest of the material, Jungheinrich recommended that we have three trucks dedicated to retrieval operations and five solely used for ingesting the material.

2.4 Forecasting demand levels Total Logistics Ltd analysed the library data on the basis that the existing material stored in closed-access stores would be moved to the BSF. They predicted annual request rates of 230,000, which was within the initial target figure of 250,000, and recommended the number of staff to meet this requirement. However, as described in chapter 1, there was a review of the collection management strategy. The review recommended that high-use books be retained in central Oxford in open-access facilities. The BSF would store only lower-use material. There was also a significant amount of effort to make more material available digitally to readers. When we included the impact of the Academic Strategy on the retrieval rates, a new prediction emerged of 176,000 requests per annum compared to the 230,000 previously assumed. This suggested a staffing model for permanent operations consisting of the following: – Bookstorage Logistics Manager x1 – Operations managers x2 – Circulations Desk Co-ordinators x 2 – Warehouse Assistants x 11 – Maps Assistant x 1 – Delivery van drivers x 2 – Administrator x 1

3 Preparing for ingest 3.1 Staffing model In the spring of 2010 we considered various staffing models for the ingest project. The options included staffing by way of temporary contracts, and by recruitment agency staff. We decided to provide staffing through a recruitment agency for

Chapter 7: Book Storage Facility ingest and retrievals operations 

 89

reasons of administrative efficiency, flexibility, and cost control. In the summer of 2010 Blue Arrow recruitment agency was contracted through a process of competitive tendering. Blue Arrow analysed the requirements of the ingest project. We needed bench processors, tray makers, movers, van-dock loaders, forklift truck drivers, and delivery van drivers. They all had to pass security clearance and demonstrate relevant skills for the roles. Blue Arrow processed over 600 applications and provided 40 staff in October 2010.

3.2 Training staff and the laying out of the ingest process The BSF building was handed over in early September 2010. The managers were recruited at this time ahead of the project team (as recommended by ReCAP at Princeton). We worked with our estates office to organize the layout of the ingest bench furniture. Project staff were inducted over three separate start periods in October 2010. This was to provide quality training from Generation Fifth Associates on how to use the warehouse management system (which we called BSFIS – Book Storage Facility Information System). We used real library books for them to practise on. The induction programme also had input from the Conservation and Collection Care department about collection care, Jungheinrich on driving the forklift trucks, and manual handling training provided by Blue Arrow. On 26 October 2010 the first books were officially ingested into the BSF. It was quite an achievement by everyone who contributed to the training to complete it on time. Fifteen days later, on 8 November 2010, the BSF went live operationally with the first books being requested. Fourteen books were requested on that first day.

4 Ingesting the material 4.1 What was achieved? Fourteen months after the first book had been ingested, the mass ingest was completed on 6 January 2012. A total of 7,049,741 items had been ingested into the warehouse management system, as well as 1.2 million maps stored separately. (Figure 7.4) All of the key milestones were met: the June 2011 deadline for the vacation of the DeepStore salt mine; the July 2011 deadline for emptying the New

90 

 Boyd Rodger

Figure 7.4: A filled chamber

Bodleian, to enable the refurbishment project to commence, and the December 2011 deadline to vacate the Nuneham Courtenay depository. This represented an average of around 25,000 items and 4,000 maps every working day. Less shelf space was used than expected, with occupancy only 72 % of what was predicted.

4.2 How books were ingested The ingest project processed the books as described above, with the three stages of tray sizing, initial data accessioning, and verification. However, the experience of the ingest project turned out to be different from what was originally intended. The aim was for the material to be shipped to the BSF via plastic totes stacked on wheeled trolleys (known as skates) . The initial material was sent to the BSF using this method, with the plastic totes and their skates being returned on the next delivery vehicle. However, we were quickly running out of totes and their skates, as so many were being used during the project. To supplement this, cardboard boxes were also used. The ingest process was adapted to suit this method, with the boxes sent on wooden pallets.

Chapter 7: Book Storage Facility ingest and retrievals operations 

 91

The problem with the cardboard boxes was that they had to be disposed of rather than returned. This caused a large amount of cardboard to accumulate in the Ancillary Hall and became known by the team as ‘cardboard mountain’. A telephone call was made to the local council who offered an immediate solution of free collection of the cardboard with the aim of recycling it. This solution proved ideal since it solved an immediate problem, it was a green alternative to destroying the cardboard, and the council paid the Libraries for the recycling. Another challenge the ingest team experienced was a problem that became known as ‘NOF’s’ (meaning Not on File). This was where the data transfer between the Libraries’ catalogue and the BSFIS had not been completed, so that at the point of ingest the books were not recognized by the system and could not be processed. The Inventory Control Project rapidly responded to this problem and trained members of the ingest project team to do some low-level problem solving, while taking back the serious problem material requiring a more skilled approach.

4.3 Capacity planning Knowing what material was being delivered to the BSF each day was important for the team to be prepared. We needed to know the type of material (i.e. size, whether legal deposit or non legal deposit, the collection code for special collections material, and the type of container). With the Bookmoves team following so closely behind the barcoders this was not always possible. Capacity planning was essential for all three core projects (barcoding, bookmoves, and ingest). After we completed the decant from the DeepStore salt mine in Cheshire in June 2011 and the New Bodleian the following month, we took stock to learn some lessons. For the remaining stage of the decant programme from the Nuneham Courtenay depository, we decided to do some scenario planning. This involved imagining different rates of barcoding/bookmoves/ingest from the target of 600 linear metres per day. Some scenarios were lower than 600 m and one was higher. We considered the logistical implications of each scenario and how we would recognize if that particular scenario was the reality. The scenario planning exercise provided the key project managers with an objective set of criteria to assess how the book decant was progressing in reality. We were also able to respond rapidly to unexpected changes of ingest rates, since we had discussed in advance what needed to happen.

92 

 Boyd Rodger

4.4 Feeding the process The ingest process was like a hungry machine that needed feeding each day. As well as the raw material of books, the ingest project required the regular supply of flat-pack trays and enough staff to keep the project going. With a ten-day lead time from ordering trays to their being delivered, this required some predictive planning based on monitoring tray usage. For example, when microfilms arrived at the BSF they were easy to ingest quickly at a rate of 1 km per day, so sufficient stock of trays had to be kept in storage for the ingest project to process whatever was sent. When material was simultaneously being delivered from the New Bodleian and the DeepStore mine, the ingest project recruited another 25 staff to work an evening shift from February to June 2011. This is another example of the flexibility required in this particular project.

5 The retrieval operation 5.1 Five retrieval periods In the early stages of the ingest the retrieval operation was only twice per day. This reflected the relatively low demand level for the small quantity of material in the warehouse. There were also two van trips to Oxford each day at 9 am and 1.45 pm. The retrieval operation worked from 9 am to 5 pm each day. When significantly more material had been accessioned into the warehouse, the full operations team was set up in August 2011. This consisted of three teams covering 7 am to 3 pm, 10 am to 6 pm, and 2 pm to 10 pm. This arrangement provided an operating window from 7 am to 10 pm Monday to Friday. The new set-up also gave us the opportunity to increase the number of retrieval operations to five each day. The start times for the new retrieval operations periods were: – 07:00 – 10:30 – 14:00 – 17:00 – 19:30 The delivery van departures are: – 07:00 – delivering the 14:00, 17:00, and 19:30 retrievals; – 12:30 – delivering the 07:00 and the 10:30 retrievals.

Chapter 7: Book Storage Facility ingest and retrievals operations 

 93

Refiling of returned books takes place when the retrievals are completed. Since the BSF went live on 8 November 2010 to May 2012, a total of 237,010 books were retrieved on request. A good service to readers was therefore maintained during the ingest project.

6 Delivering the books The ingest project and retrievals operations took place within the Book Storage Facility. There remained the issue of delivering the books to Oxford and circulating them around the libraries. Three options were considered: – Direct delivery – in which the BSF van called at all the libraries; – Hybrid delivery – in which the BSF van called at the central Bodleian Library and a hub based at Osney; – Indirect delivery to a hub only – in which the BSF van called at the hub only and two other smaller vehicles took the books to the libraries. The third option was the preferred method selected by the Libraries. Since delivering to a hub was the chosen method of delivery, a location was identified which would receive three vehicles at the same time. Space was found at Osney where many library staff are based for the hub. The Libraries’ Support Services organized the setting up of the hub and also supplied drivers for the Oxford-based delivery vans. We began to to refer to this location as the ‘hubette’ since no staff worked there. The delivery vans would be present simultaneously and the drivers would form a team to assist one another in loading and unloading their vehicles. A raised platform was also provided for the vans to load on to. Integrating the BSF van times with the two Oxford delivery vans schedules enabled a reasonably quick delivery service for the readers. If a reader makes a request for an item at the BSF before 10:30 am they will receive the item at a library of their choice by 4 pm on the same day. For items requested after 10:30 am the delivery is next day by midday at the latest. Setting up the integrated book delivery system was challenging and many obstacles were creatively resolved. However, neither the library catalogue nor BSFIS can record when an item leaves the BSF and compare it to when it is received at the destination library. The solution we have found is to select books at random and investigate the time at which they are scanned into their libraries. Some important lessons have been learnt from this investigation. For example, we found that whilst we were retrieving 100 % of requests on time,

94 

 Boyd Rodger

only 76 % were being delivered to a particular library within the promised time. The libraries were processing all the incoming totes when they arrived, but some of the delivery personnel had developed the habit of stacking the totes in the hubette, resulting in a day’s delay before their delivery at the destination library. This was raised with the libraries and the delivery team who implemented a different procedure. Random sampling continues, since it allows us to learn valuable information about how the delivery system actually performs.

7 Customer enquiries Whenever a new service is set up the obvious question asked by library colleagues is, “Who do I contact if I have a question?” In May 2011 ‘BSF Enquiries’ was set up as a contact email address. All enquiry emails were automatically circulated to the BSF management team. A spreadsheet was also created to track that enquiries had been addressed promptly and to provide data on the volume and pattern of enquiries. Since 2011, 691 enquires have been directed the BSF. The percentage of books delivered that generate complaints is 0.1 %. We are using this as a base measure against which to demonstrate improvement. This start up project at the BSF would not have been possible without the support and the creative contribution of all a variety project colleagues and specialist contributors. Special acknowledgement should be made of the 65 members of the ingest team who achieved a remarkable accomplishment.

Catríona Cannon

Chapter 8: ‘And moveth all together, if it move at all’¹: the Bookmoves Project 2010–2011 1 Introduction Between January 2010 and December 2011, the Bodleian Libraries moved over 209 kilometres of books, manuscripts, ephemera, journals, music scores, microfilms and microfiches, as well as over 1.2 million maps. This was achieved without closing any library reading rooms. Why did we do it? On account of their age, their entitlement to legal deposit copies of print publications and the generosity of their donors, the Bodleian Libraries have rich physical collections. Readers are used to being able to get access to electronic resources instantly and we wanted the physical collections to be laid out so that the majority of readers could get access to those too as quickly as possible. The introduction to this book describes the Academic Strategy, one of the principles of which was to move books which usage statistics showed were popular from closed stores – from where they had to be requested by readers – into reading rooms. If, on the other hand, they were requested relatively infrequently, they would go to the Book Storage Facility.² This strategy was the first underlying reason for the most substantial moves. The second was that the New Bodleian needed to be emptied so it could be refurbished as a dedicated library for special collections. There were many other reasons for the smaller moves, and a few of these will be described below. How did we do it? By a mixture of skills and experience – professional book movers and logistics experts; through detailed advance planning, including the use of trials for different types of material and location; by monitoring progress

1 William Wordsworth, ‘Resolution and independence’, in The complete poetical works (London, Macmillan, 1888), 9. 2 ‘Academic Strategy for Oxford University Library Services’, Oxford University Gazette (26 June 2009), accessed 29 June 2012, http://www.ox.ac.uk/gazette/2008-9/weekly/260609/ notc.htm#5Ref.

96 

 Catríona Cannon

Figure 8.1: Oxford libraries and bookstores

with detailed throughput statistics; by creative use of specialist equipment and with careful attention to reader and staff safety. Books were moved to and from a wide variety of locations in Oxford and outside, the biggest being the Old Bodleian Library, the New Bodleian, the Underground Bookstore (now the Gladstone Link), the Bodleian Law Library, the Radcliffe Science Library, the Oriental Institute Library, a warehouse in Osney, the Book Storage Facility at Swindon, the University store at Nuneham Courtenay, and two commercial companies, DeepStore and Iron Mountain (Figure 8.1).

Chapter 8: ‘And moveth all together, if it move at all’: the Bookmoves Project 2010–2011  

 97

2 Sequence of moves The sequence of moves was complex, often involving moving items back and forth to the same locations at different times and in different combinations, but they fell largely into three parts. These parts often overlapped in time. They were as in Table 8.1.

Building emptied

Linear kilometres moved

Timetable

Underground Bookstore

23

August 2010–July 2011

New Bodleian

88

March 2009–July 2011

Remote stores

35

January–December 2011

Other

63

March 2009–December 2011

Table 8.1

2.1 Emptying and refilling the Underground Bookstore, later the Gladstone Link The project to turn the Underground Bookstore, beneath the Radcliffe Camera in central Oxford, into an area open to readers for browsing and working, is described in chapter 14. The role of the bookmoves team was to remove all the materials held in the Bookstore, transport them to their agreed destinations, and then refill the space with selected books and journals once it had been transformed into a reading area. Many of the items held in the Bookstore counted as low use, and were therefore destined for the Book Storage Facility in Swindon once it was complete. Others were designated high use and would therefore be returned to the space when it became a reading area known as the Gladstone Link. The Official Papers Reading Room and staff area were also located under the Radcliffe Camera, and these were all destined to move to the Bodleian Law Library. Before this could happen, the Law Library basement had to be emptied: for many years it had been a general overspill area for the New Bodleian bookstack, and did not contain any Law material. The bookmoves team started by moving this material to a warehouse and then on to offsite storage. That freed up space for the Official Papers to move once the space had been refurbished, and that in turn contributed to the emptying of the Underground Bookstore. Once the project to refurbish the Underground Bookstore had been completed, and it had been renamed the Gladstone Link, the bookmoves team moved

98 

 Catríona Cannon

in all the high-use material that had been identified, and the area opened to readers. Work on this started in August 2010 and was completed in July 2011. The timetable was dictated by the building work to convert the Underground Bookstore and the speed at which the barcoding (Inventory Control Project) team could work.

2.2 Emptying the New Bodleian Library The project to empty the New Bodleian started in March 2009 and finished in July 2011. The redevelopment project is described in chapter 13. The role of the bookmoves team in this was to move all the library materials in the building to their agreed new home, whether this was the new temporary special collections store in the Radcliffe Science Library, another reading room or, ultimately, the Book Storage Facility. This was not as simple as it sounds, as the New Bodleian was an immense building with many caches of material which successive generations of library staff had deposited, sometimes without cataloguing or accessioning it. Working with the barcoding staff, the bookmoves team often had to find out who to talk to and to agree where the material should go. The move of the special collections store and reading room from the New Bodleian to an area in the Radcliffe Science Library is described in chapter 4. The maps and music reading room material and a small collection of other material went to Duke Humfrey’s Library, within the Old Bodleian Library. High-use items from the Indian Institute Library, which had been housed in the New Bodleian, went to the Oriental Institute Library. High-use items in other Humanities and Social Sciences subjects went back to the Gladstone Link (see above). The remainder of the collections held in the New Bodleian went to the Book Storage Facility. This was a long drawn out part of the project, necessarily so because of the size of the New Bodleian and the complexity of the collections held in it. The building has 11 floors, labelled A–L but excluding ‘I’. Floors A–H inclusive are above ground and the rest below ground. There were four lifts but not all of them served all floors. The bookmovers had dedicated use of one and occasionally used others. The lifts were the ones originally installed in the building in the late 1930s and were worn out. They broke down frequently though were very quickly repaired. The Bodleian Libraries employ an engineer who has substantial experience of the lifts and was usually able to fix them. Otherwise, as the lifts were so old, if a component part broke the replacement had to be tailor-made. The bookmoves team and the Inventory Control Project mapped everything held in the New Bodleian and agreed the order in which categories of materials

Chapter 8: ‘And moveth all together, if it move at all’: the Bookmoves Project 2010–2011  

 99

would be barcoded, then signed off for moving. Materials were delivered to the old bindery, at entrance level (see ‘People’ section below) and sent to the BSF or Gladstone Link. The deadline for emptying the building was absolute because the building contractors were scheduled to come in on 1 August 2011. Pressure during the months leading up to that date was intense. A difficulty was that some of the librarians in charge of collections were keen to work on them up to the last minute and were nervous about whether the bookmoves team would handle them carefully enough. The urgency of the situation was communicated, reassurance was given as to the safety of the materials, and the moves went ahead on time.

2.3 Emptying remote University and commercial stores The University of Oxford owns an extensive property to the south-east of Oxford, around a village called Nuneham Courtenay. This includes the Harcourt Arboretum, part of the University Botanic Gardens. On the site is a library store which was built in modules between 1975 and 2002 to house the expanding collections. We had also stored material with two commercial companies, DeepStore, in Cheshire, and Iron Mountain, mainly in Gloucestershire. Our aim with all of these stores was to move the material to the Book Storage Facility. The timetable for these moves was January to December 2011. The commercial storage companies delivered the materials they held to a warehouse in Osney, west Oxford. This warehouse is close to the offices where many of the Bodleian Libraries staff are based. During autumn 2010 it was converted into an area where material could be barcoded on an industrial scale (see chapter 9) before being moved to the Book Storage Facility. This left the bookmoves team with responsibility for moving stock from Nuneham Courtenay to Osney and from Osney to the Book Storage Facility. The first category, Nuneham to Osney, was moved by the bookmoves team themselves. A dirt track led to the store at Nuneham so the team put down a temporary roadway. They also erected scaffolding in the delivery area so that delivery vans could be more easily filled and emptied. As Nuneham is a small, quiet village and the store is surrounded by farmland, we had been concerned about causing disruption to its inhabitants, but in practice we had no complaints. Subject librarians came to visit and discuss with bookmoves staff and barcoders what needed to be done, and to alert them to any particular care that needed to be taken. This dialogue had been established at a meeting to brief subject librarians on the emptying of Nuneham, and as a result the different groups of staff involved co-operated well together.

Figure 8.2: The progress plan for the repository at Nuneham Courtenay

100   Catríona Cannon

Chapter 8: ‘And moveth all together, if it move at all’: the Bookmoves Project 2010–2011  

 101

Bookmoves staff also worked with the staff based at Nuneham, who retrieved items requested by readers. As the store was emptied fewer and fewer items were retrieved, and eventually the retrieval service stopped. The illustration (Figure 8.2) shows a plan of the repository at Nuneham, with notes and codes showing. As the storage modules were emptied, they were crossed them with black lines; it was very satisfying to see all the modules crossed out. The second category, Osney to the Book Storage Facility, was moved by a commercial removals company, Jamie Briggs Removals & Storage. Stock was put into crates by barcoders and emptied at the Book Storage Facility by the staff there. During July 2010, we had carried out a competitive tender process to find a commercial moving company who would deliver from Oxford to the Book Storage Facility. Out of eight companies who bid, Jamie Briggs was selected. We believed that it would be more efficient to contract out the transport for several reasons. Firstly, to do it ourselves we would have had to become a licensed lorry operator: this would entail specialist training, insurance and certification. Secondly and very specifically, there is a height limit on the New Bodleian roadway, so larger trucks needed to be adjusted: unlike us, a specialist would have the contacts and suppliers to do this. Finally, additional drivers were required to increase the number of runs to Swindon from three to four daily, which would have involved recruitment and management of more staff. Transport to the Book Storage Facility was an identifiable piece of work that could be reasonably contracted out without compromising on quality, so that our own staff could focus on the rest of the work. Overall, though time was spent on chasing up problems and making it work, our assessment at the end was that it was still more efficient and less time-consuming than if we had done it ourselves. In a library system such as the Bodleian Libraries there are continual requirements for books and other items to be moved. Librarians routinely relegate items that are less used to free up space for new books, and as we always keep one copy of every item, these are sent to storage if no longer required in a reading room. Also, smaller libraries are brought into bigger libraries because the space is needed for other purposes, or because it makes sense to bring collections together. The team was able to do a few of these smaller moves where they were given priority, in particular a move of stock from the former Plant Sciences and Zoology libraries to the Radcliffe Science Library and to Osney for sorting and barcoding later. The Taylor Modern Languages Faculty Library was refurbishing its reading room and building a new entrance, and relegated some stock to Nuneham Courtenay.

102 

 Catríona Cannon

3 People The bookmoves team had existed for some time, and was part of the Conservation and Collection Care department of the Libraries. It was composed of three permanent staff and several staff on fixed-term contracts, whose number varied according to the moves being undertaken at the time. Between them the team had many years of experience in moving books, particularly rare or delicate ones. They were part of the move of the special collections to the Radcliffe Science Library, described in chapter 4. When the scale of the moves which would be needed to support the emptying of the New Bodleian and the filling of the Book Storage Facility became apparent, we decided to create a temporary senior position of Project Manager for Library Logistics. We recruited a non-librarian who had experience of logistics and supply-chain management, and he in turn recruited more staff, bringing the total to three permanent staff and four temporary. Many more bookmovers were needed, and so an employment agency supplied staff as needed, with typically between fifteen and thirty working daily on the project. This gave speed of recruitment and the flexibility to increase or reduce the number of movers working, according to how much material was available to move. The Inventory Control Project (see chapter 9) had already gone through a process of choosing an agency so the Bookmoves Project was able to take advantage of this. The seven staff employed by the Libraries acted as supervisors for the teams of agency staff. The combination of people who knew the collections, the University and the library system in detail and those who were experienced in the logistics of largescale moves worked well. There was sometimes a culture clash as the University is a very different organization from a fast-moving consumer goods company, or a car factory. However, we were aiming to move more books in a shorter time than had ever been done before, and this did mean our culture needed to change. Having an external logistics specialist who could identify where time could be saved or aspects outsourced without compromising on the quality of the work we did helped us to adapt. The accompanying photographs (Figures 8.3–8.5) show members of the book moves team at work in the New Bodleian. As the project progressed and agency staff were hired, the team had two different homes. They were first based in the New Bodleian, taking over the old bindery area when the Conservation and Collection Care team moved out (see chapter 12, staff relocation). After July 2011 they moved to the warehouse at Osney, storing all their specialist equipment there. One of the staffing challenges of the project was that the supervisor of the bookmoves team, a longstanding member of staff with extensive experience of

Chapter 8: ‘And moveth all together, if it move at all’: the Bookmoves Project 2010–2011  

 103

Figure 8.3: Taking books from the shelves (picture by Nick Cistone)

Figure 8.4: Sealing the crates before transport (picture by Nick Cistone)

Figure 8.5: Loading onto the van (picture by Nick Cistone)

Figure 8.6: Bookmoves progress spreadsheet: New Bodleian

104   Catríona Cannon

Chapter 8: ‘And moveth all together, if it move at all’: the Bookmoves Project 2010–2011  

 105

the Bodleian Libraries’ collections, was unavoidably absent from July 2010 to February 2011. The Project Manager for Library Logistics worked with other members of the team to fill the gap until his return.

4 Planning and monitoring progress Throughout the project, the Project Manager planned in detail between three to four months in advance of the work being carried out. He took some time after his arrival in January 2010 to understand and plan the project as a whole before starting any of the big moves, allowing smaller moves to continue. He organized trials of each different type of material as part of the planning process, adjusting the amount of time he expected that part of the project to take according to the throughput rate achieved in the trial. For example, moving broadsheet newspapers which had been bound whole would be more time-consuming than moving microfilms in boxes. He also worked with the team to push the daily throughput rate up at this early stage, setting a standard for the project as a whole. Knowing how many items had been moved and comparing them with what we had estimated was crucial in meeting deadlines. In practice, these were measured in linear metres as the most predictable unit of work to be done. We had quite detailed statistics on what needed to be moved, and in the event these proved relatively accurate – an average variance of 12 % against estimate – given the length of time that most of the stock had been stored and that much of it was unrecorded. All members of the team recorded their throughput statistics and the Project Manager collated them daily. He then used this for planning and, most importantly, as the basis for discussion with the Inventory Control Project Manager and Book Storage Logistics Manager. They were in charge of the barcoding project (chapter 9) and the receipting of stock into the Book Storage Facility (chapter 7). All three projects depended on each other to deliver the goal: the barcoders needed to work fast enough to provide the bookmovers with stock, the Book Storage Facility staff needed the bookmovers to deliver stock for receipting, and so on. It was crucial that the project managers should have a common understanding of what had been done and what remained to be done so that we could guarantee that the project as a whole would finish on time.

106 

 Catríona Cannon

5 Project organization and communication with other interconnected projects There was a high level of interdependence between the book moves project and others. As described in chapter 2, the Bod Squad monitored progress and discussed areas of interdependence where barriers had arisen, at first twice weekly, then weekly, then fortnightly. Up to July 2011 this was supplemented with a weekly ‘BSF team’ meeting bringing together all those involved in delivering the Book Storage Facility or emptying the New Bodleian. Between August and December 2011 a smaller group met fortnightly, reduced then to those involved in emptying the Nuneham and Iron Mountain stores. At times the three project managers (bookmoves, barcoding and receipting into the storage facility) met daily, or their deputies did, in order to keep each other informed. This aspect of the project needed constant attention: it sometimes seemed that no matter how many meetings took place, people had differing understandings of what was going on or missed a piece of information which would have made things much easier. This may be a normal feature of such high-pressure projects. The times when things worked best were when problems were discussed at the level at which they arose rather than being escalated to senior management to arbitrate. Senior management were then only involved when a policy decision or more resources were needed.

6 Communication to library staff and readers As stated in the introduction, we did not close any reading rooms. However, there were times when books or other items were being moved so were not available for use. The Just-In-Time team (chapter 15) provided alternatives, but we also needed to warn readers in advance so they could plan their visits. The project had web pages within the Bodleian Libraries’ website, sitting side by side with the other projects, with a timetable, frequently asked questions and advice on finding alternatives. Books in transit showed a location of ‘book move’ on the library catalogue. We also developed a book-move application to put on the web pages but it proved too much work to keep it up to date and in fact readers were well informed through other means.

Chapter 8: ‘And moveth all together, if it move at all’: the Bookmoves Project 2010–2011  

 107

7 Equipment and health and safety It was important to keep staff and readers safe and healthy throughout the project. This meant making sure that entrances and exits to reading rooms and stores were not blocked; checking that buildings were not so full of staff that they could not be emptied quickly enough in an emergency; ensuring that lifts were in working order so that staff did not have to carry heavy items up or down stairs; using special bespoke trolleys for moving large items like maps; and, in the case of the Underground Bookstore, installing a special hoist for moving stock in and out. This was a challenge, as the buildings being emptied or filled were often very old: for example, Duke Humfrey’s Library is medieval and the Radcliffe Camera dates back to the eighteenth century. They were not intended for large numbers of people and have been adapted over the years to accommodate their current uses. The Bodleian Libraries receive legal deposit copies of broadsheet newspapers and these were covered with hard binding in the past, making them very heavy. The Project Manager visited the British Library to find out how they coped and experimented with a manual platform lift. In the end, however, two book movers together moved large items and put them into a cage.

8 Money The financial estimates for the book moves project were carried out before the arrival of the Project Manager for Library Logistics. After he arrived he revised the estimates as he was concerned that not enough had been allocated. When the University Planning and Resource Allocation Committee agreed that the Libraries could move money awarded for the projects from one to another depending on over- or underspends (see chapter 3) the Bookmoves Project was allocated extra money. However, in the event the project came in nearer the original estimate and much of the extra money was released for spending elsewhere.

9 Conclusion The project was a success: the books were moved on time and on budget. The combination of staff who knew the collections and the University well with external logistics experts was effective. If we were to do it again we would focus on the interdependencies between the bookmoves, the Inventory Control Project and the

108 

 Catríona Cannon

project to fill the Book Storage Facility at an earlier stage in the projects to agree how they would communicate and manage the overlaps. Now that the project is over, the biggest challenge for us is planning and resourcing the routine book moves that have built up in the meantime, as well as the bigger moves arising from building projects (not as big as the ones described in this chapter but still substantial). To address this, a standing committee, the Collections Deployment Group, will oversee the movement of collections and agree priorities for moves.

Michael Williams, Alasdair MacDonald and Cath Walsh

Chapter 9: Barcoding – the Inventory Control Project 1 Introduction The Bodleian Libraries’ main collections had never been barcoded, so during a period of 22 months the Inventory Control Project transformed the situation by barcoding and updating the catalogue for over seven million books. This chapter describes what was known about the collections and the various methods that were employed to barcode them. It also considers the staffing models and how it overcame the many challenges faced in such a time critical project.

2 Scale of the project During July and August 2009, a pilot project was undertaken to trial a workflow for barcoding modern monographs. The intention was that this would be scaled up to a project costing £ 1.5 million, barcoding as much material as possible within the available budget. This was intended to complement the barcoding of all newly accessioned material, which would be phased in from autumn 2009 onwards, but at this stage there was no intention of barcoding all material intended for storage in the Book Storage Facility (BSF) (see chapter 6). Following the adoption of the Generation Fifth Applications software as the BSF Information System (BSFIS) (see chapter 10) in spring 2009 and fact-finding missions to the Harvard and Research Collections and Preservation Consortium (ReCAP) depositories in October of that year, we concluded that that no hybrid system of barcode and shelfmark retrieval would be possible. As a result, with the exception of maps, every item ingested into the BSF would need a barcode associated with an entry in the Libraries’ catalogue. This decision meant that the project would have to be scaled up beyond the initial scope and budget. A report in April 2009, shared with us by the Director of the Bibliographic Control & Processing Division at Columbia University, showed that the cost of undertaking similar work at Columbia had ranged from $0.30 to $3.00 per item, depending on the condition and complexity of the material. The final cost of the Inventory Control Project was £ 3.6 million to barcode just over seven million items, giving an average cost of £ 0.51 per item across all collections.

110 

 Michael Williams, Alasdair MacDonald and Cath Walsh

3 What we knew 3.1 Introduction The Bodleian Libraries have acquired books through a variety of means throughout their lifetime. These include legal deposit, purchase, donation and exchange. Although the focus of the Inventory Control Project was on modern printed books, we knew that a certain amount of material from the Special Collections department would also be sent for temporary storage at the BSF during the renovation of the New Bodleian Library. No complete inventory of the entire collection existed. Despite eleven million records appearing on the online catalogue (which also included other libraries in Oxford) we knew that at best this only provided a collection-level description for many parts of our holdings. The principles of operation of the BSF prescribed that every item must be barcoded. Since an item inventory did not exist we could not be sure of the exact magnitude of the barcoding task. However, we had detailed measurements of the linear shelf space occupied by collections destined for the BSF. There were 192 kilometres, which enabled us to estimate that 6.5 million items would be barcoded.

3.2 Legal deposit Her Majesty’s Revenue and Customs (HMRC) were willing to give tax relief on construction costs for those parts of the BSF that were to contain legal deposit material. Early barcoding trials in the summer of 2009 showed that identifying legal deposit material by looking at the Bodleian stamp took a few extra seconds per item. The additional cost in staff time was outweighed by the savings presented by VAT relief. A decision was taken to identify every legal deposit item at the point of barcoding and to store it in separate chambers at the BSF, thus physically separating the existing mix into two distinct groups. The status of each item was indicated by the use of a barcode number starting with a seven for legal deposit material, or a six for non-legal deposit material. Considerable institutional knowledge of the Bodleian’s collections from longstanding members of staff was used to estimate the proportion of legal deposit material as part of the total. No definitive evidence existed of the legal deposit status of items received before the 1920s, although it was possible to demonstrate that some complete collections were received through this privilege before that date. We concluded – correctly, as it turned out – that 44 % of the material could be identified as legal deposit.

Chapter 9: Barcoding – the Inventory Control Project  

 111

3.3 Main Classification Schemes used on the materials barcoded Most monographs received before 1988 went into one of several in-house classification schemes, applied to the physical items as hand-written spine labels and written in pencil inside the front cover. From 1883, the scheme used was the Nicholson classification, devised by Edward Nicholson, Bodley’s Librarian from 1882 to 1912. This classification consisted of a three-digit number representing a broad subject area, which was further subdivided by up to four additional digits; a lower-case letter designating size, followed by a point and a further number representing number in sequence. Example: 170 d.15 – fifteenth ‘d’ sized book received for subject Art, General History and Collected Biography. 17005 e.4 – fourth ‘e’ sized book received for subject Art, Dutch. In addition, enumeration or sequence order within a bound volume may have been added at the end of the shelfmark, in parentheses or following a slash; this was not consistently applied and individual volumes often received individual shelfmarks. Nicholson’s classification system also contained smaller subject-specific sequences, such as the Gough Additions (Gough Adds) sequence for the topography of the British Isles. From 1988 onwards, each physical item received by the central Bodleian Library and the Radcliffe Science Library (RSL) was assigned a unique shelfmark made up of elements to indicate academic or non-academic material, year of receipt, size and number in sequence of receipt. Examples: M09.C00001 – M for academic monograph; 09 for year of receipt (2009); C for sizing; and 00001 for number in order of receipt. This would be the first ‘C’ sized academic monograph recorded as received in 2009 by the Bodleian. X89.G00034 – X for non-academic monograph; 89 for year of receipt (1989); G for sizing; and 00034 for number in order of receipt. This would be the 34th ‘G’ sized non-academic monograph recorded as received in 1989. (‘X’ material was invariably legal deposit material.)

112 

 Michael Williams, Alasdair MacDonald and Cath Walsh

This shelfmark was placed on the inside cover of the book as a machine-readable sticker containing a barcode of the shelfmark sequence and on the spine in eye-legible format. This allowed the shelfmark to act as a unique identifier linking the item to its entry on the catalogue, including uniquely identifying individual parts of multivolume sets. Special collections sequences included Old Class (three-digit classification, letter and sequence number, similar to Nicholson classification); Year Books (two-digit year of receipt, point and number in sequence); and other systems including elements of provenance, subject and sequence number. All were handwritten on the items. Most periodical issues received from 1883 to 1988 were assigned the appropriate Nicholson subject classification with the prefix ‘Per.’. For some categories of material an alternative prefix was used, such as ‘Soc.’ for the publications of learned societies, ‘N.’ for newspapers or ‘Dir.’ for directories. There was a single numerical sequence of these and the monographs in the same classification and size, though these were shelved in separate sequences except in the case of ‘Soc.’ which was interfiled with ‘Per.’ to create a single running sequence. Example: Per. 1232 e.28 – the twenty-eighth ‘e’ size item for the subject Ecclesiastical History, Scottish, being a periodical. With the exception of some Nicholson sequences such as ‘N.’, periodical issues received after 1988 were given a shelfmark consisting of the letter ‘P’, a letter to designate size and a number in sequence. No attempt was made to separate academic and non-academic material into separate sequences. Title changes may have been represented under a single shelfmark and one title may have had several distinct shelfmark sequences as sizes changed, growth space in the stacks was filled or back issues were relegated from open shelves. Titles shelved in the ‘P’ sequence may also have had back runs at other classification-based shelfmarks. Example: P.D00002 – the second ‘D’ sized periodical title.

Chapter 9: Barcoding – the Inventory Control Project  

 113

3.4 Dispersal One objective of the construction of the BSF was to rationalize the collections which were dispersed across multiple locations (Table 9.1). Historically, rapid growth in the collections had meant that sequences of material were moved to other sites to provide space for new intake. All of these dispersed collections were to be reunited at the BSF. This presented its own challenges for barcoding since the operations could not happen in a single location.

Location

Number of books

Comments

New Bodleian Bookstack

2,400,000

Main stack

Underground Bookstore

700,000

Under the Radcliffe Camera

St Cross Building

140,000

Nuneham Courtenay repository

1,280,000

SERS Warehouse, Osney

120,000

DeepStore

1,780,000

Salt mine in Cheshire

Iron Mountain

90,000

Kemble, Gloucestershire

Total

6,510,000

8.7 km (5.5 miles) from the centre of Oxford

Table 9.1: Locations and collections as estimated summer 2009

4 Operations All of the collections within the scope of the Inventory Control Project were in closed-access areas. No barcoding work was done in library reading rooms; however, as noted above, the dispersed nature of the collections meant that a number of different barcoding operations were simultaneously under way in a variety of locations. The Inventory Control Project was to make a head start on its work; the BSF was still under construction at the start of the project, and there would be no major book moves until the BSF was fully operational. This dictated that material held in the New Bodleian stacks in Oxford would be barcoded first. A separate project to refurbish the Underground Bookstore in Radcliffe Square required the books to be removed from there as soon as possible, so bar-

114 

 Michael Williams, Alasdair MacDonald and Cath Walsh

coding was prioritized to take place in this space from the beginning of operations on 8 March 2010 (see chapter 14). As this work was completed, barcoders migrated to the main bookstack in the New Bodleian and to the satellite stack in the Law Library basement at the St Cross Building about 0.8 kilometres (0.5 miles) away. All of these early collections comprised monographs and were an excellent training ground for the teams of new staff. At this time we were still designing, testing and reviewing the periodicals barcoding processes. Teams of senior barcoders were established in vacated reading rooms and staff offices to support the barcoders’ work by working directly with the online catalogue (see section 6 on Staffing). The barcoders working at the shelf were able to escalate problem items to the senior barcoders for resolution. In October 2010 the Libraries took the decision to bring forward the barcoding of the collections stored by DeepStore in the Cheshire salt mines. By uplifting the material early we would save up to a year’s worth of storage costs. We considered whether the barcoding work could be done in the cavernous salt mines or even close by, above ground, but this was dismissed on grounds of cost and concerns about working in distant locations. The Libraries had use of a warehouse in the Osney Mead area of Oxford, known as the SERS Warehouse, which was filled with temporary shelving to store more collections. Over the last three months of 2010 these collections were barcoded and shipped to the newly open BSF, where the first collections were ingested in October of that year. By emptying this space and undertaking some minor refurbishment work we were able to configure the warehouse to receive, barcode and dispatch the DeepStore collections on a daily basis. A detailed delivery schedule was devised and every weekday from the beginning of January until late July 2011 an articulated lorry brought 1,040 boxes containing between 11,000 and 15,000 books to Oxford for barcoding. The majority were barcoded on the same day and dispatched to the BSF the following day. (Figure 9.1) By this time, book moving from the New Bodleian stack was well under way. In March 2011 the last of the monographs in the New Bodleian were barcoded and all efforts were focused on the periodicals collections. Again, barcoders were working at the shelf with support from desk-based staff. This work was happening in parallel to the warehouse operation and a further 15,000 items were being barcoded each day in the New Bodleian. At the beginning of July 2011 the New Bodleian barcoding programme was three days behind schedule. We had to complete barcoding by the end of the month to vacate the building so it could be handed over to the contractors for the major refurbishment project. We increased the number of barcoders, and staff worked very hard to meet the deadline, devising new methods to speed up the rate of barcoding.

Chapter 9: Barcoding – the Inventory Control Project  

 115

Figure 9.1: SERS warehouse filled with DeepStore deliveries

At the end of July 2011, both the New Bodleian and DeepStore barcoding operations were completed. We said goodbye to a large number of agency staff as operations scaled down. We retained some staff to continue the warehouse barcoding work as collections were shipped there from the repository at Nuneham Courtenay. A small team of staff was sent to work at Nuneham Courtenay itself on the special collections because we felt that that would provide better security. We could not undertake full-scale barcoding there because high-density roller racking presented logistical limitations on the number of places at which staff could access the shelves. Work continued on these collections until the last was delivered in early December 2011. We then took delivery of the collection of foreign dissertations that were stored with Iron Mountain and the last book was barcoded on 20 December 2011.

116 

 Michael Williams, Alasdair MacDonald and Cath Walsh

5 Solutions 5.1 Use of the catalogue as an inventory tool We agreed as an operating principle that the Libraries’ catalogue would be used as the tool for representing the collections, whether or not they were catalogued at the beginning of the project. Each barcode would be a unique identifier representing a single retrievable object, represented by a single item record on the catalogue. In order to achieve this, the decision was made to work offline, using extracts of the existing catalogue data wherever possible, with daily uploads to the catalogue.

5.2 Catalogued monographs There was no machine-readable barcode associated with any physical item in the Libraries’ collections that made it suitable for ingest into the BSF until November 2009. During 2009 the application of suitably positioned external barcodes was implemented. By January 2010 all newly received monographs had an externally placed barcode added at the point of receipt. Although sequences from 1988 onwards had one-to-one mapping of item records to physical volumes, this was not always the case for older monograph sequences, where a single item record often represented multiple physical volumes for a single bibliographic record. In some cases one item record could represent thousands of individual volumes. In order for offline working to be successful and efficient for these collections, we needed to add extra item records to the catalogue to increase one-toone mapping before files of data were extracted. This was achieved by sweeping the catalogue to generate reports of records with known shelfmark indicators for multiple volumes represented by a single item record. Cataloguing staff carried out the work to add item records, often referring to the original hand-written catalogues because this was more efficient than locating the books on the shelves. Example: 170 d.16–18 – representing three volumes with consecutive shelfmarks.

Chapter 9: Barcoding – the Inventory Control Project  

 117

Cataloguers would edit the existing item record and add two more to result in: 170 d.16 170 d.17 170 d.18 In total over 14,000 new item records were added to the catalogue during preemptive work. ReBar was an in-house piece of software developed by the systems team from existing software used for the receipt of legal deposit monographs (CroBar). It worked with an extracted database of title, author and shelfmark information taken from the catalogue, allowing the barcoder to associate a new physical barcode with each individual item. When the data captured were uploaded, the barcode then became associated with the item record on the catalogue. For modern monographs with a barcode version of the shelfmark, or other unique barcode inside the cover, this process was relatively straightforward. The barcoder scanned the unique identifier, checked the title and author information, and then scanned the new barcode to be associated with the item. For this workflow the material did not need to be presented in shelfmark order. For older monographs with no machine-readable identifier, shelfmark information was often inconsistently recorded on the books, with a further problem of variation of transcription in the electronic version of the catalogue. Example: 170 d.19/1 – shelfmark written on the book 170 d.19 (v.1) – shelfmark recorded on the catalogue Nicholson classification numbers also possessed no marker to show that a subject had been subdivided. This meant that the resulting classification sequence as found at the shelf was not repeated by the internal sequencing algorithms of the library management system. This caused a problem where a classification number with a greater number of digits came before a number with fewer digits at the shelf. Example: Classification number 14024 immediately precedes 1403. There is an implicit decimal point, thus the numbers are interpreted as 140.24 and 140.3. A trial on Nicholson monographs where shelfmarks were typed into the ReBar software resulted in 25 % of items not being found in the extracted files, an unacceptable failure rate.

118 

 Michael Williams, Alasdair MacDonald and Cath Walsh

The Student Consultancy Service, a work-experience programme for University of Oxford undergraduates, visited the Bodleian and provided recommendations in the project planning stages in 2009. Following these, the systems team developed a key for sequencing both Nicholson classification numbers and the enumerative elements following the classification number to produce files that represented the order of books as they were found at the shelf. This allowed a workflow where the barcoder established a starting point by typing in a shelfmark, referring to a computer linked the library catalogue to obtain the exact shelfmark syntax if necessary, and then progressed along the shelf using a ‘Next item’ button to work through the sequence. With an estimated 2.3 million monographs in the Nicholson classification, this represented a major breakthrough for the project.

5.3 Periodicals Before 1993 for the Radcliffe Science Library and 2003 for the Bodleian, when an issue of a periodical was received, it was recorded on a paper receipt slip. To represent this in Geac Advance, the library management system in place at the start of the project, an item record was created for the first issue held at the shelfmark location, with a summary holdings statement to record all issues held there. Readers could then specify which issues they required when requesting from the stack. This system was used both for Nicholson-classified periodicals and the modern ‘P.’ sequences. Receipt of periodical issues using the serials module of the Geac software was in place for all current periodical titles by 2008. This meant that there was an electronic record for each issue received, which was used to generate a corresponding item record for the purposes of barcoding. However, the situation was not straightforward because issues of a title received before and after automated check-in could be found at the same shelfmark, with the additional complication of boxes storing many items potentially containing both checked-in and non-checked in issues. These boxes could also potentially contain issues attached to separate bibliographic records where a change of title had occurred. The following sections discuss the workflows developed for barcoding periodicals. The first two sections relate to barcoding when Geac Advance was in place. In July 2011 the Bodleian Libraries adopted Aleph, which required a complete change in periodicals barcoding workflows. This coincided with the deadline for the emptying and closure of the New Bodleian building and the removal of all material from the DeepStore facility. The third section below discusses the

Chapter 9: Barcoding – the Inventory Control Project  

 119

method for barcoding periodicals stored in the Nuneham Courtenay repository using Aleph. Two types of Microsoft Excel spreadsheet were prepared by the systems team for defined sequences: a retrospective spreadsheet and a checked-in spreadsheet. Both contained data harvested from the catalogue. Barcoders used the checked-in spreadsheet in a first sweep of the sequence. This spreadsheet listed the checked-in item records for each shelfmark. The barcoder scanned in the barcode he or she had applied to an individual issue, or the same barcode over several rows for boxed issues. The completed spreadsheet was sent to the systems team for processing and upload to the catalogue. The data for boxed issues was used to generate bound-item records with a single barcode for each box. The retrospective spreadsheet was used in a second sweep to barcode nonchecked-in issues. The spreadsheet was prepopulated with titles and system numbers relating to each shelfmark. The barcoder chose the correct bibliographic record within a given shelfmark to reflect title changes and entered caption and volume information along with the barcode. Due to time constraints for completion of the project and issues with training staff, recording enumeration and chronology had to be simplified to record the minimal number of levels to identify the physical item and distinguish it from others at the same shelfmark. Example: A bound volume containing v.1:pt.1(1977:Jan.)–v.1:pt.4(1977:Dec.) would become 1(1977). The completed spreadsheet was sent to the systems team for processing and upload to the catalogue, with the data used to generate new item records. Control and management of the spreadsheets proved a challenging task, especially given the fact that barcoders worked in three-hour shifts, with the potential for multiple staff working on one spreadsheet over the course of a day. For periodicals stored with DeepStore, the spreadsheets were tailored to reflect the order in which the material would be delivered. This will be discussed further in the section on technical challenges. At a late stage of periodicals barcoding in the New Bodleian, limitations on time and the need to prioritize limited staff resources for spreadsheet generation and processing led to the introduction of a second periodicals barcoding method. This was used in the New Bodleian stack, which housed mainly, but not exclusively, non-checked-in material in Nicholson classification.

120 

 Michael Williams, Alasdair MacDonald and Cath Walsh

Less work was required from the systems team to prepare and process the spreadsheets. Project staff requested reports from the systems team listing unique system numbers for a defined range of shelfmarks as Excel spreadsheets. Barcoders added rows to the spreadsheet for each physical item under the appropriate system number, capturing enumeration, chronology and barcodes. They removed rows where issues had been checked in or where the data structure did not allow automatic processing. The spreadsheets were processed using the automated method developed for the barcoding of publisher dumps (discussed in section 5.5 on Inventory Records). The rows which were not suitable for processing by the automated method were entered directly into the catalogue by senior barcoders. The Nuneham Courtenay repository contained both checked-in and nonchecked-in periodical sequences. This material was processed on the new library management system, Aleph, between July and December 2011. Although it was possible to adapt the retrospective spreadsheet method to allow data to be uploaded to Aleph, for technical reasons checked-in material could not be treated in the same way as before. The first stage of processing was to label checked-in issues at the shelf using a list supplied by the systems team. These were then picked and barcoded using the periodicals binding function on Aleph by senior barcoders working in situ. The remaining issues were shipped to the SERS warehouse for barcoding using the retrospective spreadsheet method.

5.4 Special Collections Although a considerable proportion of the special collections had already been relocated to the RSL, collections of manuscripts, rare books, printed music, bound maps, library records and archives remained and were destined for storage at the BSF. The majority of the John Johnson collection of printed ephemera also had to be placed under inventory control. For the rare books collections we were able to use the same barcoding processes as for the modern equivalents. However, the remaining collections were not part of the main library catalogue. Although separate catalogues and inventories existed, the only method of retrieval once the material had been moved to the BSF was through the main library catalogue. To barcode these collections, very brief inventory records were created which only included shelfmark information. This method is described briefly in the following section and in more detail in chapter 4.

Chapter 9: Barcoding – the Inventory Control Project  

 121

5.5 Inventory records Certain categories of material could not be barcoded using either the ReBar software or any of the periodicals workflows. This included material not entered in the library catalogue and material with minimal or overly complex associated metadata. The principle of these records was to include no bibliographic description, but to include a collection code and shelfmark, if available, for the purpose of identification and retrieval. These records made the material available for readers, and may in future also act as the basis for projects to improve accessibility to the Bodleian Libraries’ collections. Inventory records proved to be a pivotal tool in the timely completion of the project. Named after the complicated ‘Soc.’ periodical sequences shelved in the New Bodleian central stack and Nuneham Courtenay, the ‘Soc.’ method was developed from a workflow known as the Holmes method (after the Music Librarian) that populated the catalogue with stub records for boxed, uncatalogued music scores, which were then fed into the ReBar workflow. The representation of ‘Soc.’ periodicals on the catalogue presented a number of challenges on top of standard periodical problems such as changes of title. The ‘Soc.’ periodicals included many monographic series which were catalogued inconsistently, with sets represented by separately catalogued monograph records, periodical records, or combinations of the two. In addition, periodical titles often had substantial associated analytical cataloguing down to the article level. Finally, multiple overlapping records sometimes existed for the same title, representing different copies obtained at different times. These complications meant that the decision was taken to ignore the existing bibliographic data and to barcode using an Excel spreadsheet method which created a minimal level record for each shelfmark with associated item records for each physical volume, containing shelfmark, enumeration and chronology. The inventory record for Soc. 133 e.189, and its display on the resource discovery tool that sits over the catalogue database, Search Oxford Libraries Online (SOLO), is shown in Figure 9.2. With this development, it became possible to combine stub bibliographic record generation, item record generation and barcoding into one workflow. The end result was a parallel catalogue which allowed inventory control without the need to rationalize overly complex catalogue data. As well as being a tool for barcoding other collections with complicated associated bibliographic records such as Gough Adds and microfilms, the ‘Soc.’ method was widely used for uncatalogued material and collections only represented on card catalogues. In order to meet a crucial deadline to hand the New

 Michael Williams, Alasdair MacDonald and Cath Walsh

Figure 9.2: Inventory Control Project records on SOLO

122 

Chapter 9: Barcoding – the Inventory Control Project  

 123

Bodleian building over to contractors, the ‘Soc.’ method was also used to barcode outstanding escalated material from the ReBar and periodical workflows. Among the Nicholson-era legal deposit collections, numerous sequences of books not deemed to have academic merit (predominantly fiction, children’s books and paperback series) were arranged on the shelves according to publisher under a single shelfmark, and sequenced either by the publisher’s serial number, ISBN or order of receipt if the former were not available. These sequences were found in the Underground Bookstore, the New Bodleian stack and the Nuneham Courtenay locations. They were known as ‘publisher dumps’. Although some titles were catalogued individually, the majority were represented by bibliographic records containing only the name of the publisher or publisher’s series and a single item record sometimes representing many thousands of books. Among these non-academic sequences were many collections now seen as iconic and highly collectable, such as early Penguin books, and deemed to be part of special collections. Accordingly, many had the barcodes placed on acid free slips in order to preserve the covers. Here, the decision was made to represent each item and build on the existing publisher records. In order to assist retrieval in SOLO and prevent excessive item records on a single bibliographic record from jamming the cataloguing module, a new bibliographic record was created for each block of 80 to 100 books, to which data was uploaded from Excel spreadsheets. Example, for Penguin books at shelfmark 2704 e.45, from publisher number 873 to 1071: 245 00 $aPenguin books [2704 e.45/873–2704 e.45/1071] 500 ## $aInventory Control Project record. Because publisher numbers were captured during barcoding it may be possible in future to use them to source individual catalogue records. This workflow, of adding holdings to pre-existing records as opposed to generating new records, was given the nickname ‘speedy spreadsheets’. It became crucial in offline data collection, as it allowed senior barcoder work to be managed at times when there was insufficient physical space or staff capacity to keep up with the necessary throughput of barcoding.

5.6 Escalations Where items could not be barcoded in the main workflows, they were escalated to senior barcoders for barcoding directly onto Geac Advance, and later Aleph. For monographs, problem material included volumes of bound items, multivolume

124 

 Michael Williams, Alasdair MacDonald and Cath Walsh

Figure 9.3: Abbreviated catalogue record

sets not picked up in pre-emptive work, items either out of place on the shelf or with typographical errors in the shelfmark, material in shelfmark order but uncatalogued, and whatever else was found. Senior barcoders were trained in problem-solving and cataloguing routines, including searching for existing bibliographic and item records, correcting errors in shelfmarks, adding binding entries and creating minimal level bibliographic records. The number of uncatalogued monographs discovered was greater than could be absorbed by the Bodeian Libraries’ existing cataloguing teams, so senior barcoders were trained to produce bibliographic records to an abbreviated standard which were suitable for retrieval using keyword based searching in SOLO. The standard included complete 245 and 260 MARC fields, but did not routinely include authority control fields, which would have required more staff training and taken longer to produce. An exception was made for some academic series where a template-based cataloguing workflow was used, which contained repeated elements of the records common to each title, including a series entry. (Figure 9.3) For periodicals, problems included bound-together physical items containing distinct titles, issues which should have been checked in but were not, and uncatalogued or unexpected material. Most uncatalogued periodicals were passed to the Periodicals Cataloguing section, which created full-level bibliographic records to which item records were later added by senior barcoders. This included a substantial mini-project to catalogue a donation of Soviet newspapers and embassy despatches from the Cold War era that had been donated to the library in the 1970s. Senior barcoders usually added item records directly into the Geac or Aleph cataloguing modules, but the flexibility of Geac’s GeoCat module, which functioned as a free-text editor, allowed another way of working. Known as the ‘Notepad method,’ this workflow involved a barcoder creating a Microsoft Notepad document consisting of lines of 852 MARC fields (recording location of item) in order to barcode long runs of either periodicals or monographs. The files were then copied and pasted into the cataloguing module to create item records. This workflow allowed immediate barcoding of material that had to be moved quickly, but often required intensive preparatory work and processing by project staff directly employed by the Libraries.

Chapter 9: Barcoding – the Inventory Control Project  

 125

Figure 9.4: Barcoders at a trolley

5.7 Equipment Netbook computers equipped with a barcode scanner were used for all offline data capture, with battery-charging banks set up in the central operations areas of each site to allow recharging between shifts. In the New Bodleian operations, barcoding was undertaken away from fixed computers in confined stack areas, including roller racking and fixed shelving. A barcoding workstation was thus required to be compact, manoeuvrable and suitable for safe working over an extended period. The workstation went through several developmental stages before a final model of a two-shelved, wheeled trolley with the option of placing the netbook on an anglepoise mounting for comfortable typing and suitable screen height was adopted. Each trolley was supplied with two rolls of barcodes starting with sixes or sevens (the latter to indicate legal deposit status), acid-free paper slips, diagnostic slips for escalations, a pencil and a kick-step (Figure 9.4).

126 

 Michael Williams, Alasdair MacDonald and Cath Walsh

In the SERS warehouse barcoding was carried out seated at desks with books moving to and from the workstations in stacks of three plastic totes, placed on dollies. When required, safety clothing (dust masks, latex gloves, high-visibility jackets and steel toe-capped boots), conservation-grade ribbon and conservation-grade boxes were made available.

6 Staffing 6.1 Introduction From the outset we knew that large numbers of staff would be required to deliver the project on time. The implementation phase of the project would not start until early 2010 and was due to be completed, with the ingest of the last item at the BSF, in December 2011. Our initial calculations indicated that if we were to work a typical working week – Monday to Friday – then we had about 450 working days to barcode the estimated 6.5 million items. That meant barcoding over 14,000 items per day (a target we eventually surpassed).

6.2 Agency staff A series of trials in 2009 suggested that we would need 100 to 150 full-time equivalent (FTE) staff. It was not practically possible to fit that many people in the library stacks so we explored ways to increase daily output. Extending the working day gave us more time to achieve our target. A variety of local factors in the different locations prescribed the hours that could be worked and during the life of the project we established a nine-hour day although at some stages barcoding work was under way from 6 am to 9 pm. We expected staff turnover to be high because of the repetitive nature of the work. The Bodleian Libraries’ personnel department would not have had the resources to cope with the additional number of staff (approximately one-fifth of the existing staff numbers). With this in mind, we invited a number of local recruitment agencies to bid for a contract worth £ 1 million to supply staff for the first six months. Drawing on skills from the University’s central purchasing section we were able to negotiate a good deal that fell within our budget expectations and would enable us to achieve our targets. University of Oxford students were particularly targeted and encouraged to apply for the agency jobs, in order to cultivate a relationship between future Oxford alumni and the Bodleian Libraries.

Chapter 9: Barcoding – the Inventory Control Project  

 127

6.3 Library staff To manage, lead and support the agency barcoders we needed a team of staff employed by the Bodleian Libraries. The small group that designed and led the 2009 trials proposed two teams: – –

Supervisors to manage the agency staff, direct the daily work, and take responsibility for achieving daily targets – Inventory Control Officers (ICOs). Technical staff to provide solutions for barcoding different categories of material, process escalations, and provide bibliographic support for the catalogue – Project Officers (POs).

Initially three people were appointed to each team. The project was headed up by a project manager and supported by a senior project officer who was seconded from his technical services role within the Bibliographic Maintenance section of the Bodleian. Staff from other sections of the library contributed to the project, in particular the systems team who provided the technical management of the catalogue and developed the ReBar software. The project was also supported by existing administrative staff to deal with ordering supplies and processing agency invoices. It was very important to get this right because of the large amounts of money that were transacted every week. The variety within the collections meant we had to be able to react to sudden, unpredictable changes in workflow and productivity. For instance, moving from one classification scheme to the next meant the number of books escalated to the project officers could increase by 500 %. Learning as we went, we were able to promote some of the more skilled and experienced barcoders to desk-based work to deal with escalations by offering them a small increase in pay. This became the team of senior barcoders, who over the course of the project numbered over 50 individuals. As we started barcoding periodicals, the quantity of data to be processed increased dramatically, and we were able to put a case together to enlarge the core project team. We appointed four FTE Assistant Project Officers (APOs) for six months to support the work of the existing team, two of whom stayed on a further five months. With the planned launch of the new integrated library system in July 2011 and the very significant demands the project was placing on the systems team, we recruited a database programmer who joined that team to provide full-time support for the Inventory Control Project’s technical work. Until this point we had significantly underestimated the amount of technical support we needed especially since any hold-up in the management of the catalogue could stall the project.

128 

 Michael Williams, Alasdair MacDonald and Cath Walsh

6.4 Training Following barcoding trials, the main monograph and periodical workflows were documented. They were then revised as necessary to reflect the aims of the project and the management of different collections. All training was delivered in-house by project staff employed by the Libraries. Training of new staff and re-training of existing staff was a significant task, because of staff turnover, the sometimes unpredictable nature of the material, and, not least, the introduction of a new library management system (Aleph) mid-project. Senior barcoders were required to assess and solve problems as well as use the Geac and Aleph cataloguing modules. A manual of standard routines was created for their use and training delivered by POs and APOs using selected escalated material. Training was incremental to allow staff to become familiar with the routines and their work was reviewed by the POs until they were able to work independently. Support for senior barcoders, who frequently encountered extremely puzzling material, was provided through an email helpdesk manned by the POs. In the course of their work, senior barcoders would identify rare and valuable items, as well as finding many books flagged on the catalogue as missing. Along with the overview plan of how the collections would be managed, a certain amount of reactive work was also required as unexpected tranches of material were uncovered in the main collections and uncatalogued material came to be processed. This was to be expected when carrying out the first full inventory after several hundred years of acquisitions and required a flexible workforce, including teams of hand-picked staff trained in non-standard workflows to manage the unpredictable nature of what we found. This was particularly important in completing barcoding of the central Bodleian stacks and also the DeepStore facility, where material could not be accessed for sampling and assessment before it arrived.

7 Challenges 7.1 Introduction Despite seeking advice and using external consultants in the preliminary stages of the project, it became clear that nobody had ever embarked on a project of this scale and magnitude before, so venturing into the unknown would inevitably present us with problems and challenges.

Chapter 9: Barcoding – the Inventory Control Project  

 129

Financial and human resources were significant challenges. At the outset the project had been funded with £ 1.5 million to get started and it was not until planning was complete and the implementation had been under way for six months that we were able to make realistic budget forecasts. The ability to adjust our staffing levels to react to short-term problems was a great benefit and we would not have been able to deliver the project on time without the flexibility of agency staff and the enormous dedication and hard work of the entire project’s staff. The challenges can be broadly grouped into three distinct areas: practical, technical and inter-project.

7.2 Practical challenges All of the projects to transform the Bodleian Libraries were trying to achieve their goals with the common aim that there would be minimal disruption to existing services. Communication and building good working relationships with as many colleagues as possible proved to be the key to enable the Inventory Control Project to work alongside normal operations. Engaging with staff enabled them to understand the difficulties we faced and we found them to be adaptable and willing to help to the extent that they often provided solutions. This was backed up by tremendous support for the project at the highest level. The main communication tool was a weekly email bulletin summarizing the work done in each location and describing the next steps for the week ahead. This was sent to 70 staff in all areas that were involved or affected by the project from Bodley’s Librarian to reading room invigilators. Examples of problems that we faced were: – Health and safety: by working with existing staff in the barcoding locations and the Health and Safety Officer, we were able to overcome problems in access, fire safety and working conditions. – Building access: security of the collections was very important. The Libraries had no experience of having large numbers of agency staff or contractors working in the building. By working with facilities management and curatorial staff we were able to understand those concerns and propose new ways of working. – Space: we underestimated the amount of space required for project operations away from the books. Staff needed space for administration, training, desk-based work and break facilities. A solution to this came as staff were vacating parts of the New Bodleian Library as the project grew, and we were able to occupy that space.

130 

–

–

–

–

 Michael Williams, Alasdair MacDonald and Cath Walsh

Location of barcodes: a principle of operation for the large-scale ingest of books at the BSF was that barcodes were placed on the front cover. Concerns about defacing were dealt with robustly using communication and negotiation skills. Special collections: some collections were to be stored at the BSF temporarily so a solution to barcoding without physically affixing the barcode to the item was needed. After testing a number of solutions proposed by special collections and Conservation and Collection Care staff the answer was to apply the barcode to an acid-free paper slip inserted inside the book. Software development time: the ReBar software and the periodicals barcoding solutions enabled efficient and productive progress. The staff time required to do this work was initially at the cost of other work but by making an investment in appointing a database programmer we were able to gain full-time support in this important area. Warehouse experience: none of the existing library staff had experience of establishing or running a warehouse operation for the barcoding of the DeepStore and Nuneham collections. Identifying this at an early stage enabled us to appoint experienced staff to fill this knowledge gap.

7.3 Technical Challenges Because DeepStore material would arrive in palletized batches and was not presented on shelves, it was necessary to define what material would be arriving and when. A document known as the ’shopping list’ was drawn up to replicate as closely as possible the shelfmark sequences, with dates for delivery. In this way ReBar batches were defined in advance to match the shopping list. Although we could specify which boxes we required on which day, it was necessary for DeepStore to use the fastest picking and packaging flow possible in order to complete the move on time. This meant that we could not define the contents of individual pallets. When barcoding monographs using ReBar, barcoders could find a starting point in the file and work through boxes of monographs, but periodical sequences could arrive on different pallets or on different days. To work around this, DeepStore supplied details of the pallet contents in advance, which were used to create pallet-specific spreadsheets for periodicals barcoding work. In July 2011, the Bodleian Libraries introduced the Ex Libris integrated library system, Aleph, with two weeks’ associated downtime for the transfer of catalogue data from the Geac system. This presented a number of challenges. We had to

Chapter 9: Barcoding – the Inventory Control Project  

 131

keep agency staff working in order to complete the project on time and avoid losing highly trained and experienced staff, keep the BSF supplied with material for ingest and organize re-training of senior barcoders. The DeepStore site contained a lot of material which we knew was either uncatalogued or would need to be processed outside of the main workflows. Arrangements were made for this material to be delivered in the final two weeks of the delivery schedule, so that it could be stockpiled for processing during downtime. To keep material flowing to the ingest benches in the BSF during this time, a collection of monographs which had already been barcoded was also retrieved from DeepStore. In order to provide a seamless transition from Geac to Aleph, the Project Officer Team first attended the in-house training for the Aleph cataloguing, serials and circulation modules. The senior barcoders’ manual was then re-written to reflect the workflows of the Aleph software. Training was cascaded down to senior barcoders during downtime using the Aleph test database and client, along with shifts allocated for practice time on the new software. This proved successful and meant that the senior barcoder teams were ready when the Aleph system went live on 21 July 2011. The Nicholson sequencing key needed to be re-written by the systems team to work in Aleph, resulting in a time delay of two months between go live of Aleph and the start of barcoding monographs with AutoNext ReBar. With a little over 10 kilometres of Nicholson monographs in Nuneham Courtenay, as well as other complex shelfmark sequences that could not be handled by the internal sort algorithms on Aleph, this presented a serious challenge to keep a balanced flow of monograph and periodicals material arriving at the SERS warehouse for barcoding. This was managed in three ways. Firstly, the order of monographs retrieved from Nuneham Courtenay was changed, bringing forward collections with simple shelfmark sequences. Secondly, a large lending collection of material with barcodes used for circulation placed inside the front cover was taken, as the barcodes could be used in the same way as the machine-readable shelfmarks in other collections. Finally, sequences of Nicholson monographs were identified where a longer classification number preceded a shorter one on the shelf, as described in the section on ReBar with AutoNext. This allowed monograph barcoding to continue at the correct rate until the key was developed. When barcoded books arrived at the Book Storage Facility they were scanned against the pre-ingest file, a file of valid barcodes extracted from the library catalogue. If a physical item arrived with a barcode not present in the pre-ingest file, it was rejected as a ‘Not On File’ (NOF). This file was refreshed daily to include new barcoding work.

132 

 Michael Williams, Alasdair MacDonald and Cath Walsh

NOFs occurred for several reasons, one being when material barcoded by spreadsheet methods was shipped for ingest before barcoding data was uploaded to the catalogue. The solutions were to put a buffer of barcoded material in place to allow time for data to be uploaded to the catalogue and be harvested for preingest, and to perform a quality check on material before it was sent to the BSF. In addition, NOFs were caused by lost data files and errors made in the barcoding workflows. The NOF error rate was 1.5  %. NOFs were processed by the senior barcoders, based in the SERS warehouse and the BSF itself. Another problem dealt with at the same time as NOFs was crossed-over barcodes. These were usually flagged up by a reader attempting to order one book but being sent another with the wrong title. This required two actions: re-barcoding of the book in hand to attach the barcode to the correct bibliographic record; and an audit of the trays containing the books at the BSF to search for the book requested by the reader. Tray audits were carried out by BSF staff. The trays where the missing book might be were identified by looking for adjacent shelfmarks and barcodes. Those trays were checked to pick up any errors. An example of an error we hoped to pick up was those cases where ReBar work had become one step out from the shelfmark sequence, resulting in a series of books being associated with the following or preceding bibliographic record in the shelfmark sequence. This could be caused by missing books or books being out of shelfmark order at the time of barcoding, lack of barcoder attention and using AutoNext ReBar when machinereadable based methods of identification were available. The inventory records created by the project were unfamiliar to readers and Reader Services staff. The Senior Project Officer produced a manual for using the records and provided technical support where necessary. As the Inventory Control Project was completed, there remained an estimated 500,000 item records still mapping to now defunct stack locations and other areas of the New Bodleian building. Work will be ongoing to assess and categorize these records, many of which are for material barcoded on inventory records parallel to the catalogue, and pre-existing holdings records for periodicals. Once the above categories have been excluded, the remaining data will include all genuinely missing items as well as bibliographic duplicates. Work to resolve these records presents an excellent opportunity to complete the inventory of the Bodleian Libraries’ collections now shelved in the BSF.

Chapter 9: Barcoding – the Inventory Control Project  

 133

7.4 Inter-project challenges The barcoding, book moving and BSF projects were part of a cluster of Academic Strategy projects (see chapter 1),¹ and although run separately, they were interdependent. Other projects included in the Academic Strategy were under way concurrently, including the launch of a new integrated library system, the refurbishment of the Underground Bookstore and the large project to refurbish the New Bodleian Library. The very nature of the three interdependent projects meant we were operating as a supply chain, constantly feeding material into the BSF ingest programme. That process was always at the end of the chain but barcoding and bookmoving had different roles at different times. When tackling the New Bodleian collections, barcoding was done ahead of the moves to the BSF. By having a seven-month head start, we were able to prepare many of the collections for the move, which would proceed at a much faster rate. In the weeks leading up to the completion of barcoding in this building, the bookmovers had caught up and were always hot on the heels of the barcoders. This sometimes resulted in delays and downtime for the bookmoves team. During the Nuneham Courtenay barcoding operations in the SERS warehouse, the bookmovers were shipping unbarcoded material to the barcoders. The nature of the shelving at the Nuneham Courtenay repository limited the flow of material to the barcoders. Some collections were quicker to barcode than others which led to downtime for the barcoders. The continued uncertainty of the challenges each new collection would present made this a very difficult balancing act. For whichever operation was in the middle of the chain, be it barcoding or bookmoving, it was very difficult to maintain a steady output of material because of the variety of problems that could disrupt the input to the process. The programme for the Underground Bookstore project dictated that barcoding the collections there became our number one priority in the opening stages of the implementation of the project. This proved difficult because of the nature of the Nicholson-classified material and Official Papers stored there, our lack of experience in barcoding them and the very dense storage of collections in that space. It also quickly came to light that a considerable amount of uncatalogued material was stored there for which we had no solutions.

1 ‘Academic Strategy for Oxford University Library Services’, Oxford University Gazette (26 June 2009), accessed 29 June 2012, http://www.ox.ac.uk/gazette/2008-9/weekly/260609/ notc.htm#5Ref.

134 

 Michael Williams, Alasdair MacDonald and Cath Walsh

We increased our staff resourcing and increased our working day to 15 hours. We quickly devised new solutions which were later refined. For example, the ‘publisher dump’ method was first developed to barcode the series of Penguin and other paperback books stored in the Underground Bookstore. Barcoding of the collections stored in the DeepStore salt mine was originally scheduled as the last part of the project. When the Libraries realized that a saving of several hundred thousand pounds would be made by removing the material early, the work was brought forward to the beginning of 2011. The associated operations have already been described – our challenge was to fit this into the already busy workload. We already had experience of running parallel operations but not of the scale proposed, where we would effectively double our throughput. An opportunity to restructure and enlarge our core project team and the ability to call on additional agency staff provided the solution. We were able to put in place more rigorous selection criteria for the recruitment agency and spread our existing barcoding teams across two locations which gave us a small team of experienced barcoders in both locations, and divided the training effort into more manageable units.

8 Achievements All the planning activity including the initial analysis of the collections had estimated that there would be 6.5 million books to barcode. In early 2011 it started to become apparent that we might exceed this total given the rate of barcoding and the extent of the collections which were completely uncatalogued. The last item to be barcoded brought the total to 7,049,741. It is difficult to envisage this number of items. The achievements of the project are better described in more tangible ways. For example, there were 446 working days during the life of the project, meaning that on average 15,806 books were barcoded each day. One of the methods used to chart progress externally, and to convey it to staff and readers, was to consider the task in terms of the linear shelf space that the collections would occupy. The task was to barcode 192 kilometres of stock which is, coincidently, the same distance from Oxford to the birthplace of Sir Thomas Bodley (the Bodleian Library’s founder) in Exeter, Devon. Using Google maps and small adjustments to the figures to enable us to use Street View, we described his journey to Oxford in terms of the linear length barcoded. The journey can be seen on the Bodleian Libraries’ website, http://www.bodleian.ox.ac.uk/about/ projects/completed-projects/barcoding (See Figure 2.5).

Chapter 9: Barcoding – the Inventory Control Project  

Figure 9.5: Chart of metres of material barcoded by area

Figure 9.6: Chart of items barcoded by each method

 135

136 

 Michael Williams, Alasdair MacDonald and Cath Walsh

A variety of different collections spread across multiple locations were barcoded. The graph (Figure 9.5) summarizes our work with almost three-fifths of the material (58 %) coming from the New Bodleian bookstack and the DeepStore salt mine. The considerable investment of resources in developing the ReBar software clearly paid off: just under four million monographs were barcoded this way, representing 56 % of the total amount barcoded. The work undertaken by the senior barcoders and the employed project team was very significant. A total of 655,032 escalations were processed representing 9 % of the total. This included 288,959 item records that were added for books which were previously inadequately described, thereby enhancing the granularity of the holdings. A total of 56,598 bibliographic records were added for material that was previously uncatalogued, and during the course of the project 4,793 books that had previously been recorded as missing were found. No library had ever set out to barcode so many books in such a short period of time. We invented new processes that had to be scalable and to continuously evolve and adapt. (Figure 9.6) Pragmatic decisions were made to ensure we completed the project on time and this would sometimes compromise on the existing culture of perfectionism. In a project of this scale, there was a risk that a barcode might be mismatched with the title or volume. We set ourselves the target of an acceptable mismatch error rate of 2 % and in the event achieved a rate of less than 1 %. We achieved greater descriptive granularity in the catalogue and with every item barcoded we have improved inventory control for the Bodleian Libraries’ collections, enabling faster and more accurate retrievals. This enhanced bibliographic control will provide better access to our collections and enable us to manage them more effectively.

Andrew Bonnie

Chapter 10: The Book Storage Facility Information System 1 From WMS to the BSFIS: Why a Warehouse Management System was not the answer We recognized at the outset that the Book Storage Facility (BSF) was not typical of most warehouse operations but rather, in many aspects, an extension of library operations. Most warehouses deal with large volumes of homogeneous, transient items which are often stored for short periods and retrieved in bulk. The BSF Information System (BSFIS) needed to handle millions of individual items, most of which could spend years, or even decades in situ on the shelf, but each of which needed to have a clearly identified location to enable not only its efficient retrieval but also its return to that same location. We undertook an options appraisal to determine the best approach for managing the flow of items into and out of the BSF, and to determine the most appropriate software to support that approach. The stated objectives of the appraisal were: a) To choose a solution with the appropriate level of technology for the demand on the facility (around two deliveries/pick-ups daily), our integrated library system integration requirements and meeting our timescale; b) To work with a supplier with relevant experience of working on projects of this scale, in a library environment and able to work within our timescale; c) To implement a system that had been proven in a similar environment and required minimal, if any, development. After consideration of other suppliers, we drew up a short list of three: a) The Library Archival System (LAS) developed by Generation Fifth Applications, Inc. (GFA) and used at two similar repositories, ReCAP and Harvard; b) The Red Prairie system used by Oxford University Press (OUP); c) The FKI Logistex solution being implemented by the British Library (BL). We undertook a desk exercise to appraise each solution against a number of criteria and the best fit against the stated objectives determined.

138 

– – – –

 Andrew Bonnie

The chosen solution, the LAS from GFA, scored well on a number of criteria: It was the only solution specifically designed and built for item-level management within a high-density storage environment; The software had successfully been used in similar institutions undertaking similar exercises (e.g. ReCAP, Harvard); It would require only a little software customization to meet our needs fully; Its ability to integrate with library management systems was already proven.

2 The American model: Why barcodes are better than shelfmarks The most efficient and effective means of accessing and retrieving items from the BSF was to ensure that every item had a unique, machine-readable identifier which could be directly matched to an item entry on the main library catalogue. Barcodes offered the opportunity to avoid the manual entry of key identifier information, eliminating a potential source of error and opening the prospect of a faster and more efficient process. The challenge was that with less than a year to go until the BSF was operational, very few of the items intended to be stored in the BSF had a barcode. The items held within the old book stacks were located by their shelfmark, which placed each item at a fixed point within the sequence of items within that shelfmark sequence. Within the stacks this enabled most items to be located when requested, but also relied on the staff’s knowledge of the stack layout and the individual collections. This method of locating books did not readily translate to an operation on the scale of the BSF, operating in a Very Narrow Aisle highdensity warehouse system, with the efficiency and accuracy required of a remote storage facility. It was not reasonable to expect that an operator in the BSF, elevated over 10 metres (33 feet) above the ground would go hunting for a missing item, nor was it likely to be a successful enterprise. Items stored at the BSF would no longer be sorted in shelfmark sequence; therefore neighbouring items were less likely to have any logical relationship to any misplaced item. Nevertheless, at our request GFA, aware of the lack of barcoded items within our collections, prepared a proposal illustrating how the LAS software could be adapted to work with shelfmarks rather than barcodes, although they very strongly recommended that the collections should be barcoded. This presented us with two very clear options:

Chapter 10: The Book Storage Facility Information System 

 139

1. Amend the LAS software to enable accession and retrieval by shelf mark. However, there were a number of reasons why this offered a less than ideal solution; – The solution was unproven, and therefore carried significant risk. The LAS software’s strength is that it is a simple, efficient system that offers an established way of tracking items in storage, which results in a very high level of certainty about where items are located. All of these positive benefits stem from the fact that it is managing barcoded items and is not reliant on the manual entry of key identification data; – The solution required manual data entry and was therefore prone to error – if an item were erroneously recorded at accession, easily done given the complex and varied shelfmark schemas in operation, there would be a high probability that the item would be permanently lost; – The solution was less efficient – even if it were possible to accession by recording only the first and last shelfmark in a tray, confident that the items in between ran in sequential shelfmark order – which we weren’t! – then this was still significantly slower than scanning barcoded items. It had been demonstrated that scanning a tray of barcoded books, approximately 15 books per tray, took approximately 15 seconds, a rate which we could not hope to match accurately by typing in shelfmarks. The fact was there were large elements of some collections where there are gaps, and changes of shelfmark mid-run, that meant that it was far from certain what items lay between any two shelfmarks; – A decision to separate legal deposit material meant that recording first and last shelfmark was no longer an option. Legal deposit material was interspersed in the shelfmark sequences, and each individual item shelfmark would need to be manually recorded on accession. 2. Barcode all items – this would allow each item to be accessioned into the BSF without modifying the LAS software, enabling items to be accessioned and retrieved in the most efficient manner. This decision would save on LAS software development costs, but would require significant additional funding to achieve. A fact-finding trip to see the operations at both ReCAP and Harvard brought home the challenge that lay ahead. Using broadly the same process, supported by the LAS software, both operations demonstrated an efficient and secure operation. Workflows were clearly defined, simple and – with the use of a scanner and barcodes – offered little scope for human error. Given that we were about to embark on an ingest at a rate four or five times than that achieved by either the ReCAP or Harvard facilities, it became clear that do so using shelfmarks was not only hugely risky but in all likelihood impossible.

140 

 Andrew Bonnie

3 Check and check again: Why the BSFIS is as much about a process that works as it is the system The LAS software that underpins the BSFIS manages BSF operations by following a few simple principles: – Every item carries a unique barcode; – Each unique item is associated with a tray (or in the case of larger items, directly with a shelf) which itself is identified by a unique barcode; – Each tray is associated with a shelf address, which itself is represented by a unique barcode. In each instance the association, item to tray, tray to shelf is established by scanning the barcodes, and that association is then stored on the BSFIS. This in large part eliminates the scope for human error (none of the key identifiers has to be typed), but further surety is gained by the verification steps that are built into the workflows and governed by the BSFIS. When a tray is first filled each item barcode is scanned to make the association between the item and the tray, and a count is recorded of the number of items in that tray. This process is then repeated by another operative to ensure that the items and item-count match. Any discrepancies identified will prevent the tray from being accessioned. This may seem overkill and unnecessarily time-consuming – but if an item were to be missed and not associated with a tray then it could be lost forever.

4 Adapting to fit: We do things slightly differently this side of the pond Though one of the principal attractions of the LAS software was that it was proven and did not require major bespoke development, a critical factor given the timescales we faced, there were nonetheless some challenges in transferring the software and workflows to the UK side of the Atlantic. These arose out of a combination of local working practices, the unique nature of our collections and how they had historically been managed, and the specific design of the BSF. Few of these differences warranted changes to the software; most could be accommodated by adapting the supporting workflows. There were the obvious localization changes to support metric units of measure, millimetres rather than feet and inches. There were also date formats

Chapter 10: The Book Storage Facility Information System 

 141

that needed to be handled in European format (DD/MM/YYYY). Then there were the changes that no-one foresaw at the outset and led to some impassioned discussions as a common understanding was sought and a solution agreed upon. The area of most prolonged debate related to the format of the shelf address label. That each of the 97,132 shelves in the facility required a unique address that could be encapsulated in a barcode was not in dispute. However, the exact form that the label and barcode should take was hotly debated. Typically in US institutions shelves hold two labels: one is human-readable and is constructed in a way which reflects the actual location of the shelf e.g. it would include an aisle number, bay number and shelf number. The second label would contain a unique, but random, barcode. By disassociating the physical address of the shelf from the machine representation (i.e. the barcode) greater flexibility is offered in reconfiguring shelving within the facility without the need to re-label large numbers of shelves. Although, in common with the BSF, the US facilities had configured their shelves to accommodate the anticipated size of new intake, years of operation had shown that reconfiguration of shelving was a common and necessary practice in order to match the available shelving with the actual size of newly ingested material. Crucially, moving of the shelving within these facilities was undertaken by facility staff. The proposal for the BSF was that each shelf should have a single label, containing both the human-readable form of the address and an encoding of that same shelf address in the machine-readable barcode. The arguments in favour of this approach were that the initial shelf configuration (the ‘planogram’) was based on a very clear understanding of the current ‘sized’ collection and that therefore the need to reconfigure the shelves to accommodate an intake of material of a size different from that anticipated was unlikely for several years of the facility’s operation. What was more, such is the construction of the shelving in the BSF that specialist contractors would be required to apply any configuration changes, so we had to avoid reaching this situation if at all possible. A significant advantage of encoding the address in the barcode was that it negated the need to undertake a ‘’shelf-lighting’’ operation. Under the US model, in order to capture the shelf address within the LAS software, each shelf had to be visited by an operator and the barcode scanned and matched to the physical address in order for it to be loaded into the software. Under the approach adopted by the BSF, this wasn’t necessary. The same planogram data that was used to print the labels could also be loaded into the software without the need to visit each shelf physically to scan the shelf address. It is estimated that a shelf-lighting operation for a facility the size of the BSF would typically have taken about six months of operator time.

142 

 Andrew Bonnie

Aside from shelf addresses, barcodes continued to be the source of much discussion and an oft revisited topic as clarity was sought. Although proportionally very little of our collections had been barcoded, nonetheless the few hundred thousand items that did carry barcodes established the convention for item barcode format: – The physical item barcodes are 10 characters in length; the tenth ‘’check’’ digit is not actually entered into the system so they are stored as 9-character codes ( ‘‘check and drop’’); – For legal deposit material the first character of the barcode is a ‘7’; – For non-legal deposit material the first character of the barcode is anything other than ‘7’; – Item barcodes are encoded using the code 39 symbology. It had become clear from seeing the operations at ReCAP and Harvard that consistency in the placement of the barcode on the items, while seemingly trivial, had significant impact on the speed and efficiency with which items could be scanned and ingested and by all accounts, placing a barcode on the inside cover of a book was to be avoided. A convention of placing the barcode on the front, top left of each item was adopted to support ease of scanning once items were in a tray. Tray barcodes too broke with the convention established in the United States. Most institutions operate with two tray heights for each width to maximize the space use and ensure that items are stored to maximum density. Therefore tray labels typically consisted of 10 characters, a 2-character alphabetical prefix followed by an 8-digit number. The two characters represented the tray size, width and height e.g. ‘AH’ for a size ‘A’ High tray. Analysis of the Bodleian’s collection had determined that there was little to gain from creating tray sizes of different height for each width, therefore only a single tray height was produced for each required width. Consequently, the convention adopted for tray labels was simply to repeat the tray size. e.g. ‘AA00000001’. (Figure 10.1) Another area to which we gave much attention was the pick order. The BSFIS produces pick lists which identify the location of items to be picked for return (in most cases) to a reading room in Oxford. For high-level order-pickers (HLOPs) vertical travel is less efficient than horizontal travel, so the pick order should be designed to keep this to a minimum. In the US-based operations the pick order is shaped further by two conventions: one is that aisles are closed at one end, therefore the HLOPs have to traverse the aisle in both directions; second, the addressing of shelves is typically split vertically such that the ‘low’ addresses run contiguously from the front of the aisle to the back and then the ‘high’ addresses continue from the back of the aisle to the front. At the BSF, aisles are open at both ends with the option, therefore, that a HLOP can go down one aisle and come

Chapter 10: The Book Storage Facility Information System 

 143

Figure 10.1: Shelf, tray and book barcodes

back up another. Furthermore, the addressing of ladders within each aisle runs floor to ceiling, though the ladder is divided into five vertical sections. After careful consideration, particularly to the challenges of safely operating traffic at both ends of the aisle, we decided to operate a pick order that mirrored the US model. Sections one and two within a ladder were considered ‘low picks’. The pick order was low picks when the picker first goes into the aisle, followed by high picks as the picker comes back out of the aisle. The use of book request slips was a convention from the Libraries’ existing practice that survived the transition to the BSF, largely so as to provide continued support to reading-room operations. In the Bodleian bookstacks each request had generated a two-part book slip. When an item was picked from the shelf, one half of the book slip was put on the shelf in place of the book, the other half accompanied the book to the reading room to identify who had placed the request. The BSF does not need a book slip as a place-holder because the BSFIS maintains the ‘address’ of each item that is picked and is able to direct the operator to the correct location when it is returned to the BSF without the need of a physical place-holder in the tray. However, the reading rooms still rely on the book slip in

144 

 Andrew Bonnie

order to manage the item within the reading room and ensure it is made available to the correct reader in a timely fashion. An enhancement was added to the LAS software to enable book slips to be printed along with the pick list. The shelf barcodes labelling identification schema has the following five elements: nn-a-nn-n-nn – 1 (nn) – Aisle number: Ranges from 01 to 31: aisle 1 at the south end of the BSF. – 2 (a) – Row side: Either A or B. – 3 (nn) – Bay number: Ranges from 01 to 52: 01 starts on the east side of the BSF (ancillary block side). – 4 (n) – Section number: Range is 1 to 5 representing the vertical height in each bay split into sections. – 5 (nn) – Shelf level within the section: Range is 01 to 12. Examples: – 01-A-01-1-01 This is the bottom shelf (150mm) in the south-east corner of chamber (hall) 1 – 08-B-52-5-01 This is the top shelf in the north-west corner of chamber (hall) 1 The following items complete the specification: – No check digit is used on the label – either eye-readable or within the barcode; – The barcode contains the same location reference as the eye-readable part; – The barcode uses 39 characters/symbology.

5 Equipment (and lots of it) At face value the list of equipment required to operate the BSFIS comprises no more than a modest inventory: – A server to operate the LAS software; – A PC running a terminal session to access the BSFIS; – A wireless scanner to scan items at the ingest/retrievals desk; – A remote data collection terminal to capture data at the shelf; – A label printer to create tray labels; – A standard printer for the printing of work orders, pick list, request slips etc. However, to support the verification steps within the workflow, and the operation of up to eight ingest benches operating in parallel, this expanded into the inventory in Table 10.1.

Chapter 10: The Book Storage Facility Information System 

 145

BSFIS Operations

Items

No.

Direct Input Devices:    

Symbol LS4278 Series wireless scanner

18

Symbol LS4278 Series charge base and wedge cable (USB or P/S2)

18

Symbol LS4278 Series Intellistand

2

Label-Generating Equipment:

Zebra Z4M Printer: High-speed, thermal transfer label printer

2

Zebra DesignerPro™: Label development software

2

Remote Data Collection Equipment:               Printing:  

Symbol MC9090 portable data terminal w/ battery

6

Symbol MC9090 single-slot comm./charge cradle, power supply

3

Symbol MC9090 RS-232 comm cable (USB or Serial)

2

Symbol MC9090 four-slot charge cradle and power supply

2

Javelin application generator and data collection software – Runtime Licence Javelin application generator and data collection software – Developer Licence

2 1

Networked printer

4

Zebra Z4M Printer

2

Work Stations

BSFIS-configured workstation

Spares

MC9090 Batteries

22 4

Table 10:1: BSFIS equipment inventory

The LAS software was installed in a virtualized environment, the (physical) servers located onsite at the BSF to enable operations to continue should communications with the outside world be lost for any reason.

6 Scaling up for the mass ingest The barcoding of the collection before ingest into the BSF was a mammoth task. The pace with which it had to be achieved, and the volume of items involved, meant that it was inevitably prone to the occasional error. (See chapter 9.) In order to trap some of the potential problems before items were placed in the BSF a simple check was performed as part of the ingest. As barcodes were added to items, the item records were updated on the library catalogue. A nightly extract of the item data was then loaded into the BSFIS, so that we could verify that every item entering the BSF had a correspond-

146 

 Andrew Bonnie

ing entry on the catalogue. Without this check it would have been possible for items to be placed into the BSF without any means of subsequently retrieving them. This check alone trapped thousands of items (albeit a small percentage of the overall number of items ingested) and enabled us to reconcile the data before any item was shelved. This saved considerable effort over the post-ingest reconciliation that would have otherwise been required. It also ensured that these items remained visible to and available for request by readers. The Bodleian Libraries hold several million items courtesy of their longstanding role as a library of legal deposit. The storage of the legal deposit items was subject to some tax relief, if it could be clearly identified and separated from the non-legal deposit material. In logistical terms this meant sorting the legal deposit items from the non-legal items prior to ingest, identifying one from the other (by way of the barcode prefix) and physically housing them in separate chambers. In system terms, this was handled by the existing functionality. Legal deposit and non-legal deposit were set-up as separate ‘customers’ on the BSFIS. Validation rules were defined to permit only barcodes prefixed with a ‘7’ to be ingested for the legal deposit (LD) customer, and to prevent items prefixed with a ‘7’ barcode being ingested for the non-legal deposit customer (NL). Though assignment of the tray to the correct chamber was a manual process, the BSFIS did ensure that only trays belonging to the same customer could be assigned to a single shelf. Further use of BSFIS customer codes supported the particular needs of the special collections. A number of items formerly housed in the New Bodleian Library are destined to return to the newly refurbished Weston Library and are only temporarily housed in the BSF. The decant of this material to the Weston Library will itself be a major logistical exercise and we gave consideration to how we might make the process easier when the material was initially ingested. As well as the material being ingested under its own customer code (SC), we further subdivided the collections into over three hundred Collection Codes. We hope that the workflows that we shall develop to decant this material for return to the Weston Library will be eased by the ability to identify entire collections en masse.

7 Integration with the library management system The BSFIS contains no bibliographic information, but provides the sole source of information about where individual items are located. The library catalogue contains the bibliographic information and offers public access through the Libraries’ resource discovery platform, SOLO (‘Search Oxford Libraries Online’). However,

Chapter 10: The Book Storage Facility Information System 

 147

the catalogue holds no information on where an item in the BSF is located, other than the fact it is in the BSF. Once again, the barcode is the key link between the two systems. At ingest, shelfmark ranges were taken out of circulation on the catalogue at the point they were due to be transferred to the BSF. Once the items had been successfully ingested into the BSF, an interface file containing the item barcodes was sent back the catalogue. On receipt the location of each item was updated in the catalogue (indicating the item was in the BSF, but not with its precise location) and the item was once again made available for requesting. This in effect meant that an item could be ingested one day, and requested by a reader the next. It also meant that most material was taken out of circulation for no more than a few days while it was ingested into the BSF. Items are requested via SOLO, which generates a file of barcodes and associated request information (including requester name, reading room destination etc.). This file is fed into the BSFIS which in turn generates a retrieval work order for the item to be picked and shipped back to Oxford. When the item is returned to the BSF it is re-filed (in the same tray from which it was picked). An interface file is passed back to the catalogue containing the item barcode. On receipt, the catalogue is updated to reflect that the item is once again available for requesting. Though the underlying system which supports the catalogue changed in mid-project from Geac Advance to Ex Libris’s Aleph, the process is essentially unchanged.

Dave Price

Chapter 11: Implementation of a new integrated library system 1 Background The library community worldwide has experience of replacing an integrated library system and it usually follows a recognized process. The main interest in this study has been the radically different approach we have taken in planning the procurement and implementation. For most university libraries, the most important single IT service is their Integrated Library System or ILS (also often to referred to as a Library Management System or LMS, particularly in the UK). Comprising a catalogue, in our case, predominantly for printed books, but also for maps, music, video recordings and electronic materials, an ILS will also provide functional modules to support their management. Commonly there will be modules for cataloguing, acquisitions, circulation (i.e. lending) and serials management. There will also be an Online Public Access Catalogue, or OPAC, the interface through which readers can discover materials, reserve them, renew loans and manage their account details. In the early 1980s, Oxford ran a pilot programme to automate four lending libraries. This was immediately followed by a major project to introduce automation to the central Bodleian Library and this service went live in 1988. From the outset, the strategy was to provide a union catalogue for Oxford, known locally as Oxford Libraries Information Service (OLIS). Membership rapidly grew and there are now over one hundred OLIS libraries, comprising the University’s central, faculty and departmental libraries, most of the libraries of the colleges and permanent private halls, the libraries of several affiliated institutions and the Centre for Oxfordshire Studies (part of the local public library system). The ILS then chosen was DOBIS/LIBIS. At the time it was marketed by IBM as a ‘product offering’ rather than a full product: IBM did not take on responsibility for software development, but left this to the user community to whom they made the source code freely available. This was to have a significant impact for Oxford as it meant that our local support staff could respond rapidly to enhancement requests from users. Over the years the Oxford libraries became accustomed to an environment where changes, sometimes quite idiosyncratic, could be made to the basic software, engendering a culture where software was modified to fit existing practice, rather than reviewing business processes in the light of software

150 

 Dave Price

capabilities. Such changes and enhancements were often referred to as ‘Oxford specials’. Integrated library systems are highly complex in terms of software functionality and data structures. Inevitably they do need replacing as they become obsolete or because better functionality and performance is available. Upgrading to a new system is always a major challenge and is commonly undertaken every 7-10 years. Geac Advance was selected as a successor to DOBIS/LIBIS in 1995 and implementation completed in 1997. The software reproduced many of the Oxford specials functionality of the previous system. Over its lifetime, many more were developed, though none so extensive as the automated stack request system, developed as an addition to the circulation module to enable readers from anywhere in the world to order items from different bookstack locations for delivery in different reading rooms. This finally went into full production in 2002. By 2003, we knew that development of Advance had ceased, and though the manufacturers promised to support it until the last customer, we knew that once again we must seek a replacement. As before, we undertook a major procurement exercise in 2004-05, this time a two-stage tendering under European Union regulations. Considerable effort from many library staff and committees went into drawing up the tender document, which had 1,786 requirements. Seven of the eight interested vendors submitted proposals and gave a one-day presentation to library staff, from which three were short listed. Each vendor gave demonstrations, including complex workflow scripts under the close scrutiny of librarians, covering most aspects of the requirements document, over a period of five full days. Implementation began in 2005. There followed a period of intense activity on the part of both the vendor and librarians, which included training 700 staff in use of the new system. Despite overcoming numerous technical obstacles and developing some impressive features, in 2008 we decided to call a halt to the implementation. Why did we take this decision? If library systems are complex, then they are doubly so in the Oxford context. The system must be capable of supporting an extended, federal structure – more than a hundred different libraries, largely independent with widely differing policies, and with members able to belong to more than one library, often with different statuses. Very few systems on the market approach this level of complexity. On top of this were the aforementioned Oxford specials. The primary reason for the failure to progress to implementation was the excessive requirements demanded by Oxford. It had been unrealistic for us to expect a vendor to achieve in a short space of time the extensive development required to reproduce – and exceed – the functionality we had developed over two decades and two library systems.

Chapter 11: Implementation of a new integrated library system 

 151

2 Moving on: Resource discovery In one sense, we were back to square one regarding the ILS implementation. However, it gave us the opportunity to focus on resource discovery. We had planned independently to introduce in 2008 two ancillary library systems: firstly, to replace our Open URL resolver with Ex Libris’s SFX and, secondly, to replace our home-grown interface to electronic databases and allied resources with Ex Libris’s MetaLib. Meanwhile, a new breed of interface, referred to generically as resource discovery platforms, was beginning to emerge. They differ from the traditional library search interfaces, OPACs, in several respects. The interfaces are more modern and are more like the interfaces our users encounter elsewhere on the internet. Running independently of the ILS and using dedicated hardware and modern indexing technology, typically Apache’s Lucene/Solr, they are usually much faster than OPACs. Rather than traditional library searching methods, e.g. index browsing and precision (Boolean) searching, the new interfaces encourage users to start with a simple ‘one-box’ search, like Google, which may generate very many hits. These may then be refined by the selection of dynamically produced facets and scopes to assist the user in discovering what they need. This modern approach is augmented by relevance ranking, which attempts to present results most usefully to the user. So-called web 2.0 features provide the user with the ability to tag, review or ‘favoritize’ records, or share with other services such as reference management software and social media sites. Importantly for Oxford, where many of our collections are not included on our ILS but are listed in other catalogues and finding aids, resource discovery platforms can take catalogue records not just from the ILS but from different sources and allow users to search across them together. We determined that we could, without waiting for a new ILS, bring significant improvement to user experience by introducing rapidly a resource discovery platform. We chose Ex Libris’s Primo, badged locally as ‘SOLO: Search Oxford Libraries Online’. We had in the past provided our users with highly customized software in terms of appearance and functions. But we had learned that this came at the price of high development costs and difficulties applying upgrades to software which had diverged significantly from the mainstream product. So to install Primo rapidly and to avoid software configuration problems, we chose to minimize customization and to use it as closely as possible to ‘out of the bag’, or the ‘vanilla’ system. Behind this reasoning was a desire to send a message to staff that we have moved on from a time when software can be highly customized and perfected before entering production. We were now in a ‘beta-culture’ age when software can be released before it has been fully refined and tested to exhaustion,

152 

 Dave Price

as could be seen elsewhere on the Internet. Users nowadays understand that this is the price of an early release – but also that in the new, agile environment, small incremental improvements can be made at short notice, without inconveniencing them or having to wait, for example, until each Long Vacation before changes are introduced. SOLO was launched 1 October 2008 by which time both ‘OxLIP+’ (MetaLib) and ‘Oxford University e-Journals’ (SFX) had also been introduced. SOLO searches and displays not only records from our integrated library system (which includes e-journals and e-books), but also from our Chinese and Japanese catalogues, our electronic databases, and Refugee Studies database of grey literature.

3 Stepping back: External consultation Though providing some breathing space, SOLO did not remove the need for a replacement ILS. To help us consider the next steps, we invited a colleague from Harvard Library to work as a consultant with us, sharing knowledge of and expertise in the ILS world to provide a neutral view of the Oxford experience and to suggest how we might progress procurement of a new system. Agreeing that a replacement system was essential and urgent, he advised an ‘aggressive but not unrealistically tight timeline’ – enough time to ensure success but not unnecessarily long, since otherwise the project had the potential to consume as much time and resource as were available. This would require a highly concentrated effort for system selection and contract negotiation and good project management during the post-contract implementation phase. He warned against lack of precision in communications with the vendor regarding system configuration and database conversion and urged us to canvass other sites about their implementation experience, factoring that into our timeline. He also advised that responsibility for managing the vendor relationship should be at a high management level. The consultant also warned us against ‘prolonged decision-making around implementation decisions’, noting that Oxford libraries have a strong tradition of consultation and consensus-building, in part reflecting the decentralized nature of the library system. This hit home as several hundred librarians had had significant involvement in the last procurement and implementation. He made some acute observations regarding procurement. ‘The marketplace for ILS systems is smaller and weaker’. Whereas in the past there had been eight or nine potential suppliers, as a result of much consolidation and slow sales, there were fewer systems, perhaps just two, that could realistically provide for a large,

Chapter 11: Implementation of a new integrated library system 

 153

academic research library like Oxford. He noted, ‘the role of the ILS has shrunk’. This was not to decry its critical importance, but the ILS was not longer the single, central system in the library’s environment, but one component, albeit a major one, in an extending digital infrastructure, with many new initiatives requiring the library’s attention. Therefore the resources dedicated to the project should be considerably reduced. His view was that customization - the Oxford specials - was to be avoided whenever possible and instead we should put effort into planning how to adapt our workflows and processes in order to exploit better the software as delivered. Notwithstanding this last point, he recognized that one area of functionality, our Automated Stack Request system, would not be provided by any of the systems on the market and would have to be developed. He advised that we undertake an analysis of the service requirements of book delivery in light of the forthcoming Book Storage Facility, its internal inventory system (BSFIS) and the interface between them, in order to provide vendors with precise specifications. The consultant advised a faster, simpler and less resource-intensive selection process. Rather than the traditional exhaustive comparison of systems using a long list of functional requirements, he suggested making the decision using two elements: firstly, by appropriate site visits and by assessment of how successful an implementation has been at peer libraries; secondly, by shortlisting highlevel, truly critical factors relating to the vendor and the system. To this end, he met with the Bodleian Libraries’ Cabinet of senior managers and together they identified the following high level criteria: – Scalability – Configurability – Integration with Resource Discovery – Integration with Other University Systems – Support for Automated Stack Request – Implementation – Maintenance & Support – Supplier Relationship – Supplier Viability – Price Another outcome was the establishment of a new post of IT Programmes Manager to manage both the ILS and the BSFIS projects. With great foresight, this was made a permanent post, recognizing that once these projects had been completed, there would certainly be more major IT developments waiting in the wings.

154 

 Dave Price

4 Marching Forward: the Procurement Process Responsibility for the ILS and BSFIS implementations lay with the Integrated Library Systems Steering Group (note the plural form), itself reporting to the Bodleian Libraries’ Executive. This group had earlier overseen the implementations of SOLO, OxLIP+ and Oxford University e-Journals. Led by the Associate Director, then heading up Special Collections and IT, it included the Assistant Directors of Research & Learning Services and of Collections & Resource Description, the Head of Systems & e-Research Service, and the chair of the Committee of College Librarians. The ILS Steering Group took forward the consultant’s recommendations and in line with his advice set up a small, specialist group, the ILS Scoping Group, to run an accelerated and streamlined procurement, reporting monthly to the Steering Group. The Scoping Group was heavily weighted with members of the Systems & e-Research Service, who would be effecting the implementation, but it also included representatives of Research & Learning Services, Collections & Resource Description and the college libraries. It first met in February 2009 and its brief was to deliver recommendations to the Steering Group in July. There were several interesting aspects about this procurement which made it very different from the traditional approach we had employed in the past. University regulations allowed us to an extent to define our own procurement methodology. Rather than publishing an open invitation to tender, provided we satisfied the University’s procurement policies, we could restrict it to pre-selected vendors. As noted above, changes in the ILS marketplace meant that realistically, for our purposes, there were perhaps two leading contenders, so these went on our shortlist. We added to it the supplier of our current system because – they were extremely familiar with our current practices; – they had hitherto provided excellent service; – their replacement product looked modern with great potential; – we could be sure they could make a good job of migrating our data from the old to the new system; – they had written our original Automated Stack Request system. Our previous tender document had been 190 pages long with 1,786 requirements to which the vendors were expected to respond individually, specifying the extent to which and how they could meet them. Key to our methodology was to turn this on its head and to proceed from the premise that all three short-listed systems could meet our needs. Rather than our former approach, we chose to run the procurement as a ‘competitive dialogue’ where we could work together transparently with the vendors to ensure a common understanding of issues and help them produce their final submission.

Chapter 11: Implementation of a new integrated library system 

 155

Although Oxford had in the past been seen as a large, influential and prestigious customer that vendors would like to have on their books, it was generally understood that taking on Oxford would be challenging in terms of its complex federal structure and high level of customization. A major risk we identified was that the short-listed vendors would view Oxford as too challenging and might choose not to tender. Early discussions with the three vendors in January 2009 outlined our intended procurement procedure and emphasised Oxford’s desire to implement rapidly, to work as far as possible with standard software and to minimize customization and development. This was followed by a formal letter to the head of each company, which described the proposed process and timetable and asked if they were willing to participate. To our relief, they all were and in April they were issued with a Competitive Dialogue Briefing document just 35 pages long. The first section provided background and objectives of the project and the evaluation and selection process. The second section provided a high-level description of the functional scope of the main system modules used by Oxford, noting that consideration should be given to Oxford’s federal structure, the main areas being OPAC, resource discovery, cataloguing and authority control, acquisitions, serials management and circulation. Specific reference was also made to the Automated Stack Request requirement, now stripped to bare essentials, expecting that most vendors would need to do at least some development to enable their systems to provide this. The third section took the ten key selection criteria referred to earlier. These were then broken down into ‘evaluation points’ and vendors were asked to address all forty. So, for example, the first, scalability, was broken down into system performance, search functionality, federalization and ability to accommodate structural change. Having received the document, the vendors had one month to provide their draft responses. We invited them to contact us in that period whenever they wanted to seek any clarifications needed to help prepare their draft responses. All did, and the interchanges were fruitful on both sides. In addition, regular news bulletins were issued to them, so they each received any new information elicited from us by the others. From the viewpoint of the vendors, the next stage was two days with the Scoping Group to scrutinize the response with the express purpose of improvement before its final submission. This included a meeting with the ILS Steering Group and the Executive concentrating on strategic issues. Meanwhile, the Scoping Group conducted the equally important activity of customer site visits. For each vendor system, two sites were visited. The sites were provided with copies of the Competitive Dialogue Briefing, so they would under-

156 

 Dave Price

stand the nature of the procurement and could relate it to their own implementations and experiences. Each visit lasted two days. We should record here our immense gratitude to all of the sites who unreservedly gave their time and also unstintingly, their hospitality. Not only were the visits highly instructive, they were highly enjoyable as well and the professional contacts made have endured. The Scoping Group took a holistic approach to the evaluation process, with the vendors’ written responses considered alongside the knowledge gained from the reference site visits and vendor meetings. Sections two and three of the Competitive Dialogue Briefing were expressly designed to provide a framework for the analysis of the suitability of the candidate systems. Section two sought to test the assumption that any of the candidate systems should be broadly capable of meeting our requirements and to identify any discrepancies and gaps in this functionality. Section three took the ten key selection criteria and expanded them into a series of evaluation points against which each of the three candidate systems could be measured. The numerical scores were weighted to reflect the relative importance of criteria and evaluation points. It is feasible that any of the three vendors could have provided for us – indeed that was the premise at the outset – but the outcome was the recommendation from the Scoping Group that Oxford move forward with the acquisition of Aleph from Ex Libris as a replacement ILS. Though the Libraries accepted this recommendation in July 2009, we were not able to announce the result and commence contract negotiation with Ex Libris until we had formally secured funding from the University. This was awarded in September 2009 after the consideration of the business case. (Its preparation was the first task of our new IT Programmes Manager.) The delay this introduced was difficult for all the vendors and we were grateful for their good-humoured forbearance during this period. One nugget from several of the earlier site visits had been to allow plenty of time for contract negotiation, and this indeed proved to be the case. Once we had informed Ex Libris of our decision (and shown due diligence by following University guidelines to run a procurement exercise to appoint an external solicitor), this took a full three months. This threatened our first milestone, the project kick-off in January 2010, but happily contract signing was completed in the nick of time and good will on both sides meant we were able to work with Ex Libris during the pre-implementation phase on project planning.

Chapter 11: Implementation of a new integrated library system 

 157

5 Project planning Great care went into the overall implementation strategy. Obviously, the vendor hoped for as rapid an implementation as possible, and for a site like Oxford, they considered a year should be sufficient. Library management too were keen to have the new system running as soon as possible. On the other hand, library staff had toiled for three years during the previous implementation: they knew the scale of the detailed work that had to be accomplished and many felt three years might not be enough. But against this, it was evident that many were fatigued and it was clearly undesirable that they should suffer so soon another protracted implementation. Another factor was maintenance for the old system, which was only assured until July 2011. Accordingly a phased approach was devised. In broad terms, the first phase would last eighteen months, from January 2010 to July 2011. At the outset we would have sufficient hardware in place for Ex Libris to commence their software installation and then to work with us configuring the system, training our support staff and running test data migrations from the old ILS to the new. By January 2011, they would hand over to us a completed system, as it would be when we would go live. In effect, this was the main implementation project as far as the vendor was concerned and would last one year. It is generally best for us to introduce new major systems over the summer, after exams and before the next academic year, to minimize impact on our users, so moving the new ILS into production was planned for July 2011. This would give us a period of six months with the new ILS effectively ready to go, when final data migration issues could be resolved, staff trained, and any outstanding glitches and structural and workflow issues could be addressed. It could also serve as contingency so implementation overruns could be absorbed by this ‘buffer period’. What we expected to have at the end of the first phase was a system pretty much close to vanilla, but configured to provide for Oxford’s federal structure and to have requisite functionality to allow readers to call books from our stacks, interfacing with the BSFIS. We planned and resourced a second phase, lasting 18 months, to build up the system to its full functionality and for further ancillary development. Reporting provides a good example of the type of work allocated to phase two where over the sixteen-year lifetime of the old ILS a veritable plethora of specialized reports had been developed for the different libraries. We could provide a limited set of standard reports on day one. In phase two we are devoting development and training effort to introducing a sub-system, the Aleph Reporting Center (ARC), which will enable librarians to create many reports themselves to attain the levels of reporting they have enjoyed in the past. It will also still be necessary for systems staff to create bespoke reports to attain the full richness previously enjoyed.

158 

 Dave Price

6 Implementation The existing ILS support team comprised three applications programmers, a stock control programmer and two library systems support staff specializing in training and configuration of the circulation and acquisitions functions, including serials registration. For the duration of the two phase implementation and also to support the introduction of the BSFIS (see Chapter 10), we established five additional three-year posts: an application developer, an Oracle database administrator, a data administrator, a project administrative assistant and a systems analyst providing support and training in cataloguing. All staff worked closely together as a single ILS implementation team organized on a task basis. Phase one of the implementation proceeded as planned. After an initial installation of a test environment, in April and May 2010 Ex Libris trained our systems staff and helped them configure the software for Oxford. The full hardware was installed and the production environment built. Data migration from the old system to the new was a critical aspect of the implementation and, starting in June, three test migrations were performed in the first year, each run improving both in terms of data integrity and speed. We were able to sign an interim acceptance of the Aleph software in December 2010 on schedule, so from the viewpoint of Ex Libris they had completed the implementation project. There was still considerable work ahead for our implementation team. The principal administrative units in the Aleph structure and, for Oxford, equivalent to independent libraries, known as ADMs, needed configuration. A final test migration was performed in April 2011. We felt this was necessary because the high volume of new data created by the barcoding and book move projects meant that our database had changed considerably during 2011. Staff training began in April 2011 and continued beyond go live in July. Overall, the phase one implementation went according to plan and extremely well, and this success was due largely to the IT Programmes Manager’s excellent project management, a good relationship with Ex Libris and their project team, the skill and dedication of our implementation team, a tight governance structure and the support of senior management. Some decisions and aspects are worth highlighting for their intrinsic interest and where lessons were learned. As regards training, a strategic decision was that, rather than hiring trainers specifically for the project, we should use our own staff (so retaining knowledge) and provide less broad-brush training for staff in general, but more focused training for staff regularly using the system. Also that the trainers would produce good documentation and strong follow-up support once the new system had gone live. Another important strategic decision was to treat the management unit, the Bodleian Libraries, as a single structural unit (ADM in Aleph terminology) rather

Chapter 11: Implementation of a new integrated library system 

 159

than a large number of individual units as hitherto. In comparison with the other administrative units, this has added complexity, but it has added significantly to ease of administration in terms of support for centralized functions, circulation, reports and budgetary control. It has also provided a structure which would allow the move to harmonize policies across the Bodleian Libraries in terms of readers’ statuses and privileges if this is considered desirable. A significant step forward from the viewpoint of Oxford users with the introduction of SOLO and Aleph was the adoption of Oxford single-sign-on for access to library-provided electronic resources and patron accounts. This meant that University members are now able to use the same authentication for our services as they do for the University’s email, virtual learning environment and many other services. This improvement has been extremely well received. However, Oxford SSO is available only to University members and a high proportion of Bodleian Libraries’ membership is non-University. The corollary was that we had to develop and support a parallel authentication system for ‘Library Card’ holders (non-University members), which was a significant task. We chose to virtualize our hardware architecture, basing it on VMware ESX, in order to provide a highly available, resilient, dual-site infrastructure. This added topological complexity but Ex Libris were able to configure the Aleph software successfully to take advantage of it. Behind the Aleph software lies an Oracle database. Initially, we positioned the two mirrored Oracle database servers within our VMware infrastructure. During the implementation we experienced problems but discovered that Oracle have not certified their software to run in a virtualized environment. To mitigate this risk in the future we purchased and installed dedicated Oracle servers outside VMware. By replacing our old ILS with Aleph, using the OPAC via Primo (OvP) technology we could in principle provide our resource discovery service, SOLO, with full patron functionality in terms of the ability to place holds and reservations, renew loans and check account details. Beforehand, with SOLO acting as a front end to Aleph’s predecessor, a user would have to ‘shell out’ to the Geac Advance OPAC to do these. This raised the question of whether we should now move ahead with SOLO as the single interface to the ILS or to make the Aleph OPAC available as well. It was important to make the decision quite early in the implementation. Advantages of a single user interface included ease of navigation for users, reduced systems support and reduced user support and user education. Against this was the current lack of index browsing and date-range searching, and the difficulty experienced by a section of our readers in searching and retrieving because of the way search results are clustered. The ILS Steering Group, which comprised high-level representatives covering the Oxford library community, considered carefully the pros and cons, knowing

160 

 Dave Price

that the final decision would not suit everyone but that it was in everyone’s interest to make it rapidly. Their recommendation to Executive was to go live without a dedicated OPAC. There has been some dissatisfaction, chiefly from librarians. The SOLO User Group, made up mostly of reader services staff but including systems support staff, has achieved much by analysing SOLO issues, providing development suggestions both for our development staff and Ex Libris, and importantly providing and promoting new search strategies to tackle perceived deficits, which has been extremely helpful, particularly to those used to traditional OPACs and coming to SOLO for the first time. The ILS implementation was not the only project taking place at this time. The barcoding and bookmoving projects in particular required significant amounts of work to be done by the ILS implementation team. Here, alignment and coordination of potentially competing or conflicting project plans were dealt with most effectively by the Bod Squad, attended by the IT Programmes Manager (see chapter 2). Inevitably there has to be a period switching to the new ILS when data is frozen. Working closely with Ex Libris and refining data migration procedures, we were able to bring this down from an original estimate of three weeks to eleven days – a working week and the two weekends either side. We scheduled shutting down the old ILS at 5 pm 7 July 2011 and then exporting its data. During that very day, the old ILS server failed and we were unable to restart it. In fact it took five days to recover the old system sufficiently for us to begin the data export. In all it added four days to the migration period and we finally brought Aleph into full production 22 July 2011. At the time it felt devastating but in the context of a major project such as this, four days’ slippage is not so bad. A major business system such as an ILS should not be down for five days. At least this incident supported our view that we needed a replacement ILS! We had chosen to go live at the beginning of the summer rather than the end in order to deal with any teething problems before the start of full term in the coming academic year. In fact, this bedding in period was busier than we had expected. The sophisticated automated stack request system we used with Geac Advance had now been replaced with the Aleph closed stack management sub-system which had been enhanced by Ex Libris to match our requirements. After go live, we had to make many configuration adjustments, especially to loan limits, to fine-tune it further for our complex environment. Another area of concern was the ability of the system to handle notes describing particular copies of books within the bibliographic record, which are of particular importance in Oxford to those working on early printed books. Following some modifications by Ex Libris, the implementation team was able to devise a suitable protocol which was released over the summer.

Chapter 11: Implementation of a new integrated library system 

 161

Having dispensed with an OPAC, SOLO had to deal with much more traffic and transactions, with a consequence that we experienced performance problems. We dealt with this by installing more memory. Unlike Aleph, which is a mature product, Primo, the underlying software of SOLO, was relatively new at that time and under rapid development. Inevitably we encountered some bugs and constraints, particularly in delivering OPAC function via SOLO, that needed immediate attention, investigation and fixing. That we should have such problems was not a surprise. Where perhaps we should have invested more thought and effort was in expectation management of the Oxford librarians for the immediate post-go-live period. There was amongst many a sense of disappointment that the new system was not as good as the old. This was exacerbated by the four-day delay in going live, the teething problems mentioned above and for some, regret at losing a traditional OPAC. In retrospect we think that we had not succeeded in imparting several key messages. Firstly, that to avoid a prolonged implementation, we had broken it into two phases, the first delivering a basic but nevertheless a full production service. There would be a resourced second phase to further build and extend the system. There were mechanisms in place for library staff to influence the priorities, and report generation would certainly be one of them. Secondly, we can expect teething problems when a major business system is brought into production in a complex environment such as the Oxford libraries. The project plan allowed for three months over the summer, which for most libraries is a quiet period, to deal with them. Thirdly, we do believe that a modern resource discovery system that can incorporate different catalogues and finding aids, not just the ILS, in place of a traditional OPAC, is the right strategy. Systems like SOLO are under rapid development and their apparent deficits are being made good, or can in many cases be obviated by an open and imaginative approach. Fourthly, highly customized software is no longer sustainable. Rather than reproducing our traditional workflows, we should where necessary be examining them and redesigning them in the light of what is built into the software. Phase two began in earnest in November 2011, once the teething period was properly behind us, and it is currently resourced for a further 18 months. The ILS Steering Group, representing all areas of the Oxford library community, manages the implementation team’s priorities. A number of them have already been accomplished: a library control system which will allow a library itself to make configuration changes, such as opening hours, loan periods; ‘Scan and Deliver’, a new service to enable University members to request scans from items held in the BSF; and setting up a lending collection within the Gladstone Link in the central Bodleian Library. Other developments under way include a SOLO interface for mobile devices; roll-out of the ARC reporting system to the libraries;

162 

 Dave Price

using metadata from other catalogues and finding aids to search and retrieve in SOLO; further extension of reporting and record exports; self-return; the addition of several new member libraries; migration of the Chinese and Japanese catalogues into Aleph; and disaster recovery planning and testing.

Michael Heaney

Chapter 12: Staff relocation 1 Introduction Although the primary focus of the emptying of the New Bodleian Library was on the removal of nearly 3 million volumes from its bowels, the building was also home to more than 150 staff – over a quarter of the Libraries’ total staff complement. Some of these would reach the end of their contracts in the normal course of events before the closure of the building, but we had to find new homes for the great majority. For some of these the relocation was simply that – a standard office move. Others needed specific facilities such as conservation benches, which had to be prepared for them in advance. Yet others were inextricably linked to the building or its contents, and their moves had to be choreographed with the other elements in the timeline for the evacuation of the Library.

2 Staff occupying the New Library The staff occupying the New Library worked in several departments.

Collections and Resource Description Reader Services

48

Conservation and Collection Care

22

Imaging Services

10

Special Collections

55

Facilities Management

plus 4 vacancies

9

6

plus those who went to the Radcliffe Science Library plus 20 peripatetic staff

Table 12.1: Staff by departments

In addition, staff from the ICT Support Team, providing desktop IT and hardware support to the Libraries, had a base in the New Library as well as in Osney.

164 

 Michael Heaney

3 Identifying the destinations The Libraries had two buildings devoted primarily to administration and backoffice functions, in the eighteenth-century Clarendon Building situated in the centre of Oxford between the Old and New Libraries, and the Osney One building on the Osney Mead light industrial estate 1.3 kilometres (0.8 miles) away. The Osney One building had been acquired for the Libraries in 2006 and fortuitously the ground floor, which for the three years 2007–09 had housed Google’s staff working on the digitization of the Libraries’ nineteenth-century materials, had just become vacant upon the completion of that programme. This gave considerable scope for relocating staff. The Clarendon Building was already in full occupation but four offices in it were occupied by tenants to whom the Libraries had offered accommodation. Coincidentally, one of the tenants, representing a European organization, had just left to take up a new post so we asked that his successor be accommodated at another institution as part of the natural process of reappointment. Another was an academic project nearing the end of its lifespan. The third, a long-term non-University local history research project, relocated to the county archives. The fourth office was occupied by staff at the nearby Museum of the History of Science, and they remained in the building.

4 Chronology of the moves The initial move of special collections material and associated staff to the Radcliffe Science Library has been described in chapter 4. The maps, music and Oriental collections continued to be consulted in the New Library Reading Room. The plan now was to move them into the space in Duke Humfrey’s Library in the Old Bodleian Library partially vacated when the Special Collections Reading Room had opened in the Radcliffe Science Library. In broad terms, the sequence of staff moves was as follows – Reader services staff: After the move of special collections to the Radcliffe Science Library, the New Bodleian contained just one large reading room, for Oriental material, maps and music. These moved out in September 2010, enabling us to close the New Bodleian to readers. Also in the building were Reader Services staff working on back-room services such as inter-library loan and photocopying. The closure to readers enabled considerably more freedom of movement in the building during its final months; – Collections and Resource Description staff: Moved out at the end of September 2010, as soon as Osney had been prepared;

Chapter 12: Staff relocation 

–

–

– –

 165

Conservation and Collection Care and Imaging staff: followed Collections and Resource Description to Osney in October 2010; some elements moved to the Clarendon Building; Special Collections staff: Many continued to work on preparing the collections for barcoding and book moves, and moved out of the building piecemeal in the first half of 2011; Facilities Management staff: remained onsite until immediately before the building closed; Project teams: these were working on the collections until the building was empty.

5 Staff management and communications A survey of staff was carried out to establish their main concerns. Transport to and from Osney was a major worry, both for staff due to be based there entirely and for those expecting to move between Osney and the central site during the working day. Staff relocating to Osney were invited to visit the building to see their new offices. Where entire departments were moving to Osney, it was relatively easy to fix the timetable for relocation and to release it to the staff. The remaining moves, to a variety of smaller spaces, proved more problematic. Although we had hoped to announce the final timetable early in the process – and certainly during the first part of 2010 – in the event this proved more difficult. It was frustrating for our staff who knew that they must move, but to whom we could not provide firm information. One reason was availability of people – the architectural team was at full stretch and could not turn their attention to planning the later stages before the earlier stages were complete. The other reason was that in the early stages no single person was responsible for delivering the relocation programme. The programme was part of the wider New Bodleian refurbishment project – as a necessary precursor to it – but did not have dedicated teams of staff as did, for example, the bookmoves programme. The bulk of the work was either building work – arranged by the University’s Estates Department – or actual moving, which was contracted out to commercial firms. Individual staff members were asked take responsibility for co-ordinating actions as an additional task to their normal work, but did not feel that they had authority to act on their own initiative. Planning meetings were held at which we brought the various interest groups together, but there was no central management of actions. This situation persisted through the whole of the planning period from September 2009 to June 2010 and was not resolved until,

166 

 Michael Heaney

first, a dedicated project manager was appointed in July 2010 and, second, he was given a reporting line direct to senior staff, and thereby a voice on the Bod Squad. The project manager spent much time working individually with the staff involved, planning the logistics of each of the dozens of individual moves. Staff were also concerned about transport. A minibus service already ran between Osney and the centre of Oxford at peak hours, and we agreed to extend the service so that it ran through the day from 7:30 am to 6:10 pm, stopping at the railway station (halfway between the sites) on request. All staff remaining on the central site, not just those vacating the New Bodleian, were concerned about refreshment facilities. Staff in reading rooms and public spaces observed strict rules about eating and drinking, and saw the staff restaurant in the New Library as an essential facility both for refreshment and to relax out of the public gaze. At the same time the provision of refreshment facilities for readers was a much desired improvement in services. These two requirements complemented each other well. The smaller number of staff remaining on the central site made a service only for staff uneconomic. Extending the service to readers meant that it was feasible to continue the service to staff. By moving a small number of staff from the Old Bodleian to Osney we were able to allocate two adjacent rooms as a tea room – the main refreshment room for everybody, with a smaller separate room for staff only. The same provider was also asked to provide refreshments at Osney: the building already had professional kitchen facilities, but these had not been used since the Libraries took the building over in 2006.

6 Preparing the buildings The first stage of the main move was to prepare the Osney One building for its new occupants. The Systems & e-Research Service, already occupying the first and second floors of the building, vacated the first and reconfigured their offices entirely on the second. The ICT Support Team, who provided network and desktop support for the Libraries, had premises both at Osney and in the New Bodleian. They needed to remain close to the staff to whom they provided support, which implied retaining a base in central Oxford. The University found alternative premises for them in the middle of Oxford, which had the advantage of uniting the Team on a single site for the first time and freeing library space on the second floor for the Systems and e-Research Service. We also paid attention to the building’s infrastructure: was the data network and the link to the University network spine robust enough and capacious

Chapter 12: Staff relocation 

 167

enough to support the activity of all the new occupants? Initial plans to move the Conservation and Collection Care department to the east wing of the building were changed when we realized that most of the spare water capacity – necessary for the new conservation workshop – was located in the west wing. Staff had expressed concern about the cloakroom facilities, so additional facilities were included in the building alteration plans. The first estimation of the space required used standard office allowances for desk spaces, but members of the Collections and Resource Description department needed more space for the constant flow of books being processed. When work on the Osney building got under way in June 2010 the existing occupants of the building had to endure noise and disruption. This was a difficult situation: the systems staff’s work was essential to the continuing progress of the bookmoves and barcoding. A plan of work had been drawn up but as the construction progressed the builders began to adapt it to meet changing circumstances. Staff complained of the unexpected changes. Again, communications were the key. We set up regular meetings between the site manager for the construction and a nominated senior member of staff. Weekly emails were sent to all staff in the building notifying them of progress and of changes, with midweek updates if necessary. All staff moving to Osney were given a welcome pack on arrival, providing local information and a guide to the building. The Collections and Resource Description department moved to Osney over a weekend at the end of September 2010. For them this was a permanent move: the department would remain at Osney even after the New Bodleian’s reopening in 2015. Although some office furniture moved with the department where necessary, the opportunity was taken to re-equip with new furniture in most cases. Again, standard estimates were insufficient to cater for the amount of material to be packed for moving; 880 crates were used in the move, against an initial allocation and delivery of 96. Packing crates were delivered only a few days before the move, so the deficiency had to be remedied quickly. Some staff expressed worries about the health and safety risks of packing until the Libraries’ Health and Safety Officer came over to go through the process with them. Building works at Osney overran by a week but not so as to require us to defer the move. Heating was not available in the east half of the building for the first two weeks. The new catering service was operational from the day of the move, but because of the building works only cold food was available for the first week. The Conservation and Collection Care department moved to Osney in the first half of October 2010. This completed the two largest moves. Subsequent moves were all on a smaller scale, and the emphasis was now more on the logistics of co-ordinating the staff moves with the requirements of the other projects.

168 

 Michael Heaney

7 Smaller moves The projects which impinged upon the further moves were the barcoding project and the Bodleian Quadrangle changes. Staff in Special Collections had to remain close to the collections in the New Bodleian in order to advise on barcoding. Facilities Management staff and staff working on the commercial operations had to await the completion of stages in the reconstruction of rooms around the Quadrangle. Many of the smaller moves were plotted on a spreadsheet which related the area of each room to the occupancy. (Figure 12.1) The spreadsheet automatically calculated the current personal space each member of staff had, based on room area and occupancy, and calculated the new occupancy and personal space as staff were re-allocated to new offices. Several staff working in small offices either alone or with one other had now to share larger offices in the rooms vacated by tenants in the Clarendon Building. Care was taken to keep departments together as far as possible, and not to assign staff from different departments to the same new shared space. In one case in the Clarendon Building we found that the resulting allocation was too overcrowded. An office that had been earmarked as a hotdesking room for staff visiting from Osney was redesignated as further office space; the hot desk was moved to a desk in another office already occupied. The Historic Venues and Retail Operations teams had shared offices and staff in the Old Bodleian, but the Retail Operations also had a store and mail-order department in the New Bodleian. The mail-order operation was outsourced offsite to eliminate the need for space, and the staff moved into the Old Bodleian premises. The Historic Venues and Retail Operations staff themselves had to move within the Old Bodleian in order to accommodate the staff and readers’ tea room. The timing for the fitting out of the tea room was the first three months of 2011. However, late in 2010 it became evident that plans for the Facilities Management staff were inadequate, and this impinged upon the Old Bodleian plans. Facilities Management operated from a base on the ground floor of the New Bodleian, where deliveries were received and sent. There was a staff office of 20 square metres housing the senior team members, and a general room of 43 square metres which acted as a delivery and despatch hub and as a base for twenty further staff members on duties throughout all of the central site’s buildings. We had intended that the senior team members should relocate to a room in the basement of the Clarendon Building, where they would be joined by the two maintenance staff. (A third member of maintenance staff was redeployed as the workload diminished with the closure of the New Bodleian.) (Figures 12.2, 12.3)

Chapter 12: Staff relocation 

 169

Figure 12.1: Staff moves spreadsheet (adapted)

There were several deficiencies in the plan: – The designated room was not close to a roadway for receipt of deliveries; – It was too small to accommodate both facilities and maintenance staff; – It was already in constant use as a meeting room, the only one available to staff on the central site ; – No provision had been made for a base for the twenty further staff members. Furthermore, the implications for Facilities Management of moving large numbers of staff and operations from the central site to Osney had not been fully worked through. At the main entrance to the Bodleian Quadrangle (also known as the Old Schools Quadrangle) was a room which was in use as little more than a passageway: it consisted of a vestibule leading to a small, old lift to the reading rooms above. The lift was in temporary use for book deliveries between the closing of the conveyor belt between Old and New Bodleian in August 2010 and the opening of a large new lift in the Old Bodleian as part of the UBOB project in April 2011. The room was 23 square metres but the reduction in size was acceptable as so many activities had relocated to Osney One. We agreed that this should be refitted as a base for Facilities Management staff and as a reception point for deliveries. This work had to be fitted into a three-month window between the completion of the new lift (so freeing the room from use for book deliveries) and the closure of the New Bodleian. Nearby was an office of 22 square metres occupied by two Imaging Services staff who had remained when the rest of the department moved from the New Bodleian to Osney. This had kept them closer to the source for much of the material used by Imaging Services, the special collections in the Radcliffe Science

170 

 Michael Heaney

Figure 12.2: Clarendon Basement before the moves

Figure 12.3: Clarendon Basement initial plan

Library. We determined that this office would be better used for Facilities Management senior staff, because their need for office space on the central site was absolute. Space was found for the Imaging Services staff at Osney, but this was less convenient for access to the collections and the move was made only reluctantly. (Figures 12.4, 12.5)

Chapter 12: Staff relocation 

Figure 12.4: Bodleian (Old Schools) Quadrangle before the move

 171

Figure 12.5: Bodleian (Old Schools) Quadrangle after the move

This office had a doorway direct to the Quadrangle, which was not currently in use but had to be opened in the new office configuration. To protect the office from draughts and the weather a small new vestibule was needed, but because of the importance of the building even this small change needed Listed Building Consent from the city council before work could start, and the process took most of the three-month window. The Support Services Manager occupied an office in the basement of the Clarendon Building: he moved into a hotdesking room nearby which had been made available to staff and the maintenance staff moved into his vacated office. This enabled us to keep the meeting room, at the cost of the loss of the hotdesking space: analysis had shown that the room had only been in use for 25 % of working time, so the effect of its loss was minimal. (Figure 12.6) Facilities Management staff moved their base of operations in the third week of July. They were among the last groups of staff to leave the building. Another group of late leavers was the barcoding staff. Most barcoding staff working in the New Bodleian left employment as the barcoding task was completed. Although the barcoding project leader had indicated earlier that some staff would need to continue, based in Osney, in the second half of 2011, how many staff would need workspace had not been determined until near the end of the New Bodleian operation, when it became clear that three extra staff workspaces would have to be found. Barcoders working on the other bookmove stream were occupying, on a temporary basis, a warehouse adjacent to the Osney building. A potential office space was identified within the warehouse, but the provision of IT services to

172 

 Michael Heaney

Figure 12.6: Clarendon Basement: final configuration

it threatened to delay the move – in the end the ICT Support Team were able to provide the network infrastructure just three weeks before the closure of the New Bodleian, and the staff moved the following week. The very last staff member to move was a member of Special Collections staff who had been working on the barcoding of a particularly challenging set of material until the final week. An office had been prepared and most of the office furniture moved six weeks previously, but the staff member moved on the completion of the barcoding on 29 July, two days before the closure of the building.

8 Furnishings Linked to the staff relocation was the less sensitive, but nonetheless challenging, task of moving some of the furnishing and contents of the building. Furniture was divided into three categories: for discarding, for relocation along with staff, or for temporary removal and eventual return to the building. The New Bodleian was a Grade II Listed Building so there was a duty to retain as much of the original furniture and décor as possible. ‘Heritage’ furniture was therefore identified and put into store. The Bodleian Libraries have a large collection of portraits: many hung in the New Bodleian, and had to be expertly rehung in other buildings. One item proved especially problematic: filing cabinets used by Sir Winston Churchill during World War II and afterwards. These were strongly reinforced so as to be bombproof, so were extremely heavy. Their contents had long been removed to

Chapter 12: Staff relocation 

 173

the Churchill archives in Cambridge, and we agreed that the cabinets should be donated to the Conservative Party. The problem was that the cabinets were on the lowest floor of the building, three floors below ground, and initial calculation indicated that they could be too heavy for most of the lifts. Moreover, the strongest lifts were in constant essential use for the bookmoves. The specialist removers responsible for the bookmoves brought in special equipment to weigh the cabinets on site, and in the end we were able to extract them; but the whole process took several weeks.

9 Conclusion The staff relocation project was not a standard office move in a variety of respects. First was the interlinking with the other projects. Second was the movement to a variety of locations, some requiring preparatory work and/or a chain of moves. Third was the need to accomplish the moves with little or no disruption in service. The first lesson – obvious in retrospect, but not obvious when this was just one element within a larger project – was the need for a dedicated project manager with logistical skills. Until a manager was appointed the project struggled. Direct input to the Bod Squad for inter-project communications and decision making were also vital. Second, staff communications were crucial. Even where we could not be certain about timings and locations, staff needed to be told as much as possible about the current state of play. Communications needed to be frequent wherever activity was under way – weekly updates were very useful while building works were in progress. Communications included staff visits and surveys as well as news distributions. The staff survey was very useful in uncovering staff worries and concerns. Third, preparation of infrastructure: in our case this involved third-party elements within the University, not just the contractors, and it was important that they shared the sense of critical timing and urgency. In addition, it was important to allow for the greater than normal storage needs of most library staff. Finally, the planning had to be accompanied by frequent walk-arounds of both old and new premises, not just with the contracting staff for the preparations and move, but with the affected staff themselves. It was important to recognize that moving office is a disruptive experience for most staff and can be a source of stress.

Richard Ovenden

Chapter 13: The creation of the Weston Library 1 Construction of the New Bodleian Library The New Bodleian building was officially opened by King George VI on 24 October 1946. This event was the culmination of over 30 years’ efforts to deal with the perennial issue facing the Bodleian: that new books would arrive in the library faster than the buildings had the capacity to cope with storing them. After much soul-searching during the late 1920s and early 1930s, the University decided against the building of a large integrated library building on a green-field site on the University Parks, and instead decided to acquire a plot of land much closer to the old Bodleian buildings in the heart of Oxford.¹ A group of old buildings on the corner of Broad Street and Parks Road was purchased and demolished to make way for a new building which would primarily serve as a book depository serving the reading rooms in the Old Bodleian building and the Radcliffe Camera. A fundraising exercise was organized and a pledge was provided by the Rockefeller Foundation to provide three-fifths of the cost of the building, and some associated projects, and a more public campaign was launched to provide the remaining sum. The architect chosen at this time was Sir Giles Gilbert Scott, the latest scion of a dynasty of architects, who had already been commissioned to furnish the University of Cambridge with a new library building modelled on the scheme which Oxford rejected: that is to say a single, large integrated building, outside of the historic centre, with large space for growth. Bodley’s Librarian Sir Edmund Craster spent considerable time visiting new library buildings in the USA and in Europe together with Scott, and they jointly developed a scheme that created space for storage of collections in a combination of large underground floors, with a single ‘ziggurat’-style above-ground tower in the middle of the building providing a total of eleven floors of storage. Wrapped around the tower on three floors would be a mixture of reading rooms, staff offices, and other spaces for backroom activities. Crucially, the book storage spaces would focus on efficiency, integrating the building’s structure with the bookshelving system.

1 Library provision in Oxford: report and recommendations of the commission appointed by the Congregation of the University. (Oxford: Clarendon Press, 1931).

176 

 Richard Ovenden

Figure 13.1: The New Bodleian under construction

As the planning and design work evolved, Tudor houses and shops were purchased by the University and demolished to make way for the new building. A large empty space soon emerged at the end of Broad Street, and the opened ground was dug out, to enable the formation of the underground storage. (Figure 13.1) The superstructure was put in place by May 1939, and as Britain descended into the early years of the war, the New Bodleian building was completed and fitted out, although it was not occupied as intended by the Bodleian. Admiralty Intelligence occupied parts of the New Bodleian, working on the Bodleian’s map collection and analysing postcards and tourist photographs to prepare for the invasion of Europe; other war purposes also trumped the Bodleian’s own desire to move its own collections from the overcrowded old Bodleian buildings into the New Bodleian. Following the end of the war, however, the building was fully handed over to the Library and the King officially opened the building on 24 October 1946. The building was soon full of staff and collections, however, and a process of change began where areas such as the exhibition room had to make way for

Chapter 13: The creation of the Weston Library 

 177

increased spaces for consultation of collections. Fortunately the flexibility of Giles Gilbert Scott’s design for the outer areas of the building permitted this evolution of function. As the library world began to become more aware of the science of book and paper conservation and related issues, the understanding of the limitations of the New Bodleian also become better understood. The lack of climate control became a major issue, and a retrospective installation of a system was made in the 1970s; but because this had not been planned in the building’s original design, the infrastructure for this system – both the mechanical plant and the pipework delivering water to and from the plant – was placed among the collections themselves, creating an additional risk to the safety of the books and manuscripts stored in the building. In addition to the need to maintain a stable environment for the longterm preservation of the collections stored in the New Bodleian, the Libraries also became aware of the inherent difficulties built into the structure of the building in terms of fire safety. The unprotected steel columns that formed the main structure of the building, combined with the lack of any fire suppression system and the introduction of the air-handling plant into parts of the New Bodleian created a potential tinder-box, giving serious concerns to the Oxford fire department and to the University’s fire safety office.

2 Overcrowding The New Bodleian was designed, in the main, to serve as a storage facility, albeit one very close to the historic reading rooms of the other Bodleian central buildings. By the 1970s, the New Bodleian’s capacity had been outstripped by the rapid increase in the volume of publishing, and the growth in particular of the legal deposit collections. In 1975 the Bodleian built its first off-site storage in the gardens of the old house at Nuneham Courtenay, just under 9 kilometres (5.5 miles) south-east of the Bodleian. Storage modules were added over the subsequent decades, the last being completed in 2002, with the capacity there being in excess of 1.2 million volumes at its peak. The failure to receive planning permission for further storage modules at Nuneham Courtenay created something of a space crisis for the Bodleian, with growth space in the Bodleian soon filled to overflowing. Moreover, with increased focus from the University on health and safety, the fact that collections were being stored on the floors of the storage levels in the New Bodleian became a serious issue, which influenced the broader estates strategy that was to emerge in the period 2004–05. The immediate solution to the space crisis was, however, to rent commercial storage, first from the

178 

 Richard Ovenden

company Iron Mountain in a warehouse in Gloucestershire, then in salt mines in Cheshire run by a company called DeepStore. At the peak of the Bodleian’s use of the DeepStore facility, there were over 2 million volumes stored there.

3 Environmental conditions The Historical Manuscripts Commission made routine and regular inspections of archival repositories which held public records under the 1958 Public Records Act,² and following the transfer of the HMC to The National Archives, to become the National Archives Advisory Service, the routine inspections were assessed against a new Standard for Record Repositories, based on the British Standard 5454, but including additional elements.³ The inspections in 1999 and 2004 were especially sobering for the Bodleian as it became very clear that the Bodleian’s status as an approved repository was seriously in jeopardy, with the New Bodleian Library falling short of meeting the minimum standards in a number of key areas. The view of the TNA could be summed up as follows: the Bodleian had skipped a generation in the upgrading of its facilities for storing and preserving its outstanding collections. Something must be done to improve the conditions or a major disaster might place them at jeopardy. With the New Bodleian effectively full, and the collections stored there being subject to a variety of issues such as overcrowding, risk from fire, and lack of proper environmental controls, the University began to suggest to the Bodleian that a broader rethink of the estates strategy should be pursued.

4 NEWBOLD The idea began to emerge, following the decisive intervention of the incoming Vice-Chancellor, that the Estates strategy of the Libraries should change. Instead of incremental growth of off-site storage, they should build a large single store on a new site with space for the existing off-site collections, space for substantial growth, and space to accommodate the collections of the New Bodleian during

2 Public Records Act 1958: 1958 Chapter 51 6 and 7 Eliz 2, accessed 29 June 2012, http://www.legislation.gov.uk/ukpga/Eliz2/6-7/51. 3 Standard for Record Repositories, 1st edition (Kew: The National Archives, 2004).

Chapter 13: The creation of the Weston Library 

 179

the refurbishment of that building. This was a critical new element in thinking through the Bodleian’s problems, as it had become apparent that to undertake any major work in the New Bodleian stacks would require the emptying of those stacks. The cost of building temporary storage to do this would be wasted. A more efficient use of these resources would be to invest them in a more permanent solution which could be used for growth after the New Bodleian collections returned after its refurbishment. The subsequent development of this approach and especially the creation of the Bodleian Book Storage Facility is told elsewhere in this volume (chapter 6). With the space problems alleviated by the temporary measure of DeepStore, and the long-term issues resolved with the emergence of the new large store (to become known as the Book Storage Facility, or BSF), the focus of attention turned to the nature of the New Bodleian refurbishment itself. The earliest scheme, known internally as NEWBOLD, had focused on the refurbishment of the stacks, and became the major component of the Bodleian’s Capital Campaign, launched publicly in 2002 as the 400th Anniversary Campaign. The Campaign had failed to raise the significant sums needed to progress the refurbishment by the start of 2005, as many donors who were approached felt that the University itself should be investing in the core physical infrastructure of fire prevention and suppression, and climate control. The Bodleian Libraries were therefore given the task by the Vice-Chancellor of rethinking that scheme.

5 The new New Bodleian The concept that emerged was to combine the core element of renovation of the book stack with two new elements that were missing from the NEWBOLD scheme or not initially given prominence within it. The first element was to provide radically improved facilities for scholarship using the special collections. These would include seminar rooms for using the special collections in teaching or discussion sessions in a safe and secure environment; spaces for high-end digital scholarship, and integrated access to a major new suite of conservation laboratories and workshops. The second element was to increase dramatically in size and improve in quality the facilities for sharing the special collections with a wider public, through such elements as exhibition spaces and an auditorium. The Bodleian’s staff began the process of building up a briefing document, in combination with colleagues from the Oxford University Estates Directorate and with a consultant from the international consultancy and construction company Mace. Colleagues within the Bodleian, guided by the Bodleian’s own Estates

180 

 Richard Ovenden

Projects Officer, began to gather information to draft a design brief. This initial design brief was based on the information gathered for the NEWBOLD project, together with new information based on the radically revised scope of the project. A local firm of architects with experience of working in the constrained environment of central Oxford, Beuman Geudes Stretton, was employed to work up some initial designs in order to inform the internal planning process. At the same time a number of workshops were held with academic members of affected faculties such as History and English, and a number of key planning decisions were taken in conjunction with relevant Faculty Boards. One such was the decision to integrate the Bodleian Library of Commonwealth and African Studies at Rhodes House into the New Bodleian, enabling the sub-Saharan collections to join those of the Indian Institute, forming a ‘Commonwealth and Empire’ collection of considerable size and depth. Another key step related to obtaining a detailed understanding of the heritage nature of the building and its contents, especially important considering the architectural significance of the setting, the fact that the building had been designed by Giles Gilbert Scott (one of the most important architects of the twentieth century), and the crucial fact that the building had been given Grade II Listed status in 2002. During the second half of 2005, therefore, the University employed conservation architects Purcell Miller Tritton to undertake a detailed conservation survey of the New Bodleian building, in order to understand fully the key heritage elements, and to produce a gazetteer of the building which would inform any future design work. These initial designs and the draft Design Brief were considered by internal University planning committees and approval was given, in the autumn of 2005, to proceed toward an architectural competition and the commencement of the formal Royal Institute British Architects (RIBA) planning process. The Design Brief was taken to a new level of detail and comprehensiveness by a large team of staff working during the latter part of 2005, and the brief was signed off by internal Bodleian and University committees. The Estates Directorate formally issued the invitation to tender in the European Journal in January 2006. Almost 100 entries were received and an internal Project Sponsor Group was formed, chaired by the Pro-Vice-Chancellor for Academic Services and University Collections. In spring 2006 four architectural practices were invited to work up proposals and to come to Oxford to present their schemes, which they did in May 2006. The project was awarded, following a unanimous decision by the Project Sponsor Group, to Wilkinson Eyre Architects. Their approach was considered the most innovative use of the constraints of the building, combined with the most sensitive understanding of the issues. Key to the scheme which they presented was to shift the entrance from the small existing door on Parks Road to the main

Chapter 13: The creation of the Weston Library 

 181

Figure 13.2: Illustrative view of planned Broad Street frontage of the refurbished New Bodleian Library, courtesy of Wilkinson Eyre Architects

Figure 13.3: Illustrative view of the Blackwell Hall, courtesy of Wilkinson Eyre Architects

Broad Street frontage, where the elevation at street level would be opened up and made accessible by a series of steps and inclines. (Figure 13.2) Inside the building would be a large central atrium as entrance hall, which would allow for circulation by readers accessing research facilities and by the public coming to view exhibitions. This large and welcoming space would be lit by a series of shafts of natural light made possible by creating a sense of the central stack space ‘floating’, although in reality it would be cantilevered off two large vertical ‘cores’ which would also carry common services as well as lifts and stairs (Figure 13.3).

Figure 13.4: Illustrative view of zoning in the redesigned New Bodleian, courtesy of Wilkinson Eyre Architects

182   Richard Ovenden

Chapter 13: The creation of the Weston Library 

 183

The Wilkinson Eyre scheme also had a number of compelling attributes. The three-part zoning of the building was an excellent response to the brief: collections storage largely confined to the three large floors below ground; public access on the ground floor, and spaces dedicated to researchers on the upper floors. (Figure 13.4) The design was also favoured for a number of other reasons. Firstly, it referenced the classic Oxford topography of the quadrangle in the way that the central hall was separated from the exterior rooms. The design also brought into the New Bodleian the idea of the book galleries which are found in the 17th-century extensions to Duke Humfrey’s Library known as Arts End and Selden End, to make it abundantly clear to all entering the building that they were coming into a library, and a great library. The key fact of opening up the Broad Street frontage, to send a clear signal that the library was a welcoming place, was also of critical importance in flagging a change in philosophy within the University. Although Wilkinson Eyre were the clear winners of the competition, the Project Sponsor Group were very impressed by the submission from Shepley Bulfinch Richardson and Abbott (SBRA), a long-established architectural practice based in Boston, Massachusetts, mostly because of their excellent reputation for undertaking major library refurbishment projects in major research libraries in the US, such as those at Yale, Columbia, Duke, and Cornell. With their knowledge of trends in major US research institutions, it was agreed to include SBRA as a consultant to WEA on the project’s design phases.

6 Planning the new library Following the appointment of Wilkinson Eyre Architects, the project moved into a new feasibility stage with their ideas being worked up into more formal preliminary drawings. A professional design team was formed, with Hurley Palmer Flatt providing expertise on mechanical and electrical engineering; structural engineers Pell Frischmann, and cost consultants E.C. Harris joining Wilkinson Eyre Architects, the Oxford University Estates Directorate, and colleagues from the Bodleian. In spring 2007 the Design Team undertook a study visit to a number of libraries in the USA that had been identified as having elements which would be useful points of reference for the New Bodleian project. The visit began in Austin, Texas, with a day-long visit to the Harry Ransom Humanities Research Center at the University of Texas at Austin. Here the Team noted the mixture of public and research spaces, especially the new auditorium and large and flexible exhibition spaces, as well as the comprehensive suite of carefully integrated seminar rooms for group study and teaching using the collections.

184 

 Richard Ovenden

The team then moved to New York to visit the newly completed redevelopment of the Morgan Library and Museum, designed by Renzo Piano, where close attention was paid during the visit to the suite of exhibition galleries, and the beautifully appointed conservation workshops with their flexible design and installation. The team also visited in Manhattan the New York Public Library, where they saw the impressive facility for Visiting Scholars. The team then went by minibus to New Haven where they were joined by colleagues from Shepley Bulfinch Richardson and Abbott to view Bunshaft’s inspirational and influential Beinecke Library, as well as the refurbished Sterling Memorial Library which had been recently completed by SBRA. Particular attention was paid during the visit there to the Music Library, as this was an important element to consider as part of the New Bodleian project. Also visited in New Haven were the two great Louis Kahn buildings: the Mellon Centre for British Art and the Yale University Art Gallery. Finally, the team moved on to Boston where the Boston Public Library was visited before a workshop was held at the offices of Shepley Bulfinch Richardson and Abbott. Bodley’s new Librarian Dr Sarah Thomas, who had worked previously with SBRA, joined the workshop. Other study visits were made in the UK during the design phase, including a number of visits to the British Library building, but also to The National Archives, the Royal Institute of British Architects, the Victoria and Albert Museum and the Museum of London. In Oxford a number of college libraries were visited, as were the new auditoria at Corpus Christi College and Merton College, and Giles Gilbert Scott’s chapel at Lady Margaret Hall. As the design moved through the stages of the RIBA scheme, updates were made regularly to the Project Sponsor Group and to the University’s Buildings and Estates Subcommittee, Capital Steering Group, and Planning and Resource Allocation Committee, especially as funding was only being released to allow each phase of the design to proceed to the next, and no guarantees were being made about future progress, until University support and philanthropic funding was secured.

7 The scheme and the strategy During 2008, as the detailed design for the New Bodleian was being worked up in detail through numerous iterations and planning workshops, the Bodleian’s broader estates strategy went through major change, as the putative scheme to build a new off-site book storage facility at Osney Mead was rejected by local planners and by the appeal process, forcing the University to seek land further afield, in Swindon (as is recounted in chapter 6). The new version of the estates

Chapter 13: The creation of the Weston Library 

 185

plan involved alterations to the evolving scheme for the New Bodleian, principally the removal of the book conveyor from the scheme, allowing the budget which had been allocated to that element to be given over to other parts of the project; and space on the ground floor, which had been earmarked for sorting deliveries of books from the Book Storage Facility prior to being placed on the conveyor, to be utilized for other purposes. In order to retain some element of the conveyor in the New Bodleian, one of the conveyor stations (used for receiving and delivering books) was selected for retention in the scheme on the basement level.

8 Detailed planning The evolution of the detailed design proceeded in stages throughout the period 2008-10, with a number of key steps forward. One of these was the design by Hurley Palmer Flatt of the air-handling system. The control of temperature and relative humidity, and the need to adhere to BS 5454 and the TNA Standard, was of critical concern, and a system was worked up through a 3D computer model based on a series of plenum walls in the storage areas that would pull air through the shelving system at the required rate, avoiding pockets of stagnant air, and keeping the number of air changes to within the accepted rate. This computer model was worked up into a full-scale mock-up, which was tested through the close involvement of the Bodleian’s Conservation and Collection Care department. The mock-up had a section of shelving loaded with books and with sensors located in among the shelved area. The system was run for a period of months, with the data revealing the accuracy of the computer model, and allowing this element of the design to move forward with the confidence of all parties having been gained. The ‘floating’ stack in the middle of the interior hall was also the subject to much modelling and detailed attention to structures by Pell Frischmann. The design evolved to increase the size of the hall, and reduce the size of this floating stack, with consequent evolution of thought about the activities that would take place within it. As the Bodleian’s own plans evolved, we agreed to give more space to open-shelf collections and to study spaces for graduate students, and less to closed-stack storage. At the same time, Wilkinson Eyre began to devote considerable energies to the detailed planning for the main storage areas in the three underground floors of the building. The scheme which eventually emerged as the favoured one included the creation of a single large storage vault in the middle of the building, which would be separated by mezzanine floors, thus cre-

186 

 Richard Ovenden

ating a single four-hour fire rated compartment. Much attention was given to the method of protecting the remaining steel structure in the perimeter areas of the stack, with a combination of fire-cladding and intumescent paint finding favour with the fire consultants and the University’s fire officer and insurers. This builtin fire protection would be combined with a fire-suppressant system for the first time in this building, using a system that would generate a local ‘fog’ or fine mist of water only in the areas which were directly affected by fire. Other parts of the scheme which evolved in this period were the exhibition areas, which increased in size during the planning, at the expense of the ‘Bibliography Room’, a group of historic printing presses used as a teaching tool, which we agreed to relocate in other library buildings. The designs at this stage also included the addition of a combined staff and reader café in addition to a public café as part of the interior hall. During the Autumn of 2009 a study visit to New York City was made by the Conservation and Collection Care staff, together with staff from SBRA, to pay close attention to conservation facilities so as to inform the planning of the conservation workshops in the New Bodleian. The facilities at the Morgan Library, the Thaw Center for Paper Conservation at the Metropolitan Museum of Art and the Museum of Modern Art’s Paper Conservation Studio were all visited to the great benefit of the planning process. As 2009 progressed it became clear that major planning work was required on other elements of the Bodleian’s integrated estates strategy, so the Estates Projects Officer had his duties reallocated to concentrate solely on the planning for the Book Storage Facility and for supporting the bookmoving activities. He was replaced on the New Bodleian project for an 18-month period by a highly experienced and much admired estates administrator in the University. In the Spring of 2010 the Bod Squad and BSF Teams were also formed to provide integrated planning, monitoring, and implementation for the major projects being undertaken by the Libraries.

9 Funding The focus of attention during 2006-10 was not solely on the evolution of the detailed designs for the building. The project was only permitted to proceed through the RIBA scheme by virtue of success in fundraising. During the early phase of the project, the University, under the leadership of the Vice-Chancellor, agreed to provide £ 25 million of funds from the transfers made to the University by Oxford University Press, as the basis for ‘matched funding’. This was to prove a decisive element in the fundraising. The Bodleian was able to count this contribu-

Chapter 13: The creation of the Weston Library 

 187

tion alongside the gifts that had been forthcoming under the NEWBOLD umbrella (from George Mallinckrodt, Barrie and Deedee Wigmore, Joseph Sassoon and others), indicating confidence that the scheme would be able to continue. Toby Blackwell, the owner of the much-beloved bookselling and publishing business based next door to the New Bodleian on Broad Street, also confirmed his intention to support the project in 2007, with his gift of £ 5 million being announced to much positive publicity at the Bodleian’s annual celebration of benefactors known as Founder’s Luncheon in 2008. This was followed shortly afterward by the announcement of an even greater benefaction, from the Garfield Weston Foundation, whose outstandingly generous gift of £ 25 million – the largest single gift in the Foundation’s history – was forthcoming not only because of the scheme’s inherent merits, but also because of the leverage it gained by the £ 25 million matched funding from Oxford University Press. The Garfield Weston donation was announced as one of the lead gifts at the public launch of the University’s ‘Oxford Thinking’ Campaign at the British Academy in May 2009. In recognition of the donation the refurbished library will be named the Weston Library. Other gifts would follow in the wake of the Blackwell and Garfield Weston lead gifts: George David named the new rooftop reading room after his father, the scholarlibrarian Charles Wendell David, and the Dunard Fund generously named the music reading room in honour of Sir Charles Mackerras. The Polonsky Foundation, Sir Robert Horton (through the Emerson Company), the Scheide Fund, the Lee Foundation, the Headley Trust, and the Zvi and Ofra Meiter Family Fund all joined the ranks of the earlier donors in supporting the project. By the summer of 2012, over £ 70 million had been raised for the scheme, with only a further £ 10 million still outstanding.

10 The decant Throughout 2008, 2009 and the first half of 2010, attention began to be paid to the emptying of the building of both staff and collections, with the risk of not meeting the deadline of providing vacant possession of the building to the main contractor being the top risk on the Client Risk Register throughout this period. In the autumn of 2009 the special collections began to be moved out of the building, relocating temporarily to the basement of the Radcliffe Science Library (RSL), which had been emptied of science materials and refurbished during the first 9 months of 2009 (see chapter 4). Reader facilities for maps, music and Oriental subjects were relocated in the summer of 2010 (maps and music to Duke Humfrey’s Library and Oriental to the Oriental Institute Library).

188 

 Richard Ovenden

Simultaneous with this activity was a major project in itself to relocate the staff that had been occupying the New Bodleian for many years (see chapter 13). With staff safely relocated from the building, the focus of attention turned to the removal of collections. With the open shelf collections already gone, and the major special collections already relocated to the Radcliffe Science Library, the focus for the book moving project was to transport the bulk of the remaining collections to the Book Storage Facility in Swindon. This project is described in greater depth elsewhere in this volume (chapter 8), but it required a great deal of detailed co-ordination, with major planning meetings taking place weekly at the BSF Team Meeting, and at Bod Squad, and by smaller groups of staff on an even more frequent basis. The complications involved the co-ordination of barcoding with book moving, the necessity to have data move in time to update the library catalogue and to be uploaded to the Book Storage Facility Information System. If this process were not followed, the danger would be that books could be moved before their locations had been changed on SOLO, causing problems for readers, and also that books could arrive at Swindon to be ingested without any records for them in the BSFIS, interrupting the rapid throughput. In addition, the local coordination of staff in the building was important both for health and safety purposes, but also to ensure that the movement of large numbers of books via the lifts in the New Bodleian was well coordinated. Although there were inevitably some problems in the initial periods of working, and from time to time throughout the project, this activity was well organized and staff worked very effectively to deal with the many problems swiftly, as they emerged, showing great flexibility, creativity, and ingenuity. The building was duly handed over to the main contractor, Mace, on 1 August 2011, exactly on schedule. The last elements to be removed from the building were heritage furniture, and historic paintings (which were mostly re-hung in the Clarendon Building and the Old Bodleian Library). Also discovered during this process was a significant amount of material stored in the stacks of the New Bodleian over the years that had been overlooked. These included overstocks of past publications, a large number of artificial Christmas trees and other Christmas party decorations, and pieces of historic wood panelling. All of the non-collections items were appraised carefully, and suitable homes found for each piece, or they were sent for recycling. The handover of the New Bodleian and the provision of vacant possession of it to the main contractor, on time, and under budget, was a major achievement for the Bodleian and for the team of staff devoted to delivering it. It was the culmination of six years work from the re-envisioning of the concept to the commencement of the main project building works.

Chapter 13: The creation of the Weston Library 

 189

Figure 13.5: ‘Bodzilla’ (Simon Bentley)

11 Demolition The demolition phase began with further surveying work, and especially the completion of the final phase of surveying the building for asbestos. The presence of unknown sites containing asbestos had featured as a major risk on the project’s risk register for some time, and it was therefore a relief to begin the removal of asbestos. An unexpected area containing asbestos was found during the demolition phase, but this eventuality had been allowed for through the allocation of resources for contingency. At the close of the demolition phase in June 2012, over 100 tons of asbestos had been removed. The demolition phase commenced with the insertion of a temporary electrical transformer into the former staff canteen, and the insertion of a tower crane into the building, which was subsequently nicknamed ‘Bodzilla’ by staff. (Figure 13.5) The soft strip of listed fixtures and fittings, the removal of asbestos, and the major task of

190 

 Richard Ovenden

removing the metal shelving meant that the demolition of the structure itself did not commence in any visible way until January 2012. But this demolition proceeded at a rapid pace, so much so that the central stack was removed at the rate of a floor per week. The demolition phase was concluded on 25 June 2012 and the construction of the new structure commenced the following day. By this point the project was still on schedule and on budget. Planning was well under way with a team of Bodleian, wider University, and Estates staff working together to ensure that the structure, policies, procedures, and systems were ready for the safe, efficient, and effective occupation of the building following practical completion in July 2014.

James Legg

Chapter 14: Underground Bookstore and Old Bodleian Access Project 1 Brief history of the Old Bodleian Library, Radcliffe Camera and the Underground Bookstore The buildings of the Bodleian Library date back to 1488 when Duke Humfrey’s Library was completed above the Divinity School, whose construction had commenced in 1427. Most subsequent centuries have seen significant development, principally around the time of the founding of the Library in the seventeenth century with the construction of the Proscholium (with ‘Arts End’ above) and the north, east and south ranges of the Bodleian Library which completed the Bodleian or Old Schools Quadrangle, named after the ‘Scholae’ which occupied the ground floor. These buildings, together with Selden End at the west end of Duke Humfrey’s, form what is now known informally as the ‘Old Bodleian’ or ‘Old Library’. The Radcliffe Camera library was built in 1737–49 immediately to the south of the Bodleian Library, but was initially a rival to the Bodleian; it introduced such innovations as scientific books and evening opening hours, since artificial lighting was allowed in this building. In 1861, the books were moved to a new Radcliffe Science Library in the University Museum and the Radcliffe Camera became part of the Bodleian. At this stage, the upper floor was used as a reading room and the lower floor was converted from an open arcade to a closed bookstack to cope with the rapidly increasing flow of acquisitions, necessitating the addition in 1863 of an external staircase and new doorway. The Underground Bookstore was built between 1909 and 1912 underneath the part of Radcliffe Square which lies between the Bodleian Library and the Radcliffe Camera. It served to provide further expansion of the closed-access capacity. Its two levels were filled with the famous ‘Gladstone’ shelving, a type of mobile shelving developed from an idea of Prime Minister William Gladstone.¹ (Figure 14.1)

1 Edmund Craster, History of the Bodleian Library, 1845-1945 (Oxford:Clarendon Press, 1952), 234–35.

192 

 James Legg

Figure 14.1: The Gladstone shelving

Stairs led from the Bookstore to the lower room of the Radcliffe Camera, and a tunnel led to another staircase providing access to the Old Bodleian. The final phase of historic transformation came in 1941 at the time of the building of the New Bodleian Library and the book conveyor from the New Library to the Old. (Figure 14.2) The Underground Bookstore remained a closed stack, like the main bookstack in the New Bodleian, but the lower room of the Camera was cleared and converted into a reading room. The tunnel alongside the conveyor connected with that coming from the Camera. However, this link between the buildings remained closed to readers, who still had to enter each building separately. Apart from some minor refurbishments, this arrangement continued largely unchanged until 2011 and the completion of the Underground Bookstore and Old Bodleian Access (UBOB) Project.

Chapter 14: Underground Bookstore and Old Bodleian Access Project  

Figure 14.2: Old and New Bodleian Libraries and Radcliffe Camera (underground elements in green)

 193

194 

 James Legg

2 Context and strategy: earlier plans for the space – rethinking – interaction with other parts of the strategy After the creation of Oxford University Library Services (now the Bodleian Libraries) from the University’s centrally run libraries in 2000, and the introduction of a premises-related infrastructure space charge (PRISC) a few years later, there was a move to review and rationalize the space occupied by the various libraries which had either developed within their historic sites or had moved around the city with their respective faculties or departments. Perhaps inevitably, there was a tension between trying to continue to run heavily used services within muchloved historic buildings and ‘cutting and running’ to new buildings which could be purpose-designed for contemporary arrangements and legal requirements, or indeed for further transformations. We began to formulate an estates programme for the Libraries whose essential elements were: – the development of a major new complex for the Humanities on the recently purchased site of the former Radcliffe Infirmary; – consolidation of science libraries in the Radcliffe Science Library and of social sciences libraries in the St Cross area; – redevelopment of the New Bodleian Library as a dedicated facility for special collections; – removal of closed storage and library support services to Osney Mead; – and a withdrawal from as many as possible of the other library sites. There was much, sometimes fraught, debate about the future of the Old Bodleian and Radcliffe Camera and of the Taylor Institution Library. The view was expressed at the very highest level of the University that the last should be abandoned entirely and that, in the case of the Bodleian, the Underground Bookstore should be rendered unusable for any purpose. Otherwise the future of the Old Bodleian and the Radcliffe Camera remained unresolved, although a feasibility study had been carried out to look into improving access to the Old Bodleian. With the appointment of a new Bodley’s Librarian in February 2007 and the subsequent development of the Academic Strategy,² a fresh look was taken at the

2 ‘Academic Strategy for Oxford University Library Services’, Oxford University Gazette (26 June 2009), accessed 29 June 2012, http://www.ox.ac.uk/gazette/2008-9/weekly/260609/ notc.htm#5Ref.

Chapter 14: Underground Bookstore and Old Bodleian Access Project  

 195

central Bodleian Library and its role within the Bodleian Libraries. This look was more actively informed by a desire to retain the Old Bodleian and Radcliffe Camera as working libraries and not just to let the site drift into becoming a museum or curiosity. The aspects of the Academic Strategy relevant to this project were generally: – (3) Enhance user services through innovation; – (6) Advance a coordinated estates programme; and specifically: – (1) Increase direct access to high-demand items in the collections; – (4) Improve management of the collections. Specific mention is made of the Old Bodleian, Radcliffe Camera and Underground Bookstore in the Strategy: (under 6) The Old Bodleian and the Radcliffe Camera will also remain at the heart of OULS’s services for readers. These are iconic buildings and OULS is committed to maintaining the very special atmosphere afforded to scholars by the opportunity to study in working libraries in such historic surroundings. (under 1) Further research is being conducted on the New Bodleian collections, with an eye to moving approximately 250,000 volumes, chiefly recently published monographs in the English language, to the Underground Bookstack serving the Radcliffe Camera. The benefits of such a move would be substantial in terms of increasing user satisfaction, eliminating the delays inherent in fetching, and reducing costs. The feasibility of accommodating these titles in a renovated Underground Bookstack is still being evaluated, but the principle of giving priority on open shelves to the most heavily used stock is guiding the Libraries’ strategy.

Enhancing user services through innovation was a pervasive but not explicitly identified goal. This became a benchmark against which to test ideas as they emerged in detail, and a tool to generate new ideas. Improving management of the collections through, for example, barcoding stock and ensuring appropriately detailed cataloguing was essential to allow readers to identify and find both stock moved offsite and stock put on open access. These simple expressions of our objectives overlay a complex web of opportunities, constraints and interdependencies with other elements of the Academic Strategy and its associated estates programme. The core elements of the project became: – removal of the existing 600,000 (estimated) volumes of stock from the Underground Bookstore; – refurbishment of the Underground Bookstore as an open-access space for housing ‘high-use’ material with appropriate access arrangements for readers and a modicum of seating capacity for in-space consultation of the material; – installation of a lift in the Old Bodleian to give access to the two main reading room floors and to the tunnel connection to the Underground Bookstore and Radcliffe Camera; – creation of a Consolidated Service Point in the Old Bodleian.

196 

 James Legg

There were a number of dependencies which affected both overall planning and the timetable, requiring painstaking analysis of figures and timetables. The building and bringing into operation of the Book Storage Facility was the key enabler without which it would have been impossible to clear the closedaccess stock. Conversely, we intended to reduce the demand for requests from the Book Storage Facility by up to 50 % by identifying the high-use stock and transferring it to open access in the Underground Bookstore. This reduction was also essential since the loading bay in the New Bodleian Library (not to mention the conveyor transferring material under Broad Street to the Old Library) would no longer be available in its reincarnation as the Weston Library. The change in the pattern of use of material also had a two-way connection with plans to reduce the number of service points in the buildings by two, from five to three. Installation of a lift was essential to replace the conveyor as the means of delivering off-site material to the Old Bodleian – and would in turn provide the opportunity to make the reading rooms accessible to people with restricted mobility The only viable means of delivering to the Old Bodleian also required vehicular access across the Bodleian Quadrangle. It was fortuitous that the University’s Estates Directorate already had a plan to restore and repave the Quadrangle. Following historical evidence, the level of the Quad was raised to its original level, allowing access to all the ground-floor rooms which lead directly off it. Raising the level allowed the paving to be relaid on to a stronger substratum, sufficient to withstand regular vehicular transit. The staff canteen in the New Bodleian, though to be replaced in due course in the Weston Library, would be unavailable for four years during the rebuilding project and the UBOB Project made good this loss by creating a tea room in the Old Bodleian. This was desirable for staff morale and comfort. At the same time it allowed us to provide refreshments for readers. At the beginning of the project we had not assumed that opening the Underground Bookstore to readers would necessarily mean allowing readers to use the tunnel linking it with the Old Bodleian. As we modified this stance, with potential knock-on effects for other policies, we had to reassess the physical capacity and suitability of the tunnel. The University had been working for some years to achieve compliance with requirements of the Equality Act relating to disabled access and the Old Bodleian and Radcliffe Camera were both particularly problematic, not only in existing layout but also because they are Grade I Listed Buildings so requiring permission for every alteration. Accessibility became another pervasive theme

Chapter 14: Underground Bookstore and Old Bodleian Access Project  

 197

which informed the development of detailed planning around the core elements. The new lift in the Old Bodleian was therefore sized to take readers in wheelchairs. Perhaps the most exciting aspect of the project was that the physical changes would enable us to review and revise policies with a view to providing a better service to readers – the ultimate goal. So, for example, the ability to allow readers to move freely between the Old Bodleian, the Radcliffe Camera and the Underground Bookstore, and the expectation that the majority of readers would prefer to read their selected material in the reading rooms, led to our deciding to permit the movement of books by readers throughout the reading rooms (within certain constraints for older material). Previously, even moving material from one reading room to another in the same building had been subject to special procedures. There were inevitably some losses along the way for those with a nostalgic affection for the buildings. The principal one was the ‘Sovex’ conveyor which had served admirably to deliver stock from the New Bodleian to the Lower Reading Room and the Upper Reading Room of the Old Bodleian for sixty-five years. Transformative technology in its time and a charming auditory experience, the conveyor had become rather unreliable in recent years and the estimated cost of giving it a thorough overhaul or replacing it was approximately the same as installing a new accessibility-compliant lift. It would also be unavailable for the four years required to rebuild the New Bodleian Library. The vertical shaft of the conveyor, up through the north-west corner of the Old Bodleian, was removed to make way for the new lift. There was in fact already a lift in the Old Bodleian which had been inserted into the void space of the staircase in the north-west tower. However, as well as being too small for a trolley, or for more than one person, it served only three of the four main floor levels, bypassing the Lower Reading Room on the first floor. Its removal allowed the staircase to revert to its original form and for the flights between the ground floor and the tunnel to be rerouted to make them more easily passable by readers, as well as affording a magnificent sight line up the central stairwell (Figure 14.3). Finally, the dedicated lavatories for male readers had to make way for access to the new lift on the ground floor and for space for an accessible lavatory. Male readers now share with male staff and, although they still have two sets, women readers and staff also share.

198 

 James Legg

Figure 14.3: Old Bodleian stairwell (picture by Rosario Ivan Glorioso)

3 Increasing open access – high-use material To most libraries, open access is the touchstone of good reader service; the vast majority of stock has been individually selected and paid for, and one day will be weeded from the collection. With its more than 400-year history of legal deposit, a wondrously rich tradition of donation, and long-standing policies of retention and non-lending, the collections of the central Bodleian Library had grown and been managed in unusual ways. Crudely speaking, the provision of open-access shelving had not kept pace with the inflow of new material for at least two centuries and one can see, in the history of the major changes to the central Bodleian Library estate, an expansion of reader space and open shelving dwarfed by the addition of closed storage

Chapter 14: Underground Bookstore and Old Bodleian Access Project  

 199

in the Radcliffe Camera, then in the Underground Bookstore, the New Bodleian bookstack and most recently in the repository at Nuneham Courtenay. Before the current project, the open-shelf capacity of the non-special collections reading rooms in the Old Bodleian was about 92,000 items and the Radcliffe Camera held about another 46,000 volumes. This was complemented by deliveries of any of up to 7 million items from closed storage in various locations. The material already on open shelves was a combination of reference material supporting research in Classics (literary sources), English, History, Patristics, Philosophy and Theology, together with non-lending copies of material selected for undergraduate studies in the same subjects. We estimated that the Underground Bookstore could give open access to up to 280,000 volumes (depending on the eventual shelf layout and the size mix of the material). With this and the opportunity to connect and serve the reading rooms of the Old Bodleian and the Radcliffe Camera, we could transform a pair of primarily closed-access libraries into a single library with up to 500,000 volumes on open access, more than double the previous provision, and put this on a par with the other major research libraries of the University (compared to the Taylor Institution Library with c.450,000 volumes over two sites, c.280,000 each in the Radcliffe Science Library, Sackler Library, Social Science Library, c.260,000 in the Bodleian Law Library). Readers repeatedly expressed their desire for more open access and had identified it as one of the highest priorities in all recent reader surveys, hence its inclusion in the Academic Strategy. So, as well as reducing the number of deliveries from remote storage, the conversion of the Underground Bookstore would be a key enabler in the fulfilment of that strategy. However, success depended on our selecting the appropriate material that readers would actually want to have on open access – the ‘high-use’ material. This identification was problematic for two key reasons. Because of the particularities of the central Bodleian Library’s collection management history and policies, all general (i.e. non special collections) material in closed access enjoyed an equal status and, with some exceptions such as Oriental vernacular material, was from 1988 simply housed by size and then in order of accession. Moreover, the pattern of use of material by Humanities researchers, and even students, was not straightforward to predict by the usual parameters. Academics at any world-class university want to study the recherché rather than the mundane, and at Oxford this approach also characterizes the highly individualized undergraduate and postgraduate taught courses. Nevertheless, some early brainstorming on existing statistics suggested that this was not a pointless exercise. The two major categories of material to explore were identified as material which had already been called from closed access frequently and the ‘new academic intake’. We consulted with the relevant Faculties.

200 

 James Legg

Because most remote requests had been tracked on the library catalogue for the previous twelve years, the Systems and e-Research Service (SERS) was able to provide extensive lists of all items requested according to certain parameters. Use by multiple readers rather than repeated request by the same reader was a better (though still far from perfect) indicator of subsequent demand. From initial trawling of material acquired up to and including 2008, we established that just over 80,000 items had been requested by five or more readers over the twelve years or by three or more readers over the previous three years (to capture material which been acquired more recently). This confirmed the notion that a relatively large number of items is used very infrequently against a relatively small number experiencing repeat requests. The spread of publication date of the material confirmed that material published right through the twentieth century was still in regular use. Sorting the spreadsheets by publication date also established that, in spite of the long historical tail, the pattern of requests was weighted somewhat towards recent publications. It would therefore be worthwhile developing the idea of the new academic intake or ‘jumbo new books display’. It was also fortuitous that the Underground Bookstore would easily accommodate a significant quantity of this material along with the high-use material. With the bulk of fiction and other non-academic material acquired by legal deposit filtered out at an early stage, the remaining intake would be just under 30,000 items a year. We agreed that we could accommodate three years of intake, giving a reasonable period for the material to be on open access. Consultation with the library users’ committees elicited great enthusiasm for this overall approach.

3.1 Consultation – planning Although one of the smaller elements of the academic and estates strategy, the Underground Bookstore and Old Bodleian Access Project had complicated direct and indirect effects on many groups of people and arguably lay at the most sensitive heart of the Bodleian Libraries. Consultation and planning were therefore required on a number of fronts. We had carried out a feasibility study in 2007 leading to some initial improvements in the Proscholium (entrance hall) of the Old Bodleian and feeding into plans for improving access for readers generally and for people with restricted mobility in particular. The architects Purcell Miller Tritton prepared further studies in October 2008 and June 2009. These were updated following advice from the library planning consultants Shepley Bulfinch in September 2009 and further consultation with the University’s Safety Office and its Disability Team.

Chapter 14: Underground Bookstore and Old Bodleian Access Project  

 201

Following approval of the outline plans by the Curators of the University Libraries and the University’s Buildings and Estates Subcommittee in the autumn of 2009, we were granted a project budget of £ 5,336,000 (to which was added £ 406,783 for the associated refurbishment of the Bodleian Quadrangle). At this stage, the Underground Bookstore gained its new name – Gladstone Link – in allusion to the ‘Gladstone’ shelving and in anticipation of its new function of connecting the two historic buildings.
The application to alter a Grade I Listed Building required a public consultation. An exhibition of plans and images showing past, present and prospective arrangements was displayed for two weeks in the Proscholium and an early evening event was organized to which neighbours and other interested parties were invited. There was an inevitable tension between aspiring to keep the buildings in use as a contemporary library facility and preserving a particular point in the development of the buildings, which had already seen a great deal of change throughout their history. We therefore had to demonstrate that any changes were to good purpose, were kept to a minimum and were sensitive and of a high quality in their design. Perhaps the most contentious element was the design of the balustrade to surround the new staircase and lift from the Lower Camera reading room to the Gladstone Link. The architects had produced a very minimal and elegant design consisting almost entirely of glass with the intention of opening up the vista of Bay 1 in which it would be located, whereas a representative of an architectural society was firmly of the opinion that the structure should be panelled in oak to match the surrounding book casing. In subsequent discussions with Oxford’s City Conservation Officer and the architects, the eventual very successful final design involved ironwork, to reflect the adjacent window grilles but also to achieve a degree of the transparency promised by the glazed solution. At about the same time, discussions were held with and tours offered to the Chairs of the relevant Faculties and of the relevant Committees for Library Provision before presentation to the Committees themselves. At this stage we had discussed the broad outline of the plan but still had to elaborate and communicate the implication for each subject affected. In the mêlée of building planning and consultation with readers and external bodies, it is easy to overlook the interests of the staff affected by the changes. Clearly it was crucial to engage the staff to harness their ideas for improving the service plans but also to ensure their understanding of and commitment to the changes. To the great credit to members of staff they engaged positively with the changes. The only change which failed to gain favour and consequently has not yet been implemented was a proposal to introduce an area for drinking coffee and the like and eating cold snacks into the Gladstone Link; this remains on hold.

202 

 James Legg

Thereafter followed a cycle of reiteration in which feedback informed further detailed development of the plans and significant further development was reported out as appropriate. Further conversations were also required with the Safety Office, the Disability Team and others with a particular interest in the unfolding scheme. These included facilities management staff within the central Bodleian Library not involved face-to-face with readers in the reading rooms but nevertheless underpinning service delivery through their management and maintenance of the space.

4 Estates – construction Physical transformation was managed by the University’s Estates Directorate under governance from a Project Sponsor Group. There were already considerable constraints and the buildings threw up further difficulties as construction proceeded. Access to the site was the most problematic aspect and this would affect both the building works and the plans for the biggest single move of books into and out of the buildings since the opening of the New Bodleian, and possibly ever. None of the existing routes out of the Underground Bookstore had sufficient capacity to allow us to empty and rebuild it quickly, and so we decided that the only viable method would be to cut a hole in the roof of the Underground Bookstore, through the lawn of Radcliffe Square, and to build a temporary hoist under a structure providing weatherproofing and security. This structure required Listed Building Consent in its own right. Our Project Manager for Library Logistics had to devise a particular methodology for the site involving timetabling moves around the building work before the removal of this temporary structure. Everyone involved had to manage an extremely complex set of interactions working in various areas of the site – Underground Bookstore and Radcliffe Camera, north-west tower of the Old Bodleian, the entire Bodleian Quadrangle – whilst maintaining access for library readers and visitors and ensuring compliance with health and safety standards. Inevitably the place ‘looked like a building site’ for many months but various measures mitigated this situation. The earlier consultation both within the University and with the public meant that there was a certain level of understanding amongst readers, neighbours and other interested parties and that there was a source of texts and images to explain the works on panels, and to include on the website and in circulations. The hoist was surrounded by a neat hoarding and painted in a subdued colour to blend in as much as a large wooden box could in Radcliffe Square. The movable mesh screening used to separate readers and visitors from excavated parts of the Bodleian Quad-

Chapter 14: Underground Bookstore and Old Bodleian Access Project  

 203

rangle was used to mount information panels and directional information to help people navigate the rapidly changing routes. Some elements of the design were developed fairly late in the project. We had identified the tea room both as an aspiration for reader provision and an essential replacement for the lost staff canteen in the New Bodleian. Various options were considered on the basis of operational need and of convenience alongside early thoughts about where there might actually be plumbing and drainage, and informed by the need to keep the cost, listed building implications and disruption to a minimum. A number of these factors suggested the use of the offices of the previous Bodley’s Librarian and his deputy situated off the Bodleian Quadrangle and not directly connected to the reading rooms. We devised an arrangement for a joint reader/staff servery and seating area as well as a separate seating area and basic facility for staff only, with the plumbing eventually located in the ideal place. There were persistent concerns about the size of the spaces, especially in relation to the use by readers who had never had such a facility before, which made it difficult to gauge the take-up against the adjacent college common rooms and the commercial competition. As so often with this project, the novel elements were the most difficult to plan and assess but have proved the most rewarding once implemented. Timetabling was an abiding concern as the work would take up most of an academic year, and one in which other changes described elsewhere were also proceeding. There was certainly no question of closing. Various elements were pervasive: the Quadrangle repaving generated noise and dust in a space onto which all the reading room of the Old Bodleian face; the construction of the lift not only affected all floors of the Old Library but took out of use one of the two staircases to the reading rooms. Our intention was, of course, to keep disruption to a minimum and to restrict noisy work to the vacations. Somehow this never quite worked out and assurances that only a very severe winter would push the repaving back into Hilary (Spring) Term preceded one of the harshest winters on record. Nevertheless close liaison with the University’s Estates Directorate and the contractors meant that we knew at all times what was going on and why, so we were able to push back where appropriate and to explain to staff and readers why any disruption had occurred. Dust protection was applied extensively but the only area where specific acoustic shielding was used was round Bay 1 of the Lower Camera reading room where the staircase pit had to be substantially enlarged and a concrete box constructed for the lift and new staircase. Readers were in fact surprisingly resilient and became adept at finding the quieter parts of the building, although some were seen working on unperturbed in their usual places of study.

204 

 James Legg

5 Completion and opening Planning and listed building consent had been granted on 24 May 2010 for the temporary hoist and on 13 July 2010 for the main works including the lift, tea room and repaving of the Bodleian Quadrangle. Apart from the hoist and the new staff accommodation and service point in Bay 2 of the Lower Camera reading room which was completed on 10 September 2010, the other elements continued through the academic year 2010–11. The lift in the north-west tower of the Old Bodleian was handed back to the Libraries on 18 April 2011 as was the Gladstone Link (Underground Bookstore), at least in part. The move of books into the Gladstone Link proceeded from that point. 3 May saw completion of Quadrangle repaving and the tea room. 20 June delivered the Radcliffe Camera staircase and lift down to the Gladstone Link and the handover of the remainder of the northwest tower. Finally the removal of the hoist and associated reinstatement work was completed in the middle of July 2011. The need to carry out such a complex set of works in a delicate and peculiar building while maintaining library services throughout was undoubtedly very challenging. Adding an extra layer of operational and service changes, from relatively minor to fairly major, might be thought to be foolhardy. However, some of these changes were intrinsic to the physical transformation and some were key elements in transforming the service. It would have been very easy for the project team to focus on the building changes and to forget that users of the building, staff as well as readers, would not have the intimate knowledge of the changes and their aims. It was essential at all stages to keep the purpose of the changes and the primacy of service at the heart of all discussions. This was useful in constraining design changes from acquiring a momentum of their own rather than relating back to the original, or revised, purpose. The Head of Bodleian Library Reader Services was instrumental in maintaining the strong reader focus and translating the plans into practical arrangements on the ground. All the while, staff kept the service points running and the books coming to readers from closed access. Even with extensive planning and discussion, it had not proved possible, particularly in such a short timescale, to cover every level of detail. The book move proved to be especially difficult as we had to move the books back in while the temporary hoist was still in operation; but until the hoist was removed the Gladstone Link was still a building site. And because the books had to pass through the residual building site, maintaining effective dust proofing proved impossible. Health and safety issues threw up many last-minute glitches. It was unfortunate that a review of previous documentation threw up a previously overlooked concern about using trolleys on the metal grid flooring in the Bookstore

Chapter 14: Underground Bookstore and Old Bodleian Access Project  

 205

– unfortunate as this was one of the main historic features of the Gladstone Link and was instrumental in the ventilation strategy, as well as mitigating the effect of the very low ceilings. Some quick thinking allowed us to revise the suggested solution of reflooring the entire space and we acquired a new set of trolleys with suitably soft wheels and foam padding on the handles. Although an important element, this was a distraction from final preparations: checking the finished spaces, labelling, preparing signs and introductory leaflets, briefing staff, checking arrangements, establishing security. The lesson learned was to ensure that the health and safety issues are thoroughly covered in the early design stages. Nevertheless, the Gladstone Link was opened for business on Tuesday 5 July 2011 with a formal opening by Bodley’s Librarian and distinguished benefactor Po Chung following on Wednesday 5 October.

6 Reader (and staff) response – ongoing development – the future ‘Have you seen the tunnel? It’s really cool!’ was an encouraging comment picked up on Twitter in the first few days after opening, and on the whole feedback on the Link has been extremely positive. It is always interesting and pleasurable to show people around a new or newly refurbished building. In the Gladstone Link, the most striking reaction was astonishment first that such a contrasting space was there and second that the transitions between it and the two historic buildings that it joins are so delightful in their entirely different ways. The tunnel between the Underground Bookstore and the Old Bodleian, previously an evidently leaky and rather dismal affair, was clad in white, black and blue panels cleverly arranged to give a sense of progress down the otherwise featureless 30 metres or so from the new foot of the north-west tower staircase. (Figures 14.4 and 14.5) The staircase from the Radcliffe Camera is an altogether more luxurious affair, stone clad with understated details such as the lighting incorporated in the handrails. (Figure 14.6) Lighting is the key to the success of the Gladstone Link itself. The system mainly uses LEDs so that the fittings consume very little energy and are very compact, which is useful when the ceilings are only 2 metres above the floor. As the Gladstone Link mixes a number of distinctive areas, lighting each of them appropriately was difficult. Beyond the main through route and seating area on the upper level, seats and stacks are intermingled in various permutations. However, extensive use of passive infrared (PIR) and graduated switching

206 

 James Legg

Figure 14.4: The tunnel – before (picture by Nick Cistone)

Figure 14.5: The tunnel – after (picture by Nick Cistone)

Chapter 14: Underground Bookstore and Old Bodleian Access Project  

 207

Figure 14.6: The stairwell to the Radcliffe Camera (picture by Rosario Ivan Glorioso)

achieved two significant benefits. Instead of flashing on rather harshly and distracting readers sitting nearby, the lighting in the stack areas comes on gently and goes off gently again at the end of its cycle. In turn the gentle and irregular change of lighting level serves to give a natural, random feel which disguises the fact that no natural light at all penetrates this underground area. Harmonious contrast is evident in other aspects. The original paint of the Gladstone shelving, the variously coloured backs of some of the alcoves, the bright red ‘air socks’ (fabric ventilation tubes running across the low ceilings) and the strong colours of the loose furniture contrast with the otherwise stark white of the walls, girders and ceilings. The rigid grid of the girders contrasts with the informal layout of the furniture in the main reading area. The overall aim was to give a degree of informality which would cater to readers who might find the formality of the historic reading rooms rather overbearing. In the end, we managed to incorporate 120 reading spaces of various types, from round tables seating one, three, or up to seven people, through shared rectangular tables to a fair number of regular desks for one. (Figures 14.7 and 14.8) It has been interesting to see how often there are two chairs left at a single-person desk suggesting that readers do feel empowered to interact more freely. Quiet conversation is possible without disturbing or outraging

208 

 James Legg

other readers. Anyone who does prefer the historic reading rooms (for whatever reason) is free to collect books from the open shelves of the Gladstone Link and take them there. This brings us to most contentious aspect of the Gladstone Link, the selection and arrangement of material. After years of calls for more material on open access, this seemed like an easy gain in spite of the difficulty of testing our notions of what would be highly used when on open shelves. Although there was in fact widespread delight, it soon became apparent that the subject parameters had been set too broadly and there were for example complaints from scientists who had in fact been in the habit of calling books from closed access to read in the Radcliffe Science Library, but now had to come to the Gladstone Link to find them on the shelves. The parameters have now been adjusted for the ongoing intake to focus more narrowly on Humanities, and the opportunity of the first session of weeding has been taken to bring the existing stock into line. Classification had been a known problem: 20,000 pre-1988 acquisitions classified to the in-house Nicholson scheme and 160,000 post-1987 acquisitions shelfmarked without classification, but with a year designator, size designator and running number. Although most readers are able to work it out (and some positively celebrate the serendipity), there is a significant minority who, probably correctly, identify this as a hindrance. Again, work is ongoing to improve this situation probably by introducing Library of Congress classification to future intake, though we have ruled out a comprehensive reclassification, on the grounds that this collection will change fairly rapidly over time. Not all the details might have been right first time but by concentrating on the underlying principles, the project team was able to create new spaces and connections in the Bodleian Library and Radcliffe Camera, which have allowed a transformation in the management of the buildings and the services provided in the early twenty-first century and should allow subsequent evolutions and transformations to maintain the Library not only as a symbol of the Bodleian’s long and highly esteemed history but also as a working library able to continue to support the scholarship of future generations.

Chapter 14: Underground Bookstore and Old Bodleian Access Project  

Figure 14.7: Informal study area (showing air socks and retained Gladstone shelving) (picture by Rosario Ivan Glorioso)

Figure 14.8: Quiet study area (picture by Rosario Ivan Glorioso)

 209

Donald Mackay, Alena Ptak-Danchak and James Shaw

Chapter 15: Reader services during a time of major changes 1 Services to Readers: Just-In-Time A significant challenge during the period of the moves of library materials was to ensure that our readers continued to have access to as many of them as possible. In response to this, we established a ‘Just-In-Time’ service to assist readers. During the main move activity, between November 2010 and December 2011, over 5,000 mediated requests were processed, with the service achieving a 96 % success rate. Building on the success of the Just-In-Time service, in the late spring of 2011 as the migration of the integrated library system was quickly approaching, a live chat service known as ‘SOLO Live Help’ was launched. The aim of this new service was to provide remote and real-time assistance to readers during the period of catalogue downtime as part of the transition to the new system, Aleph.

2 ‘That’s terrific – thanks very much – what a service!’ – a reader’s comment on the Just-In-Time service in November 2010. A key challenge across 2010 and 2011 was to ensure that our readers had access to as much of our collection as possible whilst the stock moves were under way. Many other major libraries in this kind of situation have simply closed their doors or have offered very restricted services, but our plan from the start was to offer readers as ‘normal’ an experience as possible during these moves by delivering a ‘Just-In-Time’ service. The library system in Oxford is extensive. At the heart of this thriving but complex network of services are the Bodleian Libraries. Operating across multiple sites, each year the Bodleian Libraries serve more than 65,000 readers – over 40 % of them from beyond Oxford. The vast majority of the Libraries’ print collections have never been immediately accessible to our readers. To obtain an item that was not on open shelves readers or library staff ordered it via an Automated Stack Request System (ASR) for retrieval by library staff. The ASR support team worked behind the scenes to

212 

 Donald Mackay, Alena Ptak-Danchak and James Shaw

address the many complications caused by the disparate nature of the Libraries’ holdings, legacy systems and multiple classification schemes. Following the opening of the Book Storage Facility (BSF) to the Libraries in October 2010, over 7 million items were moved from dispersed stack locations to the BSF or to open-access shelving in our city centre locations. Many items were moved more than once. Inevitably this meant that there would be times when significant areas of our collections would be in transit and therefore unavailable for consultation. Under the leadership of the Project Sponsor, the Assistant Director, Research & Learning Services, a small project group was established to address this issue. The tone was set by our sponsor from the very beginning with this confident statement: Just-in-time delivery service: we will provide a service to compensate for the temporary unavailability of materials. We intend to ensure continuation of current [authors’ emphasis] levels of service, access to stock through the transitional period including, where appropriate, through alternative provisions.

The Just-In-Time project (or JIT as it soon came to be known) was envisaged as the first of a series of new service initiatives that would potentially develop and continue to deliver value-added services after the moves were completed.

3 Planning for Just-In-Time After some initial preparatory work, a small project group was assembled in June 2010. Taking lessons from earlier local moves and from examples elsewhere (such as Imperial College London and Queen’s University Belfast),¹ the basic principles and key work streams of the work were drafted and agreed and a budget was identified. We agreed that a dedicated support team would be established to provide a Just-In-Time service, the main remit of the team being to source and provide alternatives to stock otherwise inaccessible whilst being moved. We agreed also that the readers served by this initiative would include not only current members of the University but also the many thousands of external readers that the Libraries serve each year.

1 ‘Document delivery to your desktop’, Impact: your Imperial College Library newsletter, 16 (Summer 2010), 7, accessed 30 June 2012, https://workspace.imperial.ac.uk/library/Public/ Impact16.pdf; informal conversations with colleagues.

Chapter 15: Reader services during a time of major changes  

 213

The full scale of the challenge was unknown. We had figures for a whole range of metrics including the number of items requested from stacks each day during normal operation and at first glance these seemed intimidating. For example, in 2009–10 348,000 items were retrieved from stack locations for readers to consult. However, we presumed that the possible scale of the problems would be alleviated by a number of factors. For example, the bookmove project was not one monolithic move during which no stock would be accessible at all. Instead, items would be moved in discrete and (in most cases) identifiable blocks of stock, which in theory would be ingested and then made available from the BSF within just a few days. In addition, high-use materials as well as the most recent materials were to be kept in easy-to-access city centre stacks for almost all of the period of the moves. And of course Oxford had recently dramatically increased its investment in online journals and e-books (see chapter 5). Given this uncertainty, the Just-In-Time team had to be as flexible as possible, with a main Project Coordinator appointed but options left open for the recruitment of at least one full-time assistant. Back-up support from the existing Automated Stack Request (ASR) support team was confirmed from the start and staff from elsewhere in the Libraries were drafted in to support the work as required. We identified flexibility in our ability to source materials from as wide a range of sources as possible as being essential to the success of the project. We agreed a wide range of sources of content including Oxford’s colleges and academic departmental collections; Oxford’s growing electronic collections (purchased and digitized); commercial e-book suppliers; on-demand journal content providers; a wide range of available electronic resources such as the Internet Archive; Amazon, Blackwell’s and other print book suppliers; the British Library’s urgent document supply service; institutional repositories; internet archives; and, of course, other major academic and public libraries in our locality. We agreed to exclude special collections materials (in the main pre-1900 materials) from the remit of the JIT team, together with maps and music or materials associated with Oriental studies. Enquiries relating to these types of material would be directed towards the relevant subject and curatorial staff. We worked out a basic methodology and accompanying workflow to deal with reader requests, and a specific category to identify easily items in the catalogue that were in transit. All items being shipped to the Book Storage Facility were assigned the status ‘Book Moves’ and we put a process in place to ensure that any requests submitted for those items were routed directly to the JIT team. We recognized at an early stage that good communication would be key – both with our readers but also just as crucially with our staff across the Libraries and wider Oxford library community. We drew up publicity materials aimed at

214 

 Donald Mackay, Alena Ptak-Danchak and James Shaw

readers in collaboration with the Bodleian Communications Team and distributed them across libraries throughout the University. Prominent notices were placed on SOLO and on the Libraries’ web pages. We published items on the work of the team in the staff newsletter Outline.

4 As It Happened The large-scale book moves started in October 2010 and while recruitment for the JIT project staff was under way the existing ASR team provided enhanced cover and support for readers. The JIT Project Coordinator took up her post in November and started to deal with reader queries directed to her by desk staff, and orders placed on SOLO for items in transit to the BSF. She also provided valuable project support for associated pieces of work relating to the book moves. The two main paths followed by our readers and staff during the delivery phase of the project were as outlined in Figure 15.1. When a reader made a request for an item with the status ‘Book Moves’ the request was routed directly to the Just-In-Time Project Coordinator. She then ran through a number of possible stages and options to source a copy of the item, and contacted the reader with a proposed solution – for example the location of an electronic or print alternative or the offer of a free interlibrary loan. In more complex issues the reader was contacted to discuss a range of possible options. Figure 15.2 provides an overview of requests dealt with by JIT across its year of operation. The spike in May indicates a busy term-time period and the sharp rise in July was the result of the downtime caused by the migration to a new integrated library system (see chapter 11), during which Just-In-Time was used to mediate all requests, not just those in transit to offsite storage. The graph at Figure 15.3 summarizes the main sources of locating additional copies. We fulfilled 72 % of requests either by finding another print copy within the University or by finding access to an electronic copy, through subscription or openly available resources. 14 % required purchase, either through booksellers or inter-library loans, which was subsidized 100 % by the Bodleian Libraries. Overall 96 % of requests could be satisfied with additional copies, which left a troublesome 4 % that eluded our searches. This 4 % comprised unique material, such as theses, and some foreign-language materials, and particularly older or more obscure material from non-UK publishers. In these circumstances the reader was advised to re-apply for the item, which became available usually within two weeks.

 215

Figure 15.1: Just-In-Time workflow

Chapter 15: Reader services during a time of major changes  

216 

 Donald Mackay, Alena Ptak-Danchak and James Shaw

Figure 15.2: Workload of JIT requests

Figure 15.3: Sources of alternative copies

Chapter 15: Reader services during a time of major changes  

 217

4.1 Finding electronic alternatives The most effective route for providing alternative copies was to identify an electronic copy with direct access. OxLIP+, the Oxford resource discovery platform for e-journals and full text databases, was the starting point and yielded many results. If subscription e-resources produced nothing the search widened to the internet. By far the most productive of the freely available electronic resources was the Internet Archive (http://archive.org). Other university and library digitization projects, such as Gallica (Bibliothèque nationale de France: http://gallica.bnf.fr/) and Online Books (University of Pennsylvania: http://onlinebooks.library.upenn. edu/) provided valuable access. In total, 12.1 % of all requests were directed to an electronic copy.

4.2 Finding printed alternatives The search extended to printed copies if electronic versions could not be found. Oxford has almost 100 libraries so it is perhaps not surprising that 60 % of the total requests were satisfied by finding another print copy within the University. The Bodleian Libraries alone contain over 2.5 million items on their open shelves. Just-In-Time developed productive contacts both within and beyond the University library system. The college libraries, which normally restrict access to their particular cohort, for the most part agreed to allow access on request. We discovered other local collections, such as the Grandpont House Library who loaned their copy of The past & the present: problems of understanding: a philosophical and historical enquiry which was published by Grandpont House.² Oxfordshire County Libraries, particularly the Oxford Central Library, were also a notable provider.

4.3 Communicating with our readers Direct correspondence with the reader allowed the team to appreciate the immediate positive impact on the University community. For instance, the Just-In-Time team purchased Alan Partridge: every ruddy word: all the scripts, from radio to TV and back by Steve Coogan³ for an English undergraduate. It was the only version

2 Oxford: Grandpont House, 1993. 3 London: Michael Joseph, 1993.

218 

 Donald Mackay, Alena Ptak-Danchak and James Shaw

of the primary text in existence, which was urgently needed for an extended essay worth 10 % of his final degree mark. The following feedback shows the benefit to those readers who did not have time to wait for the material to be ingested: Thank you for all your kind assistance! I’m only back in Oxford for a week or two, so all the alternative resources you’ve sent me are so useful. Please pass my thanks on to your colleagues, who have also hunted down alternate copies. I’m sure I’ll be pestering you with more requests over the week, but I just wanted to let you know how much I appreciate the lengths you’ve been to.

4.4 ILS downtime and migration Until July, Just-in-Time dealt with items in transit to the BSF. On average, about 320 linear metres of material were unavailable at any one time. In July it was necessary to switch off Geac Advance (the old library management system) which meant no material could be requested from offsite. To offset this disruption the bold step was taken to use Just-in-Time to mediate access. The remit was no longer limited to in-transit items but was now extended to the 7 million items held in closed stacks. The methodology remained unchanged but more staff resources were required to maintain the service. The core team was enhanced by a four-month part-time secondment of a staff member from the Radcliffe Science Library. In addition, Just-in-Time ‘champions’ were recruited from existing front-line staff and a new workflow of managing requests devised. The champions were trained to conduct the necessary searches and the central team farmed out requests. The information was fed back to the central team who communicated with the reader. The experience of working as a champion was rewarding for staff and helped to reinforce Just-in-Time skills in libraries. To quote a member of staff, who was seconded to the team: I really enjoyed it. It was very stimulating knowing we had to find a solution for the reader in the quickest and most efficient way, and the order slips just kept on coming! It was quite intense, and I was quite exhausted by the end of each day, but with a good feeling of achievement. I slept very well, and returned each day looking forward to more challenges. I was quite pleased the downtime went on for another week, as I was enjoying it so much!

Figure 15.4 indicates the results during the downtime period. Only two requests (0.4 %) could not be mediated. In one case, the Bodleian copy was the only UK copy and the other was a thesis embargoed by the author. This success could not have been achieved without the flexibility and cooperation among individual libraries in coping with the major disruption of downtime.

Chapter 15: Reader services during a time of major changes  

 219

Figure 15.4: Results during downtime

4.5 Lessons learned Between November 2010 and December 2011 over 5,000 mediated requests were processed by the Just-In-Time team, achieving 96 % success rate. The importance and appreciation of the service are obvious from overwhelming readers’ testimonials. These results strongly suggest that the Bodleian Libraries should explore the incorporation of the Just-In-Time model into future core library services.

5 Services to readers: Live Help 5.1 Establishing the service In the late spring 2011 as the ILS migration was quickly approaching the Reader Services team decided to offer a live chat known as SOLO Live Help, a live helpdesk to assist readers in the use of SOLO (the resource discovery tool, providing

220 

 Donald Mackay, Alena Ptak-Danchak and James Shaw

access to most of our catalogues and electronic finding aids), Bodleian Libraries’ services and locations, resource discovery and other aspects of the service during the online catalogue downtime and transition to the new system Aleph. In June an email call asking librarians to volunteer for a temporary SOLO Live Help service was issued. Under the leadership of the History subject specialist librarian a group of 13 (later 18) volunteers was recruited and trained. Because of the pressing schedule, we made a decision to use Meebo, a free Web 2.0 chat and instant messaging service already in use by the Bodleian Law Library. A Bodleian Libraries’ account was created and a customized Meebo widget (using only HTML coding) was embedded on the SOLO front page. (Figure 15.5) Transition to a new library management system in the Bodleian Libraries had another significant characteristic. We made a strategic decision not to continue the online catalogue but rather to concentrate on further development of the Libraries’ discovery tool, Ex Libris’s Primo product known locally as SOLO. The SOLO Live Help team and especially the team leader had a sizable task ahead: to identify key skills required, organize documentation and information and provide a virtual forum to acquire the skills needed. The following key skills were identified, some initially, some later as a result of enquiries. – Very good SOLO skills, including basic rare-book searches; – Good knowledge of electronic journals and the discovery systems used to search and retrieve them; – Good knowledge of SOLO patron functions, knowledge of circulation and stack request services; – Very well informed about general procedures within the Bodleian Libraries (library services, printing, passwords, Wi-Fi, databases, acquisitions and legal deposit), knowledge of our websites, latest news, and staff contact information; – Confident; – Able to multi-task; – Quick typist. The original schedule was set up requiring each person to staff SOLO Live Help for one hour at a time between the hours of 9 am and 5 pm during Monday to Friday. During the month of July 2011 periods between 9–10 am, 12–2 pm, and 4–5 pm were registering the heaviest usage of the service. In the three weeks before, during and after the migration more than 370 requests were handled by the SOLO Live Help team.

Chapter 15: Reader services during a time of major changes  

 221

Figure 15.5: The Meebo live chat widget

5.2 Lessons learned The benefits of SOLO Live Help were threefold: – Readers liked it as they received instant real-time and personal support even if not physically in the library. It involved little effort on their part. The placing of the service on the front page of SOLO provided great visibility; – Staff enjoyed the experience because it brought together a varied team. They could apply their knowledge and learn from others, even in a virtual environment. Staff participating in the live chat were supported by the extensive and regularly updated Meebo ToolKit on RealSpace; – The Aleph implementation team appreciated the feedback they received relating to readers’ experiences of the new system and gathered valuable information for making further improvements and priorities. The reception of the service was so overwhelmingly positive that when the new ILS became fully operational we decided not to withdraw the service but to retain it as a permanent feature of our services for readers. In addition to the immediate benefits, the running of the SOLO Live Help provided information that will be used in planning an ongoing sustainable service. Questions related to software features, technical details, staff skills and training needed can be answered from our experience during the project.

6 Conclusion Reaching out to our readers through interactive services like the Just-In-Time service and SOLO Live Help met with overwhelming approval and appreciation from our readers. Building on this positive experience we are planning to extend and design further self services for our readers.

222 

 Donald Mackay, Alena Ptak-Danchak and James Shaw

We have striven to communicate well with readers continually throughout the process; to listen to our readers; to educate them without taking too much of their time (or our own). Our community felt they were part of the transformation process and while the transformation still continues, the majority of our readers recognize our achievement and are supportive of further developments at the Bodleian Libraries.

Catríona Cannon

Conclusion In the projects that we undertook there were many significant achievements. We believe, for example, that we moved, barcoded and added into our storage facility more items in a shorter time than any other library has ever done. We did this by identifying the problems we needed to solve and by creating a coherent strategy for dealing with them which was centred on the needs of our readers and built on the achievements of the past. Having convinced funders to pay, we were careful to be accountable to them, reporting progress and expenditure in detail. We created temporary project roles and structures that complemented the existing permanent organizational structure to ensure the projects got done, with a mixture of temporary and permanent staff involved. Senior staff gave up significant amounts of time to leading the projects, showing their high priority for the Bodleian Libraries. But beyond the specific achievements, we also believe that what we achieved was innovation in the management of large library projects, or at least in the management of large library projects in Oxford. Peter Drucker identifies three conditions for a successful innovation: Innovation is work […] To succeed, innovators must build on their strengths […] Innovation is […] a change in a process – that is, in how people work¹

We fulfilled all three of these conditions. All involved in the projects – which included all Bodleian Libraries staff, many other colleagues in the collegiate University and other organizations – put in hours of ‘hard, focused, purposeful work²’. Rather than working against the culture and strengths of the collections and the University environment, the solutions we came up with actually built on them. And finally, we changed the way we worked – we now have one store where previously we had many, so our working practices are simplified and clearer for readers to understand. There is plenty more work to be done and, inevitably, some cleaning up as a result of the projects. But through them we have built a physical and IT infrastructure that ensures our collections are well cared for, accessible and well managed.

1 Drucker, Peter F., The essential Drucker (Oxford: Elsevier, 2007), 210. 2 Drucker, 210.

Index A academic libraries 213 Academic Strategy 13–18, 21, 23, 24, 35–36, 38, 39, 40, 88, 95, 133, 194, 195, 199 – interdependencies 196 accessioning process. See ingest acquisitions 155 – fiction 200 – new academic intake 199, 200 – non-academic intake 200 acquisitions budget 13 Admiralty – Intelligence unit 176 agency staff. See also barcoding: agency staff, See also bookmoves: agency staff air handling system 185 – 3D computer model 185 – plenum walls 185 Amazon 1, 213 Apache Lucene 151 Apache Solr 151 application developer 158 asbestos 189 ASR. See Automated Stack Request ASRS. See Automated storage and retrieval system Assistant Director, Collections & Resource Description 154 Assistant Director, Research & Learning Services 154, 212 Assistant Project Officers 127, 128 Associate Director, Special Collections and IT 154 Audubon, John James – Birds of America 17 authority control 155 Automated Stack Request 86, 153, 154, 155, 160, 211, 213, 214 automated storage and retrieval system 2, 6, 8, 12, 20, 70 – maintenance cost 20

B barcodes 138–139 – code 39 symbology 142, 144 – format – for items 142 – for trays 142 – on shelves 144 – location on item 130, 142 – location on special collections 130 – of shelfmark 112 barcode scanners 125 barcoding 18, 23, 30, 38, 71, 109–136 – agency staff 102 – at Osney Mead 99, 115, 133 – Bodley's journey 33 – complexity 29 – costs 18, 21, 37–38, 40, 41 – cross-over barcodes 132 – daily rate 114, 126, 134 – data capture 117 – equipment 125–126 – escalations 114, 123–124, 125, 127 – finances 129 – health and safety 129 – Holmes method 121 – interdependency with bookmoves 31, 39, 91, 105, 106–108, 133–134, 188 – interdependency with ILS 160 – interdependency with ingest 39, 91, 105, 106–108, 133–134, 188 – item cost 109 – linear measure 110 – locations 113, 136 – maps 55 – methods 136 – metres barcoded 32 – new stock 116 – Notepad method 124 – Nuneham Courtenay 115, 133 – periodicals 118–120 – pilot project 109 – premises 129 – problem solving 123–124, 128–132 – publisher dumps 120, 123, 134 – rare books 53, 120

226 

 Index

– sequence 31, 99, 113–115, 131 – shelfmark checklists 29 – software development 130 – special collections 54, 115, 120 – speedy spreadsheets 123 – staff 30, 126–127 – trays 71 – trolleys 125 – T-shirts 33 – unmatched items 132 – use of catalogue 116–125 – use of inventory records 121–123 – use of spreadsheets 119, 120, 121, 123 – using Aleph 120 – workflow 21 – workstations 125 Beuman Geudes Stretton 180 bibliographic control – lack of 23 Billerude White paper 75 Blackwell Publishing 62 Blackwell, Toby 187 Blackwell’s 187, 213 Blue Arrow recruitment agency 89 Bodleian Law Library 97, 199, 220 – basement 97, 114 Bodleian Libraries 3, 4 – Accounts department 40 – Bibliographic Maintenance section 127 – Cabinet 63, 153 – capital campaign 179 – collections 95, 198 – Collections and Resource Description department 154, 163, 167 – Collections Deployment Group 108 – committees with readers 47 – Communications Team 214 – Conservation and Collection Care department 48, 51, 69, 89, 102, 130, 163, 165, 167, 185 – creation 194 – Document Delivery section 55 – establishment review 4 – estates 194 – Executive 30 – external readers 212 – Facilities Management department 163, 165, 168–171, 202

– – – –

general procedures 220 Historic Venues department 168 Imaging Services department 163, 170 Integrated Library Systems Scoping Group 154, 156 – Integrated Library Systems Steering Group 154, 155, 159, 161 – library records 53 – number of readers 211 – Periodicals Cataloguing section 124 – Personnel department 126 – reader services 159, 197 – Reader Services department 163, 164 – Research & Learning Services department 154 – Retail Operations department 168 – Special Collections department 130, 163, 165, 168 – strategy 194–195 – Support Services 93 – Systems & e-Research Service 154, 166, 200 – systems team 117, 118, 119, 120, 127 – Technical Services department 69 – transformation 4 – web pages 214 Bodleian Library 1, 13, 14, 30, 31, 149, 164, 175, 188, 191, 194, 195 – 400th anniversary 3 – 400th anniversary campaign 179 – Arts End 183, 191 – as Grade I Listed Building 196, 201 – Bodleian Quadrangle 168, 169, 171, 191, 196, 201, 202 – repaving 196, 203, 204 – Consolidated Service Point 195 – deliveries 196 – disabled access 23, 196 – Divinity School 1, 34, 191 – Duke Humfrey’s Library 1, 3, 4, 13, 48, 107, 164, 183, 187, 191 – Facilities Management office 170 – refurbishment 171 – former Librarian’s office 203 – Founder’s Luncheon 187 – Head of Reader Services and Customer Relations 30

Index 

– – – – – –

history 191 Imaging Services office 169 lavatories 197 link to New Bodleian Library 192 Lower Reading Room 13, 197 New Bodleian Library. See New Bodleian Library – new lift 169, 195, 196, 197, 204 – construction 203 – disabled access 197 – north-west tower 202 – old lift 197 – opening 1 – open-shelf capacity 199 – operational union with Radcliffe Camera 199 – Proscholium 191, 200, 201 – Scholae 191 – Selden End 183, 191 – tearoom 166, 196, 203, 204 – tower lift 30, 169 – tunnel – opening to readers 196 – Underground Bookstore. See Underground Bookstore – Upper Reading Room 13, 46, 197 – vestibule 169 – refurbishment 169 Bodleian Library of Commonwealth and African Studies 58 – integration into New Bodleian Library 180 Bodley’s Librarian 11, 12, 20, 23, 24, 25, 33, 129, 184, 205 – appointment 194 See also Craster, Edmund. See also Nicholson, E.A.B. Bodley, Thomas 1, 3, 134 – birthplace in Exeter 33 – journey from Exeter 134 Bod Squad 23–34, 160, 186, 188 – additional staff 29 – blog 25–28 – categories 25 – tag cloud 25 – dashboard 25 – escalations 30–31 – functioning 24

 227

– logo 32 – problem solving 29 – Project Board 30 Bodzilla 189 Bonnie, Andrew V bookmoves 30, 95–108, 186, 213 – agency staff 102 – auditing 47, 51 – Bodleian Law Library 96, 97 – Bodleian Library 96 – Bookmoving team 102 – communications 45 – contracting out 101 – costs – savings 41 – data capture 45 – deadlines 99 – DeepStore 21 – finances 107 – interdependency with barcoding 31, 39, 91, 105–108, 133–134, 188 – interdependency with ILS 160 – interdependency with ingest 39, 91, 105–108, 133–134, 188 – method 48–51, 54–55 – monitoring 105 – move crates 47, 49 – move rate 55, 56 – move teams 47–48 – New Bodleian Library 21, 29, 40, 90, 91, 96, 98–99, 114, 118, 173, 187, 188 – special collections 44 – to Book Storage Facility 53–57 – to Radcliffe Science Library 44–52 – Nuneham Courtenay 21, 90–91, 96, 99–101 – special collections 44 – Oriental Institute Library 96 – Osney warehouse 96 – planning 105 – Plant Sciences Library 101 – progress 30 – project management – special collections 44–45 – Radcliffe Science Library 19, 43, 45, 48, 62, 65, 96, 187 – security 47, 48, 50, 51

228  – – – – – – –

 Index

sequence of operations 31, 97 shipment record 51 skates 90 special collections 61, 62, 187 staff 48 staffing 102–105 Taylor Modern Languages Faculty Library 101 – to Book Storage Facility 96, 99, 101, 188 – to Gladstone Link 98, 204 – to Osney Mead 99 – to Radcliffe Science Library 187, 188 – to Weston Library 41, 44, 57–58, 146 – totes 90 – transport 50 – Underground Bookstore 21, 96–98 – construction of hoist 202 – removal of hoist 204 – Zoology Library 101 booksellers 213, 214 books published 2 Book Storage Facility V, 19, 21, 23, 30, 36, 43, 46, 61, 95, 97, 109, 110, 114, 132, 137, 153, 179, 186, 196, 212 – ancillary hall 78 – barcode scanning 85 – building materials 79 – capacity 67 – chronology 83 – construction 67–81 – construction principles 78–80 – costs 39 – cost savings 40 – customer enquiries 94 – deliveries 92–94, 185 – design 72–78 – fire compartments 79 – identification of site 19 – Information System. See Book Storage Facility Information System – ingest 21, 30, 38, 89 – costs 40 – maps 55 – ingest process 90–91 – ingest process design 85–86 – lighting 79 – mass ingest 89–90, 145–146

– – – – – – – – – – –

operational life 67 operations 83–94 pick order 142–143 pit-stop 79 planchests 55, 76 planogram 141 processing benches 78 recycling 91 refiles 147 request rates 88 retrievals 72, 92–94, 196 – request slips 143–144 – retrieval times 92 – shelf address labels 141 – shelves 72–74 – sorting of stock 74 – sprinkler system 79 – staff 21 – staffing 88–89, 92 – staff training 89 – tray audit 132 – trays 74–75 – trolleys 78 – utility costs 30 – warehouse management system. See warehouse management system – weight limits 72 – wire guidance system 77 – zero-rating for VAT 37, 38 Book Storage Facility Information System 89, 137–147, 153, 157, 158, 188 – collection codes 146 – costs – savings 41 – data check 145–146 – data load 145–146 – equipment 144–145 – Not on File 91, 131–132 – options appraisal 137–138 – server set-up 145 – shelf-lighting 141 – UK adaptations 140–142 – use of barcodes 140 Book Storage Facility team 186, 188 Book Storage Logistics Manager 105 Boolean searching 151 Boston Public Library 184

Index 

British Academy 187 British Library 1, 17, 107, 137, 184 – Additional Storage Building 2 – bookmoves 19 – document supply service 213 British Standard BS5454 16, 17, 19, 67, 70, 185 Building Research Establishment 75, 79 buildings infrastructure 173 C Cambridge 12 Cambridge University Library 1, 8 – design 175 Cambridge University Press 62 Caps Cases Ltd 75 card catalogues – Oriental collections 54 Carfax rule 9 catalogue – copy-specific notes 160 – inventory records 120–123, 132 – migration 157, 158 – migration period 131, 160, 211, 218, 220 – minimal level bibliographic records 124 – missing items 128 – multiple item records 116–117 – patron functions 220 – size 110, 134, 136 – summary item records 118 – updating 147, 188 cataloguing 155 celebration 33–34 Central Bodleian Library 3, 195 Centre for Oxfordshire Studies 149 Chinese catalogue 152 – migration into Aleph 162 Christmas decorations 188 Churchill College, Cambridge – archives 173 Churchill, Winston 172 circulation 155, 159 Clarendon Building 164, 168–169, 188 – meeting room 169 – proposed Facilities Management offices 169

 229

– Support Services office 171 Classics 199 classification schemes 208, 212 – Bodleian Library 111 – Gough Additions 111, 121 – Library of Congress 208 – modern monographs 111–112, 208 – modern periodicals 112, 118 – monographic series 121 – Nicholson 111, 117, 131, 133, 208 – periodicals 112, 118 – Old Class 112 – periodicals 112 – Radcliffe Science Library 111 – special collections 112 – year books 112 closed access – capacity 199 collections V, 1, 2, 5, 13 – in transit 212, 213, 214 college libraries 213, 217 Columbia University Library 183 commercial storage 177 Commonwealth and Empire collection 180 communications 19, 21, 23, 32–33, 47, 64, 106, 129, 161, 165–166, 173, 203, 204, 214, 217–218 – deficiencies 165–167 Conservative Party 173 consultants 23 consultation with readers 199, 200, 202 conveyor 13, 169, 185, 192, 196 – closure of 197 conveyor station 185 Cornell University Library 12, 183 Corpus Christi College 184 Craster, Edmund 175 CroBar 117 culture clash 102 Curators of the University Libraries 6, 40, 61, 201 D data administrator (ILS) 158 database programmer 127 date formats 140 David, Charles Wendell 187

230 

 Index

DeepStore 8, 11, 17, 19, 29, 30, 36, 38, 40, 43, 44, 53, 69, 89, 91, 96, 99, 114, 118, 119, 128, 130, 131, 134, 136, 178, 179 – bookmoves. See bookmoves: Deepstore – delivery schedule 114 – pallets 130 – problem solving 130 delivery times 8, 13, 14, 19 departmental collections 213 Dill, Reese 71 Divinity School. See Bodleian Library: Divinity School DOBIS/LIBIS 149–150 dreaming spires 10, 70 Duke Humfrey’s Library. See Bodleian Library: Duke Humfrey’s Library Duke University Library 183 E electronic databases 152 electronic journals 13, 213, 220 – backfiles 23, 36, 61–66 – completeness 63–65 – negotiations 63, 65 – non-UK publishers 62 – pricing 62–63 – space saving 63, 64 – subject areas 62, 63 – UK publishers 62 – usage 64–66 electronic resources 35, 37, 213, 214, 217 – backfiles 15 – book purchases 64 – evaluation of purchases 65–66 – newspaper archives 62 ELISO 61 Elsevier 65 Emerson company 187 English 199 Equality Act 196 estate 13, 35 Estates Projects Officer 180, 186 European Union regulations 150 Exeter 134 – Gandy Street 33 Ex Libris 156, 158, 160 – Aleph 118, 120, 123, 124, 128, 130, 131, 147, 156, 158–159, 161

– – – – – –

– ADMs 158, 159 – closed stack management sub-system 160 – OPAC 159 – test database 131 – training 131 Aleph Reporting Center 157, 161 MARCit 65 Metalib 151, 152 OPAC via Primo 159 Primo 151, 161, 220 – installation 151–152 SFX 65, 151, 152

F Facilities Management 25, 28 finances 35–41 – barcoding 40 – revision of budgets 39 – VAT increase 39, 41 – VAT reduction 37 – virement between funds 39–41 fire tests 75 FKI Logistex 137 flooding 10, 20, 70 foreign-language materials 214 forklift trucks. See high-level order-pickers G Gallica 217 Garfield Weston Foundation 187 Geac – Advance 118, 123, 124, 128, 130, 147, 150, 159, 160, 218 – serials module 118 – server failure 160, 161 – GeoCat 124 Generation Fifth Applications 84, 109, 137 – software 140 – software amendments 139 George VI 2, 175 Gladstone Link 15, 96, 97, 201, 204–207. See also Underground Bookstore – balustrade 201 – colour scheme 207 – costs 39 – food policy 201 – lending collection 161

Index 

– lift 204 – lighting 207 – opening 205 – open-shelf capacity 199 – readers' reactions 205 – reading areas 207–208 – selection of materials 208 – staircase 204 – tunnel 205 – use of trolleys 205 Gladstone shelving 191, 192, 201 Gladstone, William 14, 191. See also Gladstone shelving Google 1, 151 – digitization project 164 – maps 33 – Street View 134 Grandpont House Library 217 H Harcourt Arboretum 99 Harris, E.C. 183 Harvard University 20 Harvard University Library 2, 152 – depository 71, 84, 109, 138, 139, 142 Harvard VNA storage 20, 70, 71, 84, 138 Headley Trust 187 Head of Bodleian Library Reader Services and Customer Relations 204 Head of Systems & e-Research Service 154 health and safety 107, 126, 188, 204, 205 Heaney, Michael V high-demand books 14 high-level order-pickers 71, 72, 76–78, 84 high-use materials 35, 97, 98, 195, 196, 198–200, 213 Historical Manuscripts Commission 178 History 199 Horton, Robert 187 hotdesking space 168, 171 Humanities research 7 Hurley Palmer Flatt 183, 185 I ICT Support Team 163, 166, 172 ILS. See integrated library system Imperial College London 212 Indian Institute Library 98, 180

 231

ingest 84. See also Book Storage Facility: ingest – interdependency with barcoding 39, 91, 105, 106–108, 133–134, 188 – interdependency with bookmoves 39, 91, 105–108, 133–134, 188 innovation 223 institutional repositories 213 integrated library system 4, 29, 40, 127, 133, 137, 149–162 – consultant's report 152–153 – costs 38 – savings 41 – disaster recovery plan 162 – failed implementation 150 – hardware architecture 159 – hardware installation 158 – implementation 158–161 – bedding in 160–161 – interim acceptance 158 – phase II 161–162 – timescale 157 – Oxford specials 150, 153 – principles of operation 149 – procurement 152–153, 154–156 – competitive dialogue 154, 155, 156 – due diligence 156 – evaluation points 155 – site visits 155–156 – tender document 154 – vendors 153, 154, 155, 156 – production environment 158 – project manager 20 – project timeline 152 – replacement 20 – staff training 158 – stock control 53 – systems training 158 – VMware 159 Integrated Library Systems Scoping Group 156 inter-library loans 214 Internet Archive 213, 217 Inventory Control Officers 127 Inventory Control Project 91 Inventory Control Project Manager 105 Iron Mountain 69, 99, 106, 178 Item Products Ltd 75 IT Programmes Manager 38, 153, 156, 158

232 

 Index

J Jamie Briggs Removals & Storage 101 Japanese catalogue 152 – migration into Aleph 162 John Johnson collection 54, 120 Jungheinrich 87, 88 Just-In-Time 211–222 – workflow 214, 218 – workload 216 Just-In-Time Project Coordinator 213, 214 Just In Time team 106, 218 K Kahn, Louis 184 L Lady Margaret Hall 184 Lankester Room:. See Radcliffe Science Library: Lankester Room Lee Foundation 187 legal deposit 2, 13, 37, 38, 41, 63, 83, 91, 95, 110, 117, 123, 125, 139, 146, 198, 200 – barcodes 142 – tax relief 110, 146 LibQual+ 14 Library Archival System. See Generation Fifth Applications library staff V link resolver 65 Listed Building Consent 28, 202 live chat 211, 219–221 – advantages 221 – hours of operation 220 M MacDonald, Alasdair V Mace (construction company) 179, 188 Mackerras, Charles 187 Mallinckrodt, George 187 map collections 46, 55–56, 109, 176, 213 – retrieval 55, 56 Map Room. See New Bodleian Library: Map Room mass ingest 139 Meebo 220, 221 Meitar, Zvi. See Zvi and Ofra Meiter Family Fund

Mellon Centre for British Art 184 Merton College 184 metric units 140 Metropolitan Museum of Art – Thaw Center for Paper Conservation 186 microfilms 121 mobile shelving 70 Mo i Rana (Norway) 69 Morgan Library and Museum 184, 186 Museum of London 184 Museum of Modern Art – Paper Conservation Studio 186 Museum of the History of Science 164 Museums, Libraries and Archives 47 music collections 46, 56–57, 213 N National Archives Advisory Service 178 National Library of Norway – depository 69 Nature (journal) 65 netbook computers 125 New Bodleian Library 1, 11, 23, 44, 129, 164, 197 – as approved repository 178 – as depository 175 – as Grade II Listed Building 172, 180 – bindery 99, 102 – bookmoves. See bookmoves: New Bodleian Library – bookstack 2, 3, 4, 5, 8, 13, 47, 48, 192, 199 – bookstacks 98, 113, 119, 123, 128, 136, 179, 188 – mapping 98 – Broad Street frontage 181, 183 – closure 33 – construction 175–176 – environmental control 177, 179 – environmental risks 15, 19, 178 – exhibition room 176 – Facilities Management offices 168 – filing cabinets of Winston Churchill 172 – fire risk 177, 179 – former staff 34 – furnishings 172–173 – height limit 101

Index 

– – – – – – – – – – – – – – –

– – – –

heritage aspects 180 heritage furniture 172, 188 lifts 45, 50, 98, 107, 173, 188 link to Bodleian Library 192 loading bay 48, 50, 196 Map Room 55, 187 Music Room 187 NEWBOLD. See NEWBOLD opening 176 overcrowding 176–178 planchests 55 portraits 172 Project Sponsor Group 31 Reading Room 164 redevelopment 7, 35, 36, 67, 95, 133, 179–187 – auditorium 179 – Bibliography Room 186 – Blackwell Hall 181 – bookstacks 185 – café 186 – client risk register 187 – conservation workshops 186 – costs 37, 39 – demolition phase 189–190 – design brief 180–181, 183 – entrance hall. See New Bodleian Library:redevelopment: Blackwell Hall – environmental controls 185 – exhibition space 179, 181, 186 – facilities for scholars 179, 185 – fire protection 186 – floating stack 185 – funding approval 39 – handover to contractors 188 – open-shelf collections 185 – Project Sponsor Group 37, 180, 183, 184 – public engagement 179 – study visits 183–184 – tower crane 189 – zero-rating for VAT 37 – zoning 183 renovation 23, 43 Special Collections Reading Room 48 staff canteen 196 staff relocation 23, 29, 163–173, 188 – occupancy 168

 233

– project management 165–166, 173 – sequence of moves 164–165 – weight limits 45 – ziggurat style 175 NEWBOLD 178–180, 187 New Library. See New Bodleian Library New York Public Library 184 Nicholson, E.W.B. 12, 111. See also classification schemes: Nicholson NOF. See Book Storage Facility Information System: Not on File non-Roman scripts 54 Not on File. See Book Storage Facility Information System: Not on File NRLF-type high-bay shelving 70 Nuneham Courtenay 16, 99 – bookmoves. See bookmoves: Nuneham Courtenay – depository 3, 5, 7, 35, 56, 67, 123, 130, 131, 199 – capacity 68 – construction 177 – retrievals 101 – village atmosphere 99 O O’Driscoll, Denis V Official Papers 133 Old Bodleian Library. See Bodleian Library Old Library. See Bodleian Library Old Schools Quadrangle. See Bodleian Library:Bodleian Quadrangle OLIS. See Oxford Libraries Information Service Online Books (University of Pennsylvania) 217 Online Public Access Catalogue. See OPAC OPAC 149, 151, 155, 159 – lack of provision 159–161 open access 198 Open URL resolver 151 Oracle database 159 Oracle database administrator 158 Ordnance Survey 56 Oriental collections 46, 54, 58, 59, 199, 213 Oriental Institute Library 187 Osney Mead 2, 69 – delivery hub 93 – depository 6, 7–13, 15, 70, 184

234 

 Index

– planchests 70 – sinusoidal roof 70 – flooding 10 – staff visits 165 – transport 165, 166 – warehouse 30, 99, 114, 120, 126, 130, 131, 132, 171 – infrastructure 172 Osney One Building 69, 164, 169 – infrastructure 166–167 – refurbishment 166–167 – restaurant 166, 167 OULS. See Oxford University Library Services Ovenden, Richard V Oxford – Broad Street 175, 176, 181, 187, 196 – City Conservation Officer 201 – fire brigade 177 – Parks Road 175, 180 – Radcliffe Square 1, 7, 191, 202 Oxford Central Library 217 Oxford City Council – Area and Strategic Planning committee 70 Oxford Libraries Information Service 149 Oxfordshire County Archives 164 Oxfordshire County Libraries 217 Oxford single-sign-on 159 Oxford topography 183 Oxford University 223 – Audit and Scrutiny Committee 40 – Botanic Gardens 99 – Buildings and Estates Subcommittee 184, 201 – Capital Steering Group 184 – central purchasing section 126 – college libraries 154, 184 – Committee of College Librarians 154 – Council 39 – Disability Team 200, 202 – Estates Directorate 180, 183, 196, 202, 203 – fire safety office 177 – Libraries Board 67 – ‘Oxford Thinking’ campaign 187 – Planning and Resource Allocation Committee 31, 39, 107, 184 – Safety Office 200, 202 – Security Service 47, 50

– Student Consultancy Service 118 – University Museum 191 Oxford University Archives 58 Oxford University e-journals 152, 154 Oxford University Library Services 3. See also Bodleian Libraries Oxford University Press 137, 186, 187 OxLIP+ 152, 154, 217 P Patristics 199 patron authentication 159 – non-University members 159 Pell Frischmann 183, 185 Penguin books 123, 134 PerBar 120 philanthropic funding 184, 186–187 Philosophy 199 Piano, Renzo 184 planning permission 5, 10, 11, 21, 68 – appeal 11, 12 Po Chung 205 Polonsky Foundation, The 187 predicting reader demand 88 premises-related infrastructure space charge 194 Price, Dave V Prince2Lite 45 Princeton University Library 84 project administrative assistant (ILS) 158 project delivery V project management – focused view 28 – procedures 31 – strategy 19 Project Manager for Library Logistics 102, 105, 107, 202 Project Officers 127, 128 projects – capital funding 23 – interdependency 23, 29 Pro-Vice-Chancellor for Academic Services and University Collections 180 public libraries 213 Public Records Act 178 publisher dumps 123 Purcell Miller Tritton 180, 200

Index 

Q Queen’s University Belfast 212 R Radcliffe Camera 1, 13, 14, 97, 107, 175, 194, 195, 202 – as Grade I Listed Building 196, 201 – as part of Bodleian 191 – disabled access 196 – history 191 – Lower Reading Room 192, 204 – as bookstack 191, 199 – balustrade 201 – construction 203 – lift 204 – staircase 204 – Official Papers Reading Room 97 – open-shelf capacity 199 – operational union with Bodleian Library  199 – Upper Reading Room 191 Radcliffe Infirmary site 7 – Humanities and Area Studies library 7, 12, 23, 35 Radcliffe Observatory Quarter. See Radcliffe Infirmary site Radcliffe Science Library 19, 43, 45, 163, 191, 199, 208 – basement 16 – bookmoves. See bookmoves: Radcliffe Science Library – bookstack 17, 19, 47, 48 – circulation 15 – environmental controls 43 – journals 61, 62 – Lankester Room 17 – refurbishment 43, 44 – lifts 45 – refurbishment 43 – renovations 23, 36 – costs 39 – Special Collections Reading Room 19, 43, 44, 45, 164 – weight limits 45 rare books 46, 53, 120, 220 reader services 35 – maintenance 204

 235

readers’ requests – mediated requests 211 readers’ requests 147 – alternative sources 213–217 ReBar 117–118, 121, 127, 130, 132, 136 – with AutoNext 117–118, 131, 132 ReCAP depository 84, 89, 109, 138, 139, 142 Red Prairie system 137 refreshment facilities 166 relevance ranking 151 request ratios 14 resource discovery 151–152, 155 resource discovery tools 151. See also SOLO restricted services 211 retrievals 147, 213 Rhodes House Library. See Bodleian Library of Commonwealth and African Studies RIBA planning process 180, 184, 186 Rockefeller Foundation 175 Royal Institute of British Architects 184. See also RIBA planning process S Sackler Library 199 salt mine. See DeepStore Sassoon, Joseph 187 Scan and Deliver 161 scenario planning 91 Scheide Fund 187 scientific periodicals 17 Scott, Giles Gilbert 2, 175, 177, 180, 184 Search Oxford Libraries Online. See SOLO security 129 senior barcoders 114, 120, 124, 127, 128, 131, 132, 136 Senior Project Officer 132 serials management 155 SERS warehouse. See Osney Mead:warehouse SFX. See Ex Libris:SFX shelfmarks 12, 138–139, 199 shelf-sequence 49, 50 Shepley Bulfinch Richardson and Abbott 183, 184, 186, 200 Smurfit Kappa 75 Social Science Library 199 SOLO 123, 124, 146, 151, 152, 154, 159, 188, 214, 220

236 

 Index

– interface for mobile devices 161 – Live Help. See live chat – patron functionality 159 – performance problems 161 SOLO Live Help team 220 South Marston 20, 67, 71 special collections 7, 8, 16, 19, 35, 43–59, 91, 110, 169, 213 – identification 46 – inventory 53–54 – Radcliffe Science Library. See Radcliffe Science Library:special collections reading room – staff 29 – unusual formats 51 – vulnerable materials 51 Special Collections Move Co-ordinator 47, 48, 51, 53 special collections staff 172 staff 126–127 – agency staff 34, 115, 126, 129, 131 – barcoders 171–172 – end of contract 163 – library staff 127 – maintenance staff 171 – project teams 165 – stress 173 – tension 24, 31–32 – training 128 staff survey 165 Stationers’ Company 2 St Cross Building. See Bodleian Law Library storage systems – United Kingdom 69 – United States 69 – USA 71, 84 storage trays 84, 85, 88, 92 subject librarians 99 sub-Saharan collections 180 Support Services Manager 171 Swindon 184 – Borough Council 21 systems analyst (ILS) 158 T Taylor and Francis 62 Taylor Institution Library 194, 199

the Book Storage Facility – funding 23 – Project Sponsor Group 31 The National Archives 4, 16, 47, 178, 184. See also TNA Standard Theology 199 theses 214, 218 Thomas, Sarah. See Bodley’s Librarian TNA Standard 185 Total Logistics Ltd 70, 87, 88 Twitter 205 U UBOB Project 40, 133, 169, 191–209 – construction 202–203 – consultation with readers 200–202 – consultation with staff 201 – exhibition 201 – Project Sponsor Group 31, 202 – timescale 203, 204 uncatalogued collections 46, 53, 58, 131 uncatalogued material 98 uncatalogued monographs 124 uncatalogued periodicals 124 Underground Bookstore 1, 7, 14, 30, 31, 113, 123, 133, 134, 192, 194, 195, 196, 199, 202 – bookmoves. See bookmoves: Underground Bookstore – capacity 15 – Gladstone shelving 14, 201 – hoist 107 – link to Bodleian Library 192 – link to Radcliffe Camera 192 – open-shelf capacity 199 – refurbishment 195 – renaming as Gladstone Link 201 – renovation 23, 36 – costs 39 Underground Bookstore and Old Bodleian Access Project. See UBOB Project uninterrupted access 23 University Church 1 University of Austin – Harry Ransom Humanities Research Center 183 usage analysis 200

Index 

V Vatican Library 17 Very Narrow Aisle storage. See Harvard VNA storage Vice-Chancellor 3, 11, 20, 178, 186 Victoria and Albert Museum 184 visual impact 70 W Wainwright, Bruce V Walsh, Cath V warehouse management system 20, 29, 30, 38, 71, 84. See also Book Storage Facility Information System warehouse management system – project manager 20 weather conditions 30 web 2.0 151 Web of Science 61 western manuscripts 46, 53 Weston Library V, 187, 175–190, 196

 237

Wigmore, Barrie and Deedee 187 Wikipedia 1 Wiley 62 Wilkinson Eyre Architects 180, 183, 185 Williams, Michael V Wolvercote 1 World War II 172, 176 – invasion of Europe 176 Y Yale University – Art Gallery 184 Yale University Library 183 – Beinecke Library 184 – depository 71, 84 – Sterling Memorial Library 184 – Music Library 184 Z Zvi and Ofra Meiter Family Fund 187