Proceedings of the 2nd Human Engineering Symposium: HUMENS 2023, Pekan, Pahang, Malaysia (Lecture Notes in Mechanical Engineering) 9819968895, 9789819968893
This book acts as a compilation of papers presented in the 2nd Human Engineering Symposium (HUMENS 2023), held at Pekan,
162 92 16MB
English Pages 487 [471] Year 2024
Preface
Contents
About the Editors
Developing a Survey Tool to Measure Psychosocial Risk and Work Performance at a Workplace
1 Introduction
2 Material and Method
3 Results and Discussions
3.1 Reliability Analysis of the Instrument
3.2 Demographic Data
3.3 Exploratory Factor Analysis (EFA)
4 Conclusion
References
Enhancing Mental Health Through Ambient Lighting
1 Introduction
1.1 Ambient Lighting
1.2 Biosignals: Galvanic Skin Response (GSR)
1.3 Stressor: Digit Span
1.4 Questionnaire: State-Trait Anxiety Inventory (STAI) and Reactance
2 Methodology
2.1 Experimental Set-Ups
3 Results and Discussion
3.1 Participants
3.2 Hypotheses Testing
4 Conclusion
References
Parameter Extraction of Muscle Contraction Signals from Children with ASD During Fine Motor Activities
1 Introduction
2 Methodology
3 Results and Discussion
4 Conclusion
References
Design of Hose Roller for Firefighter: A Fatigue Study
1 Introduction
2 Materials and Methods
2.1 Fabrication
2.2 Testing
2.3 Fatigue Analysis
3 Results and Discussion
3.1 Fabrication
3.2 Data Analysis
3.3 Data Comparisons Between Manual Handling and Hose Rolling Method
4 Conclusions
References
Noise Risk Assessment on Noise Exposure Among Urban Rail Maintenance Workers Using Personal Monitoring Method
1 Introduction
2 Methods
2.1 Noise Risk Assessment (NRA)
2.2 Procedure to Conduct Personal Monitoring
3 Method Results and Discussion
3.1 Noise Risk Assessment (NRA)
3.2 Personal Monitoring
3.3 Noise Reduction
4 Conclusion
References
Bangla Text Summarization Analysis Using Machine Learning: An Extractive Approach
1 Introduction
2 Related Work
3 Overview of Text Summarization Techniques
3.1 Extractive Text Summarization
3.2 Similarity Matrix
3.3 Word2vec
3.4 Word Count
4 Methodology for Text Summarization
4.1 Preprocessing
4.2 Calculating the Word Frequency
4.3 Sentence Weighting
5 Summary Generation
6 Result and Evaluation
6.1 Result
6.2 Evaluation
7 Conclusion
References
Human Factors: Drivers’ Speed Choice on Relatively Low-Speed Limit Roads
1 Introduction
2 Literature Review
2.1 Speed and Risk of Accidents
2.2 Factors Influencing Speed Choice
3 Methodology
3.1 Sample Size
3.2 Site Selection
3.3 Data Collection
4 Result and Discussion
4.1 Distribution of Spot Speed Data and Speed Limit Violation
4.2 Drivers’ Perceived Safe and Hazardous Speeds
5 Conclusion
References
Topology Optimization for Custom Bed-Resting Ankle Foot Orthosis
1 Introduction
2 Methodology
2.1 3D Scanning of Foot
2.2 Conceptual Design of AFO
2.3 Topology Optimization
2.4 Finite Element Analysis
3 Results and Discussion
3.1 Topology Optimization
3.2 Effects of Different Load
4 Conclusion
References
Influence of Environmental Factors and Road Characteristics in Commuting Accidents Among Public University Staffs
1 Introduction
2 Methodology
2.1 Document Review
2.2 Semi-structured Interview Technique
3 Results and Discussion
3.1 Descriptive Analysis
3.2 Inferential Analysis
4 Conclusion
References
Effects of Material Properties in Developing the Ear Prosthetics
1 Introduction
2 Methodology
2.1 Phase 1: 3D Scanning Model of an Artificial Ear
2.2 Phase 2: Finite Element Analysis for Artificial Ear
3 Results and Discussion
3.1 Effect of Material Properties on Artificial Ear (Axial Force)
3.2 Effect of Material Properties on Artificial Ear (Pressure)
4 Conclusion
References
Study of Primary Stability of Hip Implant for Semi Hip Replacement by Using Finite Element Analysis
1 Introduction
2 Methodology
2.1 Development of a Three-Dimensional of Femur Bone
2.2 Development of a Three-Dimensional Hip Implant
3 Results and Discussion
4 Conclusion
References
Investigation of Mental Health Condition Among Factory Worker During Covid pandemic–A Cross-Sectional Study
1 Introduction
2 Methodology
2.1 Participants
2.2 Instrumentation
2.3 Data Analysis
3 Result
3.1 Demographic Analysis
3.2 Depression, Anxiety, and Stress
3.3 Depression, Anxiety, Stress and Demographic Factors
3.4 Depression, Anxiety, Stress and Socioeconomic Factors
3.5 Depression, Anxiety and Stress Among Workers in Relation to the Organizational Factors
4 Discussion
5 Conclusion
References
The Influence of Body Balance Towards the Golf Putting Performance
1 Introduction
2 Methodology
2.1 Arm Movement
2.2 Weight Band
2.3 Putting Performance
3 Results and Discussion
3.1 Putting Outcome
3.2 Arm Movement
3.3 Center of Gravity (COG)
4 Conclusion
References
Risk Assessment for Manual Handling Activities in a Dairy Industry
1 Introduction
2 Methodology
2.1 Field Observation
2.2 Manual Handling Hazards Identification at the Workstations
2.3 The Manual Activities Assessment Chart (MAC Tool)
3 Results
3.1 Manual Handling Hazard Identification
3.2 Manual Handling Risk Assessment (MAC Score Sheet)
4 Discussion
5 Conclusion
References
Brief Review of Recent Study on Fluid–Structure Interaction Modeling of Blood Flow in Peripheral Arterial Disease
1 Introduction
2 Methods
2.1 Search Strategy and Eligible Criteria
2.2 Flow Diagram of the Study
3 Results and Discussion
3.1 Study Characteristic
3.2 Peripheral Artery Geometry Construction
3.3 FSI Modeling Platform
3.4 Blood Viscosity
3.5 Arterial Wall Material Models
3.6 Validation of Fluid–Structure Interaction Models
4 Conclusion
References
Head Injury During Heading of Two Types of Sepak Takraw Balls: Analytical Approach
1 Introduction
2 Methodology
2.1 Drop Ball Test
2.2 Mathematical Model
3 Result and Analysis
4 Conclusion
References
Fluid–Structure Interaction (FSI) Modelling in Stenotic Carotid Artery Bifurcation
1 Introduction
2 Methodology-Modeling and Computational Setup
2.1 Geometry for Idealized Healthy and Carotid Artery Stenosis Bifurcation
2.2 Flow Simulation and Boundary (CFD)
2.3 Mesh Independence Test—Idealized Healthy Carotid Artery
2.4 Mesh Sensitivity Analysis
3 Material Properties and Boundary
4 Result and Discussion
4.1 Flow Profiles
4.2 FSI Result—Total Wall Deformation
4.3 FSI Result for Wall Stress Equivalent (von Mises)
4.4 FSI Result for Wall Displacement—Comparison Between Wall Thickness 1.0 and 0.7 mm
4.5 Discussion
5 Conclusion
References
Prediction of Atherosclerosis in Peripheral Arterial Disease Using Computational Fluid Dynamics Modelling
1 Introduction
2 Materials and Methods
2.1 Geometry Construction
2.2 Meshing
2.3 Boundary Conditions
2.4 Flow Simulation
3 Results
3.1 Mesh Independence Study Healthy Model
3.2 Prediction of Potential Region for Atherosclerosis to Occur with Laminar Flow Condition
3.3 Comparison of High Potential Region to Develop Atherosclerosis with Laminar Flow Condition
4 Conclusion
References
Impact Analysis of Motorcycle Helmet: Finite Element Modeling
1 Introduction
2 Methodology
2.1 Design
2.2 Simulation
3 Results and Discussion
3.1 Total Deformation
3.2 Von Mises Stress
4 Conclusion
References
The Protective Performance of Different Types of Motorcycle Helmets in Terms of HIC and BrIC
1 Introduction
2 Methodology
2.1 Pendulum Test Rig
2.2 The Hybrid III Anthropomorphic Test Device
2.3 Sensor
2.4 Signal Processing
2.5 Experiment Procedure
2.6 Output Calculation
3 Results and Discussion
4 Conclusion
References
Measuring Running Performance Through Technology: A Brief Review
1 Introduction
2 Running Technologies
2.1 Methods and Model
2.2 Measurement for Running Performance and Technique
2.3 Quantifying the Workload
3 The Effectiveness of Running Technologies in Improving Performance and Technique
3.1 Technology that Detects Technique
4 Conclusion
References
Experimental Study of Gait Monitoring on Wearable Shoes Insole and Analysis: A Review
1 Introduction
2 Methodology
2.1 Development of In-Shoe Wearable Pressure Sensor Using an Android Application [2]
2.2 In-Sole Shoe Foot Pressure Monitoring for Gait Analysis [3]
2.3 SmartStep: A Fully Integrated, Low-Power Insole Monitor [4]
2.4 SmartStep 2.0: A Completely Wireless, Versatile Insole Monitoring System [5]
2.5 Piezoelectric Pressure Sensing Shoe Insole [6]
2.6 Development of the RT-GAIT, a Real-Time Feedback Device to Improve the Gait of Individuals with Stroke [7]
2.7 Smart Insole: Wearable Sensor Device for Unobtrusive Gait Monitoring in Daily Life [8]
2.8 Design and Development of Integrated Insole System for Gait Analysis [9]
2.9 A Reliable Gait Phase Detection System [10]
2.10 A Shoe-Integrated Sensor System for Long-Term Center of Pressure Evaluation [11]
3 Result and Discussion
3.1 Finding
3.2 Limitation
3.3 Similarities
4 Conclusion
References
Preliminary Ergonomics Analysis of Sit-Stand (STS) Desk on the Patient with Lower Back Pain Problem: A Case Study
1 Introduction
2 Methodology
2.1 Observational Method and Case Study
2.2 Digital Human Modelling (DHM)
3 Result and Discussion
3.1 Anthropometric Measurement
3.2 Analysis of Body Posture of the Subject During Work-Standing
3.3 Analysis of Body Posture of the Subject During Work-Sitting on the Bedside
3.4 Analysis of the Body Posture of the Subject During Work-Sitting on the Bed
4 Conclusion
References
Developing a Survey Tool to Measure Human Factors Constructs for Personal Hearing Protector (PHP) Use Among Industrial Workers—First Phase
1 Introduction
2 Materials and Methods
2.1 Survey Tool Development
2.2 Conceptual Framework on the Human Factors and Personal Hearing Protector (PHP) Use Among Industrial Workers
3 Results and Discussion
3.1 Sociodemographic
3.2 Analyses of Human Factors Constructs for PHP Use
4 Conclusion
References
A Review on the Pedal Error Cases Among Car Drivers in Malaysia
1 Introduction
2 Aim and Methods
3 Analyses
3.1 Pedal Error Cases in Malaysia Reported by News Media from 2016 to 2022
3.2 Driver Age
3.3 Driver Gender
3.4 Crash Location
3.5 Pre-crash Maneuver
3.6 Injuries and Fatalities
4 Discussion
5 Conclusions
References
Study of Anxiety Parameters and Sensors Related to Monitoring the Anxiety Concentration Index Level Among Archer Athletes: A Review
1 Introduction
2 Methodology
2.1 Anxiety Among Archer Athletes
2.2 Heart Rate (HR)
2.3 Blood Pressure (BP)
2.4 Electromyography (EMG)
3 Result and Discussion
3.1 Research Finding
3.2 Limitation
3.3 Similarities
4 Conclusion
References
EEG and EMG-Based Multimodal Driver Drowsiness Detection: A CWT and Improved VGG-16 Pipeline
1 Introduction
2 Method and Materials
2.1 Continuous Wavelet Transform
2.2 Improved-VGG16 Model
2.3 Experimental Data
2.4 Performance Evaluation
3 Results and Discussions
4 Conclusion
References
Rehabilitation and Gamification Technology Device for Lower Extremities Patient: A Review
1 Introduction
2 Methodology
2.1 Ankle Sprain
2.2 Virtual Reality, Augmented Reality, and Gamification
2.3 Wearable Robots and Platform-Based Robots
3 Result and Discussion
3.1 Effectiveness of Rehabilitation Robot
3.2 Discussion
3.3 Limitation
3.4 Similarities
4 Conclusion
References
The Importance of Proper Motorcycle Helmet Buckling: A Scientific Study
1 Introduction
1.1 Motorcycle Helmet Legislation
1.2 Motorcycle Helmet Certification
1.3 Motorcycle Helmet Parts
1.4 Motorcycle Helmet Retention Status
2 Methodology
2.1 Monorail Impact Test Machine
2.2 The Hybrid III Anthropomorphic Test Device
2.3 Shimmer Sensor
2.4 Signal Processing
2.5 Test Procedure
2.6 Parameters Calculation
3 Results and Discussion
4 Conclusion
References
A Short Review on Development of Table Tennis Robotic Launcher
1 Introduction
2 Customer Needs
3 Design Specification
4 Design Concept
5 Conclusion
References
Reusability Study of 3D Printing Mould and Resin Casting for Takraw Ball Launcher Wheel
1 Introduction
1.1 An Overview of Sepak Takraw
2 Literature Review on Sepak Takraw Training Device
3 Methodology
4 Results and Discussion
5 Conclusion and Recommendation
References
Development of Noise Risk Assessment (NRA) and Management System
1 Introduction
2 Methodology
2.1 Framework Development
2.2 System Development by Microsoft Software
2.3 System Usability Scale (SUS) Test
2.4 Expert Validation
3 Results and Discussion
3.1 Manual Documentation
3.2 Noise Risk Assessment and Management System
3.3 Development of Noise Risk Assessment and Management System
3.4 Validation and Usability of the Developed System
4 Conclusion
References
Framework of Safety Helmet Compliance Detection and Employee Tracking by Using Quick Response (QR Code) Technology
1 Introduction
1.1 Challenges in Construction Industry
2 Research Methodology
3 Result and Discussion
3.1 Framework of Safety Helmet Compliance Detection and Employee Tracking
3.2 Safety Helmet and Employee Tracking Information Prototype System
4 Conclusion
References
e-HIRARC Tool for Brick Laboratory in Civil Engineering Department at TVET (Technical and Vocational Education Training) Campus
1 Introduction
2 Methodology
2.1 e-HIRARC Framework
2.2 e-HIRARC Tool Development
2.3 Validation Test and Case Study
3 Results and Discussion
3.1 e-HIRARC Framework
3.2 e-HIRARC Tool Development
3.3 Content Validation
3.4 Case Studies and System Usability Scale (SUS)
4 Conclusions
References
A Review of Biomechanical and Psychosocial Risk Factors Among Workers
1 Introduction
2 Material and Method
3 Results and Discussions
3.1 Ergonomic Risk Factors Evaluation Method
3.2 Ergonomic Risk Factors and MSD
3.3 Musculoskeletal Disorders (MSD) and Its Effects
4 Conclusion
References
Knowledge and Awareness of Road Safety Among University Students
1 Introduction
2 Methodology
3 Results and Discussion
3.1 Demographic Analysis
3.2 Knowledge of Road Safety
4 Conclusion
References
Riding Towards Safety: Examining the Patterns of Motorcycle Accidents in Malaysia
1 Introduction
1.1 Malaysian Traffic Fatalities
1.2 Motorcyclist Issue in Malaysia
2 Objective and Methodology
2.1 The Screening Method
3 Results and Discussion
4 Conclusion
References
Development of Automatic Cervical Brace for Neck Pain Rehabilitation
1 Introduction
2 Methodology
3 Results and Discussion
4 Conclusion and Recommendations
References
Recommend Papers
![Intelligent Manufacturing & Mechatronics: Proceedings of Symposium, 29 January 2018, Pekan, Pahang, Malaysia (Lecture Notes in Mechanical Engineering) [1st ed. 2018]
9789811087882, 9781615839490, 9811087881](https://ebin.pub/img/200x200/intelligent-manufacturing-amp-mechatronics-proceedings-of-symposium-29-january-2018-pekan-pahang-malaysia-lecture-notes-in-mechanical-engineering-1st-ed-2018-9789811087882-9781615839490-9811087881.jpg)
File loading please wait...
Citation preview
Lecture Notes in Mechanical Engineering
Mohd Hasnun Arif Hassan Mohd Nadzeri Omar Nasrul Hadi Johari Yongmin Zhong Editors
Proceedings of the 2nd Human Engineering Symposium HUMENS 2023, Pekan, Pahang, Malaysia
Lecture Notes in Mechanical Engineering Series Editors Fakher Chaari, National School of Engineers, University of Sfax, Sfax, Tunisia Francesco Gherardini , Dipartimento di Ingegneria “Enzo Ferrari”, Università di Modena e Reggio Emilia, Modena, Italy Vitalii Ivanov, Department of Manufacturing Engineering, Machines and Tools, Sumy State University, Sumy, Ukraine Mohamed Haddar, National School of Engineers of Sfax (ENIS), Sfax, Tunisia Editorial Board Francisco Cavas-Martínez , Departamento de Estructuras, Construcción y Expresión Gráfica Universidad Politécnica de Cartagena, Cartagena, Murcia, Spain Francesca di Mare, Institute of Energy Technology, Ruhr-Universität Bochum, Bochum, Nordrhein-Westfalen, Germany Young W. Kwon, Department of Manufacturing Engineering and Aerospace Engineering, Graduate School of Engineering and Applied Science, Monterey, CA, USA Justyna Trojanowska, Poznan University of Technology, Poznan, Poland Jinyang Xu, School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai, China
Lecture Notes in Mechanical Engineering (LNME) publishes the latest developments in Mechanical Engineering—quickly, informally and with high quality. Original research reported in proceedings and post-proceedings represents the core of LNME. Volumes published in LNME embrace all aspects, subfields and new challenges of mechanical engineering. To submit a proposal or request further information, please contact the Springer Editor of your location: Europe, USA, Africa: Leontina Di Cecco at [email protected] China: Ella Zhang at [email protected] India: Priya Vyas at [email protected] Rest of Asia, Australia, New Zealand: Swati Meherishi at swati.meherishi@ springer.com Topics in the series include: . . . . . . . . . . . . . . . . .
Engineering Design Machinery and Machine Elements Mechanical Structures and Stress Analysis Automotive Engineering Engine Technology Aerospace Technology and Astronautics Nanotechnology and Microengineering Control, Robotics, Mechatronics MEMS Theoretical and Applied Mechanics Dynamical Systems, Control Fluid Mechanics Engineering Thermodynamics, Heat and Mass Transfer Manufacturing Engineering and Smart Manufacturing Precision Engineering, Instrumentation, Measurement Materials Engineering Tribology and Surface Technology
Indexed by SCOPUS, EI Compendex, and INSPEC. All books published in the series are evaluated by Web of Science for the Conference Proceedings Citation Index (CPCI). To submit a proposal for a monograph, please check our Springer Tracts in Mechanical Engineering at https://link.springer.com/bookseries/11693.
Mohd Hasnun Arif Hassan · Mohd Nadzeri Omar · Nasrul Hadi Johari · Yongmin Zhong Editors
Proceedings of the 2nd Human Engineering Symposium HUMENS 2023, Pekan, Pahang, Malaysia
Editors Mohd Hasnun Arif Hassan Faculty of Mechanical and Automotive Engineering Technology Universiti Malaysia Pahang Al-Sultan Abdullah Pekan, Pahang, Malaysia Nasrul Hadi Johari Faculty of Mechanical and Automotive Engineering Technology Universiti Malaysia Pahang Al-Sultan Abdullah Pekan, Pahang, Malaysia
Mohd Nadzeri Omar Faculty of Mechanical and Automotive Engineering Technology Universiti Malaysia Pahang Al-Sultan Abdullah Pekan, Pahang, Malaysia Yongmin Zhong School of Engineering RMIT University Melbourne, VIC, Australia
ISSN 2195-4356 ISSN 2195-4364 (electronic) Lecture Notes in Mechanical Engineering ISBN 978-981-99-6889-3 ISBN 978-981-99-6890-9 (eBook) https://doi.org/10.1007/978-981-99-6890-9 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore Paper in this product is recyclable.
Preface
Technological advancements have significantly benefited humans. Technology has led to the development of ergonomic tools and equipment that improve human comfort, reduce strain, and enhance overall productivity. From adjustable office chairs to ergonomic keyboards, these innovations promote proper posture and reduce the risk of musculoskeletal disorders. When it comes to road safety, technology has played a pivotal role in saving lives and preventing accidents. Advanced driver assistance systems (ADAS) equipped with sensors, cameras, and artificial intelligence algorithms help detect potential hazards, warn drivers, and even intervene if necessary. In the realm of sports technology, advancements have revolutionized training methodologies and performance analysis. Athletes now have access to wearable devices that monitor their biometric data, providing insights into their physical condition, performance metrics, and injury prevention. Further, technological advancements have led to sophisticated tools and methods for studying the human body’s mechanics and movement. High-speed cameras, force sensors, and motiontracking systems enable researchers to gain deeper insights into human locomotion, joint mechanics, and muscle activation patterns. These findings help design better prosthetics, rehabilitation programs, and ergonomic solutions tailored to individual needs. The “Unlocking Human Potential: The Future of Human Engineering” symposium seeks to delve into the cutting-edge field of human engineering, exploring the possibilities of augmenting and optimizing human capabilities through advancements in science, technology, and design. This symposium brings together experts from various disciplines to discuss and showcase innovative approaches, methodologies, and ethical considerations in the realm of human engineering. From neuroenhancement to prosthetics, cognitive augmentation to genetic engineering, this symposium aims to stimulate insightful discussions and inspire the creation of a future where human potential knows no bounds. Pekan, Malaysia
Mohd Hasnun Arif Hassan
v
Contents
Developing a Survey Tool to Measure Psychosocial Risk and Work Performance at a Workplace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nuruzzakiyah Mohd Ishanuddin, Hanida Abdul Aziz, and Ezrin Hani Sukadarin Enhancing Mental Health Through Ambient Lighting . . . . . . . . . . . . . . . . Ilhamy Isyraq bin Ahmad Fadzil, Aimi Shazwani Ghazali, Farahiyah Jasni, and Muhammad Hariz bin Hafizalshah Parameter Extraction of Muscle Contraction Signals from Children with ASD During Fine Motor Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nor Zainah Mohamad, Nur Azah Hamzaid, and Muhammad Haziq Ahmad Fauzi Design of Hose Roller for Firefighter: A Fatigue Study . . . . . . . . . . . . . . . . Mohammad Luqman Hakim Mustapha, Salwa Mahmood, Helmy Mustafa El Bakri, Ismail Abdul Rahman, Noorul Azreen Azis, and Mohd Rizal Buang Noise Risk Assessment on Noise Exposure Among Urban Rail Maintenance Workers Using Personal Monitoring Method . . . . . . . . . . . . M. Mifzal-Nazhan, J. Azlis-Sani, A. Nor-Azali, Y. Nur-Annuar, S. Shahrul-Azhar, and A. Mohd-Zulhelmi Bangla Text Summarization Analysis Using Machine Learning: An Extractive Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mizanur Rahman, Sajib Debnath, Masud Rana, Saydul Akbar Murad, Abu Jafar Md Muzahid, Syed Zahidur Rashid, and Abdul Gafur Human Factors: Drivers’ Speed Choice on Relatively Low-Speed Limit Roads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Othman Che Puan, Azlina Ismail, Khairil Azman Masri, and Muhammad Shafiq Mohd Rozainee
1
19
33
43
57
65
81
vii
viii
Contents
Topology Optimization for Custom Bed-Resting Ankle Foot Orthosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Amir Mustakim Ab Rashid, Effi Zuhairah Md Nazid, Muhammad Hazli Mazlan, Azizah Intan Pangesty, and Abdul Halim Abdullah
95
Influence of Environmental Factors and Road Characteristics in Commuting Accidents Among Public University Staffs . . . . . . . . . . . . . 107 Mohd Najib Yaacob and Khairiah M. Mokhtar Effects of Material Properties in Developing the Ear Prosthetics . . . . . . . 119 Abdul Halim Abdullah, Mohd Noor Asnawi Mohd Noordin, Suziana Ahmad, Nor Fazli Adull Manan, and Shahrul Hisyam Marwan Study of Primary Stability of Hip Implant for Semi Hip Replacement by Using Finite Element Analysis . . . . . . . . . . . . . . . . . . . . . . . 133 Haslina Abdullah, Mohamad Shukri Zakaria, and Norfazillah Talib Investigation of Mental Health Condition Among Factory Worker During Covid pandemic–A Cross-Sectional Study . . . . . . . . . . . . . . . . . . . . 145 Irna Syahira Hassan, Nur Fazhilah Abdul Razak, Junaidah Zakaria, and Ezrin Hani Sukadarin The Influence of Body Balance Towards the Golf Putting Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Abdul Raouf Abas, Mohd Nadzeri Omar, Nasrul Hadi Johari, and Mohd Hasnun Arif Hassan Risk Assessment for Manual Handling Activities in a Dairy Industry . . . 173 Khairulhafiy Muhammad Ruzairi, Ezrin Hani Sukadarin, Mirta Widia, and A. Alaman Brief Review of Recent Study on Fluid–Structure Interaction Modeling of Blood Flow in Peripheral Arterial Disease . . . . . . . . . . . . . . . . 185 M. Firdaus M. Fauzi, Nasrul Hadi Johari, and M. Jamil M. Mokhtarudin Head Injury During Heading of Two Types of Sepak Takraw Balls: Analytical Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 Nik M. Haikal M. Hassan, Nasrul Hadi Johari, Mohd Hasnun Arif Hassan, Idris Mat Sahat, Mohd Nazderi Omar, and Zulkifli Ahmad Fluid–Structure Interaction (FSI) Modelling in Stenotic Carotid Artery Bifurcation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209 A. Rusydan Alias and Nasrul Hadi Johari Prediction of Atherosclerosis in Peripheral Arterial Disease Using Computational Fluid Dynamics Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 Ukasyah Zulfaqar Shahrulakmar, Nasrul Hadi Johari, Juhara Haron, Chandran Nadarajan, and M. Nadzeri Omar
Contents
ix
Impact Analysis of Motorcycle Helmet: Finite Element Modeling . . . . . . 239 N. Aimi Huda and M. S. Salwani The Protective Performance of Different Types of Motorcycle Helmets in Terms of HIC and BrIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249 N. Q. Radzuan, M. H. A. Hassan, M. N. Omar, and K. A. Abu Kassim Measuring Running Performance Through Technology: A Brief Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263 Siti Rabiatull Aisha Idris Experimental Study of Gait Monitoring on Wearable Shoes Insole and Analysis: A Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273 Nur Wahida Saadion and Mohd Azrul Hisham Mohd Adib Preliminary Ergonomics Analysis of Sit-Stand (STS) Desk on the Patient with Lower Back Pain Problem: A Case Study . . . . . . . . . . 289 Muhammad Rafli Salim Hasan Raza, Mohd Azrul Hisham Mohd Adib, and Nurul Shahida Mohd Shalahim Developing a Survey Tool to Measure Human Factors Constructs for Personal Hearing Protector (PHP) Use Among Industrial Workers—First Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 Nur Syafiqah Fauzan, Mirta Widia, and Ezrin Hani Sukadarin A Review on the Pedal Error Cases Among Car Drivers in Malaysia . . . 313 Nursya Mimie Ayuny Ismail, Mohamad Zairi Baharom, Zulkifli Ahmad, Mohd Hasnun Arif Hassan, Juffrizal Karjanto, Zulhaidi Mohd Jawi, and Khairil Anwar Abu Kassim Study of Anxiety Parameters and Sensors Related to Monitoring the Anxiety Concentration Index Level Among Archer Athletes: A Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327 Nur Khalijah Kamarudin, Wan Nurlisa Wan Ahmad, and Mohd Azrul Hisham Mohd Adib EEG and EMG-Based Multimodal Driver Drowsiness Detection: A CWT and Improved VGG-16 Pipeline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339 Mamunur Rashid, Mahfuzah Mustafa, Norizam Sulaiman, and Md Nahidul Islam Rehabilitation and Gamification Technology Device for Lower Extremities Patient: A Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351 Mohd Adib Syazwi Ismail and Mohd Azrul Hisham Mohd Adib The Importance of Proper Motorcycle Helmet Buckling: A Scientific Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363 N. Q. Radzuan, M. H. A. Hassan, M. N. Omar, N. A. Othman, M. A. Mohamad Radzi, and K. A. Abu Kassim
x
Contents
A Short Review on Development of Table Tennis Robotic Launcher . . . . 377 Irlina Jazlin Jamaludin, Zulkifli Ahmad, and Mohamad Zairi Baharom Reusability Study of 3D Printing Mould and Resin Casting for Takraw Ball Launcher Wheel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389 Idris Mat Sahat and Nasrul Hadi Johari Development of Noise Risk Assessment (NRA) and Management System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399 Kirubalini Asok Kumar, Nur Syafiqah Fauzan, Mirta Widia, Ezrin Hani Sukadarin, Nor Liyana Man, and Mohd Ikhwan Mohd Ibrahim Framework of Safety Helmet Compliance Detection and Employee Tracking by Using Quick Response (QR Code) Technology . . . . . . . . . . . . 415 Nuraini Wahidah Rusli, Hanida Abdul Aziz, and Naz Edayu Mat Nawi e-HIRARC Tool for Brick Laboratory in Civil Engineering Department at TVET (Technical and Vocational Education Training) Campus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425 Masita Hassan, Hanida Abdul Aziz, and Mohd Zahidi Rahim A Review of Biomechanical and Psychosocial Risk Factors Among Workers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437 Khairulhafiy Muhammad Ruzairi, Ezrin Hani Sukadarin, Mirta Widia, and A. Alaman Knowledge and Awareness of Road Safety Among University Students . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445 Nur Nadhirah Najwa Musni, Wan Norlinda Roshana Mohd Nawi, and Mirta Widia Riding Towards Safety: Examining the Patterns of Motorcycle Accidents in Malaysia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455 N. Q. Radzuan, M. H. A. Hassan, M. N. Omar, N. A. Othman, and K. A. Abu Kassim Development of Automatic Cervical Brace for Neck Pain Rehabilitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469 M. Z. Ahmad Fazril, Nur Haizal Mat Yaacob, Norsuhaily Abu Bakar, Mohamad Shaban AlSmadi, and Nasrul Hadi Johari
About the Editors
Dr. Mohd Hasnun Arif Hassan earned his first degree in Mechanical Engineering from the Technische Hochschule Bingen, Germany, in 2010. During the final year of his undergraduate study, he was offered a scholarship by Universiti Malaysia Pahang (UMP) to pursue a Master’s degree in Mechanical Engineering at the University of Malaya in Kuala Lumpur, which he graduated with distinction in 2012. After that, he embarked on his Ph.D. journey at UMP where he studied about the head injury sustained by soccer players due to heading manoeuvre. He completed his Ph.D. study in 2016 and then continued to serve UMP as a senior lecturer. His research interests include finite element modelling of the interaction between human and sports equipment, instrumentation of sports equipment, and injury prevention particularly with regards to sports and traffic accidents. His work aims to apply engineering principles in sports not only to enhance the performance of an athlete but also to prevent injuries. Dr. Mohd Nadzeri Omar received the B.Eng. (Hons) and Ph.D. degrees from RMIT University, Melbourne, Australia, in 2013 and 2017, respectively. He is a senior lecturer with the Faculty of Mechanical and Automotive Engineering Technology, Universiti Malaysia Pahang. He is also attached to the Human Engineering Research Group which focuses on research, development, and innovations in human-centered technology and products. His research interests include soft tissue modelling, sports technology, biomechanical engineering, and mechatronics. Dr. Nasrul Hadi Johari obtained his Ph.D. in Biofluid Mechanics from Imperial College London, United Kingdom. He is currently a senior lecturer at the Faculty of Mechanical and Automotive Engineering Technology, Universiti Malaysia Pahang. Dr. Johari’s research activities include computational modeling of blood flow, tissue mechanics, and mass transport in the cardiovascular system, with applications ranging from evaluating the hemodynamic performance of medical devices to predict the outcome of endovascular interventional procedures. He is also interested in computational and experimental modeling of the interaction between human
xi
xii
About the Editors
and sports equipment particularly in improving training aid systems and injury prevention. Dr. Yongmin Zhong is currently an Associate Professor with the School of Aerospace, Mechanical and Manufacturing Engineering, at RMIT University, Australia. His research interests include computational engineering, haptics, soft tissue modeling, surgical simulation, aerospace navigation and control, intelligent systems, and robotics.
Developing a Survey Tool to Measure Psychosocial Risk and Work Performance at a Workplace Nuruzzakiyah Mohd Ishanuddin, Hanida Abdul Aziz , and Ezrin Hani Sukadarin
Abstract This paper aims to develop a survey tool for psychosocial risk and work performance in the manufacturing industry in Malaysia. A cross-sectional study was conducted among 258 respondents from the manufacturing plant. The validity and reliability of a set questionnaire adapted from the Copenhagen Psychosocial Questionnaire (COPSOQ III), NIOSH Generic Job Stress Questionnaire and Individual Work Performance Questionnaire (IWPQ 1.0) instruments were tested using Exploratory Factor Analysis (EFA) and reliability analysis. The results showed that the originated ten construct measures of psychosocial risk factors and work performance were reduced into eight construct measures understudy after conducting factor analysis by Principal Component Analysis as a dimensional reduction method. This current study is essential to explore the presence of psychosocial risk factors that underlying in the manufacturing industry which might affect worker performance and well-being. Also, for future research purposes, this study can be utilised as the main tool to explore the psychosocial risk factors and work performance in other sectors. Keywords Exploratory factor analysis · Psychosocial risk factors · Work performance
N. M. Ishanuddin Department of Occupational Safety and Health, DSH Institute of Technology, Kuala Lumpur, Malaysia H. A. Aziz Faculty of Industrial Sciences and Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, Pahang, Malaysia E. H. Sukadarin (B) Department of Chemical Engineering Technology, Faculty of Engineering Technology, Universiti Tun Hussein Onn Malaysia, Pagoh Campus, Johor, Malaysia e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_1
1
2
N. M. Ishanuddin et al.
1 Introduction The Rapid changes in industrial technology concerning elevated demands in production and resources lead to automation of machinery—more advanced systems being implemented in the plant has led to new types of risk, such as psychosocial risk— generated from the organisational working environment, which negatively impacts the mental health of worker [1, 2]. Nowadays, automation has taken over most manufacturing jobs in the plant. Global development in technology consecutively, alteration in the psychosocial work environment and work intensity could not be avoided as the enhancement corresponds with current technologies and demands [3, 4]. In Malaysia, the manufacturing sector is one of the most significant contributors to the country’s revenue. Department of Occupational Safety and Health (2018) [5], Malaysia, reported across five years from 2015 until 2019 that the manufacturing sector had the highest number of occupational accidents compared to the other sectors. Working in the manufacturing sector exposes workers to many types of physical and mechanical hazards. At least to know that workers are emotionally drained and mentally exhausted from working for long, laborious working hours [6, 7]. Working in a manufacturing plant makes psychosocial risk an unseen hazard [8]. The emergence of psychosocial risk should not be seen as less priority over other types of hazards. Lately, studies regarding psychosocial and mental health aspects in work settings have arisen due to the adverse effect that has been latent over the years, especially in the working community [9–11]. Extensive research and management should be considered to ensure the safety and well-being of the workers [12]. Psychosocial risk is determined as the potential psychosocial hazards to cause harm to the workers [13, 14]. While psychosocial risks at work refer to the specific aspects of work design and organisation and management of work, also the social context can result in negative physical, psychological and social outcomes such as violence and high job demand [15]. Determination of psychosocial risk can enhance the wellbeing of workers and improve the working environment. One study suggested that controlling psychosocial hazards may prevent an accident at work [16]. To investigate the emergence of psychosocial risk, the psychosocial work environment, which includes the organisational aspect that incorporates the work nature needs to be in consideration. One of the indicators of inadequate safety at work is the multiple occurrences of accidents at work. An increasing number of mental health problems at work with relatively detrimental consequences follow concerning major mental health issues at work [17]. This issue leads directly to the deterioration of the work performance of the workforce and organisation revenue [18]. The presence of risk at the workplace might interrupt workers’ performance and organisational productivity since performance at work is measured through the competency and proficiency of the job task performed at work. Performance at work has been a significant measure in occupational health studies [19]. Eurofound and European Agency for Safety and Health at Work in 2014 reported that work performance also related to psychosocial risk factors other than adverse health outcomes. A poor working environment with psychosocial
Developing a Survey Tool to Measure Psychosocial Risk and Work …
3
risk exposes workers towards mental health deterioration and compromises work performance and productivity [18, 20, 21]. Another study found that psychological risks such as negative work behaviour can influence work performance in terms of technicality central to the job [22]. Psychosocial stressors encompass a few organisational aspects at the workplace were outlined by the World Health Organization (WHO), Health and Safety Executive (HSE), European Agency for Safety and Health at Work and the International Labour Organization (ILO) to govern the mental health well-being of the workers adequately [2, 6, 23, 24]. Some of the risk factors emphasised by these agencies include job demand, interpersonal relationships at work, job control, career development and others. The Malaysian government also took the initiative by using law enforcement to improve workplace safety and health (OSHA 1994). This law embodied the fact that every workplace ensures the safety and well-being of the workers at the workplace. Globally, psychosocial risk is becoming an issue concerning safety health and public health practitioner [25]. Obvious impacts on the working population include poor performance, unreliable decisions, impaired judgement, accidents, missed deadlines and increased costs in business [26]. Other than that, in terms of work performance wise, it can cause low motivation and commitment, a dispute among colleagues, human error and poor decision-making skills [27]. Lack of awareness among developing and underdeveloped countries contributes to harming workers’ health [28]. Active prevention to manage the intangibility of this type of risk is essentially vital. This paper intends to present the process of developing a survey tool of psychosocial risk factors and work performance in a manufacturing plant. The aforementioned dimensional construct of psychosocial risk will be determined. Then, using the firstgeneration statistical analysis technique—Principal Component Analysis (PCA)—a more robust study construct is designed. Finally, the paper discussed the result of the analysis.
2 Material and Method A pilot study was conducted before the actual data collection and the instrument was found reliable (Cronbach alpha = 0.729). During the pilot study, the electronic questionnaire version was distributed to the workers, and the constructed questionnaire was tested in terms of reliability. After improving the comprehension of the questions and suitability of the work context in that particular plant, the questionnaire set is ready to distribute for actual data collection. Respondents in this study are the workers working in one selected manufacturing plant. A purposive sampling technique has been employed in distributing the survey questionnaires to respondents in a manufacturing plant. Inclusion criteria include mental health workers with at least 1 year of working experience. While exclusion criteria include using any prescribed medication and illicit drug usage. The questionnaires were printed and distributed directly to the respondent during a training organised by the plant. Workers selected
4
N. M. Ishanuddin et al.
for this study were asked for voluntary participation, with participation being taken as consent. The consent form was given before the workers answered the questionnaire. An explanation about the survey was given prior to completing the survey. Upon completion, a total of 267 completed questionnaires were returned and yielded a response rate of 95%. There are a total of 650 workers in the plant. The sample size representative of the workers in this study is 242. It is determined based on Krejcie and Morgan’s sample size determination table. Krejcie and Morgan’s sample size calculation was based on p = 0.05 where the probability of committing a type I error is less than 5% or p < 0.05. The questionnaire was adapted from the Copenhagen Psychosocial Questionnaire (COPSOQ) [29], the NIOSH Generic Job Stress Questionnaire (NGJSQ) [30] and Individual Work Performance Questionnaire (IWPQ) [19]. A set of questionnaires was formed from 3 different questionnaires. The questionnaire is administered to all of the respondents. This questionnaire consists of 3 main parts. Part A consists of demographic questions, including gender, age, nationality, marital status, educational level, departments, years of working and health status-related questions. Part B of the questionnaire consists of 7 psychosocial risk factors: interpersonal relationships at work, job demands, job control, career development, environment and equipment, job content and role in the organisation. While part C consists of questions related to the work performance of the workers which are task performance, contextual performance and counterproductive work behaviour. There are items total of 63 questions which included 10 factors in this study before conducting PCA. Data from the questionnaires were keyed in manually and before that it was coded into different values for each of the responses. For example, 1 = always, 2 = often, 3 = sometimes, 4 = seldom and 5 = never. A reliability analysis was conducted and the instrument was found to be reliable. Data were examined for normality using the Kolmogorov–Smirnov test. The validity of the questionnaire was analysed using EFA by PCA to reduce the constructed measure into more accurate and precise measurements to investigate psychosocial risk factors and work performance in the manufacturing industry.
3 Results and Discussions 3.1 Reliability Analysis of the Instrument Internal consistency of the instrument using Cronbach Alpha value indicated the reliability of the items used to measure the factors under study. To determine the instrument’s internal consistency, Cronbach Alpha for each psychosocial risk factor and work performance factor were analysed. Table 1 shows the internal consistency total of 10 factors included in this study. From Table 1, job control, environment and equipment, job content and role organisation have low Cronbach’s alpha values with α = 0.618, α = 0.634, α = 0.596 and α = 0.608 respectively. Job content
Developing a Survey Tool to Measure Psychosocial Risk and Work …
5
Table 1 Research construct and Cronbach’s alpha value Psychosocial risk factors
Cronbach’s alpha value
Interpersonal relationships at work
0.796
No of items 4 12
Job demand
0.747
Job control
0.618
6
Career development
0.836
4
Environment and equipment
0.634
8
Job content
0.596
6
Role in organisation
0.608
5
Work performance factors
Cronbach’s alpha value
No of items
Task performance
0.915
5
Contextual performance
0.902
8
Counterproductive work behaviour
0.904
5
had the lowest value among all, with α = 0.596. One of the assumptions regarding low alpha value is due to the low number of items for the factors [31]. Besides, low Cronbach’s alpha can also indicate that the questions administered to the respondents are interpreted differently in which there is a need for improvement of the more understandable context of the questions. Interpersonal relationships at work and job demand had acceptable Cronbach’s alpha values with 0.796 and 0.747, respectively. While the career development factor had a good Cronbach’s alpha value of 0.836, Cronbach’s alpha scores between α = 0.60 and α = 0.70 could be considered borderline, but in general, they did not consider poor [32]. Task performance, contextual performance and counterproductive work behaviour factors all had excellent Cronbach’s alpha values with 0.915, 0.902 and 0.904, respectively. The excellent Cronbach’s alpha value indicated that the respondents agree or disagree on the items collectively for work performance factors. This is supported by Hoekstra et al. [33], the preferable score or answer for each participant, it has to produce the same result when the questionnaire is once again administered under the same test conditions, which is referred to as the high reliability of the score test. Thus, some of the instruments used in the current study show relatively acceptable Cronbach’s alpha value, and some need appropriate improvement.
3.2 Demographic Data From Table 2, most of the respondents are male, which encompasses 89.5%, and the rest are female, contributing only 10.5% of the respondents. The majority of the respondents that joined in this study are 26–35 years old age group (69%). This was followed by 16–25 years and 36–45 years with 14.7% and 13.2% respectively. The minority age group of the respondents is the 46–65 years age group with a
6
N. M. Ishanuddin et al.
Table 2 Demographic information of the respondents Demographic
Items
Frequency (N = 258)
Percentage (%)
Gender
Male
231
89.5
Female
27
10.5
16–25 years
38
14.7
26–35 years
178
36–45 years
34
46–65 years
8
3.1
255
98.8
Age
Nationality
Malaysian Other
Educational level
69 13.2
3
1.2
Certificate
120
46.5
Diploma
112
43.4
25
9.7
Bachelor degree Master
1
0.4
Years of
≤Five years
177
68.6
working
6–10 years
68
26.4
11 years and above
13
5
percentage of 3.1%. For nationality status, most of the respondents with 98.8%, are Malaysian, and the rest, 1.2%, are foreign workers. In terms of education level, only 1 (0.4%) of the participants have a Master degree, and 25 respondents, with 9.7%, have a bachelor’s degree. The majority of the respondents, 120 (46.5%), have a basic education level which is a certificate level, followed by 112 respondents, with 43.4% with a diploma (academic qualification). For the distribution of years of working, most of the respondents, 68.6%, are in the shortest working period, which is ≤ five years of working experience. There are 68 respondents with 26.4% in the group with 6–10 years of working experience. The remaining 13 respondents with 5% having the most extended working period in the company with 11 years and above, considered senior workers in the plant.
3.3 Exploratory Factor Analysis (EFA) EFA is performed by PCA in this study. It is a method to extract and reduce the number of items to a smaller number of variables. The PCA is done to extract such factors as it may allow for the loss of information as little as possible [34]. Also, a smaller set of construct measures is simpler to understand and be used in further analysis [35]. An EFA was conducted on 63 items with a Varimax rotation with Kaiser Normalization. The selection of factors to retain in the study takes into account Kaiser Criteria (eigenvalues greater than one), scree plot analysis, criteria based on the number of total variances explained (at least more than 50%), and KMO [25]. To determine
Developing a Survey Tool to Measure Psychosocial Risk and Work …
7
the suitability of the EFA analysis, the sampling adequacy test was conducted by Kaiser–Meyer–Olkin (KMO) and Bartlett’s test [26]. KMO values of more than 0.7 show that the factor analysis for the data is significant [36]. The result has presented Bartlett’s Test of Sphericity as significant with Chi-Square = 8459.107 and p-value < 0.001, KMO of sampling adequacy has appeared as 0.812, which is indicated as an excellent value because it surpasses the suggested value of 0.7. These two methods are vital so that the construct of data is relevant to proceed with factor analysis. Table 3 shows the changes in the dimensional construct under study. Aforementioned, the selection of the constructed measure to be retained after conducting dimensional reduction (PCA) are few criteria, including Kaiser Criteria (eigenvalues greater than one), scree plot analysis, criteria based on the number of total variances explained (at least more than 50%), and KMO value. The purpose of the dimensional construct into a smaller set of variables is to reduce random variables into a significant purpose-driven construct under study. In other words, PCA is a dimensionality reduction technique that condenses the original variables to a smaller number of significant principal components [37, 38]. Table 4 shows the psychosocial and work performance factors present in the manufacturing industry. Task and Contextual performance are combined as Factor 1. Factor 1 is found to be the strongest factor that influences workers. In contrast, the weakest factor that can affect the manufacturing industry workers is Job control. The second factor, followed after Factor 1, is Job Demands and Counterproductive Work Behaviour, which then become consecutive of Factor 2 and Factor 3. Job demand is part of psychosocial factors while counterproductive work behaviour is from work performance factors. Next, Factors 4, 5 and 6 are demonstrated by Environment and Table 3 Changes in the dimensional construct of the factors under study Original construct Factor
Source
Construct retained for the current study
Item
Factor
Item
Psychosocial risk
7
S1, S2, S3, S4, F1, F2, F3, COPSOQ F4, F5, F6, F7, F8, F9, F10, [29] and F11, F12, J1, J2, J3, J4, J5, NGJSQ [30] J6, C1, C2, C3, C4, E1, E2, E3, E4, E5, E6, E7, E8, R1, R2, R3, R4, R5, R6, O1, O2, O3, O4, O5,
6
S1, S2, S3, S4, F4, F5, F10, F11, J2, J3, J5, C1, C2, C4, E2, E3, E4, E7, E8, R1, R2, R4, R5
Work performance
3
T1, T2, T3, T4, T5, P1, P2, P3, P4, P5, P6, P7, P8, B1, B2, B3, B4, B5
IWPQ 1.0 [19]
2
T1, T2, T3, T4, T5, P2, P3, P4, P5, P6, P7, P8, B, B2, B3, B4, B5
63
–
8
40
Total
10
Note S = Interpersonal relationships, F = Job demands, J = Job control, E = Environment, R = Job content, O = Role in organisation, T = Task performance, P = Contextual performance, B = Counterproductive work behaviour
8
N. M. Ishanuddin et al.
Equipment, Job Content and Career Development factors. Interpersonal Relationships at Work emerged as the 7th Factor that contributes to the factors that affect the well-being of the workers in the manufacturing industry. Therefore, from the result achieved, in the manufacturing industry, excellent construct to conduct the psychosocial study will be job demands, environment and equipment, job content, career development, interpersonal relationships at work and job control. Instead, the role in the organisation construct is weak to use in the study. While, for work performance study, task performance, contextual performance and counterproductive work behaviour are the precise construct that can be applied. Table 5 shows the result of the factor analysis, which presented the retained items in the final construct using PCA. The items that are retained have an eigenvalue of more than 1. Thus, eight factors with eigenvalues of more than one have accounted for 51.37% of the total variance. From the results, Factor 1 is termed task and contextual performance with 12 items ranging from 0.608 to 0.768 of factor loading. Factor 1 originated from two factors which are task performance and contextual performance factors which then merged and became one factor after conducting dimensional reduction using PCA. Both Factor 2 (Job Demand) and Factor 5 (Job Content) consist of 4 items which range from 0.610 to 0.703 and 0.603 and 0.658, respectively. Also, the four items construct applied to Factor 7, which is termed as Interpersonal Relationships at Work. The items have factor loading ranging from 0.699 to 0.759. Next, Factor 3 and Factor 4, termed as Counterproductive Work Behaviour and Environment and Equipment, consist of 5 items construct. Factor 3 has factor loading ranging from 0.780 to 0.828, while Factor 4 has factor loading ranging from 0.660 to 0.802. Lastly, Career Development Factor and Job Control Factor, known as Factor 6 and Factor 8, are included with 3 item construct. Three items in Factor 6 have factor loading 0.766, 0.799 and 0.853. While three items in Factor 8 have factor loading of 0.672, 0.677 and 0.693. In short, the PCA had reduced the original ten factors understudy with 63 items in total into eight factors that construct a measure of psychosocial risk and work performance study with 40 items. The final research construct and the mean and standard deviation values can be seen in Table 6. Table 4 Contributing psychosocial risk and work performance factors Contributing construct
Factors
Total % of variance
51.37%
Factor 1
Task and contextual performance (12.704%)
Factor 2
Job demands (6.971%)
Factor 3
Counterproductive work behaviour (6.671%)
Factor 4
Environment and equipment (5.995%)
Factor 5
Job content (5.162%)
Factor 6
Career Development (5.080%)
Factor 7
Interpersonal relationships at work (4.747%)
Factor 8
Job control (4.039%)
0.72
0.608
T4
P8
0.61
F11
B4
0.647
0.635
F10
F5
0.703
0.721
P2
Job Demands
F4
0.737
0.735
P3
0.74
T5
T2
0.756
0.755
T3
P5
0.769
0.761
P4
T1
0.786
0.77
P7
Task and Contextual Performance
Component
P6
Items
Rotated Component Matrixa
0.828
Counterproductive Work Behaviour
Environment and Equipment
Table 5 Result of factor analysis by Principle Component Analysis (PCA)
Job Content
Career Development
InterPersonal Relationship at work
(continued)
Job Control
Developing a Survey Tool to Measure Psychosocial Risk and Work … 9
0.653
R2
0.688 0.66
E8
E4 0.658
0.699
E2
Job Content
R1
0.802 0.712
E7
0.786 0.78
B5
B2
Environment and Equipment
E3
0.82
Counterproductive Work Behaviour
0.79
Job Demands
B3
Task and Contextual Performance
Component
B1
Items
Rotated Component Matrixa
Table 5 (continued)
Career Development
InterPersonal Relationship at work
(continued)
Job Control
10 N. M. Ishanuddin et al.
Environment and Equipment
Job Content
0.806
0.803
0.869
0.709
3.739
4.562 2.874
4.314 2.718
Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization a Rotation converged in 8 iterations
8.71
5.487
17.31
10.905
% of Var
Eigen-value
5.936
2.621
4.16
2.093
3.322
1.925
3.055
0.587
0.904
α
0.697
0.677 0.672
J2
J5
0.693
0.701 0.699
S3
S4
Job Control
J3
0.759 0.709
S2
0.766
C1
InterPersonal Relationship at work
S1
0.853 0.799
C2
Career Development
C4
0.939
0.634
Counterproductive Work Behaviour
0.603
Job Demands
R5
Task and Contextual Performance
Component
R4
Items
Rotated Component Matrixa
Table 5 (continued)
Developing a Survey Tool to Measure Psychosocial Risk and Work … 11
12
N. M. Ishanuddin et al.
Table 6 Finalise research construct Psychosocial risk factors
Mean
S.D.
Interpersonal relationships at work (S) S1
How often do you get help and support from your colleagues?
2.17
0.951
S2
How often do you get help and support from your immediate superior?
2.30
0.951
S3
Is there a good atmosphere and good cooperation between you and your colleagues?
1.66
0.794
S4
Is there good cooperation between the colleagues at work?
1.60
0.733
3.44
1.080
Job demand (F) F1
Is your workload unevenly distributed so it piles up?
F2
Do you have enough time for your work tasks?
2.50
0.896
F3
Does your work require you to remember a lot of things?
2.48
0.926
F4
Does your work require you to make quick and difficult decisions?
2.66
0.899
F5
Do you have to make very important decisions at your workplace?
2.53
0.996
F6
Does your work put you in emotionally disturbing situations?
3.56
1.085
F7
Does your work demand a great deal of concentration or constant attention or high level of precision?
2.34
1.015
F8
Does your work require that you have very clear and precise eyesight?
2.12
0.906
F9
Could your work injure other people or affect the well-being of others if you make mistakes in your work?
3.12
1.286
F10
Could it cause financial losses if you make mistakes in your work?
2.85
1.335
F11
Does your work demand you to come up with new ideas?
2.27
0.885
F12
How many break times between heavy workloads do you have?
2.92
0.772
Job control (J) J1
Do other people make decisions regarding your work tasks?
3.27
0.932
J2
Do you have a say in choosing who you work with?
3.09
0.956
J3
Can you influence the amount of work assigned to you?
3.16
0.957
J4
Can you decide whenever to take a break?
2.65
0.956
J5
Do you have any influence on your work environment?
2.91
0.966
J6
If you have some personal business, is it possible for you to leave 3.84 your place of work for half an hour without special permissions?
1.143
Career development (C) C1
Do you have the possibility of learning new things through your work?
1.93
0.820
C2
Does your work give you the opportunity to develop your skills? 1.90
0.909
C3
Are you certain regarding the opportunities for promotion and advancement in the next few years?
1.011
2.50
(continued)
Developing a Survey Tool to Measure Psychosocial Risk and Work …
13
Table 6 (continued) Psychosocial risk factors C4
Mean
Are you certain about whether your job skills will be used and valued five years from now?
S.D. 2.08
0.961
Environment and equipment (E) E1
How often does your job expose you to verbal abuse or confrontations with clients or the general public?
2.05
1.130
E2
How often does your job expose you to the threat of physical harm or injury?
2.63
1.084
E3
The level of noise in the area in which I work is usually high
2.66
0.941
E4
The level of lighting in the area in which I work is usually poor
2.47
1.154
E5
The temperature of my work area during the hot weather is usually comfortable
2.83
0.995
E6
The air in my work area is clean and free of pollution
3.16
2.152
E7
The overall quality of the physical environment where I work is poor
2.45
1.105
E8
My work area is awfully crowded
2.36
1.065
3.35
0.918
Job content (R) R1
How often does your job require you to work very fast?
R2
How often does your job require you to work very hard?
3.30
0.874
R3
Do you have too little to do at work?
2.36
0.911
R4
How often is there a marked increase in the workload?
3.00
0.788
R5
How often is there a marked increase in the amount of concentration required on your job?
3.18
0.814
R6
How often can you use the skills from the previous experience and training?
3.40
1.071
Role in organisation (O) O1
Do you know exactly which areas are your responsibility?
2.14
0.749
O2
Do you know exactly what is expected of you at work?
2.47
0.834
O3
Are contradictory demands placed on you at work?
3.37
0.972
O4
Do you sometimes have to do things, which ought to have been done in a different way?
3.06
0.857
O5
Do you sometimes have to do things, which seem to you to be unnecessary?
3.35
1.009
Task performance (T) T1
I managed to plan my work so that it was done on time
3.79
0.896
T2
My planning was optimal
3.82
0.869
T3
I kept in mind the results that I had to achieve in my work
3.96
0.891
T4
I was able to separate main issues from side issues at work
3.90
0.896
T5
I was able to perform my work well with minimal time and effort 3.67
0.873 (continued)
14
N. M. Ishanuddin et al.
Table 6 (continued) Psychosocial risk factors
Mean
S.D.
Contextual performance (P) P1
I took on extra responsibilities
3.24
0.895
P2
I started new tasks myself, when my old ones were finished
3.59
0.835
P3
I took on challenging work tasks, when available
3.45
0.838
P4
I worked at keeping my job knowledge up-to-date
3.83
0.863
P5
I worked at keeping my job skills up-to-date
3.88
0.897
P6
I came up with creative solutions to new problems
3.60
0.858
P7
I kept looking for new challenges in my job
3.56
0.915
P8
I actively participated in work meetings
3.49
0.856
Counterproductive work behaviour (B) B1
I complained about unimportant matters at work
2.08
1.030
B2
I made problems greater than they were at work
1.70
0.965
B3
I focused on the negative aspects of a work situation, instead of on the positive aspects
1.87
1.112
B4
I spoke with colleagues about the negative aspects of my work
2.01
1.106
B5
I spoke with people from outside the organisation about the negative aspects of my work
1.65
1.049
High performance at work is important to ensure the job is delivered timely as per required. Having bad relationships can cause one to develop poor work performance at work. Poor work performance is expressed as violence, insulting, complaining and others. According to, different types of interpersonal relationships can cause significant changes in negative work behaviour. Interpersonal arguments among peers at work are certainly associated with counterproductive work behaviour [39]. High workers’ performances are basically because of the high motivation to work. Career advancement opportunities might be one of the reasons that impact the workers’ motivation [40]. Thus, undoubtedly low career development might lead to negative emotional responses among workers that turned into counterproductive behaviour too. Abrey and Smallwood [41] also highlighted that physical work environment correlates with performance in terms of productivity which in this study factors of physical work condition and negative behaviour are included together. Not only those, but researchers also believe that social interaction at work may also affect worker’s contextual performance. This hypothesis is also supported by the study of Shaukat, Yousaf and Sanders [42].
Developing a Survey Tool to Measure Psychosocial Risk and Work …
15
4 Conclusion Finally, the factor analysis by PCA to find the constructed measure of psychosocial factors and workers’ performance in a manufacturing industry is studied. This study managed to develop 8-factor constructs consisting of 40 items which include (1) task and contextual performance, (2) job demands, (3) counterproductive work behaviour, (4) environment and equipment, (5) job content, (6) career development, (7) interpersonal relationships and (8) job control. The eight construct measure of psychosocial risk factors with work performance factors is valid to be used in a manufacturing setting. However, the current study only validated eight factors which, for future study, other latent factors might be present. Thus, the found construct measure can be used to conduct a psychosocial and work performance study in the manufacturing sector. Generally, the determined Cronbach’s alpha value is within the acceptable limit, indicating the instrument is found to be reliable for the study. KMO and Bartlett’s test is also found significant, thus indicating the suitability of data for structure detection. For future research purposes, this study can be utilised as the main tool to explore the psychosocial risk factors and work performance in other sectors.
References 1. Sousa, C., Magalhaes, J., Neves, D., Neves, M., (2019). Study of Psychosocial Risk Factors in the Portuguese Navy Machine Drivers. In SHO2019: International Symposium On Occupational Safety And Hygiene (pp. 204–207). Portugal: Portuguese Soc Occupational Safety & Hygiene, Dps-Universidade Minho, Guimaraes, 4800–058, Portugal Categories /Classification. 2. Isha ASN, Javaid, MU, Zaib Abbasi A, Bano S, Zahid M, Memon MA, …, Imtiaz N (2020) Malay validation of Copenhagen psychosocial work environment questionnaire in context of second generation statistical techniques. Biomed Res Int 11.https://doi.org/10.1155/2020/768 0960 3. Boucekkine R, Hritonenko N, Yatsenko Y (2014) Health, work intensity, and technological innovations. J Biol Syst 22(2):219–233. https://doi.org/10.1142/S0218339014400038 4. Cascio WF, Montealegre R (2016) How technology is changing work and organisations. Annu Rev Organ Psych Organ Behav 3(1):349–375. https://doi.org/10.1146/annurev-orgpsych-041 015-062352 5. DOSH (2016) website department of occupational safety and health Malaysia—DOSH profile. Retrieved September 25, 2018, from http://www.dosh.gov.my/index.php/en/occupational-acc ident-statistics/by-sector 6. Cockburn W, Milczarek M, Irastorza X, Rial González E (2012) The management of psychosocial risks across the European union: findings from ESENER. In: contemporary occupational health psychology: global perspectives on research and practice, vol 2, pp 162–183, https:// doi.org/10.1002/9781119942849.ch10 7. Engelbrecht GJ, de Beer LT, Schaufeli WB (2020) The relationships between work intensity, workaholism, burnout, and self-reported musculoskeletal complaints. Hum Factors Ergon Manuf 30(1):59–70. https://doi.org/10.1002/hfm.20821 8. Zakaria J, Che Hassan CR, Hamid MD, Sukadarin EH (2019) Safety climate factors at selected chemical manufacturing plant in Malaysia. Process Safety Progress, (August), 1–10. https:// doi.org/10.1002/prs.12096. Berg BL, Lune H, Lune H (2004) Qualitative research methods for the social sciences, vol 5. Pearson, Boston, MA
16
N. M. Ishanuddin et al.
9. Balderrama-Armendariz CO, Valadez-Torres SG, Maldonado-Macías AA, Avelar-Sosa L, Camacho-Alamilla M del R, Garcia-Alcaraz JL (2017) Analysis of burnout syndrome, musculoskeletal complaints, and job content in middle and senior managers: case study of manufacturing industries in Ciudad Juárez, Mexico. Work 58(4):549–565. https://doi.org/10.3233/wor172642 10. Draksler K, Hafner ND, Arneri´c N, Fikfak MD (2018) Restructuring of a textile manufacturing company and workers’ health. New Solut 28(1):131–150. https://doi.org/10.1177/104829111 8755755 11. Clays E, Casini A, Van Herck K, De Bacquer D, Kittel F, De Backer G, Holtermann A (2016) Do psychosocial job resources buffer the relation between physical work demands and coronary heart disease? A prospective study among men. Int Arch Occup Environ Health 89(8):1299– 1307. https://doi.org/10.1007/s00420-016-1165-z 12. Sukadarin EH, Suhaimi S, Abdull N (2012) Preliminary study of the safety culture in a manufacturing industry. Int J Humanit Soc Sci 2(4):176–183 13. Di Tecco C, Jain A, Valenti A, Iavicoli S, Leka S (2017) An evaluation of the impact of a policy-level intervention to address psychosocial risks on organisational action in Italy. Saf Sci 100:103–109. https://doi.org/10.1016/j.ssci.2017.05.015 14. Nuruzzakiyah MI, Ezrin Hani S, Hanida AA (2020) The Correlation between psychosocial risk factors and work performance in manufacturing industry Malaysia. J Public Health Med, Spec 1:23–29 15. Eurofound & European Agency for Safety and Health at Work (EU-OSHA) (2014) Psychosocial risks in Europe: prevalence and strategies for prevention (A joint report from the European Foundation for the Improvement of Living and Working Conditions and the European Agency for Safety and Health at Work). Luxembourg: Publications Off. https://doi.org/10.2806/70971 16. Guadix J, Carrillo-Castrillo J, Onieva L, Lucena D (2015) Strategies for psychosocial risk management in manufacturing. J Bus Res 68(7):1475–1480. https://doi.org/10.1016/J.JBU SRES.2015.01.037 17. Dupret E, Teherani M, Feltrin M, Bocéréan C, Pejtersen JH (2012) Psychosocial risk assessment: French validation of the Copenhagen Psychosocial Questionnaire (COPSOQ). Scandinavian Journal of Public Health 40(5):482–490. https://doi.org/10.1177/140349481245 3888 18. Schaufeli WB (2017) Applying the Job Demands-Resources model: A ‘how to’ guide to measuring and tackling work engagement and burnout. Organ Dyn 46(2):120–132. https:// doi.org/10.1016/j.orgdyn.2017.04.008 19. Koopmans L, Bernaards C, Hildebrandt V, Van Buuren S, Van Der Beek AJ, de Vet HCW (2012) Development of an individual work performance questionnaire. Int J Prod Perform Manag 62(1):6–28. https://doi.org/10.1108/17410401311285273 20. Katsuro P, Gadzirayi CT, Taruwona M, Mupararano S (2010) Impact of occupational health and safety on worker productivity : A case of Zimbabwe food industry. J Bus 4(13):2644–2651 21. Rasool SF, Wang M, Zhang Y, Samma M (2020) Sustainable work performance: the roles of workplace violence and occupational stress. Int J Environ Res Public Health 17(3). https://doi. org/10.3390/ijerph17030912 22. Wang L, Zhang J (2020) The effect of psychological risk elements on pilot flight operational performance. Hum Factors Ergon Manuf 30(1):3–13. https://doi.org/10.1002/hfm.20816 23. Parent-Thirion A, Vermeylen G, Houten G van, Lyly-Yrjänäinen M, Biletta I, Cabrita J, Niedhammer I, (2012) Fifth Eurofound Working Conditions Survey. Retrieved from https://www. eurofound.europa.eu/sites/default/files/ef_publication/field_ef_document/ef1182en.pdf 24. Leka S, Jain A (2010) Health impact of Psychosocial Hazards at Work: An Overview. World Health Organization. World Health Organization (WHO). Retrieved from http://apps.who.int/ iris/bitstream/10665/44428/1/9789241500272_eng.pdf 25. Rosário S, Azevedo LF, Fonseca JA, Nienhaus A, Nübling M, Da Costa JT (2017) the Portuguese long version of the Copenhagen psychosocial questionnaire II (COPSOQ II)—a validation study. J Occup Med Toxicol 12(1). https://doi.org/10.1186/s12995-017-0170-9
Developing a Survey Tool to Measure Psychosocial Risk and Work …
17
26. International Labour Office (2012) Integrating Health Promotion into Work place OSH Policies: Participant’s Workbook Pub. Retrieved from https://books.google.com.my/books?id=Cr8 ws7Pp5FwC&pg=PA131&lpg=PA131&dq=indirect+cost+of+psychosocial+at+work+are+ insurance+cost,retirement+funds,+safety+and+health,+medical+assistance,+counselling& source=bl&ots=OE46i2KCxE&sig=x7mmAELJn1teK-bQiVtHOxfzOaM& 27. Chylova M (2019) Psychosocial factors in work environment and employee mental health. In: 17th international conference on work and organisational psychology—past, present, and challenges to the future. pp 176–186 28. Javaid MU, Bano S, Mirza MZ, Isha ASN, Nadeem S, Jawaid A, …, Kaur P (2019) Connotations of psychological and physiological health in the psychosocial work environment: an industrial context. Work 64(3):551–561. https://doi.org/10.3233/WOR-193016 29. Pejtersen JH, Kristensen TS, Borg V, Bjorner JB (2010) The second version of the Copenhagen Psychosocial Questionnaire. Scandinavian Journal of Public Health 38(SUPPL. 3):8–24. https://doi.org/10.1177/1403494809349858 30. National Institute of Occupational Safety and Health US (2017) CDC—Organization of Work: measurement tools: NIOSH generic job stress questionnaire—NIOSH. Retrieved October 18, 2020, from https://www.cdc.gov/niosh/topics/workorg/detail088.html 31. Kvalheim SA, Antonsen S, Haugen S (2016) Safety climate as an indicator for major accident risk: can we use safety climate as an indicator on the plant level? Int J Disaster Risk Reduct 18:23–31. https://doi.org/10.1016/j.ijdrr.2016.05.011 32. George D, Paul Mallery W (2003) SPSS for windows step by step a simple guide and reference fourth edition (11.0 update) answers to selected exercises. Retrieved from https://wps.ablong man.com/wps/media/objects/385/394732/george4answers.pdf 33. Hoekstra R, Vugteveen J, Warrens MJ, Kruyen PM (2019) An empirical analysis of alleged misunderstandings of coefficient alpha. Int J Soc Res Methodol 22(4):351–364. https://doi.org/ 10.1080/13645579.2018.1547523 34. Chan APC, Wong FKW, Hon CKH, Lyu S, Javed AA (2017) Investigating ethnic minorities’ perceptions of safety climate in the construction industry. J Safety Res 63:9–19. https://doi. org/10.1016/j.jsr.2017.08.006 35. Kasim H, Che Hassan CR, Hamid MD, Emami SD, Danaee M (2018) Determination of factors affecting safety practices in Malaysian radiation facilities. Saf Sci 104:70–80. https://doi.org/ 10.1016/j.ssci.2017.12.031 36. Aminian M, Dianat I, Miri A, Asghari-Jafarabadi M (2016) The Iranian version of the Copenhagen Psychosocial Questionnaire (COPSOQ) for assessment of psychological risk factors at work. Health Promot Perspect 7(1):7–13. https://doi.org/10.15171/hpp.2017.03 37. Tomaszewski JE, Hipp J, Tangrea M, Madabhushi A (2014) Machine vision and machine learning in digital pathology. pathobiology of human disease: a dynamic Encyclopedia of disease mechanisms. Elsevier Inc. https://doi.org/10.1016/B978-0-12-386456-7.07202-6 38. Mendlein A, Szkudlarek C, Goodpaster JV (2013) Chemometrics Encyclopedia of forensic sciences: second edition, 646–651.https://doi.org/10.1016/B978-0-12-382165-2.00259-2 39. Kundi YM, Badar K (2021) Interpersonal conflict and counterproductive work behavior: the moderating roles of emotional intelligence and gender. Int J Confl Manag 32(3):514–534. https://doi.org/10.1108/IJCMA-10-2020-0179 40. Sulea C, Virga D, Maricutoiu LP, Schaufeli W, Zaborila Dumitru C, Sava FA (2012) Work engagement as mediator between job characteristics and positive and negative extra role behaviors. Career Dev Int 17(3):188–207. https://doi.org/10.1108/13620431211241054 41. Abrey, M., Smallwood, J. J. (2014). The effects of unsatisfactory working conditions on productivity in the construction industry. In Procedia Engineering (Vol. 85, pp. 3–9). Elsevier Ltd. https://doi.org/10.1016/j.proeng.2014.10.522. 42. Shaukat R, Yousaf A, Sanders K (2018) Examining the linkages between relationship conflict, performance and turnover intentions: role of job burnout as a mediator. Electron Libr 34(1):1–5
Enhancing Mental Health Through Ambient Lighting Ilhamy Isyraq bin Ahmad Fadzil, Aimi Shazwani Ghazali , Farahiyah Jasni , and Muhammad Hariz bin Hafizalshah
Abstract Untreated stress is one of the prominent near-to-lead mental health problems. To address the mental health issue, the effect of ambient lighting was investigated in the study to determine its effect on a person’s stress and relaxation conditions. The coloured ambient illuminations consisting of red, green, and blue (RGB) hues were chosen with care to meet the study’s purpose. Digit Span Test and a piece of loud music were chosen as stressors to induce stress while calm music was used to stimulate relaxed conditions during the experimental session. A physiological parameter was monitored via a Galvanic Skin Response (GSR) sensor in classifying the stress and relax states. At the end of the experiment, the participants were interviewed verbally by the experimenter on their feelings after being exposed to the ambient illuminations. Using within-subject design as the experimental setups, the sensor’s output was then quantified using a series of developed questionnaires. As a result, the participants experienced higher anxiety and reactance under the stress condition compared to the relaxed condition through the questionnaire. Interestingly based on GSR reading and quantitative analysis using questionnaires, data showed that the participants experienced the most relaxed state when being exposed to greencoloured ambient illumination (66%) compared to other colours. In contrast, bluecoloured ambient lighting elicited the most stress state through the analysed data collected from GSR reading (33%) and red-coloured ambient lighting (66%) from quantitative analysis using questionnaire and qualitative analysis (interview). Future studies might utilize different coloured ambient illumination for the same settings. Keywords Galvanic skin response (GSR) · State-trait anxiety inventory (STAI) · Psychological reactance · Ambient lighting · Mental health
I. I. A. Fadzil · A. S. Ghazali (B) · F. Jasni · M. H. Hafizalshah Department of Mechatronics Engineering, International Islamic University Malaysia, Jln Gombak, Kuala Lumpur, Malaysia e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_2
19
20
I. I. A. Fadzil et al.
1 Introduction People’s emotional, psychological, and social well-being all fall under the category of mental health which affects human thoughts, feelings, and actions. Additionally, it influences how people deal with stress, interact with others and make choices. Mental health is critical at all stages of life, including childhood, adolescence, and maturity. In the study covering multinational citizenship, a significant degree of stress, depression, anxiety, psychological distress, post-traumatic stress symptoms, poor sleep quality and insomnia were experienced by the frontline population of healthcare workers worldwide, especially during the pandemic era [1]. Due to the social stigma attached to mental health issues, many people prefer not to discuss them with anyone. They may also be in denial or less likely to seek care than the general population due to a variety of factors such as a lack of awareness, geographic proximity to the nearest hospitals, and a lack of self-confidence to confront the truth [2]. Also, lack of motivation and hectic life as students have hindered Malaysian youth from getting the benefits of a healthy routine such as doing simple exercises [3]. Importantly, an appropriate ambient light also may provide a sufficient and comfortable environment to human beings in alleviating stress, no study has been conducted in Malaysia to prove whether Malaysians, especially youth, will be affected by the lighting [4, 5]. It has also remained unknown which ambient lighting colours can aid to create a relaxing atmosphere. The main objective of the research is to investigate the influence of ambient lighting on enhancing Malaysian youth’s mental health. The objective can be divided into several sub-objectives which are: (1) To design a ubiquitous experimental setup in creating a room with three colours of ambient lighting which are blue, red and green. (2) To investigate the effect of ambient lighting on human mental health concerning stress and relaxation conditions and compare it with neutral ambient lighting (benchmark). (3) To establish a relationship between ambient lighting and human psychological and physiological effects through a questionnaire, Galvanic Skin Response (GSR) sensor and interview. The research will cover these hypotheses for the sensor reading, quantitative and qualitative analysis: Hypothesis 1. There is a main significant effect of conditions (based on phases) on GSR reading. Hypothesis 2. There is a significant interaction effect of ambient lighting and sequence for the ambient lighting exposure on GSR reading. Hypothesis 3. There is a main significant effect of benchmark condition (neutral ambient lighting) and all coloured ambient lighting conditions on quantitative analysis using a questionnaire.
Enhancing Mental Health Through Ambient Lighting
21
Hypothesis 4. There is a main significant effect of blue, red and green-coloured ambient lighting conditions on qualitative analysis. Hypothesis 5. There is a significant effect of coloured ambient lighting conditions on the GSR reading, quantitative analysis using a questionnaire and qualitative analysis (interview).
1.1 Ambient Lighting Ambient lighting, which is also known as general lighting is a primary source of light in a room. It can improve the sense of warmth and depth of a room or space and provide a comfortable and energetic feeling to users. Light is a common ambient medium to express additional information not only peripherally and calmly, but it is also environmentally stimulant to create atmosphere, evokes moods, and provide immersive experiences [5]. There are three main qualities of light concerning colour. Firstly, brightness is the amount of light given off by a light source. It is expressed in terms of lumen or lux. Brighter light can intensify emotion while low light does not remove the emotion but keeps them steady. Next, the saturation of light is the intensity of a colour. The more saturated hue the colour is, the more the emotion will be amplified. On the other hand, muted colours such as grey, green, and brown can dampen emotion. Hue is defined as a colour or shade, and it has been proven that natural light makes a happier environment. However, colours created by artificial light can also summon different emotions and psychology in a human [6]. The study used colours that are measured in Correlated Color Temperature (CCT) and differentiated into three main categories which are neutral, warm, and cool ambient white light. As specified in [7], CCT is a one-dimensional metric used to assess the visual quality of nominal white light sources. Its ease of usage makes it a popular proxy for light source colour quality. However, CCT lacks colour fidelity for research purposes. Due to the loss of information, while converting the spectral power distribution of light sources into a one-dimensional metric, two light sources with identical CCTs can appear perceptually different. The relation between fundamentals of CCT chromaticity and colour rendering are two important aspects to measure the colour quality of light sources for general illumination purposes [8]. The research focused on three colours of ambient lighting that can be seen from the CCT chromatic diagram. Blue, red, and green (RGB) colours are chosen as it can be said the primary light beam when observed in the CCT. The mixture of all three colours in various ways can reproduce a broad array of colours. As the classifications of cool and warm ambient lighting, blue light is considered cool ambient lighting while red colour is classified as warm ambient lighting. The green colour is the extra colour to be studied in this research as in the visible light spectrum, in which the colour is placed between blue and red colour. Moreover, neutral white ambient light also was used in this study for the benchmark phase and relax phase.
22
I. I. A. Fadzil et al.
1.2 Biosignals: Galvanic Skin Response (GSR) Several physiological sensors are being put under consideration to measure the mental conditions of the participants during the study (stress vs relax) under different ambient lighting such as Photoplethysmography (PPG) sensor [9] and temperature sensor [10]. Other than that, Galvanic Skin Response (GSR) sensor is a device which can be used to detect stress levels by measuring skin conductance [11]. It measures the electrical conductivity of the skin’s surface. Intense emotions would likely trigger the sympathetic nervous system, causing the sweat glands to produce more sweat. The method simply attaches the two electrodes of the GSR sensor to two fingers in one hand. There are many sweat glands in the hands and feet, making them ideal locations for GSR signals [12]. Its examination distinguishes between its skin conductance level (SCL) and its skin conductance response (SCR). The SCR amplitude and delay, as well as the average SCL value, are usually retrieved from the signal. Heart rate variability and GSR are by far the most accurate representations of stress levels, whereas respiration is seen as a reliable supporting signal [5]. Commonly, the conductivity of the skin is measured in micro- or millisiemens per centimetre (uS/cm or mS/cm). With the help of two classifiers which are Lazy IBk and Random Forest, the normalized datasets of the GSR sensor proved that the accuracy of the mentioned sensor can go up to 97.8% accuracy [11]. In the relaxed condition of the person, sweat is not secreted. Hence, the resistance of the skin is higher than in stress conditions. This results in lower conductive voltage. In the physical arousal condition of the person, sweat is secreted and causes the resistance of the skin to be low. This results in higher conductive voltage.
1.3 Stressor: Digit Span To induce stress and relax conditions, stressors will be used in the study. The stressor is a scenario, requirement, or situation that has the potential to cause stress, a biochemical change in behavioural, physiological, and/or psychological health. Examples of stressors that are commonly used in accessing human mental health are Stroop Color and Word Test (SCWT) [13] and Digit Span (DGS) [14]. In studying the ambient lighting effect on mental health, SCWT is not selected as a stressor since a huge problem will occur due to the lighting conditions used in the experimental set-ups. That is, the stressor involves displaying words in a variety of colours while in our study itself, the lighting will be varied. This situation most likely will alter the conditions and display of SCWT on the monitor. Therefore, DGS will be used in the study on enhancing mental health through ambient lighting. DSG is an assessment of verbal short-term and working memory that is available in two formats: forward and reverse. This is a verbal activity used as a stressor in the study, involving auditory stimuli and participant responses that are automatically
Enhancing Mental Health Through Ambient Lighting
23
assessed by the software. In an earlier study [11], participants were given a random sequence of digits and asked to repeat them in either the order in which they were presented (ahead span) or in reverse order (backward span). While forward and backwards-span tasks appear to be similar on the surface, they need relatively distinct cognitive capacities: the easier forward-span task involves verbal working memory and attention, whereas the backwards-span task also assesses cognitive control and executive function. These activities have been proven successful in eliciting stress conditions [11] for experimental works.
1.4 Questionnaire: State-Trait Anxiety Inventory (STAI) and Reactance State-Trait Anxiety Inventory (STAI) [15] focuses on adults to evaluate the degree of a person’s anxiety through a questionnaire delivered to adults. An individual’s responses to access state and trait anxiety questions are used to generate both short and reliable assessments. People who have higher STAI scores also have higher levels of stress and anxiety. Concern, tension, and stress are usually labelled as anxiety. It is typically accompanied by a stressful event, such as a big test or interview. Other than STAI, a set of questionnaires known as reactance also can be used to detect stress [16]. Psychological reactance is the driving state believed to occur when a person’s freedom is revoked or threatened with revoked. It is unpleasant motivational arousal (reaction) to offers, individuals, rules, or constraints that threaten or destroy specific behavioural rights. A reaction is an unpleasant form of motivational arousal that occurs when people’s liberty is threatened or taken away. It catalyzes reclaiming one’s liberty. The magnitude of the reaction is influenced by the relevance of the threatened freedom and the perceived threat’s scope. When an individual perceives that someone or something is removing or restricting their alternatives, they react [17].
2 Methodology 2.1 Experimental Set-Ups The experimental set-ups were created in such a way that it repetitively placed the participants in a state of mental and emotional distress. The arrangement was done to examine the differences in GSR readings when participants were in stress-induced physiological parameters known as “stress” or “relax” conditions while being under the influence of the blue, red and green colours of ambient lighting. The participants wore the GSR sensor throughout the entire experimental process from the rest phase in the beginning until the last psychological quantitative analysis
24
I. I. A. Fadzil et al.
Fig. 1 Flow of the experiment
was submitted to monitor the changes in their skin conductance. They only interacted with the monitor screen and music played by the headphones. Figure 1 shows the full process of the experiment. Each participant was greeted upon arrival and sat on a chair facing the laptop. A laptop was placed near the participant, for filling in the pre-and post-experiment questionnaires and the stressor, Digit Span Test was readily displayed on a dedicated laptop. This experiment consisted of three phases: (1) Introduction [10 min] (2) Experiment [31 min] (3) Closing [10 min]. For the study, there were four phases of engagement. The first one was the benchmark phase, and the other three phases were stress conditions under randomized order of colours of ambient lighting. The randomization was done to ensure the counterbalanced of different colours of ambient lighting in a controlled environment. Before starting the experiment, the participants were briefed about the experimental procedures and needed to read and sign a consent form. In the benchmark phase, the participants were asked to sit in a controlled environment with neutral ambient lighting which used a white colour light for three minutes in the rest phase. As a note, the participants were asked to put away their cell phones to avoid any distractions during the experiment while listening to calming music. The “Weightless” music by Macaroni Union has been used in this study in all resting phases as it has been proven that music acts as a therapy in reducing anxiety levels [18]. Then, the participants were asked to answer a series of psychological quantitative questionnaires using a Likert scale of 4 in Google Forms. It is important to know the current state of the participant which is likely to be relaxing was used as a benchmark for each stress level before stressors were placed. Before the stress conditions, the participants undergo a rest phase for three minutes and straight away first stress induction phase under randomized ambient lighting colour which required the participant to do IQ tests provided by the experimenter to induce “stress” condition for five minutes. To make sure stress was induced during the tests, the participants needed to listen to noisy and metal background music which was proven that loud music could induce stress to a person [19, 20]. The participants were equipped with Sony WH-1000XM3 wireless noise-cancelling headphones so that no external factors are disrupting their concentration besides helping them to fully occupy themselves with the provided background music along with the experiment.
Enhancing Mental Health Through Ambient Lighting
25
Then, they were asked to fill out again the psychological quantitative questionnaire and undergo relax phase same as before. In Digit Span test software, participants were given a random sequence of digits and asked to repeat them in which they were presented (ahead span). When participants enter the correct digit sequence, the following digit sequence is increased by one digit, and when they enter the incorrect digit sequence, the next digit sequence is decreased by one digit. All the participants will begin with a four-digit sequence. In other words, the higher the level, the longer the digit sequence that participants had to recall and enter as they were presented (ahead span). The next stress induction phase would be the same procedure as the first stress induction phase but with different ambient lighting colours. Lastly, the experimental procedure ended with the third and last induction stress of the last ambient lighting colour and the psychological quantitative analysis needed to be filled in again. In the end, participants were debriefed by the experimenter in oral form and received a small monetary reward or research credits for their participation at the end of the experiment. Note that, neutral white ambient lighting was used in the rest phase while the participants answered the psychology quantitative questionnaires. These included every coloured ambient lighting in the stress induction phase, in which the neutral white ambient lighting was quickly switched once the five-minute stress condition was over. One of the E27 LED bulbs was placed on top and middle of the room to be the main light while the other E27 LED bulb was placed behind the monitor to create enough light-colour intensity when the participants see the screen during the experiment. Moreover, the RGB strip was placed under the table to increase the illumination of the floor. These set ups can be seen in Fig. 2 under all lighting conditions. For this study, the “Light Meter—Lux & Kelvin” android application developed by Trajkovski Labs was used to identify the colour temperature and brightness of the bulb. The four colours were fixed based on the colour temperature. For the blue colour, the colour temperature was fixed at 7265 K, the red colour at 2340 K, the green colour at 5675 K and the neutral white ambient light at 4500 K. This four-colour temperature was fixed for all ambient lighting sources and in every experimental process. Based on Fig. 2 (neutral, red, green, and blue coloured ambient lighting),
Fig. 2 Experimental setup with coloured lighting conditions
26
I. I. A. Fadzil et al.
the lux of every coloured ambient lighting room also was fixed between 200 and 400 lx for all four coloured conditions which explained the suitable illumination light level for an office [21]. After each phase, participants completed a set of quantitative analytic questions to ascertain their psychological well-being. Four questions from the Psychological Reactance scale [22] and six questions from STAI [15] were used in the study. The sequence of questions enabled the construction of a valid measure of stress. The higher the scores, the greater the participants’ anxiety and reactance, and hence as hypothesised, the higher the GSR sensor readings should be. After the experiment was done, the participants were interviewed verbally by the experimenter about their opinions throughout the experimental process. They were asked which lighting colour influenced their mood and made them feel anxious during the stress test and otherwise.
3 Results and Discussion 3.1 Participants Thirty participants (15 males and 15 females) aged 21–25 years old participated in this research. The experiment lasted for 51 min in which at the end of the study, the participants were given drinks and snacks as a small token of appreciation. The participants were purposely sampled among students from Kulliyah of Engineering, International Islamic University Malaysia.
3.2 Hypotheses Testing In the SPSS, the reliability analysis of the scale analyzer was used to ensure that the GSR reading was reliable. For the analysis, the mean of normalised data for every 30 s was used. As results, some outliers have been removed from the analysis so that Cronbach’s Alpha (α) is more than 0.85 for all phases to have a good internal consistency [16]. For the baseline, data of GSR readings from 210 to 390 s were used in the analysis at α = 0.861. For blue-coloured ambient lighting stress, 480–690 s were taken resulting in α = 0.874. For the rest phase after blue-coloured ambient lighting stress, 810–900 s were taken resulting in α = 0.856. For the red-coloured ambient lighting stress, 1020–1290 s were taken resulting in α = 0.871. For the rest phase after red-coloured ambient lighting stress, 1380–1500 s were taken resulting in α = 0.870. For the green-coloured ambient lighting stress, 1620–1830 s were taken resulting in α = 0.865. Overall, Cronbach’s Alpha for all phases was maintained at α > 0.85.
Enhancing Mental Health Through Ambient Lighting
27
Also, the reliability analysis of the scale analyzer was used to check and make sure the quantitative analysis (psychological questionnaire) is reliable. The questionnaire for baseline and all stress phases was combined for all 30 participants. The questionnaire consisted of 10 items and the value for Cronbach’s Alpha for the questionnaire was α = 0.7. Hypothesis 1. A one-way repeated-measures ANOVA test was used for all phases in the experiment by comparing three relax and three stress conditions as independent variables and the GSR reading as dependent variables. It showed that there was a significant main effect of conditions (based on phases) on GSR reading, F(5, 25) = 2.117, p = 0.0097, partial ŋ2 = 0.297. The linear test of within-subject contrast indicates a significant relation between both independent and dependent variables, F(5, 145) = 3.472, p < 0.005, partial ŋ2 = 0.107. Further analysis was done, to check whether each of the phases has a significant difference in GSR reading. As a result, Bonferroni post hoc tests showed that several phases have significant difference readings of the GSR sensor which are: Benchmark and Rest Blue. p < 0.1, with a 90% confidence interval [−0.158, − 0.003]. Benchmark and Rest Red. p < 0.05, with a 90% confidence interval [−0.224, − 0.029]. Blue and Rest Blue. p < 0.05, with a 90% confidence interval [−0.249, −0.032]. Blue and Rest Red. p < 0.05, with a 90% confidence interval [−0.296, −0.077]. Red and Rest Blue. p < 0.05, with a 90% confidence interval [−0.211, −0.034]. Red and Rest Red. p < 0.05, with a 90% confidence interval [−0.267, −0.071]. Green and Rest Red. p < 0.05, with a 90% confidence interval [−0.239, −0.023]. As a result, hypothesis 1 was accepted through the findings shown earlier in finding the significant difference between relax and stress conditions regardless of the coloured ambient lighting. Hypothesis 2. To further investigate the effect of ambient lighting on stress conditions, another repeated measure analysis (ANOVA) was done by omitting the GSR reading from the baseline and rest phases. Also, the sequence of the coloured condition has been used as a between-subject factor. As an output, there is a significant effect of ambient lighting and sequence for stress condition on GSR reading with F(5, 24) = 4.598, p = 0.004, partial ŋ2 = 0.489. The finding is in line with the hypothesis. Based on the mean value of GSR reading for red-coloured ambient lighting, it recorded the lowest stress experienced (highest GSR reading) at green-blue-red and blue-green-red sequences while the highest stress experienced (lowest GSR reading) at red-blue-green, red-green-blue and green-red-blue sequences. Next, for green-coloured ambient lighting, it detailed the lowest stress experienced (high GSR reading) at blue-red-green and red-blue-green sequences and there was no data recorded at the highest stress experienced (lowest GSR reading) in all sequences for stress conditions for green-coloured ambient lighting. For the blue-coloured ambient
28
I. I. A. Fadzil et al.
lighting, it recorded the lowest stress experienced (high GSR reading) at red-greenblue and green-red-blue sequences while the highest stress experienced (lowest GSR reading) at blue-red-green, blue-green-red and green-blue-red sequences. Hypothesis 3. Based on the questionnaire, the participants experienced the lowest stress in the benchmark phase (M = 1.22, SD = 0.43) compared to other conditions. For red coloured ambient lighting stress induction phase, the participants were not relaxed and experienced the highest stress level (M = 2.03, SD = 0.97) compared to green (M = 1.86, SD = 0.81) and blue (M = 1.90, SD = 0.91) coloured ambient lightings. A repeated one-way measure of ANOVA test for quantitative analysis using a questionnaire was run by comparing the benchmark and three coloured ambient lighting as independent variables and the quantitative analysis using questionnaire scores as dependent variables. The results of the one-way repeated-measures ANOVA showed that there was a significant main effect of benchmark and all coloured ambient lighting conditions on quantitative analysis using a questionnaire, F(3, 27) = 19.416, p = 0.000, partial ŋ2 = 0.683. The linear test of within-subject contrast indicates a significant relation between both independent and dependent variables, F(3, 87) = 12.430, p < 0.005, partial ŋ2 = 0.3. Further analysis was done, to check whether the benchmark and all coloured ambient lighting conditions have significant differences in quantitative analysis using a questionnaire. As a result, Bonferroni post hoc tests showed that several phases have significant differences in quantitative analysis using a questionnaire which is: Benchmark and blue-coloured ambient lighting conditions. p < 0.05, with a 90% confidence interval [−0.895, −0.439]. Benchmark and red-coloured ambient lighting conditions. p < 0.05, with a 90% confidence interval [−1.040, −0.558]. Benchmark and green-coloured ambient lighting conditions. p < 0.05, with a 90% confidence interval [−0.782, −0.438]. This evidence supports the hypothesis that the relatedness of quantitative analysis using a questionnaire on benchmark and all coloured ambient lighting conditions affects participants’ ability in handling stress and there are significant differences between the benchmark and all coloured ambient lighting. Hypothesis 4. For the qualitative analysis, each participant was asked about their opinions on which coloured ambient lighting has the most effect on their moods during the stress test. Surprisingly, twenty-three participants indicated that redcoloured ambient lighting made them feel the most stress. In contrast, blue and green coloured ambient lighting induce relaxed feelings while answering the Digit Span test and listening to a piece of loud music. The results of the one-way repeated-measures ANOVA showed that there was a significant main effect of blue, red and green-coloured ambient lighting conditions on qualitative analysis (interview), F(2, 28) = 12.326, p = 0.000, partial ŋ2 = 0.468. The linear test of within-subject contrast indicates a significant relation between both independent and dependent variables, F(2, 58) = 12.627, p < 0.005, partial ŋ2 = 0.303.
Enhancing Mental Health Through Ambient Lighting
29
Further analysis was done, to check whether the blue, red and green coloured ambient lighting conditions have significant differences in qualitative analysis (interview). As a result, Bonferroni post hoc tests showed that several phases have significant differences in qualitative analysis (interview) which are: Blue and red coloured ambient lighting conditions. p < 0.05, with a 90% confidence interval [−1.346, −0.654]. Green and red coloured ambient lighting conditions. p < 0.05, with a 90% confidence interval [−1.310, −0.490]. This evidence supports the hypothesis that the relatedness of qualitative analysis (interview) on the blue, red and green-coloured ambient lighting conditions affects participants’ ability in handling stress through their opinions. But there was no evidence of a significant difference between blue and green-coloured ambient lighting conditions in qualitative analysis (interview). Hypothesis 5. The results of the one-way repeated-measures mixed ANOVA showed that there was a significant effect of all coloured ambient lighting conditions on the GSR reading, quantitative analysis (psychological questionnaire) and qualitative analysis, F(6, 24) = 4.353, p = 0.004, partial ŋ2 = 0.521. The linear test of within-subject contrast indicates a significant relation between both independent and dependent variables, F(2, 58) = 12.627, p < 0.005, partial ŋ2 = 0.303. Based on statistical analysis, it showed that the participants experienced the highest stress (lowest GSR reading) in blue-coloured ambient lighting (M = 0.4359, SD = 0.22273) while the participants experienced the lowest stress (highest GSR reading) in green-coloured ambient lighting (M = 0.4917, SD = 0.21508). For participant stress, red-coloured ambient lighting (M = 0.4537, SD = 0.20700) was not as high as green ambient lighting and not as low as blue ambient lighting. Next, the single mean value quantitative analysis using the questionnaire of all result from all coloured ambient lighting were analysed. The result showed that redcoloured ambient lighting (M = 2.022, SD = 0.74598) caused the participants to experience the highest stress while green-coloured ambient lighting (M = 1.8333, SD = 0.57642) induced the participants to experience the lowest stress. Blue-coloured ambient lighting (M = 1.8903, SD = 0.70757) was neither as high nor as low as red or green-coloured ambient lighting. In the single mean value of qualitative analysis (interview), the participants experienced the highest stress in red-coloured ambient lighting (M = 2.6333, SD = 0.71840) while experiencing the lowest stress in blue-coloured ambient lighting (M = 1.6333, SD = 0.61495). Green-coloured ambient lighting (M = 1.7333, SD = 0.73968) was not as high as red-coloured ambient lighting or as low as blue-coloured ambient lighting.
30
I. I. A. Fadzil et al.
4 Conclusion In conclusion, this study demonstrated the influence of coloured ambient lighting on mental health. The goals of the study were to find out whether ambient lighting can elicit stress and relaxation conditions through the GSR readings. In supporting the physiological sensor, a psychological questionnaire together with an interview have been conducted for additional quantitative and qualitative measures. Thirty participants were used in this project. The first objective has been achieved in designing a ubiquitous experiment in enhancing mental health using three colours of ambient lighting which are blue, red, and green. Next, the second objective which was to investigate the effect of ambient lighting on human mental health regarding stress and relaxation conditions and compare it with neutral ambient lighting (benchmark) had been achieved through descriptive analysis. That is, the findings in this study proved that all the hypotheses were accepted. In detail, the SPSS analysis showed that the greencoloured ambient lighting helped the participants to induce relaxed feelings through GSR and psychological questionnaire readings. While through the interview, the blue-coloured ambient induced a relax feeling the most. Lastly, establishing a relationship between ambient lighting and human psychological and physiological effects has been achieved through statistical analysis using the SPSS analysis. Recommendations for future works include the increment of the number of participants in each order of the coloured ambient lighting phase to make sure the data collected are more reliable and accurate. Next, varieties of coloured ambient lighting should also be used in future works as there is a wide range of coloured ambient illumination. Lastly, more advanced physiological signal sensors should be taken into consideration in detecting the stress of a person so that the motion artefact of participants will not interfere with the obtained data.
References 1. Sahimi HMS et al (2021) Depression and suicidal ideation in a sample of Malaysian healthcare workers: a preliminary study during the COVID-19 pandemic. 12:658174 2. Janko MR, Smeds MR (2019) Burnout, depression, perceived stress, and self-efficacy in vascular surgery trainees. 69(4):233–1242 3. Jacquart J et al (2020) Using exercise to facilitate arousal reappraisal and reduce stress reactivity: a randomized controlled trial. 18:100324 4. Jin C et al (2015) The effect of color light combination on preference for living room. In: 2015 12th China international forum on solid state lighting (SSLCHINA). IEEE 5. Yu B et al (2018) DeLight: biofeedback through ambient light for stress intervention and relaxation assistance. 22(4):787–805 6. Küller R et al (2006) The impact of light and colour on psychological mood: a cross-cultural study of indoor work environments. 49(14):1496–1507 7. Durmus A (2022) Correlated color temperature: use and limitations. Light Res Technol 54(4):363–375
Enhancing Mental Health Through Ambient Lighting
31
8. Ohno Y (2017) Color quality of white LEDs. In: III-nitride based light emitting diodes and applications. Springer, pp 457–480 9. Castaneda D et al (2018) A review on wearable photoplethysmography sensors and their potential future applications in health care. 4(4):195 10. Sano A et al (2018) Identifying objective physiological markers and modifiable behaviors for self-reported stress and mental health status using wearable sensors and mobile phones: observational study. 20(6):e9410 11. Amiruldin AA, Ghazali AS (2022) You are too loud! Classification of psychological conditions for stress detection system using galvanic skin response. In: Enabling industry 4.0 through advances in mechatronics. Springer, pp 51–60 12. Ladakis I, Chouvarda I (2021) Overview of biosignal analysis methods for the assessment of stress. Emerg Sci J 5(2):233–244 13. Lingaraj V et al (2022) Design of expert active KNN classifier algorithm using flow stroop colour word test to assess flow state. 1–14 14. Surya ND et al (2022) Subclinical hypothyroidism and digit span test performance in children: a systematic review and meta-analysis. 62(5) 15. Lavedán Santamaría A et al (2022) Diagnostic concordance between the visual analogue anxiety scale (VAS-A) and the state-trait anxiety inventory (STAI) in nursing students during the COVID-19 pandemic. 19(12):7053 16. Gupta AS, Mukherjee J (2022) Decoding revenge buying in retail: role of psychological reactance and perceived stress. Int J Retail Distrib Manag (ahead-of-print) 17. Brehm SS, Brehm JW (2013) Psychological reactance: a theory of freedom and control. Academic Press 18. Rather JA, Shrivastava Y (2019) Effect of music therapy on pre-competition anxiety in college level soccer players of Kashmir 19. Asif A et al (2019) Human stress classification using EEG signals in response to music tracks. 107:182–196 20. Krahé B, Bieneck S (2012) The effect of music-induced mood on aggressive affect, cognition, and behavior. J Appl Soc Psychol 42(2):271–290 21. ToolBox E (2004) Illuminance—recommended light level. https://www.engineeringtoolbox. com/light-level-rooms-d_708.html 22. Steindl C et al (2015) Understanding psychological reactance: new developments and findings. 223(4):205
Parameter Extraction of Muscle Contraction Signals from Children with ASD During Fine Motor Activities Nor Zainah Mohamad, Nur Azah Hamzaid , and Muhammad Haziq Ahmad Fauzi
Abstract This study was performed to extract meaningful parameters from Muscle contraction (MC) sensor signals, which could be used as biofeedback parameters for children with autism spectrum disorder (ASD) performance during fine motor activities. Four male children with ASD and 2 typical development (TD) male children participated in the experiment. Participants were asked to perform repetitive hand grips on the hand dynamometer while simultaneously recording measurements from the MC sensor on the forearm (flexor digitorum profundis). Because MMG signals are inherently mechanical, signal acquisition can be performed without separate circuitry to eliminate electrical noise interfaces, particularly 50 Hz noise. Handgrip strength using the MC sensor was 0.98 ± 0.69 V in children with ASD and 3.32 ± 0.56 V at TD. Comparison between the measured peak signal MC and hand dynamometer torque revealed a strong linear relationship with a high degree of agreement R = 98%, P < 0.0001. This study demonstrates that the MC sensor can accurately measure the contraction of a small muscle, the flexor digitorum profundis, in children. Keywords Parameter extraction · Muscle sensor · Autism · Fine motor · Grip strength
1 Introduction A study by Kern et al. [14] has found that the severity of the disorder was correlated with handgrip strength in participants diagnosed with ASD. Further investigation is required to determine the muscle weakness’ degree and consistency and also its possible treatments. Mechanomyography (MMG), or the mechanical muscle signal is widely used in clinical assessment and experimental evaluation to study muscle properties, including muscle function (MF), physiological movement, signal processing, N. Z. Mohamad · N. A. Hamzaid (B) · M. H. A. Fauzi Department of Biomedical Engineering, Faculty of Engineering, Universiti Malaya, Kuala Lumpur, Malaysia e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_3
33
34
N. Z. Mohamad et al.
prosthetic and/or switch control, and medical rehabilitation [11, 16, 23]. By observing the MMG pattern, therapists were able to determine the appropriate training for children with autism spectrum disorders (ASD). Compared to typical development (TD) individuals, ASD individuals tend to show lower upper extremity strength [15], handgrip [14], and pincer grip [6]. More study is required to compare the differences in motor strength between people with ASD and TD control participants to demonstrate fine motor impairments in ASD and to increase our understanding of ASD tests and therapies [2, 12, 25, 26]. There is no previous research that has examined the ability of the MC sensor to measure muscle movement in ASD children compared with TD children during fine motor activities [18, 20]. One of the mechanomyography sensors, i.e., MC sensor, can be used to compare muscle activity signals during this fine motor skills between ASD and typical development children to help therapies determine the relevant training for those with autism spectrum disorder (ASD). Among many parameters extracted from MC sensor include muscle forces of muscle contraction. Because MMG signals are inherently mechanical, signal acquisition can be facilitated because the signal can be acquired without the need for separate circuitry to eliminate electrical noise interfaces, particularly 50 Hz noise [10]. Signal peak and area under the curve of muscle contraction signals from MC could be used as MC features extraction [7, 18].
2 Methodology Six male children; 4 ASD and 2 typical developments (TD) with age 7 ± 0.69 years old participated in the experiment. Participants were seated on a suitable chair and their upper extremities were positioned so that the shoulder was adducted and neutrally rotated, the elbow was flexed 90 degrees, the forearm was neutral, and the wrist was extended 30 degrees. As presented in Fig. 1, an MC sensor was placed on the arm to measure the flexor digitorum profundis muscle signal. Participants were instructed to grasp a hand dynamometer (Fig. 2) and squeeze the dynamometer handle as hard as they could for 2–5 s. Three trials were performed for each hand, and the highest force value of the three trials was taken as the final value while simultaneously recording measurements from the MC sensor. Analysis of the signals from MC was performed using Sensmotion and MATLAB software. The MC-voltage signal was sampled at 5 kHz and low-pass filtered with a cutoff frequency of 10 Hz using MATLAB (Mathworks Inc., Natick, MA, USA). To determine the relationship between the torque measured on the dynamometer and the MC signal voltage during both experiments, the signal peak (SP) was compared with the peak torque for each contraction. Before performing descriptive statistical analyses and regression modeling, data for hand dynamometer torque and MC tension were normalized to each subject’s maximum muscle stimulation contraction to analyze the hand dynamometer torque and MC tension signals simultaneously.
Parameter Extraction of Muscle Contraction Signals from Children …
35
Fig. 1 The participant was asked to grab the hand dynamometer handle
Fig. 2 MC sensor placement on the flexor digitorum superficialis muscle
3 Results and Discussion The linear MC signal observed in Fig. 3 (the muscle contraction signal for an ASD child). The enveloped signals displayed a clearer illustration of the performance of the children during hand grip activities. Figure 4 shows that the measured peak signal MC and the hand dynamometer torque have a strong linear relationship with a high degree of association R = 98%, P < 0.0001. It also shows that the MC sensor could measure muscle contraction in a small muscle area, flexor digitorum profundis. The Bland Altman plot (Fig. 5) shows that the measured peak MC signal and hand dynamometer torque had a small standard deviation, SD = 0.21 with Bland Altman bias −2.377. All measured parameters are presented in Table 1.
36
N. Z. Mohamad et al.
MC Signal during hand grip - ASD Children 2.4
2.2
MC-signal (V)
2.0
1.8
1.6
1.4
1.2
1.0 0
2
4
6
8
10
12
14
Time (s)
Fig. 3 Linear muscle contraction signal for ASD child Fig. 4 Linear regression of peak hand dynamometer torque and MC signal peak
10 9
MC-signal peak (V)
8 7 6 5 4 3 2 1 0 -1 0
2
4
6
Dyna-torque peak (Kg)
8
10
Parameter Extraction of Muscle Contraction Signals from Children …
37
Bland-Altman Graph Difference Dyna-torque peak - MC-signal peak
-0.5 Mean + 1.96SD
-1.0 -1.5 -2.0 -2.5
Mean
-3.0 -3.5 -4.0
Mean - 1.96SD
-4.5 0
1
2
3
4
5
6
7
Difference Dyna-torque peak - MC-signal peak Fig. 5 Bland Altman plot for two methods data; using MC sensor and hand dynamometer
Table 1 Signal peaks (SP) relationships between the torque of the hand dynamometer and the torque of MC Parameter
Mean strength ASD (N = 6)
Mean strength TD (N = 2)
Coefficient of determination (R2 )
Standard deviation (SD)
Bland Altman bias
MC peak signal (V)
0.98 ± 0.69
3.32 ± 0.56
96%
0.21
−2.377
Hand Grip strength (KG)
3.09 ± 1.16
6.95 ± 1.01
MC sensor signal could represent the measurement of muscle contraction strength, although the ASD children have low grip torque as shown in Fig. 6. TD Children have higher torque values during the handgrip experiment compared to ASD children, which could be detected with both the hand dynamometer and MC sensor. The average handgrip torque handgrip strength peak was 6.95 ± 1.01 kg for children with ASD, and 3.09 ± 1.16 kg for TD. While for MC, 0.98 ± 0.69 V for children with ASD, and 3.32 ± 0.56 V for TD. The findings corroborate earlier research and show that TD children’s handgrip strength is much lower than that of children with ASD [13]. The study of [8] suggests
38
N. Z. Mohamad et al.
Fig. 6 A graph of peak average torque for both hand dynamometer and MC signal in ASD and TD child
that motor synergy is impaired in ASD. The consequences of motor control can be far-reaching. These findings also support the study by Odeh et al. [21] which found that children with ASD have significant differences in complex motor skills, balance skills, and global motor performance compared to their peers (TD) in all three domains. Measurement of handgrip strength in adults and children is reliable and has been successfully used in autism research [1, 9]. MC sensor could be used as a replacement for other instruments such as EMG [22, 24], and hand dynamometer to measure muscle contraction force in autism. MC sensor has been used in various groups of individuals, such as paraplegics [18], and athletes [17]. To further this study, the MC sensor will be used in a study with a larger number of participants and different instructions, such as a pinch test, on different muscle targets. This study adds several novel elements when compared to other research on muscular contraction monitoring in people with ASD. Many previous studies have examined motor impairments in individuals with ASD using various techniques such as electromyography (EMG) [3–5, 19, 24] or other clinical assessments. However, this study’s use of the mechanomyography (MMG) sensor—more particularly, the MC sensor to measure muscular contraction in kids with ASD—represents a novel method. The application of the MC sensor to assess muscle strength in children with ASD during fine motor activities is a departure from traditional methods like hand dynamometers or manual strength assessments. While hand dynamometers have been widely used in research and clinical settings to evaluate grip strength, the MC sensor offers a different modality by directly measuring the mechanical vibrations
Parameter Extraction of Muscle Contraction Signals from Children …
39
associated with muscle contraction. This allows for a non-invasive and potentially more objective assessment of muscle activity. This study also shows the precision and dependability of the MC sensor as a tool for monitoring muscular contraction in kids with ASD by establishing a link between the MC signal and hand dynamometer torque. This comparison further supports the possibility of the MC sensor as an additional or alternative assessment tool for evaluating motor deficits in this population.
4 Conclusion This study shows that the MC sensor can be used to measure muscle signals in a person with low fine motor skills (grip torque). The results suggest that handgrip strength is significantly lower in children with ASD compared to TD children. The results of this study are consistent with previous research. The findings of this study help explain motor deficiencies in people with ASD, particularly in the area of fine motor skills. We provide insight into possible areas for targeted interventions and therapies by identifying differences in muscular contraction between children with ASD and children who are typically developing. The possibility of employing biofeedback to improve motor abilities in children with ASD while engaging in fine motor tasks has been made possible by the extraction of useful parameters from the MC sensor outputs. Acknowledgements The authors acknowledge all the Autism and Typical development children who participated in this study. Also the teachers from Makmal Pembelajaran Autism. This research was supported by the Knowledge Transfer Programme (KTP) through grant MRUN-2020-1A.
References 1. Alaniz ML, Galit E, Necesito CI, Rosario ER (2015) Hand strength, handwriting, and functional skills in children with autism. Am J Occup Ther 69(4):6904220030p1–6904220030p9. https:// doi.org/10.5014/ajot.2015.016022 2. Alsaedi RH (2020) An assessment of the motor performance skills of children with autism spectrum disorder in the gulf region. Brain Sci 10(9):607. https://doi.org/10.3390/brainsci1 0090607 3. Brahim S, Ahmad AH, Safri NM (2017) Case study on fine motor skills of special children when using light and sound tool. J Telecommun, Electron Comput Eng 9:41–45 4. Cattaneo L, Fabbri-Destro M, Boria S, Pieraccini C, Monti A, Cossu G, Rizzolatti G (2007) Impairment of actions chains in autism and its possible role in intention understanding. Proc Natl Acad Sci 104(45):17825–17830. https://doi.org/10.1073/pnas.0706273104 5. Cheng YTY, Tsang WWN, Schooling CM, Fong SSM (2018) Reactive balance performance and neuromuscular and cognitive responses to unpredictable balance perturbations in children with developmental coordination disorder. Gait Posture 62:20–26. https://doi.org/10.1016/j. gaitpost.2018.02.025
40
N. Z. Mohamad et al.
6. David FJ, Baranek GT, Wiesen C, Miao AF, Thorpe DE (2012) Coordination of precision grip in 2–6 years-old children with autism spectrum disorders compared to children developing typically and children with developmental disabilities. Front IntegrNeurosci 6.https://doi.org/ 10.3389/fnint.2012.00122 7. Dordevi´c S, Stanˇcin S, Megliˇc A, Milutinovi´c V, Tomažiˇc S (2011) MC sensor-a novel method for measurement of muscle tension. Sensors 11(10):9411–9425. https://doi.org/10.3390/s11 1009411 8. Emanuele M, Nazzaro G, Marini M, Veronesi C, Boni S, Polletta G, D’Ausilio A, Fadiga L (2021) Motor synergies: Evidence for a novel motor signature in autism spectrum disorder. Cognition 213:104652. https://doi.org/10.1016/j.cognition.2021.104652 9. G˛asior J, Pawłowski M, Jele´n P, Rameckers E, Williams C, Makuch R, Werner B (2020) Testretest reliability of handgrip strength measurement in children and preadolescents. Int J Environ Res Public Health 17(21):8026. https://doi.org/10.3390/ijerph17218026 10. Ibitoye M, Hamzaid N, Zuniga J, Hasnan N, Wahab A (2014) Mechanomyographic parameter extraction methods: an appraisal for clinical applications. Sensors 14(12):22940–22970. https:// doi.org/10.3390/s141222940 11. Islam MdA, Sundaraj K, Ahmad RB, Ahamed NU (2013) Mechanomyogram for muscle function assessment: a review. PLoS ONE 8(3):e58902. https://doi.org/10.1371/journal.pone.005 8902 12. Kaur M, Srinivasan SM, Bhat AN (2018) Comparing motor performance, praxis, coordination, and interpersonal synchrony between children with and without Autism Spectrum Disorder (ASD). Res Dev Disabil 72:79–95.https://doi.org/10.1016/j.ridd.2017.10.025 13. Kern JK, Geier DA, Adams JB, Troutman MR, Davis GA, King PG, Geier MR (2013) Handgrip strength in autism spectrum disorder compared with controls. J Strength Cond Res 27(8):2277– 2281. https://doi.org/10.1519/JSC.0b013e31827de068 14. Kern JK, Geier DA, Adams JB, Troutman MR, Davis G, King PG, Young JL, Geier MR (2011) Autism severity and muscle strength: a correlation analysis. Res Autism Spectr Disord 5(3):1011–1015. https://doi.org/10.1016/j.rasd.2010.11.002 15. Lourenço C, Esteves D, Corredeira R, Seabra A (2015) Children with autism spectrum disorder and trampoline training. Wulfenia J 22(5):342–351 16. Meagher C, Franco E, Turk R, Wilson S, Steadman N, McNicholas L, Vaidyanathan R, Burridge J, Stokes M (2020) New advances in mechanomyography sensor technology and signal processing: Validity and intrarater reliability of recordings from muscle. J Rehabil Assist Technol Eng 7:205566832091611. https://doi.org/10.1177/2055668320916116 - c S, Belušiˇc G, Megliˇc A, Uršiˇc M, Škorjanc A, 17. Megliˇc A, Uršiˇc M, Škorjanc A, Ðordevi´ - c S, Belušiˇc G (2019) The piezo-resistive MC sensor is a fast and accurate sensor for Ðordevi´ the measurement of mechanical muscle activity. Sensors 19(9):2108. https://doi.org/10.3390/ s19092108 18. Mohamad NZ, Hamzaid NA, Davis GM, Abdul Wahab AK, Hasnan N (2017) Mechanomyography and torque during FES-evoked muscle contractions to fatigue in individuals with spinal cord injury. Sensors (Switzerland) 17(7):1–15. https://doi.org/10.3390/s17071627 19. Mohd Nor MN, Jailani R, Tahir NM (2020) Feature selection of electromyography signals for autism spectrum disorder children during gait using Mann-Whitney test. Jurnal Teknologi 82(2). https://doi.org/10.11113/jt.v82.13928 20. Mohd Nordin A, Ismail J, Kamal Nor N (2021) Motor development in children with autism spectrum disorder. Front Pediatr 9.https://doi.org/10.3389/fped.2021.598276 21. Odeh CE, Gladfelter AL, Stoesser C, Roth S (2022) Comprehensive motor skills assessment in children with autism spectrum disorder yields global deficits. Int J Dev Disabil 68(3):290–300. https://doi.org/10.1080/20473869.2020.1764241 22. Pascolo P, Ragogna P, Cremaschi S, Mondani M, Carniel R, Corubolo M, Budai R (2010) Autism and motor acts: Experimental analysis on mylohyoid muscle EMG recordings during grasping-to-eat action. Biomed Sci Instrum 46 23. Mohamad Ismail RM, Lam CK, Sundaraj K, Rahiman MHF (2019) Hand motion pattern recognition analysis of forearm muscle using MMG signals. Bull Electr Eng Inform 8(2):533– 540. https://doi.org/10.11591/eei.v8i2.1415
Parameter Extraction of Muscle Contraction Signals from Children …
41
24. Safri NM, Sheng RTY (2019) Surface Electromyographic signals of special needs children during fine motor task. Int Biomed Instrum Tech-Nology Conf (IBITeC) 2019:17–20. https:// doi.org/10.1109/IBITeC46597.2019.9091717 25. Sung Y-S, Loh SC, Lin L-Y (2021) Physical Activity and motor performance: a comparison between young children with and without autism spectrum disorder. Neuropsychiatr Dis Treat 17:3743–3751. https://doi.org/10.2147/NDT.S343552 26. Wang Z, Magnon GC, White SP, Greene RK, Vaillancourt DE, Mosconi MW (2015) Individuals with autism spectrum disorder show abnormalities during initial and subsequent phases of precision gripping. J Neurophysiol 113(7):1989–2001. https://doi.org/10.1152/jn.00661.2014
Design of Hose Roller for Firefighter: A Fatigue Study Mohammad Luqman Hakim Mustapha, Salwa Mahmood , Helmy Mustafa El Bakri , Ismail Abdul Rahman , Noorul Azreen Azis , and Mohd Rizal Buang
Abstract In a working environment, worker’s safety and health are the most critical considerations. Previous study discovered that firefighters are exposed to a great deal of ergonomic risk factors (ERF). ERF exposure during hose rolling includes awkward posture and forceful exertion. Therefore, the primary goal of this research paper is to fabricate an ergonomic hose roller for firefighters and conduct a fatigue analysis to determine the efficiency of the tool designed to safeguard firefighters against the risk of low back disorder (LBD). Hose roller testing is necessary to guarantee that it can withstand the weight of fire hoses while still being comfortable for users’ bodies. Fatigue analysis was conducted using Industrial Lumbar Motion Monitor (i-LMM) equipment to evaluate LBD risk during hose rolling. Manual handling contributes 57.67% to the total average percentage value used to compute LBD risk results, while utilizing a roller tool, the hose rolling procedure yields a 27% LBD risk limit value. The design of experiment (DOE) method should be used in future studies to gather more information for the LBD risk assessment. Keywords Ergonomic design · Low back disorder · Fatigue analysis · Ergonomic risk factor
M. L. H. Mustapha · S. Mahmood (B) · H. M. E. Bakri Faculty of Engineering Technology, Universiti Tun Hussein Onn Malaysia, Pagoh Education Hub, KM 1, Jalan Panchor, 84600 Pagoh, Johor, Malaysia e-mail: [email protected] I. A. Rahman · N. A. Azis Ergonomics Excellence Centre, National Institute of Occupational Safety and Health, Taman Teknologi Johor, 81400 Senai, Johor, Malaysia M. R. Buang Johor State Fire Department Headquarter, Jalan Kangkar Tebrau, 81110 Johor Bahru, Johor, Malaysia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_4
43
44
M. L. H. Mustapha et al.
1 Introduction Before the use of mechanical tools, firefighters used to conduct the two-man hose roll method, which is believed to protect the firefighters’ lower backs [1]. A fire station operates for 24 h, and usually involves two shifts which are 12 h shift and 24 h shift. The job scope during daytime and nighttime included maintenance of vehicles, rescue equipment, emergency training, rescuing victims, and also attempting to salvage the content of buildings [2]. According to data obtained from the Social Security Organization (SOCSO), there were 10 recorded cases of MSD among Malaysian employers and employees in 2005 and 675 cases in 2014 [3]. From 2009 to 2014, there was an increase in the total number of cases of musculoskeletal disorders (MSDs) as a result of incident involving manual handling tasks [4]. This trend is anticipated to increase awareness among Malaysian businesses and employees. To create awareness among firefighters regarding MSDs, this research paper proposed the development an ergonomic hose-rolling equipment for firefighters. This paper aims to design a lightweight, easy-to-carry and ergonomic hose roller for firefighters. As a result, having a good choice of tools and a design that is durable and easy to use is required to improve the fabrication process. This paper also aims to evaluate the efficiency of the hose rolling system during fatigue analysis by utilizing appropriate tools and applying Ballet Software. The efficiency and safety measures of firefighting tools are the most important factors that influence firefighters’ performance. Delivering a fire extinguishing supply necessitates the use of high-efficiency firefighting equipment; otherwise, it could possibly lead to strain, injuries, and potentially dangerous circumstances. The use of gear and equipment designed to reduce the physical exertion required to manage heavy lifting and carrying duties is one of the approaches that can be used to reduce the danger faced by firefighters during fire extinguishing operations. However, the hose rolling tools’ durability can have an impact on the firefighting process, as long-lasting tools are required to avoid performance failure during the operation period [5]. Firefighters have also encountered an issue where the tools used during operations were inefficient and took a long time to use. Additionally, the hose rolling tools need to be of the right size and weight in order to be convenient to carry and store [6]. Previous research has shown that aluminium is recommended as the main material for the research due to its lightweight, durable, and easy to cast factors [7]. This paper offers the possibility of fabricating hose rolling equipment, which could help firefighters reduce the risk during firefighting operations and hose reel inspections. However, most previous research analyses did not focus on fatigue analysis after the fabrication [1]. As part of the fatigue analysis, this research paper analyzed the efficiency of the fire hose roller by obtaining the value of LBD risk in terms of percentage using iLMM tools. This ensures that the fire hose roller is adequate for firefighting operations while also reducing the ERF specifically for firefighters while rolling the hose, which is one of the critical tasks.
Design of Hose Roller for Firefighter: A Fatigue Study
45
2 Materials and Methods The selected materials and procedures were used to fabricate the hose roller tool. However, testing of the developed tool has been carried out to ensure that the hose roller tool performs appropriately and smoothly for fatigue analysis. The testing process also ensures that the hose roller tool can accommodate the sturdiness of the fire hose. Fatigue analysis was carried out to determine the average probability of possible LBD risk for firefighters. The methodology flowchart for this paper is illustrated in Fig. 1.
2.1 Fabrication The fabrication process of the hose roller is mainly based on metal fabrication since most of the materials and equipment are made of steel. The procedures of casting or molding a raw or semi-finished metal workpiece is referred to as a metal fabrication. Forming, shaping, bending, and cutting are all parts of the processes involved. It is preferable to use the fitting and shaping process rather than welding when fabricating hose rollers. This is due to the possibility that welding will create a permanent junction, making hose roller maintenance virtually impossible. The Failure Mode Effect Analysis (FMEA) and proposed design from previous work were used to develop the hose roller outer and inner systems [8–10]. The fabrication process of Fig. 1 The methodology flowchart of the project
46
M. L. H. Mustapha et al.
Fig. 2 The result of an aluminum sheet after being shaped using grinder during fabrication
Fig. 3 The fitting task of shaped aluminum sheet using a bracket and a rivet
the hose roller was applied which are shaping, fitting and application of bolts and nuts respectively (see Figs. 2, 3 and 4).
2.2 Testing The hose roller tool testing method is critical for ensuring that the fatigue analysis runs smoothly and correctly. The hose roller tool has been tested in order to determine its ability and efficiency in fatigue analysis, besides its ability to accommodate the weight and size of the fire hose during the hose rolling task. The fire hose used in the measuring process is 10 m long, 10.5 cm wide, and weighs 4.9 kg. The testing process for the hose roller conducted at Pagoh Fire Station is depicted in Fig. 5.
Design of Hose Roller for Firefighter: A Fatigue Study
47
Fig. 4 The application of bolts and nuts respectively during the fabrication process
Fig. 5 The testing process of hose roller at Pagoh fire station
2.3 Fatigue Analysis After the fabrication process, a fatigue analysis was performed to determine the hose roller’s capabilities and durability, as well as the probability of a firefighter being exposed to an ergonomic risk while using the hose roller tool or manual handling hose rolling. Fatigue analysis is the study of a model’s or material’s tendency to fracture. For conditions when the stress cycles are regular and lower than the normal strength, this can be achieved by applying constant amplitude loading [11]. The National Institute of Occupational Safety Health (NIOSH) conducts a fatigue investigation of the hose roller, in which data on body posture will be acquired using i-LMM tools. As the fatigue analysis progresses, it will become clear whether the hose roller requires changes or adjustments to its materials or equipment, which may affect body posture while using hose rolling instruments. This method is necessary in order to demonstrate how the hose roller tool can decrease the risk of a firefighter being exposed to an ergonomic risk, particularly when it comes to body posture and repetitive action. i-LMM is a set of tools working in collaboration with the Ergonomic Excellence Centre (EEC), NIOSH. It is a set of tools that allows assessing LBD risk in industrial
48
M. L. H. Mustapha et al.
Fig. 6 The application of i-LMM tool
employment in terms of body postures and dynamic movement effects. The i-LMM is a tri-axial thoracolumbar goniometer that measures lumbar spine motion [12]. It also includes the BALLET software as well as a set of harnesses with two motion sensors attached. The respondent must wear harnesses, with one sensor directly in line with the spine and the other on the same level as the pelvis. The use of the i-LMM tool and harness was applied on a respondent before the data collection process (see Fig. 6). The lift rate, average twisting velocity, maximum moment, maximum sagittal flexion, and maximum lateral velocity are all detected by these motion sensors. The sensors capture position data at 60 Hz, which can be accessed via a desktop or notebook computer with BALLET software. After that, the BALLET software will determine the average risk of LBD in percentages. The data collected will then appear on the BALLET program start-up screen (see Fig. 7).
3 Results and Discussion 3.1 Fabrication The fabrication procedure for this research is based on a concept proposed by previous research [13]. The dimensions of the produced hose roller tool differ slightly from the initial design. The testing process also resulted in improvements based on fatigue analysis which will be explained in the Data Analysis Section below. The dimensions of the hose roller tool have been extended broader and appeared to be larger on the hose roller hook in order for it to perform well during fatigue analysis. This also serves as a balancing mechanism for the hose roller tool’s movement. The design of the fabricated hose roller tool was drafted using SolidWork software in both isometric and exploded views. The isometric view and exploded view of the hose roller tool
Design of Hose Roller for Firefighter: A Fatigue Study
49
Fig. 7 Graphical user interface of BALLET software version 3.1 Fig. 8 Isometric view of hose roller tool
were extracted from SolidWork software respectively (see Figs. 8 and 9). The hose roller tool has been fabricated according to the extracted design (see Fig. 10).
3.2 Data Analysis The data and results of the fatigue analysis were collected by two people, one of whom performed the hose rolling task and the other who monitored the data collection using
50
M. L. H. Mustapha et al.
Fig. 9 Exploded view of hose roller tool Fig. 10 A complete fabricated hose roller tool
BALLET software. The two methods for collecting hose rolling data are manual handling and the use of a hose rolling tool. Each method is repeated three times, and three average probabilities of LBD risk are generated in percentages. The values such as the specifications of the hose rolling tools and the distance between tools and respondent must be determined earlier in order to perform the data collection in the BALLET software (see Fig. 11). The information will then be used by the BALLET software to compute the average risk of LBD accurately. Three distinct runs were recorded using both methods. Three independent data sets with an average likelihood of LBD risk from the manual handling approach are
Design of Hose Roller for Firefighter: A Fatigue Study
51
Fig. 11 The values required for hose rolling tool specification in BALLET software
given (see Figs. 12, 13 and 14). The average likelihood can be categorized as high, according to the data, increasing slightly from 55 to 59%, respectively. For three distinct runs, the data for the hose roller tool approach was also recorded. According to the statistics from the hose rolling tool method, the average likelihood is lower than the manual handling approach, which ranges from 18 to 32%. The data obtained using the hose roller tool method are demonstrated in Figs. 15, 16, and 17.
Fig. 12 Data on the average probability of LBD risk on the first run from the manual handling method
52
M. L. H. Mustapha et al.
Fig. 13 Data on the average probability of LBD risk on the second run from the manual handling method
Fig. 14 Data on the average probability of LBD risk on the third run from the manual handling method
Design of Hose Roller for Firefighter: A Fatigue Study
53
Fig. 15 Data on the average probability of LBD risk on the first run from the hose rolling tool method
Fig. 16 Data on the average probability of LBD risk on the second run from the hose rolling tool method
54
M. L. H. Mustapha et al.
Fig. 17 Data on the average probability of LBD risk on the third run from the hose rolling tool method
3.3 Data Comparisons Between Manual Handling and Hose Rolling Method In this paper, data comparison is used to compare and evaluate the efficiency of the fabricated hose rolling tool. The manual handling method and the hose roller tool method of hose rolling method data have been compared. The data comparison approach was used to determine the efficiency of the hose roller tool. Table 1 depicts a data comparison table of the average LBD risk for each approach. According to the data shown, the outcome of LBD risks from BALLET software can be compared between manual handling and the usage of a hose rolling tool. When the average likelihood of LBD hazards reaches 30%, the work method is exposed to Table 1 Data comparison of total average LBD risk
Method
LBD risk (%)
Run
Total average (%)
Hose roller tool
18
1
27.00
31
2
32
3
55
1
59
2
59
3
Manual handling
57.67
Design of Hose Roller for Firefighter: A Fatigue Study
55
LBD risks. The average LBD risk percentage value for the manual handling method is higher on average, ranging from 55 to 59%, with a total average LBD risk value of 57.67%. Meanwhile, the percentage value of the hose rolling tool approach has increased slightly from 18 to 32%. Despite the value range exceeding 30%, the overall average is below the 27% LBD risk limit value. According to fatigue research, the risk of LBD increases as body posture, flexure, and twisting are increased during hose rolling work. The manual handling strategy is more vulnerable to LBD hazards than the hose rolling method, which only recorded 27% because the overall average percentage number exceeds the LBD risks limit defined in the BALLET software, which is 57.67%. Utilizing a hose roller tool that requires less flexure, twisting, and moment on body posture to perform the hose rolling activity is essential for the response on the hose rolling task to maintain a favorable body posture condition. Employing a hose roller tool can help reduce the percentage of time that workers are exposed to LBD risks, according to the statistics [1].
4 Conclusions The hose roller tool is intended to lessen the risk danger of ERF and LBD, particularly among firefighters when rolling the hose. The main objectives of this paper are to develop the hose roller as an ergonomic tool for firefighters and apply a fatigue analysis to evaluate the hose roller tool’s efficiency in preventing ERF and LBD risks. The fatigue analysis was conducted with the aid of the i-LMM tools, which allowed for a comparison of the average chance of LBD risk when performing hose rolling activities using the manual handling method and the hose roller tool method. Based on data obtained from fatigue studies, it is shown that the development of a hose roller tool in the hose rolling work may minimize the ERF and LBD risk factors. When applying a hose roller tool for hose rolling, the average likelihood of LBD risk is lower than when using a manual handling method; 27% average LBD risk. Hence, from the analysis, it can be concluded that it can also avoid ERF caused by MSD which originates from awkward posture, forceful exertions, pressure points, and static postures [14, 15]. For future recommendations of this research, the hose roller tool has to be improved in terms of its fabrication and fatigue study. When replacing rivets, it is strongly suggested that the bolts and nuts are fully applied for fitting and joining parts. Besides, it is recommended that the fire hose is evaluated in two different situations namely in wet and dry conditions for improvement in the data collection stage. This could also determine the hose roller tool’s efficiency in various fire hose conditions. As a result, future research may need to use the design of experiment (DOE) method to find the best, suitable condition to use as a factor in DOE. Moreover, another ergonomic design principle can be used, such as a motor installation for force reduction or a handle grip design for comfort. Additionally, a comparison
56
M. L. H. Mustapha et al.
between manual handling and the use of newly developed tools requires the conduct of an ergonomics risk assessment using the appropriate techniques. Acknowledgements The authors wish to thank the Universiti Tun Hussein Onn Malaysia for their involvement in this research. This research is supported by the Ministry of Human Resources National Institute of Occupational Safety and Health (NIOSH), under Geran Kontrak Kementerian Vot K447, and University Tun Hussein Onn Malaysia (UTHM) through GPPS Vot Q213.
References 1. Afzal M, Kiar M, Nazri M et al (2019) Initial ergonomic risk assessment on unrolling and rolling fire hose activity among firefighters at Putrajaya fire and rescue station. Human Factors Ergon J 4:53–56 2. Osman S, Bahari I, Arifin K et al (2012) Accident risk indices of Malaysia’s firefighters working in 12 and 24 hours shift works. J Occup 3. Manap M, Amat H, Sharif S et al (2017) Guideline on ergonomic risk assessment at workplace. Department of Occupational Safety and Health Ministry of Human Resources of Malaysia, Putrajaya 4. Manap M, Amat H, Sharif S et al (2018) Guidelines for manual handling at workplace. Department of Occupational Safety and Health Ministry of Human Resources of Malaysia 5. Svensson S (2002) The operational problem of fire control technology. University. http://etudes project.org/nonfpdata/cog/FS344/files-linked/The-Operational-Problem-of-Fire-Control.pdf 6. Benfer ME, Forssell E, Scheffey J, Hughes J (2017) Determination of fire hose friction loss characteristics. Fire Technol 53:1059–1075 7. Stojanovic B, Bukvic M, Epler I (2018) Application of aluminum and aluminum alloys in engineering. Appl Eng Lett 8. Luo C, Jin J (2012) Design feature analysis and pilot ergonomic evaluation for protective fire gear. Procedia Eng 43:374–378 9. Examiner P, Jillions JM (2004) Reeling device for fire hoses. United States Patent 2 10. Gentzler M, Stader S (2010) Posture stress on firefighters and emergency medical technicians (EMTs) associated with repetitive reaching, bending, lifting, and pulling tasks. Work. In: Proceedings of the human factors and ergonomics society annual meeting 11. Cyprien R (2017) A guide to fatigue analysis. http://feaforall.com/wp-content/uploads/2017/ 07/Fatigue-analysis-Guide.pdf. Accessed 21 February 2022 12. Biodynamic Solutions I (2011) Operator’s manual for Acupath lumbar motion monitor with BALLET software. 1–72 13. Isamudin A, Mahmood S (2021) Design of an ergonomic portable fire hose roller: a simulation study. Progr Eng Appl Technol 2:1016–1025 14. Nunes IL, Bush PM (2012) Work-related Musculoskeletal disorders assessment and prevention. In: Ergonomic—a system approach, pp 1–26 15. Kodom-Wiredu JK (2019) The relationship between firefighters’ work demand and workrelated Musculoskeletal disorders: the moderating role of task characteristics. In: Safety health at work
Noise Risk Assessment on Noise Exposure Among Urban Rail Maintenance Workers Using Personal Monitoring Method M. Mifzal-Nazhan , J. Azlis-Sani , A. Nor-Azali , Y. Nur-Annuar , S. Shahrul-Azhar, and A. Mohd-Zulhelmi
Abstract Urban Rail maintenance work typically involves a lot of daily maintenance work as proper maintenance must always be a priority for the rail industry to ensure that the passengers are always comfortable and safe. These maintenance activities expose the technicians to noise as they handle a lot of hand tools and machinery while performing their tasks. Noise-Induced Hearing Loss (NIHL) is a major compensable occupational disease in Malaysia due to excessive noise exposure above the permissible daily noise exposure limit. This study aimed to conduct Personal Monitoring as part of Noise Risk Assessment (NRA) to measure the level of noise exposure received by maintenance technicians in the excessive noise area. This study focused on the areas and tasks involved in the Track Network Maintenance Hall (TNMH) and Track Vehicle Storage Building (TVSB) at one of the urban rail companies. Noise Risk Assessment was the selected method to measure the level of noise exposure among maintenance technicians in the region with excessive noise. The Personal Monitoring method was conducted as suggested in the ICOP provided by DOSH Malaysia. Recorded data showed that a total of 2 out of 3 technicians were exposed to a daily noise exposure limit exceeding 85 dB(A) for their 8-h working shifts. The exposure level is currently controlled by the practise of using Personal Hearing Protector (PHP). Thus, this study confirmed that maintenance workers are exposed to high noise levels when performing their maintenance tasks. Keywords Maintenance worker · Noise risk assessment · Excessive noise · Railway · Transportation
M. Mifzal-Nazhan (B) · J. Azlis-Sani · A. Nor-Azali · Y. Nur-Annuar Faculty of Mechanical and Manufacturing Engineering, Universiti Tun Hussein Onn Malaysia, 86400 Batu Pahat, Malaysia e-mail: [email protected] S. Shahrul-Azhar · A. Mohd-Zulhelmi Safety, Health and Environment Department, Rapid Rail Sdn Bhd, No 1, Jalan PJU 1A/46, Off Jalan Lapangan Terbang Subang, 47301 Petaling Jaya, Selangor, Malaysia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_5
57
58
M. Mifzal-Nazhan et al.
1 Introduction Light Rapid Transit (LRT) is one of the public transport options that has been used by people to travel around the areas of Kuala Lumpur and Selangor. In Malaysia, only a few rail operators offer urban rail services, including commuters, Light Rail Transit (LRT), monorails, and Mass Rapid Transit (MRT). In order to guarantee that the passengers are always comfortable and safe while receiving excellent service, the rail industry must always prioritise proper maintenance. Usually, railway industries involve a lot of daily maintenance work, including track maintenance, overhead lines, signalling systems, power supplies, security systems, and inspection of rail assets [1]. Other than that, the railway maintenance industry involves complex human–machine interactions and safety–critical operations with considerable risks to the health and safety of its employees during the maintenance process [2]. Recently, high levels of noise exposure are one of the most common occupational hazards that happen in Malaysian industries [3]. In addition, noise-induced hearing loss (NIHL) was the highest notifiable occupational health issue among Malaysian workers in 2014 [4]. As noise-induced hearing loss (NIHL) usually happens in various industries, more studies have been carried out to analyse, evaluate, and receive information about these problems. Previous research in the rail sector has stated that maintenance crew are exposed to high noise during preventive maintenance activity which involves the usage of compressed air to remove dust and particulates from the RPM or also known as the blowing activity [5]. According to research from the aircraft sector, which is another industry, the aircraft maintenance workers in hangars are exposed to noise levels greater than 85 dB(A), which can cause noise-induced hearing loss (NIHL) [6]. Thus, it is important to study the possible causes of high noise exposure to prevent noise-induced hearing loss cases among maintenance workers in the rail industry.
2 Methods 2.1 Noise Risk Assessment (NRA) Noise Risk Assessment is a new method under the Industry Code of Practice (ICOP) for Management of Occupational Noise Exposure and Hearing Conservation 2019 that was published by the Department of Safety and Health (DOSH) Malaysia and applied to all workplaces where persons are employed in any industrial sector that is covered under the Occupational Safety and Health Act 1994 [Act 514] [7]. Hence, the purpose of this assessment is to identify any instances of excessive noise that can expose employees to it at work. It can also identify the effectiveness of the existing measures taken to reduce noise exposure. For this study, a noise exposure personal monitoring methodology was used to measure the level of noise exposure among maintenance technicians in an excessive noise area.
Noise Risk Assessment on Noise Exposure Among Urban Rail …
59
2.2 Procedure to Conduct Personal Monitoring As stated in the guideline for conducting Noise Risk Assessment (NRA), the noise measuring equipment shall comply with the requirements of IEC 61672-1, IEC 61252, and any other relevant standards. In this study, the Noise Dosimeter LARSON DAVIS model 706 was used according to the requirements stated in the guideline. Other than that, the noise dosimeter was set up as follows; Criterion Level: Lc = 85 dB(A), Threshold Level: Lt = 80 dB(A), Exchange Rate: q = 3 dB, Time Constant = ‘slow’, Peak Level = 140 dB(C). Following the identification of the Similar Exposure Group (SEG), a sample of technicians from each working group was selected, and the daily noise exposure was measured during their working shift. The personal monitoring sampling duration for all identified technicians was 9 h, starting from 0800 until 1700, without pausing the dosimeter. The monitoring sampling was conducted in a different date due to the limited number of dosimeters in possession. Before conducting the monitoring, the noise dosimeter was calibrated at the workplace. Then, the dosimeter was installed on the subject’s body and a microphone was mounted on the top of the shoulder, approximately 0.1 m from the entrance of the external ear canal at the side of the most exposed ear and approximately 0.04 m above the shoulder. Finally, a cable was attached neatly and safely as long as the subject felt comfortable and free to move while conducting any kind of work or activity (Fig. 1).
Fig. 1 Example for position of microphone attached on subject
60
M. Mifzal-Nazhan et al.
3 Method Results and Discussion 3.1 Noise Risk Assessment (NRA) The identification of the area and task was made through a discussion with the management of the Track Network Maintenance Department (TNMD). From the discussion, the department’s two main halls, Track Network Maintenance Hall (TNMH) and Track Vehicle Storage Building (TVSB), were both implicated. Other than that, the Track Maintenance Group and Rail Equipment Maintenance Group are the two key technical groups within this department. Track Maintenance Group was divided into two sub-groups, which are Depot Group and Mainline Group. Depot Group was divided into four sub-groups which are Heavy Maintenance, Special Trackwork, Power Rail, and Switch Machine, as shown in Fig. 2. Switch Machine Group was selected among the others to conduct this assessment because their tasks were located inside the Switch Machine Room at the Track Vehicle Storage Building (TVSB). Other than that, the Rail Equipment Maintenance Group was divided into two sub-groups which included the Rail Grinding Group and the Rail Borne Group. For the Rail Grinding Groups, the tasks that were conducted inside the hall involved air blowing activity and preventive maintenance for the Rail Grinding Unit (RGU) Vehicle. For the Rail Borne Group, the preventive maintenance and inspection for all Rail Borne Vehicles were conducted inside the halls and involved a lot of noise sources generated by different vehicle engines. Therefore, the Noise Risk Assessment (NRA) was conducted by focusing on the three selected groups, which included Switch Machine Group, Rail Grinding Group, and Rail Borne Group. Figure 2 illustrates every group and sub-group that is involved in the Track Network Maintenance Department (TNMD) and the selected groups for assessments after the employer had determined that there was a potential risk of noise during working hours for a particular working group.
3.2 Personal Monitoring Table 1 displays the overall results from the personal monitoring assessment. Subject 1 was a technician from the Switch Machine Group, while Subject 2 represented the Rail Grinding Group, and Subject 3 represented the Rail Borne Group. Subject 1 was monitored while performing maintenance activity and installing the Switch Machine Box onto the mainline Track in the depot. Meanwhile, Subject 2’s task was blowing air activity and conducting preventive maintenance on the Rail Grinding Unit (RGU) vehicle. Subject 3, on the other hand, performed preventive maintenance on a rail-borne vehicle in TVSB Hall. Overall, the daily exposure limit for Subject 1 was 84 dB(A), which is below the noise exposure limit as specified under OSH (Noise Exposure) Regulations
Noise Risk Assessment on Noise Exposure Among Urban Rail …
61
Fig. 2 Group of workers in track network maintenance department (TNMD)
Table 1 Results of personal monitoring Subjects
Subject 1
Subject 2
Subject 3
Working area
Track vehicle storage building
Track network maintenance hall
Track vehicle storage building
Sampling duration (h)
9
9
9
Dose T e (%)
103.1
93.5
93.2
L eqT e , dB(A)
84.0
93.4
100.6
L E X,8h , dB(A)
84.0
93.4
100.6
Max level, dB(A)
120.2
119.2
117.3
Peak level, dB(C)
152.4
148.8
150.0
2019. However, Subjects 2 and 3’s daily noise exposure levels were 93.4 dB(A) and 100.6 dB(A) respectively, while performing their task. Subject 1 did not exceed the high noise exposure level as the task involved was the basic installation of switch machine box at the mainline track. This task did not involve high-level noise-generation machines in the workplace. For Subject 2, the technician was exposed to a noise exposure limit over the permissible level because of the air pressure from the air blowing gun to clean the RGU vehicles from dust and impurities. The activities would cease until all the dust and impurities were completely cleaned from the RGU vehicle, which typically took 45–60 min. The noise was observed to be produced by the high pressure of air flowing from the gun
62
M. Mifzal-Nazhan et al.
nozzles with high velocity at the air gun tip. Previous research has shown that the noise is generated by the air acceleration that varies from near zero velocity in the reservoir to peak velocity at the nozzle’s exit. Furthermore, the typical sound pressure level at 1 m from a blow-off nozzle can reach up to 105 dB(A) [8]. After cleaning, the technician continued to perform preventive maintenance on the RGU vehicle while the engine was switched on. During the noise survey, the noise generated reached approximately 85 dB(A) at 5 m and 82 dB(A) at 7.5 m, respectively. Therefore, proper personal hearing protectors must be used to perform these tasks. For Subject 3, the technician was exposed to excessive noise due to the noise emitted from the rail-borne vehicle’s engine while conducting preventive maintenance activities. This scenario was similar to Subject 2 but riskier because the worker was involved with various types of vehicles to perform preventive maintenance, compared to Subject 2 which only focused on RGU vehicles only. In addition, preventive maintenance was performed inside the maintenance hall, which might have increased the value of noise exposure, especially when two or more types of engines were switched on. Overall, the values on maximum sound pressure level of 115 dB(A) and maximum peak level of 140 dB(C) are shown in Table 1. All subjects were exposed and exceeded the permissible limit during the assessments. However, all these values were recorded in a short time which was less than one minute when performing the tasks. Hence, the noise was assumed to be emitted due to unidentified sources such as false contributions. These include wind or suddenly knocking on the microphone with a cloth or body parts. In conclusion, Subjects 2 and 3 were exposed to excessive noise compared to Subject 1, thus further actions are suggested for noise reduction.
3.3 Noise Reduction Personal Hearing Protectors (PHP) Personal Hearing Protectors will be the last option for noise control when engineering and administrative control measures do not reduce the exposure to noise below the Noise Exposure Limit (NEL) specified in the Occupational Safety and Health (Noise Exposure) Regulations 2019. The results of personal monitoring showed that the technicians were exposed to noise that exceeded the noise exposure limit of 85 dB(A); hence the Noise Reduction Rating (NRR) (ICOP, 2019) was determined to obtain the estimated noise exposure after wearing the PHP. Table 2 shows the Personal Noise Exposure Monitoring Result after using Personal Protection Protectors. Based on Table 2, the Noise Reduction Rating (NRR) for Subjects 1 and 2 was reduced after wearing PHP and the noise exposure value was below the noise exposure limit of 85 dB(A). Meanwhile, Subject 3 still received a high level of noise exposure after wearing the PHP, thus he is recommended to use dual hearing protection with a combination of minimum NRR = 29 earplugs and NRR = 27 earmuffs to reduce the noise below daily noise exposure level of 85 dB(A). The new estimated exposure,
Noise Risk Assessment on Noise Exposure Among Urban Rail …
63
Table 2 Personal noise exposure monitoring result after using personal protection protectors Subjects
Subject 1
Subject 2
Subject 3
Working area
Track vehicle storage building
Track network maintenance hall
Track vehicle storage building
Job category
Switch machine group technician
Rail grinding group technician
Rail borne group technician
Exposure level without PHP, dB
84.0
93.4
100.6
NRR
9
10
9
Attenuation level, dB With PHP
84.0
93.4
100.6
Exposure level, dB
75
83.4
91.6
Exceed NEL?
NO
NO
YES
dB(A) after wearing dual protection is 84.6 dB(A), as shown in the calculation by using the formula below: N R Rh − 7 +5 2 29 − 7 +5 100.6 (dB(A)) − 2
L E X,8h (dB(A)) −
84.6 (dB(A))
(1) (2) (3)
4 Conclusion In conclusion, the Personal Monitoring from Noise Risk Assessment was conducted to measure the level of daily noise exposure among maintenance technicians in the excessive generation area. From the results, two out of the three subjects were exposed to excessive noise exposure limits for a working duration of eight hours. Finally, all the data were analysed and discussed according to the previous research and the Industrial Code of Practice (ICOP) provided by DOSH Malaysia. At the end of this study, suggestions were offered to the company to reduce noise exposure to the workforce, especially hearing protection control. Finally, this study has strengthened the understanding and importance of noise exposure towards workers in the workplace.
64
M. Mifzal-Nazhan et al.
References 1. Curcuruto M, Griffin MA, Kandola R, Morgan JI (2018) Multilevel safety climate in the UK rail industry: a cross validation of the Zohar and Luria MSC scale. Saf Sci 110(July 2017):183–194. https://doi.org/10.1016/j.ssci.2018.02.008 2. Wilson AMJ, Norris B (2005) Rail human factors: supporting the integrated railway 3. Zare S, Baneshi MR, Hemmatjo R, Ahmadi S, Omidvar M, Dehaghi BF (2019) The effect of occupational noise exposure on serum cortisol concentration of night-shift industrial workers: a field study. Saf Health Work 10(1):109–113. https://doi.org/10.1016/j.shaw.2018.07.002 4. Sam WY, Anita AR, Hayati KS, Haslinda A, Lim CS (2017) Prevalence of hearing loss and hearing impairment among small and medium enterprise workers in Selangor, Malaysia. Sains Malays 46(2):267–274. https://doi.org/10.17576/jsm-2017-4602-11 5. Jacquelyne A, Azlis-Sani J, Nor Azali A, Nur-Annuar M, Shahrul Azhar S, Mifzal-Nazhan M (2020) Task analysis on maintenance worker (Rail Grinder) of light rail transit (LRT). Malays J Publ Health Med 20(Special issue 1):223–230. https://doi.org/10.37268/MJPHM/VOL.20/NO. SPECIAL1/ART.704 6. Noweir MH, Zytoon MA (2013) Occupational exposure to noise and hearing thresholds among civilian aircraft maintenance workers. Int J Ind Ergon 43(6):495–502. https://doi.org/10.1016/j. ergon.2013.04.001 7. Department of Occupational Safety and Health Malaysia (2019) Industry code of practice for management of occupational noise exposure and hearing conservation 2019. Industry Code of Practice. http://www.dosh.gov.my/index.php/legislation/codes-of-practice/industrial-hyg iene/3286-industry-code-of-practice-for-management-of-occupational-noise-exposure-andhearing-conservation-2019/file 8. Barron RF, Barron RF (2019) Noise sources. In: Industrial noise control and acoustic. Pfeiffer, pp 162–224. https://doi.org/10.1201/9780203910085-5
Bangla Text Summarization Analysis Using Machine Learning: An Extractive Approach Mizanur Rahman, Sajib Debnath, Masud Rana, Saydul Akbar Murad, Abu Jafar Md Muzahid, Syed Zahidur Rashid, and Abdul Gafur
Abstract A notable expansion of information in digital space provides quick access to these massive amounts of news, literature, and so on. Although maximum news is available in English, people are interested in reading in their language. Almost all Bangla newspapers now have their online version, so people prefer to read newspapers online. Without reading the whole news, it is not possible to obtain significant information and a summary from a news article. A news summary is a process of extracting the most significant information from a news article in a precise manner. In this case, we require an effective Bangla text summarizer tool to get insight from the news easily. Text summarizers use linguistic methods to generate and interpret text before uncovering natural concepts and phrases to represent the content by compressing the original text document into a shorter text that contains the most relevant information. Summarizing lengthy and important text by enforcing specific guidelines is a challenging task. Therefore, a method based on intelligent machine learning is widely anticipated in this content. This paper uses extractive automatic text summarization techniques to condense the source text into a shorter version by preserving the significant information of the original text. Here we have adopted two machine learning-based techniques (i.e., word2vec, and similarity matrix) and a rule-based technique (word count) for Bangla text summarization. The results show M. Rahman (B) · S. A. Murad · A. J. M. Muzahid Faculty of Computing, Universiti Malaysia Pahang Al-Sultan Abdullah, 26600 Pekan, Pahang, Malaysia e-mail: [email protected] S. Debnath Department of Computer Science and Engineering, United International University, Dhaka, Bangladesh M. Rana Information and Communication Engineering, Noakhali Science and Technology University, Noakhali, Bangladesh S. Z. Rashid · A. Gafur Department of Electronic and Telecommunication Engineering, International Islamic University Chittagong, Chittagong, Bangladesh © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_6
65
66
M. Rahman et al.
that the word2vec technique performs very well compared to the summary performed by other techniques. Keywords Word2vec · Word Count · RNN · Bangla text summarization · NLP · Similarity Matrix
1 Introduction Throughout the Internet, there is a vast volume of written content available, comprising web pages, news articles, blogs, etc. Nevertheless, the information in it is unstructured, and the most we can do is scan and skim the findings to access it. As a result, there is a growing need to reduce this text material, summarizing it to capture the relevant elements so that we can access it more quickly, as well as to verify whether the larger file contains the information we are looking for. Text summarization is the process of automatically creating a shorter version of one or more text documents [1]. Text summarization techniques can be categorized as extractive and abstractive techniques. The extractive summarization technique deals with choosing and concatenating significant phrases to make a summary from the original document [2]. Abstractive text summarization comprises the understanding of the initial text and making a new summary on its own in fewer words [3]. An effective study on Bangla Text Summarization needs to be formulated by selecting the crucial sentences from the original text in the way to form the final summary. Therefore, measuring the performance of the summarized text could be a challenging task. Bangla language research is difficult for the following reasons. 1. According to extensive contemporary investigation, automatic methods for the Bangla language are scarce. 2. WordNet [4] and other lexical databases are not available in Bangla due to a lack of research. A restricted version of such a tool will be developed for Bangla [5]. 3. The subjects and objects of all sentences must be identified for correct recognition of sentence structures, which are also more complex in Bangla than in English. The subject might come at the beginning, end, or middle of a Bangla phrase in both active and passive voices (before or after the verb). We have an overflow of Bangla electronic news data in Bangladesh. However, it is a source of great regret since there are so few research studies on Bangla news summaries [6–8]. In a Bangla statement, determining the subject and object is quite tough. Furthermore, there is an excessive amount of inconsistency in this language’s grammatical rules. The study aims to gain summaries of a single document based on extractive summarization techniques. Text summarization relates to the process of producing a brief, accurate, and consistent overview of a longer text document. The following are the key concerns of this paper:
Bangla Text Summarization Analysis Using Machine Learning …
67
1. We looked into the approach and the significance of the preprocessing stage of the natural language processing pipeline during extractive automatic text summarization as a critical point in the process. 2. The purpose is to identify the correct word tokens early in the process so that sentences for the final summary can be chosen. 3. The work tries to improve the summarization process by implementing Machine learning-based approaches compared to other approaches. The following is how the rest of the paper is arranged: Sect. 2 summarizes previous research in this area. Sections 3 and 4 discusses the methodology used in this study. Section 5 introduces the simulation verification platform for evaluating the efficacy and dependability of the best technique. Finally, the Conclusions and future works are provided in Sect. 6.
2 Related Work Text summarization is a well-known technique in text compression. It has been used in many works [9–11] to create a concise and precise summary that represents the most important or relevant information of the original content. Ghosh [12] proposed an extraction-based summarization method consisting of four significant steps: preprocessing, stemming, sentence ranking, and summary generation. Along with this extraction technique, Rahimi [10] has presented a relationship between text mining with text summarization, where various text mining applications and text preprocessing steps are also described. Following that, a review of some of the summarization methodologies and their key parameters was carried out, with the main steps of the summarizing process being determined by tracing predominant sentences. The most significant extraction criteria are presented, along with the most fundamental proposed evaluation methods [13]. Madhuri [11] suggested a cutting-edge statistical technique to show extractive text summarization on a single document. Sentence extraction is a method that presents the idea of the supplied text in a condensed form. The weights that are assigned to sentences determine how they are ranked. To create a high-quality summary of the input document and store the summary as audio, highly scored sentences are taken from the input document. Ferreira [1] described and performed a quantitative and qualitative assessment of 15 algorithms for sentence scoring and the algorithms were evaluated using three different datasets. They also suggested multiple common criteria to improve sentence scoring, which include cue phases, sentence inclusion of numerical data, sentence length, sentence position, sentence centrality, and sentence resemblance to the title. This research [14] described a method for summarizing news articles in Bangla that extracts relevant sentences from single or several texts. They employed a sentence clustering approach to creating summaries from single and multi-document sources. The authors [15] offer a phrase frequency and sentence clustering-based Bangla
68
M. Rahman et al.
news summarizing technique. It also takes into account numerical data. It separates the sentences into two clusters for summary creation and takes half of the summary sentence from each cluster. Manju and David Peter said [16] by excluding the irrelevant and superfluous sentences, the extractive text summary technique was used to make sense of the sensitive part of the content. Gambhir [17] presented a comprehensive survey of recent text summarization extractive approaches along with a few abstractive and multilingual text summarization approaches. They analyzed extractive approaches like statistical-based approaches, discourse-based approaches, approaches based on machine learning, etc. Cengzhang [18] has represented words in an article as vectors trained by Word2vec, the sentence vector and the weight of each sentence are calculated by combining word sentence relationship with the graph-based ranking model, the weight of each word, the sentence vector and the weight of each sentence are calculated by combining word–sentence relationship with the graph-based ranking model.
3 Overview of Text Summarization Techniques 3.1 Extractive Text Summarization In a nutshell, extractive summarization builds summaries by simply extracting key sentences from the original text document without affecting the original text’s substance [2]. Extractive text summarization involves selecting expressions, sentences, and wordings from the source document to generate a new summary. This method works by identifying the important sections of the text and generating them precisely so that they depend only on the extraction of sentences from the original text. One of the strategies to get reasonable sentences is to assign a few numerical measures of a sentence which is called sentence scoring, for generating the summary, and after that, select the higher score sentences to create a record based on the compression rate. The higher the compression rate, the larger the summary is obtained along with increased insignificant content. And if we decrease the compression rate, a shorter summary is obtained, and valuable information is lost. The quality of the summary is acceptable when the compression rate is within 5–30%.
3.2 Similarity Matrix A similarity matrix, also known as a distance matrix, allows us to understand how similar or far apart each pair of items is from the user’s perspective. The cosine of two non-zero vectors can be acquired by using the Euclidean dot product formula: A.B = ||A||||B||cosθ
(1)
Bangla Text Summarization Analysis Using Machine Learning …
69
Fig. 1 Continuous Bag of Words (CBOW) Architecture
Given the two vectors of attributes, A and B, the cosine similarity, cosθ , is represented using a dot product, and the magnitude obtained is: cosθ =
∑n A.B i=1 Ai Bi / = /∑ ||A||||B|| n 2 ∑n 2 A i=1 i i=1 Bi
(2)
3.3 Word2vec Word2Vec uses the CBOW method as the default for building the model. CBOW generally uses Bag of Word, the most common feature extraction approach for Natural Language Processing (NLP). CBOW works as follows (Fig. 1): Suppose V is the vocabulary size and N is the hidden layer size. Input is defined as X i−1 , X i−2 , X i+1 , and X i+2 . We obtain the weight matrix by multiplying V * N. For sentence scoring, the score of each word for each sentence generated by the Word2Vec model is summed up, and from the summed value, we developed a score for each sentence using the following formula: S = α ∗ STF + β ∗ PV + γ + λ
(3)
Here, STF = Summation of term frequency, V = Vocabulary size.
3.4 Word Count Word count allows us to compare documents and measure their similarities for applications such as document classification, topic modeling, etc. The higher the score is obtained when a word occurs, the more recurrent the number of times in a text.
70
M. Rahman et al.
The number of the appearance of each word in the whole document is counted first using the following formula: number of occurance =
n ∑
word
(4)
i=1
The frequency of each word is then measured by dividing the number of occurrences of each word by the number of occurrences of the highest appeared word.
4 Methodology for Text Summarization In our experiment, we have taken a text document as input and summarized the input using a constructive rigorous process. The process is shown in Fig. 2. We will take a single document as input to find the word frequency and determine the sentence similarity. A weight will be assigned to each sentence, which will be followed by sorting the sentences according to their rank. Finally, we will select sentences with a higher rank for generating the summary.
4.1 Preprocessing In our approach, a user first inputs a Bangla news document, which is then preprocessed. Each sentence has been tokenized into words. In Fig. 3 you find the token version of our input text. A list of stop words is retained in the system for identifying
Fig. 2 Process of extractive text summarization
Bangla Text Summarization Analysis Using Machine Learning …
71
Fig. 3 Tokenization
stop words. After that, we remove the stop word, which is less important to represent a document. After that, word frequency had measured.
4.2 Calculating the Word Frequency Word scoring features define the importance of a word about all of the words in a sentence. The most important word in a sentence is the one with the highest score. The most important part is choosing features for a summarization method. Figure 4 shows how word frequency is represented in the program.
4.3 Sentence Weighting After the total score calculation is completed, each sentence of a news document will have a score. Sentences will be sorted in descending order based on the sentence
72
M. Rahman et al.
Fig. 4 Word frequency of input text
score assigned to them. This sorted list represents the order of the sentences in that news document. Figure 5 shows the sentences along with the total weight of the sentences.
5 Summary Generation The number of sentences that will be present in the summary is determined using the following formula: /√ / / / Summary Frequency = / Top Ranked Sentences/
(5)
Bangla Text Summarization Analysis Using Machine Learning …
73
Fig. 5 Sentences score of input text
6 Result and Evaluation 6.1 Result We have built a Bangla text summarizer using three different techniques to generate a summary from a single document. To demonstrate these summarization methods, we take a sample input text from a Bangla newspaper and generate three different summaries using Word Count, Word2Vec, and Similarity matrix. In Fig. 6 the sample input is shown and the summary generated by three techniques is shown in Figs. 7, 8, and 9.
6.2 Evaluation In this study, summarize Bangla text is compared to three models. Precision, recall, and F-measure are employed here since they have long been utilized as significant evaluation measures in the field of information retrieval If ’A’ denotes the number of sentences recovered by the summarizer and ’B’ denotes the number of relevant sentences compared to the target set.
74
Fig. 6 Sample input text
Fig. 7 Summary using Word Count
M. Rahman et al.
Bangla Text Summarization Analysis Using Machine Learning …
75
Fig. 8 Summary using Similarity Matrix
Fig. 9 Summary using Word2Vec
Precision(P) = (A ∩ B)/A
(6)
Recall(R) = (A ∩ B)/B
(7)
F − measure = (2 × P × R)/(P + R)
(8)
76
M. Rahman et al.
In Fig. 10 we can observe that average precision, average recall, and average Fmeasure shows that Word Count shows good results for Bangla text summarization compared to other techniques. Comparison considering the compression rate of the different documents was done based on summary using Word2vec as shown in the Table 1: Table 1 shows the result of the summary generated using Word2Vec by testing on ten documents. This experiment was conducted by calculating the compression rate of the number of sentences present in the original document and the number of sentences present in the summary using Word2Vec. By considering the compression rate of the ten documents, we can say that, on average, 21.4% compression is done using Word2vec.
Fig. 10 Step-by-step improvement of performance for Precision, Recall, and F-measure
Table 1 Statistical result of a summary using Word2vec SL no
No. of sentences in Original document
Compression rate by Summary by program
Word2vec
Average percentage (approx) 21.4%
1
21
4
19.0
2
10
3
30.0
3
9
2
22.2
4
47
7
14.0
5
13
3
23.0
6
35
6
17.1
7
17
4
23.5
8
62
13
20.9
9
28
5
17.8
10
11
3
27.2
Bangla Text Summarization Analysis Using Machine Learning …
77
Table 2 shows the statistical result of the summary obtained using the Similarity Matrix by experimenting on ten documents. This experiment was conducted by calculating the compression rate of the number of sentences present in the original document. The number of sentences present in the summary using the Similarity Matrix 29.45% compression was done using the Similarity Matrix. Table 3 shows the comparison considering the compression rate of different documents based on the summary using Word Count. This experiment was conducted by calculating the compression rate of the number of sentences present in the original document and the number of sentences present in the summary using Word Count. By considering the compression rate of the ten documents, we can say that, on average, 26.93% compression was done using Word Count. Table 2 Statistical result of a summary using the Similarity Matrix SL no
No. of sentences in Original document
1
21
Compression rate by Summary by program 5
Similarity matrix
Average percentage (approx)
23.8
29.45%
2
10
4
40.0
3
9
4
44.4
4
47
10
21.2
5
13
4
30.7
6
35
8
22.8
7
17
5
29.4
8
62
13
20.9
9
28
7
25.0
10
11
4
36.3
Table 3 Statistical result of a summary using Word Count SL no
No. of sentences in Original document
Compression rate by Summary by program
Word Count
Average percentage (approx) 26.93%
1
21
5
23.8
2
10
3
30.0
3
9
3
33.3
4
47
9
19.14
5
13
4
30.7
6
35
8
22.8
7
17
5
29.4
8
62
14
22.5
9
28
6
21.4
10
11
4
36.3
78
M. Rahman et al.
Table 3 shows the statistical result of the summary obtained using Word Count by experimenting on ten documents and getting 26.93% compression rate. From the above discussion, we can say that word2vec techniques provide a better summary based on the compression rate. We took Bangla input text from ‘The Daily Prothom Alo’ one of the renowned Bangla news portals in Bangladesh, which has both online and offline versions. We evaluate 100 input texts to measure our model performances and compare ten input texts to the generated summary by their techniques in Fig. 11. Figure 12 we can observe that in the case of summary generation based on compression rate, Word2vec is significantly better than the other two techniques.
Fig. 11 Number of sentences of three techniques
Fig. 12 Average of three techniques
Bangla Text Summarization Analysis Using Machine Learning …
79
7 Conclusion This paper has developed a Bangla text summarization method based on three techniques Word Count, Similarity Matrix, and Word2Vec. As a result, a Bangla reader can save a lot of time by following our method of reading only the most important news. We’ll improve sentence grading features to bring the systemgenerated summary closer to the one created by humans. Word2vec produces a more compressed summary compared to the other two techniques by keeping the significance of the content. We focused on producing summaries more effectively using machine learning-based models. Though many studies work for an English summary, due to the difficulties of the Bangla language, these may not be immediately relevant to Bangla.
References 1. Yadav D, Desai J, Yadav AK (2022) Automatic text summarization methods: a comprehensive review. arXiv preprint arXiv:2204.01849 2. Gaikwad DK, Mahender CN (2016) A review paper on text summarization. Int J Adv Res Comput Commun Eng 5(3):154–160 3. Verma R, Lee D (2017) Extractive summarization: limits, compression, generalized model and heuristics. Computación y Sistemas 21(4):787–798 4. McCrae JP et al (2020) English WordNet 2020: improving and extending a WordNet for English using an open-source methodology. In: Proceedings of the LREC 2020 workshop on multimodal WordNets (MMW2020) 5. Pal AR et al (2021) In search of a suitable method for disambiguation of word senses in Bengali. Int J Speech Technol 24(2):439–454 6. Sarkar K (2014) A keyphrase-based approach to text summarization for English and Bengali documents. Int J Technol Diffusion (IJTD) 5(2):28–38 7. Jahan B et al (2021) Construction of an automatic Bengali text summarizer using machine learning approaches. J Data Anal Inform Process 10(1):43–57 8. Haque MM, Pervin S, Begum Z (2017) An innovative approach of Bangla text summarization by introducing pronoun replacement and improved sentence ranking. J Inform Process Syst 13(4):752–777 9. Islam M et al (2020) Hybrid text summarizer for Bangla document. Int J Com- put Vis Sig Process 1(1):27–38 10. Rahimi SR, Mozhdehi AT, Abdolah M (2018) An overview on extractive text summarization. In: 2017 IEEE 4th international conference on knowledge-based engineering and innovation (KBEI) 11. Madhuri JN, Ganesh Kumar R (2019) Extractive text summarization using sentence ranking. In: 2019 international conference on data science and communication (IconDSC). IEEE 12. Ghosh PP, Shahariar R, Khan MAH (2018) A rule-based extractive text summarization technique for Bangla news documents. Int J Modern Educ Comput Sci 11(12):44 13. Kadhim AI (2018) An evaluation of preprocessing techniques for text classification. Int J Comput Sci Inform Secur (IJCSIS) 16(6):22–32 14. Akter S et al (2017) An extractive text summarization technique for Bengali docu- ment (s) using K-means clustering algorithm. In: 2017 IEEE international conference on imaging, vision and pattern recognition (IVR). IEEE
80
M. Rahman et al.
15. Haque MM, Pervin S, Begum Z (2015) Automatic Bengali news documents summarization by introducing sentence frequency and clustering. In: 2015 18th international conference on computer and information technology (ICCIT). IEEE 16. Manju K, David Peter S, Idicula SM (2021) A framework for generating extractive summary from multiple malayalam documents. Information 12(1):41 17. Gambhir M, Gupta V (2017) Recent automatic text summarization techniques: a survey. Artif Intell Rev 47(1):1–66 18. Chengzhang X, Dan L (2018) Chinese text summarization algorithm based on Word2vec. J Phys Conf Ser 976(1)
Human Factors: Drivers’ Speed Choice on Relatively Low-Speed Limit Roads Othman Che Puan, Azlina Ismail, Khairil Azman Masri, and Muhammad Shafiq Mohd Rozainee
Abstract The human factor in speed selection while driving is always associated with road accidents. Drivers who drive too slow or too fast are both at risk of being involved in road accidents. This paper presents the findings of a study conducted to determine the drivers’ perceived safe speeds and the hazardous speeds while driving on single carriageway roads under free-flowing traffic conditions and subject to relatively low-speed limit regulation, i.e., 60 and 70 km/h. Speed and its associated data were collected at three locations, i.e., Jalan Kulai–Kota Tinggi, Jalan Skudai– Pontian, and Jalan Pekan Nenas–Kulai (i.e., Jalan Sawah). The speed data for more than 2,300 vehicles were gathered using a speed radar meter. The data was analysed based on two categories of vehicles, namely light vehicles and commercial vehicles. The threshold for the hazardous speed is represented by the speed at the 15th percentile of a speed curve, and the 85th percentile speed represents the perceived safe speed. The analysis results show that the 85th percentile speed for all vehicles at each road segment is 81–85 km/h, which is well above the speed limits posted at each road segment involved in the study, i.e., 60 and 70 km/h, respectively. The hazardous speed on road segments with 70 km/h posted speed limit based on the aggregated data at each road segment is 57 km/h, i.e., about 13 km/h below the speed limit posted at the corresponding road segments. The hazardous speed due to commercial vehicles on the same road segments is 46–49 km/h, i.e., 21–24 km/ h below the posted speed limit of 70 km/h. For road segments with a speed limit of 60 km/h, the data showed that most light vehicles travelled at speeds higher than the 60 km/h speed limit. Consequently, the hazardous speed representing all types of vehicles is about 57 km/h, which is slightly lower than the posted speed limit of 60 km/h. On the other hand, the hazardous speed due to commercial vehicles is just 10 km/h below the posted speed limit of 60 km/h, i.e., 50 km/h. The study showed that most drivers appeared challenging to adapt to a speed limit of 70 km/h or lower. O. Che Puan (B) · A. Ismail · K. A. Masri Faculty of Civil Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, Kuantan, Pahang, Malaysia e-mail: [email protected] M. S. Mohd Rozainee Faculty of Civil Engineering, Universiti Teknologi Malaysia, Skudai, Johor, Malaysia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_7
81
82
O. Che Puan et al.
As a result, the range of safe speed for the drivers to travel without violating the speed limit is relatively small because the hazardous speed due to commercial vehicles is relatively high compared to the respective speed limits. Keywords Hazardous Speed · Spot Speed · 85th percentile speed
1 Introduction The percentage of motorists involved in road accidents due to factors associated with speed, such as speeding, following the leading vehicle too close, and dangerous overtaking, has increased from 6.7% in 2019 to 12% in 2020 [1, 2]. Speed is directly related to humans or, specifically, to drivers’ behaviour. As such, speed (particularly the spot speed) is one of the critical factors considered in road traffic safety analysis. Two important parameters of interest in analysing spot speed data are the 85th percentile and 15th percentile speeds. By definition, the speed at or below which 85% of vehicles travel is the 85th percentile. On the other hand, the 15th percentile speed is defined as the speed at or below which 15% of vehicles travel. In road safety management practices, if the spot speed study was conducted correctly, the 85th percentile speed may be used to select and implement the maximum speed limit for a particular stretch of a road segment. The 15th percentile speed, on the other hand, is considered the minimum safe speed for the given road and traffic conditions. A vehicle travelling at speed lower than the 15th percentile is considered to increase the risk of an accident due to travelling at a speed far below the average speed of the traffic stream. A vehicle with such a speed obstructs the smooth flow of traffic [3]. The Texas Department of Transport [4] suggested that “…whenever minimum speed zones are used, the minimum posted speed limit should be within 5 miles per hour of the 15th percentile value”. This paper reports a study conducted to determine the 85th percentile (i.e., driver’s perceived safe speed) and 15th percentile speeds (i.e., hazardous speeds) of drivers on single carriageway roads. In addition, this paper focuses on the speeds of vehicles on roads where relatively low-speed limits are imposed.
2 Literature Review 2.1 Speed and Risk of Accidents Driving at high speed is commonly considered one of the main factors increasing the risk of road accidents. The ability of a motorist to respond timely when necessary to do so reduces at high speeds. The European Commission [5] stated that for a
Human Factors: Drivers’ Speed Choice on Relatively Low-Speed Limit …
83
120 km/h road, the number of road accidents would change by 2% if the average speed changes by 1 km/h. For a 50 km/h road, the number of road accidents will change by 4% for a similar average speed change. Undeniably, driving at a speed inappropriate for the given road and traffic conditions will significantly threaten road safety. European Road Safety Observatory [6] reported that speeding or inappropriate speed has resulted in about 10 to 15% of all road accidents. Such inappropriate speed has also been a significant factor in 30% of all fatal road accidents. The difference in vehicle speed is also an essential factor influencing road accident risk. The presence of low-performance vehicles in a relatively high-speed traffic stream creates a more considerable speed difference. The road accident risk increases when the speed difference between vehicles in the traffic stream increases. This is because speed differences will result in more encounters with other vehicles and more frequent close-following or tail-gaiting, lane-changing, and overtaking manoeuvres. Many studies [7] have shown a strong relationship between speeding and accident risk. Although speeding is known to be one of the leading causes of road accidents, there are cases where accidents occur due to driving too slowly. For example, the Telegraph in 2018 [8] reported that, in Britain, there was a 31% increase in total casualties from accidents due to slow drivers in one year. It was also reported that driving too slowly, i.e., driving at or lower than the hazardous speed, has caused 175 people to be injured and two people killed in road accidents in 2017. Slow drivers could affect other drivers’ behaviour on the road. A driver following a slow-moving vehicle tends to initiate and remain in a close-following manoeuvre while waiting for an opportunity to overtake. Consequently, he or she tends to change lanes and overtake the slow-moving vehicle upon identifying a perceived safe gap in vehicles ahead. The risk of road accidents is high when more than one driver has the same idea and initiates the same manoeuvres. This practice will increase the frequency of vehicles changing lanes, which exposes drivers to a high risk of accidents. In Malaysia, 5,407 fatal accidents occurred during overtaking or changing lanes, representing about 10.8% of total fatal accidents from 2011 until 2018 [9]. The risk of accidents on a single carriageway road is significantly higher than on a dual-carriageway highway. This is mainly because, on a single carriageway road, there are no barriers between opposing lanes and a lack of lay-by areas on the side of the road. Therefore, any overtaking manoeuvres will require the drivers to use the opposing lane, which increases the risk of head-on collisions. Mwesige et al. [10] suggested that overtaking on two-lane two-way rural roads is associated with the risks of an accident due to incomplete overtaking manoeuvres by the drivers.
2.2 Factors Influencing Speed Choice World Health Organization [11] reported that vehicle speed is influenced by some factors such as:
84
O. Che Puan et al.
. Driver-related factors such as age, gender, and alcohol level; . Road-and vehicle-related factors such as road layout, surface quality, and vehicle power; . Traffic- and environment-related factors such as traffic density and composition, prevailing speed, and weather conditions A large gap between vehicles exists due to the vehicles that move significantly slower than the other vehicles in the same lane. Slow-moving vehicles in a traffic stream will lead to close-following or tail-gaiting and subsequently overtaking manoeuvres to occur more frequently. In a survey comparing Malaysian and Vietnamese drivers, Mohamad et al. [12] reported that 32.2% of Malaysian and 50.0% of Vietnamese drivers tend to speed due to slow drivers. This means one-third of Malaysian and half of Vietnamese drivers would speed up and try to overtake whenever there is a slow vehicle ahead of them. Another survey by the same authors [13] on young Malaysian drivers suggests that 29% agreed that slow drivers influenced them to start speeding. These results show that slow driving would negatively impact other drivers speeding behaviour.
3 Methodology The primary data required for the study are speed and type of vehicles associated with it. Speed in this study refers to a specific vehicle type’s spot speed. It is an instantaneous speed of a vehicle at a point on the road. Vehicles were categorised into two major types, i.e., light and commercial. For this study, a light vehicle refers to a 2-axle vehicle with a total of not more than four wheels. A commercial vehicle refers to medium and heavy goods vehicles with an unladen weight exceeding 1.5 tonnes or a vehicle having more than two axles or more than two wheels on the rear axle.
3.1 Sample Size It is a common practice to conduct a speed study based on a sampling approach because it is not practical to observe the entire population of vehicles at a particular location over a long time. In this context, the accuracy of the analysis relies on the size of the sample to be used and the sampling techniques. Therefore, the larger the sample size (i.e., the number of observations), the higher the accuracy or the confidence that can be placed on the estimates of attributes, such as mean, median, etc., for the entire population of vehicles at that road segment. For a spot speed study, the Center for Transportation Research and Education [14] suggested a minimum sample size of 50. However, a sample size of 100 vehicles is usually preferred. According to Federal Highway Administration [15], a sample size
Human Factors: Drivers’ Speed Choice on Relatively Low-Speed Limit …
85
of 100 vehicles will generally yield an estimate of 85th percentile speed within an error of between 1 and 2 mph (i.e., 1.6 and 3.2 km/h) at a 95% level of confidence and standard deviation of 5 mph (i.e., 8.0 km/h). This study used Eq. 1 [16] to estimate the minimum required sample size. The estimates were based on the 95% confidence level with an acceptable 1.5 km/h error. ( N=
Zασ e
)2 (1)
where; N = minimum sample size Zα = normal statistic for α confidence level σ = standard deviation e = permitted error of estimate (km/h)
3.2 Site Selection The three road segments where the speeds of vehicles were collected for the study are (a) Jalan Kulai–Kota Tinggi, (b) Jalan Skudai–Pontian (the single carriageway section), and (c) Jalan Sawah (Pekan Nenas–Kulai). The location of each road section is shown in Fig. 1. Figure 2a–c shows the general layout of the road segments where the spot speed data was collected. In speed studies involving field data collection, gathering sufficient data to achieve sound statistical analysis and conclusions regarding drivers’ speed choices is essential. Therefore, in order to minimise errors or bias in the data sets, each of the road segments was selected with the following characteristics: (i) A relatively straight and flat road segment to ensure that road curves and gradients do not influence speed; (ii) Availability of a practically safe place for locating the observation instrument with an unobstructed view but not visible to motorists; (iii) The road is free from roadworks or maintenance activities that might affect the operating speeds of drivers; (iv) The condition of road pavement is free from severe defects such as potholes or other deformations; and (v) The road segment is free from major junctions.
86
O. Che Puan et al.
Fig. 1 Study locations marked as 1, 2, and 3
3.3 Data Collection The speed and types of vehicle data were collected manually using a hand-held speed radar meter (i.e., Bushnell) and the manual data recording sheets. The handheld speed radar meter is simple and can be operated easily. However, the accuracy of the recorded speed data depends on the excellent line of sight. Therefore, during data collection exercises, the position of the enumerator with the speed radar meter by the roadside is arranged so that he has a clear view of the vehicles to be measured while his presence is not apparent to the motorists to avoid errors in the data set. An example of the spot speed data sheet used during data collection exercises is shown in Fig. 3. The study requires the speed data to be collected where the drivers’ choice of speeds is not influenced by factors other than vehicle characteristics and road geometry conditions. Therefore, the data was collected during off-peak periods under good weather conditions to meet such a requirement. Data were also collected at different times of day to include variations in drivers’ population in the data set.
Human Factors: Drivers’ Speed Choice on Relatively Low-Speed Limit … Fig. 2 General characteristics of the road sections studied
(a) Road segment 1: Jalan Kulai–Kota Tinggi
(b) Road segment 2: Jalan Skudai–Pontian
(c) Road segment 3: Jalan Sawah
87
88
O. Che Puan et al.
Fig. 3 Example of spot speed data sheet
4 Result and Discussion A total of 2,366 spot speed data was collected at three different locations of road sections. The data collected meets the minimum sample requirement for a 95% confidence level with a 1.5 km/h acceptable error. The distribution of the sample size is summarised in Table 1.
Human Factors: Drivers’ Speed Choice on Relatively Low-Speed Limit …
89
Table 1 Sample size Road segment no
Location
Light vehicles
Commercial vehicles
1
All vehicles
Minimum required size
Actual size
Minimum required size
Actual size
Minimum required size
Actual size
Jalan Kulai–Kota Tinggi
241
649
191
198
295
847
2
Jalan Skudai–Pontian
236
482
197
268
293
750
3
Jalan Sawah
282
546
201
223
341
769
4.1 Distribution of Spot Speed Data and Speed Limit Violation The graphical presentations of the spot speed distribution at each road segment involved in the study are shown in Fig. 4a–c. As can be seen from Fig. 4a–c, the spot speeds at each of the road segments are somewhat normally distributed, and the speeds of commercial vehicles appear on the lower side of the respective figures when compared with the distributions of the speeds for the light vehicles. Such characteristics are as expected. The mean speeds of vehicles at each road segment are summarised in Table 2. The data shows that on the road segments marked as No. 1, i.e., Jalan Kulai–Kota Tinggi, and No. 3, i.e., Jalan Sawah, the mean speed of light vehicles is about 7.1% and 11.1% higher than the speed limit of 70 km/h, respectively. On the road segment marked as No. 2, i.e., Jalan Skudai–Pontian, the mean speed of light vehicles is about 29.6% higher than the 60 km/h speed limit. The data representing the commercial vehicles show that on Jalan Kulai–Kota Tinggi and Jalan Sawah, the mean speed of this type of vehicle was about 10.3% and 13.9% below the speed limit of 70 km/h, respectively. On the other hand, Jalan Skudai–Pontian’s mean speed is about 7.2% above the speed limit of 60 km/h. From this scenario, on a relatively reasonably good standard of single carriageway roads, a speed limit of 70 km/h and below appeared too low for the drivers of light vehicles. On the other hand, a speed limit of 60 km/h or below is considered too low by the drivers of commercial vehicles. The data shows that on road segments where a 70 km/h speed limit is imposed, the violation rate by the drivers of light vehicles is 55–65%, and the violation rate by commercial vehicles is 10–15%. However, on road segments with a 60 km/h speed limit, the violation rate by light and commercial vehicles is 95% and 55%, respectively. Therefore, the speed limit violation rate on road segments with a 60 km/ h speed limit is higher than that on road segments with a 70 km/h speed limit.
90
O. Che Puan et al.
(a) Road segment 1: Jalan Kulai–Kota Tinggi
(b) Road segment 2: Jalan Skudai–Pontian
Fig. 4 Distributions of speed data at each of the road segments
4.2 Drivers’ Perceived Safe and Hazardous Speeds In practice, the speed at the 85th percentile, i.e., at or below which 85% of vehicles are observed to travel under free-flowing conditions, is considered a reasonably safe speed for the given road and traffic conditions. In other words, the speed at the 85th percentile can also be regarded as the safe speed perceived by the drivers for a particular road segment. Drivers travelling above this speed are considered to violate the average speed.
Human Factors: Drivers’ Speed Choice on Relatively Low-Speed Limit …
91
(c) Road segment 3: Jalan Sawah
Fig. 4 (continued)
Table 2 Mean speeds Road segment no
Location
Posted speed limit
Light vehicles
Commercial vehicles
All vehicles
Mean speed
SD
Mean speed
SD
mean speed
SD
1
Jalan Kulai–Kota 70 Tinggi
75.0
11.9
60.3
10.6
71.6
13.1
2
Jalan Skudai–Pontian
60
77.8
11.8
64.3
10.7
73.0
13.1
3
Jalan Sawah
70
77.8
12.9
62.8
10.9
73.4
14.1
The 15th percentile speed, i.e., the speed at or below which 15% of vehicles are observed to travel under free-flowing conditions, on the other hand, is considered a hazardous speed. A vehicle travelling at speed lower than the 15th percentile will lead to platooning to develop at a faster rate, close car–following situations, and increase the need to overtake, increasing the risk of accidents. Both the 85th percentile and 15th percentile speeds can be determined using the cumulative plot of the spot speed data. For example, the graphical presentation of the spot speed distributions at each of the road segments involved in the study is shown in Fig. 5. The speeds of interest that can be extracted from Fig. 5 are summarised in Table 3. Table 3 shows that the average hazardous speed for light vehicles on all road segments is 61–64 km/h and, for commercial vehicles, is 46–50 km/h. The hazardous speed for all types of vehicles at each of the road segments is 57 km/h. This is 13 km/
92
O. Che Puan et al.
Fig. 5 Cumulative curves of spot speed data at three road segments
Table 3 Perceived and hazardous speed Road segment
Speed limit (km/h)
Lowest speed 15th percentile speed recorded (km/h) (km/h)
85th percentile speed (km/h)
LV
CV
LV
CV
All
LV
CV
All
Jalan Kulai–Kota Tinggi
70
43
28
61
46
57
83
68
81
Jalan Skudai–Pontian
60
53
38
64
50
57
86
74
82
Jalan Sawah
70
48
33
62
49
57
87
71
85
Note LV—Light vehicles; CV—Commercial vehicles
h lower than the speed limit of 70 and 3 km/h lower than the speed limit of 60 km/h. Such a slight difference between the maximum speed limit and the hazardous speed implies that on roads with a relatively low posted speed limit, there will be a high tendency for many vehicles to travel at speeds considered hazardous to other users. In terms of safe speed as perceived by the drivers, i.e., the 85th percentile speed, the data as summarised in Table 3 shows that, on roads where the posted speed limit is 70 km/h or lower, drivers of light vehicles appear to perceive speeds of higher than 80 km/h as safe and comfortable for the given road. On the other hand, the commercial vehicles’ drivers appear to agree that the speed limit of 70 km/h or higher is safe and acceptable for their vehicles.
Human Factors: Drivers’ Speed Choice on Relatively Low-Speed Limit …
93
5 Conclusion This paper highlighted the result of a study on one of the essential human factors in road traffic safety: drivers’ behaviour in terms of speed choice while driving on single carriageway roads with relatively low-speed limits. The findings from this study can be summarised as follows: (a) A speed limit of 70 km/h or lower tends to be violated by most drivers; (b) Most drivers expect the safe speed for single carriageway roads to be higher than the posted speed limit of 70 km/h or lower; and (c) under free-flow traffic conditions on a single carriageway, vehicles below 60 km/ h may be hazardous to other motorists. The result of the study also shows that in order to improve traffic safety factors, the practicality of a speed limit of 70 km/h or lower to be imposed on rural roads or roads passing through undeveloped areas requires a thorough assessment to avoid negative impacts in terms of traffic safety.
References 1. Polis DiRaja Malaysia (2019) Laporan Perangkaan Kemalangan Jalan Raya (statistical report on road traffic accidents) 2019. Jabatan Siasatan dan Penguatkuasaan Trafik, Ibu Pejabat Polis Bukit Aman, Kuala Lumpur 2. Polis DiRaja Malaysia (2020) Laporan Perangkaan Kemalangan Jalan Raya (statistical report on road traffic accidents) 2020. Jabatan Siasatan dan Penguatkuasaan Trafik, Ibu Pejabat Polis Bukit Aman, Kuala Lumpur 3. Rohaizan YR, Mashros N (2016) Effect of posted speed limit on drivers speed choice during off-peak period. Final Year Project Report. UTM (unpublished) 4. Texas Department of Transport (2015) Manual notice 2015-1. http://onlinemanuals.txdot.gov/ txdotmanuals/szn/determining_the_85th_percentile_speed.htm 5. European Commission (2022) Speed and accident risk. European road safety observatory. Brussels, European Commission, Directorate General for Transport. http://www.dacota-pro ject.eu/Links/erso/knowledge/Content/20_speed/speed_and_accident_risk.html 6. European Commission (2021) Road safety thematic report—speeding. European Road Safety Observatory. Brussels, European Commission, Directorate General for Transport 7. Zaman RN, Hoong APW, Arif STMST, Abdul Manap AR, Goonting K (2020) A review of the effects of changing speed limits on roads in Malaysia. Malaysia Institute of Road Safety Research (MIROS) 8. The Telegraph (2018) Casualties from crashes caused by slow drivers increase, as cautious elderly drivers are partly blamed. Retrieved from https://www.telegraph.co.uk/news/2018/12/ 29/casualties-crashes-caused-slow-drivers-increase-cautious-elderly/ 9. Ministry of Home Affairs Malaysia (2019) Causes of fatal accidents statistics 2011–2018 10. Mwesige G, Farah H, Koutsopoulos HN (2016) Risk appraisal of passing zones on two-lane rural highways and policy applications. Accident Anal Prevent 90:1–12 11. World Health Organization (WHO) (2004) Road safety—facts. Retrieved from https://www. who.int/violence_injury_prevention/publications/road_traffic/world_report/speed_en.pdf 12. Mohamad et al (2018) Understanding of speed behaviour in relation to road traffic accident: a comparison between Malaysian and Vietnamese drivers. Malays J Civil Eng 30(1):23–36
94
O. Che Puan et al.
13. Mohamad et al (2017) Do young Malaysian drivers involve in speeding behaviour: an investigation at selected accident hotspots in peninsular Malaysia. J Adv Res Soc Behav Sci 9(1):75–83 14. Iowa State University Center for Transportation Research and Education (CTRE) (2002) Handbook of simplified practice for traffic studies. Iowa DOT Project TR-455 15. Federal Highway Administration (FHWA) (2012) Methods and practices for setting speed limits: an informational report. FHWA Safety Program, FHWA-SA-12-004. U.S. Department of Transportation, Washington D.C. 16. Currin TR (2001) Introduction to traffic engineering: a manual for data collection and analysis. Brooks/Cole. Canada
Topology Optimization for Custom Bed-Resting Ankle Foot Orthosis Amir Mustakim Ab Rashid, Effi Zuhairah Md Nazid, Muhammad Hazli Mazlan, Azizah Intan Pangesty, and Abdul Halim Abdullah
Abstract The ankle-foot orthosis (AFO) is a widely used aid for people who suffer from weakness of the dorsiflexion muscles of the ankle due to peripheral or central nervous system problems. Both conditions are caused by a lack of dorsiflexion assistance due to muscle weakness. The mobility of the surrounding joints is affected by the deformity and muscle weakness of a lower limb joint, so corrective measures are required. The aim of this study is, therefore, to observe the effects of a topology optimized AFO model on pressure distribution. The product was designed in Solidworks and subjected to finite element analysis using polylactic acid (PLA) to determine the maximum value of Von Mises stress (VMS) and displacement. The design is then optimized using a topology study to analyze the iteration of the component design that meets a specific optimized target and geometric constraints. The final product is expected to have the desired function of the AFO for bedridden patients. Keywords Ankle foot orthosis · Topology optimization · Bed-resting
A. M. A. Rashid · E. Z. M. Nazid · A. H. Abdullah (B) Biomechanical & Clinical Engineering (BioMeC) Research Group, School of Mechanical Engineering, College of Engineering, Universiti Teknologi MARA, 40450 Shah Alam, Selangor, Malaysia e-mail: [email protected] M. H. Mazlan Microelectronics and Nanotechnology - Shamsuddin Research Centre, Universiti Tun Hussein Onn Malaysia, Persiaran Tun Dr. Ismail, 86400 Parit Raja, Johor, Malaysia A. I. Pangesty Department of Metallurgical and Material Engineering, Faculty of Engineering, Universitas Indonesia, Depok City, West Java 16424, Indonesia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_8
95
96
A. M. A. Rashid et al.
1 Introduction Bedridden is caused by a medical condition such as a stroke, cerebral palsy, brain injury, or spinal cord injury. It is projected that by 2025, 13.4% of Malaysia’s population will be over 65 years old [1], and 0.2% of them will be bedridden [2]. Formation of heel ulcers, muscle tension, and stiffness are common issues among the bedridden, but limb support, an active routine, and passive activity can help maintain muscle tone and flexibility. Inactive muscle on a body part can cause muscle shortening and atrophy. Thus, bedridden patients do need ankle foot orthosis (AFO) in order to maintain the position of patients’ ankles and feet, relieving or redistributing weight-bearing forces and preventing ankle dorsiflexion [3]. AFOs, which are closely fitted to the anthropometry of the wearer, are ideal for treating patients with lower limb impairments. Despite their low cost, most massproduced AFOs on the market do not meet the following standards. Existing AFOs that fit individual anthropometry, on the other hand, are costly due to their complicated geometry and require the input of qualified and experienced orthopaedic technicians. The long lead time required to manufacture an orthosis adds to these costs [4]. Many versions of the AFO device have been developed to assist patients with drop foot and gait conditions to regain their normal walking condition [5]. 3D printing technology allows manufacturers to overcome these limitations and has been used widely to fabricate assistive devices for people living with disabilities [6]. The benefits of 3D printed AFO encompass fast fabrication, as most traditional orthoses take a long time to make since the individual pieces must be merged manually [4]. Moreover, 3D printing technology also allows high customizability of a product [7]. Also, since the parts can easily be reproduced, refabricating the parts with the very same value can be done at any time [8, 9]. Topology optimization has the potential to significantly improve the performance of structures in a variety of engineering applications. One of the most important aspects to consider while creating a product is the stiffness. Optimization of structure aims to maximise a structure’s performance while adhering to multiple conditions like a limited amount of material. Due to the limitation of material resources, impact on the environment, and technological rivalry, all of which demand lightweight, low-cost, and efficient structures, optimal structural design is becoming vital [10]. This study was conducted with the aim of observing the effect of topology optimized AFO model on pressure distribution.
2 Methodology 2.1 3D Scanning of Foot A point cloud reproduction of the lower leg and foot anatomy of an actual patient must be acquired before anything else can be done to build a custom-fitted anklefoot orthosis (AFO). This must be done before anything else can be done. Using 3D
Topology Optimization for Custom Bed-Resting Ankle Foot Orthosis
97
scanning technology and a Creality CR-scan 01 3D Scanner, this can be performed. In order to construct a single model, many scans had to be manually merged and registered first. This model was then registered globally to improve the anatomical data. After that, the scanned data was kept intact while the noise and outlying areas were mitigated with the use of the CR Studio 2.0 (Shenzhen Creality 3D Technology Co. Ltd.) program. Before it could be imported into Solidworks SolidWorks 2021 (Dassault Systems SolidWorks Corp., USA) for conceptual design, the final stereolithography (.stl) file had to be developed. This was done so that three different ideas could make use of the same-sized foot mold.
2.2 Conceptual Design of AFO The idea and concept for creating the desired product were related to existing AFOs. A total of three conceptual designs were sketched and developed using SolidWorks 2021 (Dassault Systems SolidWorks Corp., USA) software based on the 3D scanned foot model. Two of the conceptual designs are solid AFOs, while the other is a foldable AFO. The foldable AFO consists of a separate foot shell and shaft. All three designs have the same function, which is to keep the ankle in a neutral position but allow some restriction of movement to prevent muscle shortening or atrophy. Figure 1 shows the sketches of the design concept of the AFO.
2.3 Topology Optimization The design has been optimized using topology optimization in SolidWorks 2021 (Dassault Systems SolidWorks Corp., USA) to analyze iterations of the component design that meet a specific optimization objective, namely the best stiffness-to-weight ratio, and its geometric limits are symmetry and thickness control. Symmetry control is semi-symmetric, while thickness is fixed between 0.6 and 1.2 cm. Topology studies are used to define a desired outcome, such as the best stiffness-to-weight ratio, mass reduction, or displacement of a part. With topology-optimized structures, 3D printing allows engineers to overcome the limitations of traditional manufacturing processes and focus on developing lightweight, high-performance structures. The result of the topology study is the best ratio between stiffness and weight. Therefore, the weight reduction is calculated between the original AFO design model and the optimized AFO design model (Fig. 2).
98
A. M. A. Rashid et al.
Fig. 1 Design concepts
2.4 Finite Element Analysis Once the design was finalized, finite element analysis (FEA) was conducted using SolidWorks 2021 (Dassault Systems SolidWorks Corp., USA) software, a method of predicting how a product will behave under certain conditions. Polylactic acid (PLA) with Young’s modulus of 3500 MPa and Poisson ratio of 0.36 was chosen as the material because it is inexpensive and widely available on the market [11, 12]. The FEA study was performed for each optimized AFO design based on different applied loads, which were determined based on data from previous research work that measured the weight of each part of a human body [13]. The results obtained from the analysis are the Von Mises stress and displacement. Table 1 shows the mean body weight segment of a respective part of the human body.
Topology Optimization for Custom Bed-Resting Ankle Foot Orthosis
99
Fig. 2 Parametric settings for topology study
Table 1 Mean body weight segment [13]
Segment
Mass (%) Female
Male
Foot
4.81
4.33
Shank
1.29
1.37
3 Results and Discussion 3.1 Topology Optimization Figure 3 shows the result of topology optimization of the design to get the best stiffness-to-weight ratio with the manufacturing controls of symmetry and thickness control. The figure shows the area that can be removed to achieve its goal of getting the best stiffness by reducing the mass of the design. Figure 4 shows the design optimization model after the area that can be removed according to the topology study has been removed. The weight of the designs decreases according to what is removed. The weight reduction of the model can be seen in Table 2. Table 2 shows that concept 3 has the highest weight reduction of 55.89% from its initial weight of 0.560 to 0.247 kg, followed by concept 1, which reduces 17.17%
100
A. M. A. Rashid et al.
Fig. 3 Result of topology optimization
Fig. 4 Design after topology optimization Table 2 Weight reduction of AFO Design concept
Weight before optimization (kg)
Weight after optimization (kg)
Weight reduction (%)
Concept 1
0.594
0.492
17.17
Concept 2
0.273
0.259
5.13
Concept 3
0.560
0.247
55.89
Topology Optimization for Custom Bed-Resting Ankle Foot Orthosis
101
of its initial weight of 0.594 kg and is 0.492 kg after optimization. Concept 2 has the most minor weight reduction, with only 5.13% of its initial weight from 0.273 to 0.259 kg. Through topology optimization, weight reduction can contribute to lower production costs, especially when fabrication is done through the 3D printing method [14].
3.2 Effects of Different Load The loads applied are 140, 280, and 560 N. For concept 1, as can be seen in Figs. 5 and 6, the biggest load applied, which is 560 N, has the highest value of maximum VMS with the value of 34.11 MPa, and its maximum displacement is 3.289 mm. The smallest value of maximum VMS is 8.53 MPa, and the displacement is 0.822 mm which is when 140 N is applied to it. When 280 N load is applied, the maximum VMS is 17.06 MPa, and the displacement is 1.645 mm.
Fig. 5 VMS visual analysis for concept 1
Fig. 6 Displacement visual analysis for concept 1
102
A. M. A. Rashid et al.
For concept 2, it can be seen in Figs. 7 and 8 that when 140 N load is applied, the maximum VMS is 16.46 MPa, and the displacement is 0.831 mm. Then, when 280 N is applied, the maximum VMS and displacement increase to 32.94 MPa and 1.646 mm, respectively. The maximum VMS is 65.98 MPa when a 560 N load is applied, and the displacement is 3.232 mm. As shown in Figs. 9 and 10, for concept 3, the maximum VMS and the displacement when 140 N load is applied are 14.18 MPa and 3.501 mm, respectively. The maximum VMS increases to 28.35 MPa, and the displacement also increases to 7.002 mm when a 280 N load is applied. When the biggest load of 560 N is applied, the maximum VMS is 45.41 MPa, and the displacement is 17.000 mm. Table 3 shows the values of maximum VMS and maximum displacement for all three concepts when different loads of 140, 280, and 560 N are applied to the foot. As can be seen in Fig. 11, concept 2 has the highest maximum VMS for every load applied, followed by concept 3, then concept 1, which has the lowest maximum VMS compared to the other two. Figure 12 illustrates that concept 3 has the highest maximum displacement compared to the other two concepts, which have a similar value of maximum displacement with only slight differences between those two
Fig. 7 VMS visual analysis for concept 2
Fig. 8 Displacement visual analysis for concept 2
Topology Optimization for Custom Bed-Resting Ankle Foot Orthosis
103
Fig. 9 VMS visual analysis for concept 3
Fig. 10 Displacement visual analysis for concept 3
concepts. From Table 3, it was proven that all designs do not fail upon loads applied as the highest VMS values of all designs do not exceed the yield strength of the material, which is 70 MPa [11]. Concept 2 shows the highest VMS and displacement values compared to other designs, respectively, which are 65.98 MPa and 3.232 mm. This can be caused by the slim design of concept 2, which reduces the ability to distribute the stress and brings down the stiffness. Concept 2 has a property of trimlined AFO, which had been proven by Surmen et al. to have lower stiffness [15]. Table 3 Maximum VMS and displacement for each concept Concept 1
Concept 2
Concept 3
Load (N) Max. VMS Max. Max. VMS Max. Max. VMS Max. (MPa) displacement (MPa) displacement (MPa) displacement (mm) (mm) (mm) 140
8.53
0.822
16.46
0.831
14.18
3.501
280
17.06
1.645
32.94
1.646
28.35
7.002
560
34.11
3.289
65.98
3.232
45.41
17.000
104
A. M. A. Rashid et al.
Fig. 11 Load versus maximum VMS for each concept
Fig. 12 Load versus maximum displacement for each concept
Trimlined AFO is used to correct abduction or adduction and plantarflexion or dorsiflexion of the foot. AFO stiffness can be affected by the patient’s biomechanical conditions, such as the weight and the level of deformity of the patient. For example, for a more flexible AFO, a trimline design needs to be applied at the ankle section; meanwhile, for adduction and abduction correction, the medial and lateral need to be thickened, respectively [16]. In designing and developing AFO, key parameters that need to be considered are the geometrical shape, stiffness, and material type, as these parameters can affect the effectiveness of the AFO [17].
Topology Optimization for Custom Bed-Resting Ankle Foot Orthosis
105
4 Conclusion The results of the topology optimization show that concept 3 has the highest weight reduction. This is because most of the area that is not loaded can be removed after the topology optimization. The FEA study conducted after optimizing the design shows that the more load applied to the footrest, the higher the value of maximum VMS and displacement. Even though the VMS increases after optimizing the design, the results show that the stress values are still acceptable as they do not exceed their tensile strength of 70 MPa. Some improvements can be made to the AFO model. For example, the AFO can be made according to the patient’s foot size. To obtain the measurements of the patient’s feet, a 3D scanner can be used to scan the patient’s feet. Also, an automatic mechanism can be installed to help the patient move his ankle quickly. This can be a great help for the patient to avoid weakening or atrophy of the muscles, as a patient who is confined to bed does not use his muscles often. Muscles that are not used regularly degenerate very quickly. Acknowledgements This study was supported by the Universiti Teknologi MARA, Malaysia, and Universitas Indonesia under the Strategic Research Partnership (SRP) UI-UITM BISA 2021 Grant, Grant No: 100-RMC 5/3/SRP (037/2021). We thank and acknowledge the College of Engineering, UiTM, and our colleagues from Universitas Indonesia, who provided insight and expertise in the research work.
References 1. Koris R et al (2019) The cost of healthcare among Malaysian community-dwelling elderly. Jurnal Ekonomi Malaysia 53(1):89–103 2. Tantirat P et al (2020) Projection of the number of elderly in different health states in Thailand in the next ten years, 2020–2030. Int J Environ Res Public Health 17(22):8703 3. Khan S, Radzmi I (2021) Design and analysis of various thermoplastic for optimized ankle foot orthosis. J Phys Conf Ser (IOP Publishing) 4. Dal Maso A, Cosmi F (2019) 3D-printed ankle-foot orthosis: a design method. Mater Today: Proc 12:252–261 5. Ab Wahid AM et al (2022) Development of ankle-foot orthosis with the integration of IoT controller. Int J Emerg Technol Adv Eng 12(5):49–55 6. Hashim NM et al (2022) Creating an inclusive ecosystem through healthcare in disability management: Malaysians’ experience. Disabil CBR Incl Dev 33(1) 7. Mazlan MA et al (2021) 3D printed assistive writing device for phocomelia patient. Malays J Med Health Sci 17:7–11 8. Cha YH et al (2017) Ankle-foot orthosis made by 3D printing technique and automated design software. Appl Bion Biomech 2017 9. Chen RK et al (2016) Additive manufacturing of custom orthoses and prostheses—a review. Addit Manuf 12:77–89 10. Huang X, Xie M (2010) Evolutionary topology optimization of continuum structures: methods and applications. Wiley
106
A. M. A. Rashid et al.
11. Farah S, Anderson DG, Langer R (2016) Physical and mechanical properties of PLA, and their functions in widespread applications—a comprehensive review. Adv Drug Deliv Rev 107:367–392 12. Zhang Y et al (2020) A physical investigation of wear and thermal characteristics of 3D printed nylon spur gears. Tribol Int 141:105953 13. Chauvie S et al (2012) A method for the visual analysis of early-stage Parkinson’s disease based on virtual MRI-derived SPECT images. Int J Imaging Syst Technol 22(3):172–176 14. Raj R et al (2022) Numerical and experimental mechanical analysis of additively manufactured ankle-foot orthoses. Materials 15(17):6130 15. Surmen HK, Arslan YZ (2021) Evaluation of various design concepts in passive ankle-foot orthoses using finite element analysis. Eng Sci Technol Int J 24(6):1301–1307 16. Surmen HK, Akalan NE, Arslan YZ (2019) Design, manufacture, and selection of ankle-footorthoses. In: Advanced methodologies and technologies in artificial intelligence, computer simulation, and human-computer interaction, IGI Global, pp 250–266 17. Russell Esposito E et al (2014) How does ankle-foot orthosis stiffness affect gait in patients with lower limb salvage? Clin Orthop Relat Res® 472:3026–3035
Influence of Environmental Factors and Road Characteristics in Commuting Accidents Among Public University Staffs Mohd Najib Yaacob and Khairiah M. Mokhtar
Abstract The occurrence of commuting accidents is associated with various factors; individual factors, workload demand factors, environmental factors and road characteristics. These factors result in not only pay out but productivity loss and negative psychological consequences. Hence, this study aims to explore the influence factors of commuting accidents among staff in a public university, University A. A questionnaire is distributed among staff who were involved in accidents in University A. Data from the questionnaires are analysed by using Statistical Package for the Social Sciences (SPSS), and the Spearman’s Rho method is applied in identifying the correlation between environmental factors and road characteristics with commuting accidents. The findings demonstrated a strong correlation between the time of day and the commuting accidents as reflected by the p-value of 0.006. A collaboration between various government agencies and employers is vital in providing training, experience sharing, and safety and defensive riding education to reduce the number of commuting accidents. Keywords Road transport safety · Commuting accidents · Environmental factors · Road characteristics
1 Introduction Commuting accidents are another type of occupational accident workers can be involved in. A workplace accident is any physical injury experienced by an individual attributed to working, while a commuting accident occurs on the way to and from the place of work [1]. According to Section 24 of the Employees Social Security Act 1969, a commuting accident is defined as an accident happening while travelling on a route between a place of residence to a place of work, while travelling for any M. N. Yaacob · K. M. Mokhtar (B) Faculty of Industrial Sciences and Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, 26300 Kuantan, Pahang, Malaysia e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_9
107
108
M. N. Yaacob and K. M. Mokhtar
reason related to their employment, or while travelling during an authorised break [2]. There are various influence factors associated with commuting accidents including individual/human factors [3, 4], workload demand factors [3, 5], road characteristics [6–9] and environmental factors. The risk of accidents is increased by individual factors like speeding and using a phone while driving due to loss of concentration [4, 5]. Similarly, workload demand factors impact driving focus due to physical and emotional exhaustion [3, 5]. Additionally, road condition factors such as road geometry, road markings and roadside installations (i.e. barriers and posts) contributed to road accidents, particularly those involving motorcyclists [6]. Literature demonstrated the risks associated with road geometry are curves, straight road sections, the horizontal curve’s radius and length, and shoulder width [7, 8]. Meanwhile, another study [9] highlighted unevenness, slippery surfaces, repaired patches on the road, longitudinal parallel grooves, drain covers and gratings. Additionally, environmental factors such as bad weather are also significant contributors to commuting accidents [10]. According to the Road Transport Department of Malaysia, the number of commuting accidents reported in 2017 shown in Fig. 1 is 6% higher than that of reported cases in the previous year [11]. In the 2019 annual report published by Social Security Organisation (SOCSO), 77,642 accident cases were reported, reflecting a 6.9% increase from the reported cases in 2018. Of these, 48.74% were associated with work-related commuting accidents, which results in a significant payout annually [12]. Besides medical costs, commuting accidents also cause productivity loss, leisure time, permanent invalidity-related costs and detrimental psychological effects [13]. SOCSO’s statistics, which also demonstrate two out of every three daily fatalities are caused by commuting accidents, reflect the need for a proactive mitigation strategy. Hence, the purpose of the present study is to investigate the factors contributing to commuting accidents among staff in a public university, University A.
2 Methodology All methods involved in this study were illustrated in Fig. 2. This study was conducted to determine the influence of environmental factors and road characteristics on commuting accidents among University A staff.
2.1 Document Review All relevant documents from Malaysian Institute of Road Safety Research (MIROS), the Royal Malaysian Police report and the Department of Transport of Malaysia have been reviewed to gather statistical data on commuting accidents. Technically, the Why
Influence of Environmental Factors and Road Characteristics …
109
Fig. 1 Total of commuting accidents in Malaysia (2003–2017). Source Official Portal of Road Transport Department of Malaysia (2017)
Fig. 2 Research flow
Because Analysis was applied to gather all the necessary information including time of day, travel purpose, weather, road surface, road condition, road lane, the presence of traffic lights, divided existence, the police report and vehicle insurance claim.
110
M. N. Yaacob and K. M. Mokhtar
2.2 Semi-structured Interview Technique The Head of the Police Traffic Station Kuantan has been interviewed to gain information on the accident statistics and common causes of accidents. Furthermore, a one-on-one interview session with the identified University A staff members has been held. The inclusion of respondents with and without commuting accidents history was also ensured. Informed consent was sought before the interview session began, and the confidentiality of collected data was guaranteed. The interviews which lasted for 30–45 min were recorded with the respondent’s consent, and some notes were collected for verification at the end of the interviews to enhance the validity of the information gathered. The first part of the interview contained questions on demographics and driving profiles. The respondents were then questioned regarding five different categories of road characteristics and environmental factors. The respondents were questioned about whether they had ever been involved in commuting accidents, how long it took, and the extent to which these circumstances enhanced the probability of a crash based on a Likert scale from “0” to “5” (extremely dangerous). The interview also includes input on the correlation of commuting accidents with the identified factors.
3 Results and Discussion 3.1 Descriptive Analysis Figure 3 illustrates the purposes of travelling for University A staff who experience road accidents. According to The Royal Society for the Prevention of Accidents’ definition, commuting and returning, break and performing official duties are all included under commuting accidents [14]. Referring to this definition, commuting accidents account for 73% of reported accidents, while personal reasons such as managing children’s school matters or so-called work-family conflict (WFC) account for only 27% [15, 16]. As shown in Fig. 4, accidents occur most frequently between 6.00 a.m. and 9.00 a.m., when people commute to work and the rush hour for school. Meanwhile, in terms of weather factor, the majority of the accidents (85%) occurred in fine weather as illustrated in Fig. 5. This result supports the finding from the literature [17] that clear weather, in addition to dry road surfaces, was found to increase accident magnitude. The road characteristics comprising of road surface, road condition, road lane, traffic light and road divider are studied, and the findings are presented in Figs. 6, 7, 8 and 9, respectively. Based on Fig. 6, 91% of commuting accidents occurred on dry road surfaces rather than wet ones, which may be because drivers are more likely to pay attention and slow down on slippery roads. Meanwhile, the road condition factor illustrated in Fig. 7 shows that curve and T-junction roads both had 25% and 19% of all accidents, respectively, whereas straight roads had 50% of all accidents.
Influence of Environmental Factors and Road Characteristics …
111
Fig. 3 Travel purpose
Fig. 4 Travel time when commuting accidents occur
This result is consistent with the findings in the literature [6] that the frequency of accidents is greatly influenced by straight roads, which are classified as part of road geometrics. Drivers may be exceeding the speed limit since there are lesser traffic lights to act as a deterrent and curves on the straight road compared to T-Junction and 4-Junction. Drivers must slow down around these obstacles or they risk skidding. Additionally, as shown in Fig. 8, two lanes caused more commuting accidents than broader roads with four lanes. It is reported that two lanes accounted for 73% of accidents. Another factor categorised under road characteristics; road divider as demonstrated in Fig. 9 was a contributing factor in 55% of accidents. Meanwhile, Fig. 10 illustrates that the majority of the victims were car drivers compared to motorcycle riders. Concerning mode choice, researchers [17–20] have reported that
112
M. N. Yaacob and K. M. Mokhtar
Fig. 5 Weather conditions during the occurrence of the commuting accident
Fig. 6 Road surface during the occurrence of the commuting accident
weather has a significant influence in determining the choice between open-air transport modes (such as motorcycling or bicycling) and in-vehicle modes (such as driving a car). Therefore, it can be presumed that the staff of University A drives rather than rides a motorcycle to work each day.
3.2 Inferential Analysis The relationship between environmental factors and road characteristics, and commuting accidents is demonstrated in Table 1. The p-value of the finding (0.006) shows a significant correlation between commuting accidents and the time of day or
Influence of Environmental Factors and Road Characteristics …
113
Fig. 7 Road condition during the occurrence of the commuting accident
Fig. 8 Number of lanes at the location of the commuting accident
respondents’ travel time under environmental factors. This result is consistent with the research showing that 88% of accidents happened while people were travelling to and from their work [13]. Previous studies [21–23] have proved that the time of day significantly affects the vulnerability of accident-injury severity in road accidents. Moreover, the study on eye-tracking experiments [24, 25] demonstrated the significance of pupil diameter and gaze time in determining the level of concentration among drivers. Most nighttime drivers may require a higher level of concentration than daytime drivers, due to the driver’s gaze points on the road. Without a doubt, the driver’s concentration directly increases the risk of fatigue, which results in commuting accidents.
114
M. N. Yaacob and K. M. Mokhtar
Fig. 9 Presence of road divider at the location of the commuting accident
Fig. 10 Mode of transport used by the commuting accident victims
Influence of Environmental Factors and Road Characteristics … Table 1 Correlation between influence factors and commuting accidents
Factors
Correlation coefficient
115
P-value
Environmental Time of day
0.467
0.006**
Weather
−0.042
0.817
Mode of transport
−0.42
0.817
Road surface
−0.083
0.645
Road condition
−0.52
0.774
Road lane
−1.88
0.296
Traffic light
−0.283
0.110
0.052
0.775
Road characteristics
Road divider
Note ** Correlation is significant at the 0.01 level (2-tailed)
4 Conclusion The relationship between environmental factors and road characteristics, and commuting accidents has been explored. The findings show that the time of day classified under environmental factor is significant for the occurrence of commuting accidents, particularly when workers commute to and back from work. However, the factors categorised under road characteristics, such as road surface, road conditions and road divider, are not giving significant impacts. Effort in decreasing the risk of commuting accidents requires actions from all parties involving government institutions, employers as well as staff. With coordinated efforts from government organisations like the Department of Occupational Safety and Health and the Malaysian Institute of Road Safety Research (MIROS), much more may be accomplished. Unquestionably, a combination of governmental regulations, employee attitudes and excellent practices introduced by employers is essential in developing a positive safety culture at work. The provision of training by employer is an effort in enhancing awareness and knowledge about road safety. Further, experience sharing, safety and defensive riding education to society are other initiatives in diminishing the number of commuting accidents. Acknowledgements Acknowledgements are made to the following UMP research funding, RDU210359, for granting permission to produce this kind of research.
116
M. N. Yaacob and K. M. Mokhtar
References 1. International Labour Organization (ILO) (2010) World Social Security Report 2010/ 11. http://www.ilo.org/wcmsp5/groups/public/%2D%2Ddgreports/%2D%2D-dcomm/%2D% 2D-publ/documents/publication/wcms_146566.pdf 2. Social Security Organization (SOCSO) (1969) Employees’ Social Security Act 1969 3. Zamani-Alavijeh F, Niknami S, Bazargan M, Mohammadi E, Montazeri A, Ahmadi F, Ghofranipour F (2009) Accident-related risk behaviours associated with motivations for motorcycle use in Iran: a country with very high traffic deaths. Traffic Inj Prev 10(3):237–242 4. Regan M, Lee J, Young K (2008) Driver distraction: theory, effects and mitigation. CRC Press, New York, USA 5. Oxley J, Yuen J, Ravi MD, Hoareau E, Mohammed MAA, Bakar H, Venkataraman S, Nair PK (2013) Commuter motorcycle crashes in Malaysia: an understanding of contributing factors. Ann Adv Automot Med 57:45 6. Manan MMA, Várhelyi A, Çelik AK, Hashim HH (2018) Road characteristics and environmental factors associated with motorcycle fatal crashes in Malaysia. IATSS Res 42(4):207–220 7. Clarke DD, Ward P, Bartle C, Truman W (2007) The role of motorcyclists and other driver behaviour in two types of serious accidents in the UK. Accid Anal Prev 39(5):974–981 8. Schneider WH IV, Savolainen PT, Moore DN (2010) Effects of horizontal curvature on singlevehicle motorcycle crashes along rural two-lane highways. Transp Res Rec 2194(1):91–98 9. Preusser DF, Williams AF, Ulmer RG (1995) Analysis of fatal motorcycle crash: crash typing. Accid Anal Prev 27(6):845–851 10. Yaacob SS, Ismail KI, Shaarial SZM, Noor NM, Selvaraju R, Ab Ghani H (2018) Commuting accidents among health care workers working in Malaysia government hospitals. KnE Life Sci 79–87 11. Road Transport Department of Malaysia (2017) Official Portal of Road Transport Department of Malaysia 12. Social Security Organization (SOCSO) (2019) 2019 Annual Report. https://www.perkeso.gov. my/images/laporan_tahunan/AR_2019_FINAL.pdf 13. Aertsens J, de Geus B, Vandenbulcke G, Degraeuwe B, Broekx S, De Nocker L, Liekens I, Mayer I, Meeusen R, Thomas I, Torfs R (2010) Commuting by bike in Belgium, the costs of minor accidents. Accid Anal Prev 42(6):2149–2157 14. The Royal Society for the Prevention of Accidents (2012) Driving for work: safer speeds, 1(July):1–12 15. Turgeman-Lupo K, Biron M (2017) Make it to work (and back home) safely: the effect of psychological work stressors on employee behaviour while commuting by car. Eur J Work Organ Psy 26(2):161–170 16. Parkes LP, Langford PH (2008) Work-life balance or work–life alignment? A test of the importance of work-life balance for employee engagement and intention to stay in organisations. J Manag Organ 14(3):267–284 17. Lee J-Y, Chung J-H, Son B (2008) Analysis of traffic accident size for Korean highways using structural equation models. Accid Anal Prev 40(6):1955–1963 18. Heinen E, Maat K, van Wee B (2011) Day-to-day choice to commute or not by bicycle. Transp Res Rec 2230(1):9–18 19. Bocker L, Dijst M, Prillwitz J (2013) Impact of everyday weather on individual daily travel behaviours in perspective: a literature review. Transp Rev 33(1):71–91 20. Cools M, Creemers L (2013) The dual role of weather forecasts on changes in activity-travel behaviour. J Transp Geogr 28:167–175 21. Liu C, Susilo YO, Karlstrom A (2017) Weather variability and travel behaviour – what we know and what we do not know. Transp Rev 37(6):715–741 22. Plainis S, Murray IJ (2002) Reaction times as an index of visual conspicuity when driving at night. Ophthalmic Physiol Opt 22(5):409–415 23. Hao W, Kamga C, Wan D (2016) The effect of time of day on driver’s injury severity at highwayrail grade crossings in the United States. J Traffic Transp Eng (English edition) 3(1):37–50
Influence of Environmental Factors and Road Characteristics …
117
24. Jägerbrand A, Sjöbergh J (2016) Effects of weather conditions, light conditions, and road lighting on vehicle speed. Springerplus 5(1):1–17 25. Konstantopoulos P, Chapman P, Crundall D (2010) Driver’s visual attention as a function of driving experience and visibility. Using a driving simulator to explore drivers’ eye movements in day, night and rain driving. Accid Anal Prev 42(3):827–834
Effects of Material Properties in Developing the Ear Prosthetics Abdul Halim Abdullah, Mohd Noor Asnawi Mohd Noordin, Suziana Ahmad, Nor Fazli Adull Manan, and Shahrul Hisyam Marwan
Abstract A person may develop ear problems because of an accident or a birth condition. Such flaws damage a person’s appearance and feelings of confidence, forcing them to shun social situations. To improve their appearance, patients may choose to have a prosthetic ear implanted. This project entails a 3D scanning model of an artificial ear as well as analyzing various material properties such as thermoplastic polyurethane (TPU), thermoplastic elastomer (TPE), and nylon (PA6). These are the alternatives chosen to replace the current use of silicone material to produce the prosthetic ear. This project uses a 3D scanner to scan an artificial ear model that was made from silicone material. The aim of analyzing various materials on the artificial ear model is to decide the best material that can be used for the patient’s head. The material is analyzed using finite element analysis on SOLIDWORKS software. The material that is suitable for the patient to be used as a prosthetic ear is Nylon 6 due to its low deformation value and lowest stress value. This prosthetic ear allows the patient to participate more fully in their social and family lives, making them happier and more at ease. Keywords Thermoplastic polyurethane · Thermoplastic elastomer · Nylon 6 · Finite element analysis
A. H. Abdullah · M. N. A. M. Noordin · N. F. A. Manan · S. H. Marwan (B) Biomechanical & Clinical Engineering (BioMeC) Research Group, College of Engineering, Universiti Teknologi MARA (UiTM), 40450 Shah Alam, Selangor, Malaysia e-mail: [email protected] S. Ahmad Faculty of Electrical and Electronic Engineering Technology, Universiti Teknikal Malaysia Melaka, Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia S. H. Marwan School of Mechanical Engineering, College of Engineering, Universiti Teknologi MARA (UiTM) Terengganu Branch, Bukit Besi Campus, 23200 Dungun, Terengganu, Malaysia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_10
119
120
A. H. Abdullah et al.
1 Introduction Ear defects can be due to accidents or birth defects such as microtia. Surgical reconstruction is not always possible for the patient who has the defects. In such instances, prosthetic rehabilitation is the optimal course of therapy since it restores the patient’s natural anatomy and appearance while also providing significant psychological advantages. Based on the study of the incident that was done in Italy, France, Sweden, Finland, and the United States, microtia happens in the ranges of 0.83 and 4.43 per 10,000 births [1]. A prosthetic ear is an artificially made ear that is used to restore a portion of the entire natural ear. It can be created to look like a normal ear, which is suitable for the patient. It can be retained with adhesives or with bone-anchored implants [2]. The material that is commonly used for artificial prosthetic ears is silicone due to its flexibility, lightweight, and lifelike appearance [3]. Prosthetic ears made from silicone would cost between $2000 and $7000 for a pair [2]. Thus, 3D printing technology was chosen in this study as an alternative to silicone by using cheaper materials and fabrication methods. Materials chosen for the study are thermoplastic polyurethane (TPU), thermoplastic elastomer (TPE), and nylon (PA6). These materials are chosen because they have a soft filament, which is the most important factor for efficient soft 3D printing manufacturing. Thermoplastic polyurethane (TPU) is a subset of robust, multi-purpose thermoplastic elastomers that are well known for their durability, flexibility, and appropriateness for a wide variety of high-performance applications. The TPU material has properties that fall between rubber and plastic. Due to the material’s thermoplastic nature, it has several advantages over other elastomers [4]. The special features of TPU material are its great tensile strength, excellent load-bearing capacity, and high elongation at break. Thermoplastic elastomers (TPE) are materials that combine the production characteristics of thermoplastics with the mechanical qualities of vulcanized rubbers [5]. TPE is a polymer blend or compound that exhibits thermoplastic behavior above its melt temperature, allowing it to be shaped into a fabricated article, and that exhibits elastomeric behavior within its design temperature range without crosslinking during fabrication. Due to the reduced density of most TPEs compared to traditional rubber compounds, their volume cost is frequently cheaper [6]. TPEs are accessible as extremely soft gel materials with hardness values ranging from 20 Shore OO to 90 Shore A, at which time they reach the Shore D range and may be manufactured to have hardness values as high as 85 Shore D, which denotes a very hard material. Nylon is one of the most commonly used engineering thermoplastics due to its flexibility. Nylon 6 is a very durable and abrasion-resistant material. In comparison with nylon 6/6, it offers a better surface look and processability. Additionally, because it is slightly less crystalline, it may be molded at temperatures as low as 27 °C with reduced mold shrinkage [7].
Effects of Material Properties in Developing the Ear Prosthetics
121
3D scanning technology was introduced in 1960, and the technology eventually revolutionized the science of anthropometry [8]. 3D scanning is a method that can generate a three-dimensional model of an object based on an acquisition system specifically configured to capture images in real-time. This system is designed based on the acquisition system to capture a multitude of images of the object and compute them in depth maps [9]. 3D printing is a method of manufacturing in which materials such as metal or plastic are deposited onto one another in layers to produce a three-dimensional model [10]. Nowadays, 3D printing has been widely used in engineering, not only to produce engineering prototypes but also to produce objects rather than traditionally manufacture them. 3D printing technology has benefits in medical applications. The first one is that it gives the freedom of customization to produce custom-made medical products or equipment [11]. Next, the ability to produce items cheaply has made 3D printing technology preferable to traditional manufacturing. It is cost-efficient, especially for small-sized standard implants and prosthetics [11]. 3D printing also enhances productivity, as a product can be made within several hours [12]. Thus, in this research, a static analysis of the ear prosthetic was conducted to predict the effects of different material properties selected in the fabrication of the ear prosthetic using the 3D printing technique. The Solidworks simulation software was used in the computation analysis. The best material selection will be proposed to replace the silicone rubber material that is currently on the market.
2 Methodology To develop the artificial ear, the study was broken down into two phases. The first phase involved 3D scanning, which was done to get the exact three-dimensional ear model. Meanwhile, in phase two, several steps were done: a literature review of thermoplastic polyurethane (TPU), thermoplastic elastomer (TPE), and nylon 6 (PA6), and a finite element analysis (FEA) of the ear model. Figure 1 shows the flowchart of this project.
Fig. 1 Flowchart of the progress of the project
122
A. H. Abdullah et al.
2.1 Phase 1: 3D Scanning Model of an Artificial Ear Phase 1 was done to get a 3D ear model by using a silicone ear model. For this process, a 3D scanner, an Intel RealSense camera SR300, was used to scan the silicone ear model, as illustrated in Fig. 2. It is a portable depth camera capable of providing a VGA-size depth map at 60 fps and 0.1255 mm depth resolution [13]. To use the 3D scanner, software called 3D System Sense was installed. Before starting the scanning process, the ear model was placed on a flat surface covered by black cloth, which contrasts with the silicone color. The ear model used in this study is made of soft silicone with a real human ear size and is a suitable model for science and education purposes. Then, the 3D scanner was plugged into the laptop using a USB cable to connect to the 3D System Sense software. Start scanning by pressing the scan button on the software. Make sure to position the scanning device looking down at a 45° angle to the surface so that the ear model is visible on the screen. Move slowly around the object until you complete a full 360° scan. If the scanner moves too rapidly, the notification box will indicate that the user should slow down the scanning process. The result of the scanning process is shown in Fig. 3. After finishing the scanning process, the 3D model must undergo the editing process, which will solidify and trim the unwanted areas of the scanned model. The editing process started with enabling the wireframe and removing the color function through the setting function on the top-right window. This process involves the crop function, which selects the targeted area. Other than that, I used the trimming function to remove the big area and the erase function to remove small areas and refine the model. Then, continue with the solidify function, which fills the hole in the model. Fig. 2 Intel RealSense 2 camera SR300
Effects of Material Properties in Developing the Ear Prosthetics
123
Fig. 3 3D scanning process of a silicone ear model using sense 3D scanner
Fig. 4 Final ear model using sense 3D scanner
Lastly, click the finish button once the editing process is complete. The model was saved and exported into 3D printing format STL. The result can be seen in Fig. 4.
2.2 Phase 2: Finite Element Analysis for Artificial Ear Finite element analysis was widely used in predicting the performance of the developed product, especially in medical engineering [14–16]. For this phase, the FEA
124
A. H. Abdullah et al.
Table 1 Material properties for 3D printed parts [19] Material
Young’s modulus (MPa)
Poisson ratio
Density (kg/m3 )
Yield strength (MPa)
TPU
2410
0.3897
1240
53.8
TPE
1622
0.394
960
Nylon 6
2950
0.42
1120
29 103.649
engineering software that was used was SOLIDWORKS. SOLIDWORKS Simulation can produce precise, dependable findings for a broad variety of study types, ranging from simple linear static analysis to more sophisticated nonlinear and dynamic analysis. The design will undergo one of the finite element analyses, which is a static analysis. Two case studies have been done: the effect of axial load and the effect of pressure on the ear prosthetic. The outer rim of the ear, also known as the helix, will be the most critical part of the ear prosthetic as it is the most vulnerable. The materials used are thermoplastic polyurethane (TPU), thermoplastic elastomer (TPE), and nylon 6 (PA6). For finite element analysis, the material data are set up according to Table 1. The selected material filament used in this study was similar to previous studies that developed adaptive assistive devices [17, 18]. The axial load and pressure were set to 5 N and 2000 MPa, respectively. The data for the material properties were obtained from the Materials Data Book by the Cambridge University Engineering Department. The study is performed to determine the von Mises stress and deformation for different material settings, therefore determining the optimal fabrication process setting. Fixed geometry is defined on the supports, and loading is applied to the surfaces shown in Fig. 5. The green indicators represent fixture supports, while the red arrows shown in Fig. 6 indicate the load or pressure directly applied to the surface of the ear model. Next, the ear model will be divided into a smaller unit called meshing, separating it into finite components. Mesh processing will be used to split the CAD model of the ear prosthetic into smaller regions called elements. All components were constructed using tetrahedral elements. The number of elements meshing for the ear prosthetics is 8431. The element size used is 2.34 mm. Figure 7 depicts a mesh ear prosthetic model.
3 Results and Discussion In the static analysis, the type of material used for the artificial ear model will be the manipulated variable through the response to the von Mises stress and the static deformation of the artificial ear model. Two case studies have been done: the effect of axial force and pressure on the artificial ear model.
Effects of Material Properties in Developing the Ear Prosthetics
125
Fig. 5 Setting the fixed geometry support
Fig. 6 Setting the external load
3.1 Effect of Material Properties on Artificial Ear (Axial Force) Figure 8 shows the material TPU, TPE, and Nylon 6 deformation when the axial force is applied to the artificial ear prosthetic. Based on Fig. 9, the deformation values for TPU, TPE, and Nylon 6 are 4.194 × 10−6 mm, 6.011 × 10−6 mm, and 2.568 × 10−6 mm, respectively. The result shows that TPE has the highest deformation at
126
A. H. Abdullah et al.
Fig. 7 Meshing of ear model
6.011 × 10−6 mm when 5 N of force is applied to the artificial ear model. Nylon 6 has the lowest deformation value at 2.568 × 10−6 mm. Figure 10 shows von Mises stress on materials TPU, TPE, and Nylon 6 when the axial force is applied to the artificial ear prosthetic. For comparison in Fig. 11, it shows that Von-Misses stress for TPU and TPE have almost the same value of maximum stress, which is 0.001795 MPa and 0.001787 MPa, respectively. Nylon has the lowest value of stress, which is 0.001739 MPa. The summary of the results of the artificial ear model can be identified in Fig. 12.
Fig. 8 Deformation on different materials: TPU, TPE, and Nylon 6 (from left)
Effects of Material Properties in Developing the Ear Prosthetics
127
Fig. 9 Comparison of deformation between TPU, TPE, and Nylon 6
Fig. 10 von Mises stress on different materials: TPU, TPE, and Nylon (from left)
3.2 Effect of Material Properties on Artificial Ear (Pressure) Figure 13 shows the material TPU, TPE, and Nylon 6 deformation when pressure was applied to the artificial ear prosthetic. Based on Fig. 14, the deformation values for TPU, TPE, and Nylon 6 are 5.689 mm, 8.154 mm, and 3.483 mm, respectively. The result shows that TPE has the highest deformation (8.154 mm) when 2000 MPa of pressure are applied to the artificial ear model. Nylon 6 has the lowest deformation value at 3.483 mm.
128
A. H. Abdullah et al.
Fig. 11 Comparison of Von-Misses stress between different materials
Fig. 12 Summary of deformation and von Mises stress on different materials
Figure 15 shows the von Mises stress on materials TPU, TPE, and Nylon 6 when pressure is applied to the artificial ear model. For comparison, Fig. 16 shows that VonMisses stress for TPU and TPE has almost the same value of maximum stress, which is 2435 MPa and 2424 MPa, respectively. Nylon 6 has the lowest value of stress,
Effects of Material Properties in Developing the Ear Prosthetics
129
Fig. 13 Deformation on different materials: TPU, TPE, and Nylon (from left)
Fig. 14 Comparison of deformation between different materials
which is 2358 MPa. The summary of the results when an axial force is applied to the artificial ear model can be identified in Fig. 17. According to the results of the static analysis on the impact of pressure and axial force, as shown in Figs. 12 and 17, Nylon 6 is the most suitable material for printing prosthetic ears since it had the lowest deformation value and lowest von Mises stress among the tested materials.
130
A. H. Abdullah et al.
Fig. 15 von Mises stress on different materials: TPU, TPE, and Nylon (from left)
Fig. 16 Comparison of Von-Misses stress between different materials
Effects of Material Properties in Developing the Ear Prosthetics
131
Fig. 17 Summary of deformation and von Mises stress on different materials
4 Conclusion In conclusion, the objective of 3D scanning the ear model and analyzing the ear model using finite element analysis for different materials has been achieved. By analyzing the ear model using SOLIDWORKS software using different materials, which are TPU, TPE, and Nylon 6, it was discovered that Nylon 6 is the most suitable material to be used for the ear prosthetic. It is due to Nylon 6 obtaining the lowest von Mises stress and deformation values compared to TPU and TPE. The project has given an understanding of the importance of CAD-CAE software in the development of a new product. The advantage of this technique is that it may be produced anywhere on the globe concurrently with the availability of a 3D printer, allowing individuals living in remote regions to rapidly obtain their face prosthesis without having to visit a doctor. For future research, it is recommended to do further research on the flexible filament material that is suitable for real patient conditions so that the best result can be obtained. Acknowledgements This study was supported by the Universiti Teknologi MARA, Malaysia. We thank and acknowledge the College of Engineering at UiTM, which provided insight and expertise in our research work.
132
A. H. Abdullah et al.
References 1. Gendron C, Schwentker A, van Aalst J (2016) Genetic advances in the understanding of microtia. J Pediatr Genet. https://doi.org/10.1055/s-0036-1592422 2. Saadi R, Lighthall JG (2017) Prosthetic reconstruction of the ear. Oper Tech Tolaryngol - Head Neck Surg. https://doi.org/10.1016/j.otot.2017.03.013 3. Chinnasamy A, Gopinath V, Jain AR (2018) Ear prosthesis for postburn deformity. Case Rep Otolaryngol. https://doi.org/10.1155/2018/2689098 4. Prasad A, Fotou G, Li S (2013) The effect of polymer hardness, pore size, and porosity on the performance of thermoplastic polyurethane-based chemical mechanical polishing pads. J Mater Res 28(17):2380–2393. https://doi.org/10.1557/jmr.2013.173 5. Lu YM, Kutka J (2002) Transparent and highly heat-resistant TPE materials. Int Polym Sci Technol 29(7):11–14. https://doi.org/10.1177/0307174x0202900703 6. Drobny JG (2014) Introduction. In: Handbook of thermoplastic elastomers, vol 1, pp 1–11. https://doi.org/10.1016/b978-0-323-22136-8.00001-6 7. Liu FF, Marchildon EK, McAuley KB (2019) Modeling equilibrium behavior of Nylon 6, Nylon 6,6 and Nylon 6/6,6 copolymer. Macromol React Eng 13(2):1–16. https://doi.org/10. 1002/mren.201800078 8. Macgillivray M, Domina T (2007) 3D laser scanning: a model of multidisciplinary research. Accessed 16 January 2021. https://www.researchgate.net/publication/228625999 9. Negru N, Leba M, Rosca S, Marica L, Ionica A (2019) A new approach on 3D scanning-printing technologies with medical applications. IOP Conf Ser Mater Sci Eng 572(1):012049. https:// doi.org/10.1088/1757-899X/572/1/012049 10. Schubert C, Van Langeveld MC, Donoso LA (2014) Innovations in 3D printing: a 3D overview from optics to organs. Br J Ophthalmol 98(2):159–161. https://doi.org/10.1136/bjophthalmol2013-304446 11. Banks J (2013) Adding value in additive manufacturing: researchers in the United Kingdom and Europe look to 3D printing for customization. IEEE Pulse 4(6):22–26. https://doi.org/10. 1109/MPUL.2013.2279617 12. Mertz L (2013) Dream it, design it, print it in 3-D: what can 3-D printing do for you? IEEE Pulse 4(6):15–21. https://doi.org/10.1109/MPUL.2013.2279616 13. Zabatani A et al (2020) Intel® RealSenseTM SR300 coded light depth camera. IEEE Trans Pattern Anal Mach Intell 42(10):2333–2345. https://doi.org/10.1109/TPAMI.2019.2915841 14. Dagang WNFSW, Marwan SH, Mahmud J, Manan NFA, Abdullah AH (2021) The influence of pore size and material properties on biomechanical analysis of parietal-temporal implant. EVERGREEN Joint J Novel Carbon Resour Sci Green Asia Strategy 08(04):750–758 15. Yusof MS, Aznan N, Manan NFA, Marwan SH, Mazlan MH, Abdullah AH (2021) Effects of varus and sagittal implant positioning to the stress adaptation in cementless hip arthroplasty. Malays J Med Health Sci 22–27 16. Dagang WNFSW, Hamdi NHQN, Marwan SH, Mahmud J, Manan NFA, Ramlee MH, Abdullah AH (2021) Effects of pentagonal pore sizes in the zinc hydroxyapatite parietal-temporal implant. Int J Emerg Technol Adv Eng 11:76–85 17. Ab Wahid AM, Tardan G, Pangesty AI, Rashid H, Abdullah AH (2022) Development of anklefoot orthosis with the integration of IoT controller. Int J Emerg Technol Adv Eng 12:49–55 18. Mazlan MA, Hashim NM, Che ZA, Abdullah A (2021) 3D printed assistive writing device for phocomelia patient. Malays J Med Health Sci 17:7–11 19. MatWeb Homepage. http://matweb.com/search/DataSheet.aspx?MatGUID=8d78f3cfcb6f49d 595896ce6ce6a2ef1. Last accessed 30 September 2021
Study of Primary Stability of Hip Implant for Semi Hip Replacement by Using Finite Element Analysis Haslina Abdullah , Mohamad Shukri Zakaria , and Norfazillah Talib
Abstract One factor contributing to the failure of hip arthroplasty or hip surgery is the loosening of the hip implant. Loosening of the hip implant is assessed by primary stability that is associated with the relative displacement occurring at the interface between the bone and the implant. The geometrical of hip implant significance influences the primary stability. Hence, this paper investigated the effect of the geometry of the implant to the primary stability. A three-dimensional of femur was constructed based on the computed tomography dataset acquired from a Malaysian patient. In contrast, the type of hip implant was produced based on the dimension of the bone. The finite element method was implemented to simulate the primary stability based on normal walking conditions. Then, the primary stability is defined based on the differences of displacement at the interface of the bone and implant interface. From the analysis, it was found that rectangular hip implants led to the better stability at the proximal area and the tips distal end of the implant. It can be concluded that the finite element method predicted the implant’s primary stability and enhanced the surgery’s performance. Keywords Hip implant · Stability · Finite element analysis
1 Introduction Hip surgery is a practice to substitute the injured bone in the hip joint with an implant known as hip stem. Through this surgery, it is expected to reduce the pain in a patient. To ensure the success and longevity of hip surgery, several factors need to H. Abdullah (B) · N. Talib Faculty of Mechanical Engineering and Manufacturing, Universiti Tun Hussein Onn Malaysia, 86400 Parit Raja, Johor, Malaysia e-mail: [email protected] M. S. Zakaria Fakulti Kejuruteraan Mekanikal, Universiti Teknikal Malaysia Melaka, Hang Tuah Jaya, Durian Tunggal, Melaka, Malaysia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_11
133
134
H. Abdullah et al.
be considered, for example, distribution of the femur bone and the primary stability of the hip implant. Therefore, studying stress distribution and stability of the hip implant is essential. The finite element technique has been utilised to simulate the distribution of stress induced by hip implants. By using this simulation, it will help to design a better hip implant. There were many studies that have been conducted to perform the simulation of stress distribution in a hip implant. For example, Sahai et al. [1] developed a hollow lightweight hip implant model with biocompatibility material, Ti-6Al-4V material. In this study, there were three models of hip implants: design without hole, design with 3-mm hole and design with 4.5-mm hole. The feature of a hole is expected to produce lightweight of hip implant design. It was found that a design with 3-mm hole produced better stress distribution than the design with 4.5mm hole, while Chethan et al. [2] performed simulation to analyse the consequences of implant geometry on the stress distribution. Several types of geometry have been developed: oval, circular, trapezoidal-ellipse and shaped stem designs. Based on the simulation, it was found that all the design had produced less stress than its yielded strength. The implant material plays an essential role in ensuring the longevity of the implant. In the previous study by Bhawe et al. [3], two various of materials were studied. For the first set of designs, UHMWPE for the acetabular cup, Ti-6Al-4V was nominated as the backing cup material, CoCr has been chosen for the femoral head and Ti-6Al-4V was used for hip implant. For the second type, CoCr has been used for the backing cup, CoCr for the femoral head, UHMWPE used for the acetabular cup and CoCr material was used for the stem of the implant. A simulation has been executed to define the stress distribution, and it was found that the best arrangement of material was Ti-6Al-4V for stem and a backing cup of CoCr and an acetabular cup of UHMWPE that produced a lower von Mises stress. Meanwhile, Faris et al. [4] studied the effect of titanium-niobium-hydroxyapatite (Ti-Nb-HA) material weightage difference on stress distribution. However, the design of hip implant was fixed. There were five set of materials, such as Ti-6Al-4V, CoCr, Ti0%NbHA, Ti10%NbHA, Ti20%NbHA, Ti30%NbHA and Ti40%NbHA. It was found that the highest equivalent von Mises stress and maximum contact pressure are produced by the implant with Ti30%NbHA material. Besides the geometry and materials, other types of activity also influenced the performance of the implant. Therefore, various researches have been performed to determine the effect of activity. A study by using finite element analysis (FEA) was conducted by Rosli et al. [5] to evaluate the different types of cycle on deformation of the implant based on different activities, such as slow walking, tripping, climbing and climbing down. Based on the simulation, it showed that tripping affects to produce larger stress and strain distribution in the hip joint, with the biggest total deformation occurring on the acetabular cup. At the same time, slow walking had the lowest parameters, while a study conducted by Putra et al. [6] made a comparison between three types of activities, such as normal walking, walking down of stairs and jumping. The study showed that the activity walking down the stairs produced higher principal stress. Similar results were also obtained based on the study by Annanto et al. [7], whereby jumping activity produced higher stress in the artificial hip implant. It can
Study of Primary Stability of Hip Implant for Semi Hip Replacement …
135
be concluded that activities with more movement influenced the stress and strain in hip implant. Besides the stress and strain distribution, many authors have agreed that primary stability is one of the circumstances that played a role to the continuing longevity [8, 9]. Camine et al. [10] constructed a collarless implant and collared type to study the effect on primary stability based on loading conditions, with good press-fit. It was observed that the effect of collar gave no significant difference in primary stability of the implants. Both designs produced good stability, with lower micromotion below the decided threshold, whereas the collar did not influence subsidence or micromotion. On the other hand, the alteration of the Zweymüller stem with smaller proximal design did not significantly affect the axial stability but affecting the rotational stability [11]. To achieve a better stability, the condition of the surface between bone and the implant is also considered as a factor that influenced the stability. In a research performed by Ismail et al. [12], the impact of the value of interference fit (δ) on the stability has been performed by defined coefficient of friction, such as 0.15, 0.40 and 1.00. It was discovered that the 0.50-mm interference fit produced a better primary fixation. In a research by Kanaizumi et al. [13], the stability was determined based on the value of micromotion. The implant design with a short stem, rectangular cross-section and with fins. The investigation showed that the rectangular stem and finned stem did not affect the primary stability. However, the highest micromotion had occurred at the proximal and tip of hip stem. Besides, Hosny et al. [14] and Lomami et al. [15] measured the stability of hip implants by using a different method known as EBRA-femoral component analysis (FCA) software, as a gauge of accomplishment of initial stability. Primary stability indicated the total of micromotion occurring at the interface between bone and the implant stimulated by the physical loading instead of the biological development. Though, secondary stability is the micromotion at the boneimplant surface as soon as the biological process is completed [16]. Many aspects affect the initial stability, for instance, geometrical aspects and properties of implant material, quality of the bone, and the types of patient’s activity. Several methodologies have been applied in assessing the stability of either by using a simulation or experimental work. These methods were essential to increase the fixation of hip implant and the effectiveness of the hip surgery. Therefore, this study’s motivation is to evaluate and verify the displacement at the interface of bone and implant for the cementless type of implant by considering the effect of the geometry of the implant by focusing on the impact of shoulder at the proximal region of the hip implant. Then, the difference in micromotion between cylindrical and tapered rectangular designs was studied. In this paper, an investigation on the effect of cylindrical, trapezoid and rectangular hip implant by using finite element analysis was performed. The main purpose is to estimate the displacement between the surface of bone and implant based on normal walking situations. A three-dimensional femur bone created from the CT dataset was acquired from a Malaysian patient. Then, the implant was produced to fit the size of the femur bone.
136
H. Abdullah et al.
(a)
(b)
(c)
Fig. 1 Process of development of bone: a CT dataset, b 3D model of bone and c mesh bone
2 Methodology 2.1 Development of a Three-Dimensional of Femur Bone In this study, the three-dimensional femur bone was created established from a computer tomography (CT) dataset acquired from a Malaysian patient. This is because the design of the implant should follow the size of the Malaysian patient. To construct the bone, the slice of the two-dimensional dataset was collected repeatedly to create the triangular surface of the femur bone by using AMIRA software. Figure 1 shows the finalised of the completed femur bone with a triangle meshing.
2.2 Development of a Three-Dimensional Hip Implant To ensure the design of the hip implant is fixed with the bone when inserted into the cortical area, the implant was constructed based on the dimension of the femur bone extracted from morphological data. The size and the dimension of the hip implant must be related to the geometric of the bone as demonstrated in Fig. 2a, b. Several geometric constraints must be measured to construct a good hip implant, such as femoral head diameter, angle of femoral neck shaft, length of femoral head offset and size of the isthmus. In this study, there were three types of the hip implant constructed, as shown in Fig. 3a–c. For the first design, the characteristics were cylindrical, straight, double tapered and collarless. Then, for the second design, the implant was designed with
Study of Primary Stability of Hip Implant for Semi Hip Replacement …
(a)
137
(b )
Fig. 2 Relationship between the size: a femoral bone, b implant
a larger cross section at the proximal section as illustrated in Fig. 3b, while for the third design, the cross section of the implant was altered to rectangular cross section with a larger area at the proximal area. The purpose of constructing these design was to examine the influence of cylindrical and trapezoid section on the primary stability. Both the 3D femur and the hip implant were saved as STL files to be imported into the finite element software for the simulation procedure. However, only a halfmodel femur bone was considered for the part of the bone. With half femur, it would reduce the time of the simulation. For the simulation contact analyses, inserting the implant into the canal space was assumed to be seamlessly fit. Table 1 illustrates the mechanical properties that implemented in the finite element simulation, which is adopted from Singh and Harsha [17]. In this analysis, the materials were presumed elastic, homogenous and isotropic. The initial or primary stability of the hip implant is evaluated by defining the differences value of displacement between the node of the bone and the implant. Therefore, contact analyses were implemented in the simulation. The interface between the bone and implant was assumed the frictional contact. In the contact analysis, the master surface needs to be defined for the implant and slave surface of the bone. Static loading was applied on the node on the femoral bone based on normal walking conditions and load value, as depicted in Table 2. The magnitude and position of force gained from [18] are shown in Fig. 4.
138
H. Abdullah et al. θ = 60°
θ = 60°
θ = 60°
48mm
48mm
48mm
20mm
20mm
20mm
120mm
120mm
120mm
(a)
(b)
(c)
Fig. 3 Construction of 3D model of hip implant: a cylindrical, b cylindrical with larger proximal and c rectangular tapered with larger proximal Table 1 Properties of material Materials
Elastic modulus (GPa)
Yield strength (MPa)
Poisson ratio (ν)
Cortical bone
17.26
115
0.29
Titanium alloy
110
485
0.3
Table 2 Loading condition of normal walking Force (N)
X
Y
Z
Point ()
Joint contact force
433.8
263.8
−1841.3
1
Abductor force
−465.9
−34.5
695
2
Tensor fascia lata, distal part
−4
−5.6
−152.6
2
Tensor fascia lata, proximal part
57.8
93.2
106
2
Vastus lateralis
7.2
−148.6
−746.3
3
Fixed
0
0
0
4
Study of Primary Stability of Hip Implant for Semi Hip Replacement …
139
Fig. 4 Position of forces 2 1
3
4
3 Results and Discussion The amount of relative displacement at the interface of bone and implant has been assessed in evaluating the primary stability on the effect geometrical of hip implant design. The displacement denotes to the primary stability of the implant attained by investigating the movement of nodes at the contact surface of implant and in x, y and z directions. Better stability is represented by a lesser relative displacement. The acceptable threshold value of relative displacement must be within 40–150 μm. In addition, the value of less than 40 μm will encourage osteointegration and enhance the bone growth rate to the implant surface. In contrast, the magnitude of relative displacement more than 150 μm indicated that the development of membrane tissue at the interface of bone and implant interface and led to the loosening of the implant. Figure 5 illustrates the graph of relative displacement for the first design occurring at the lateral and medial areas. The graph shows that the relative displacement on the lateral side was between 2 and 33 μm. Since the displacement was lower than 40 μm, there was no significant difference and will not cause any failure of the implant. On the other hand, the magnitude of relative displacement along the medial side was between 2 and 146 μm. As this magnitude was within the threshold value, it would not contribute to any failure. It can been seen that normal walking activity produced the highest relative displacement at the proximal area along medial and lateral areas. This result findings agreed with the previous research conducted by Kanizumi et al. [13] and Chanda et al. [19]. To examine the effect of proximal shoulder and rectangular tapered design on the stability of implant, contact node has been defined along the lateral side as shown
140
H. Abdullah et al.
Location of node on the prosthesis (mm)
140 A
120
Proximal C
100 80 60 40 Lateral 20
Medial D
0 0
30
60 90 120 Relative Displacement (μm)
150
B
Distal end
180
Fig. 5 Relative displacement for initial design along lateral and medial sides of initial design
in Fig. 6 and a graph of relative displacement was plotted as depicted in Fig. 7. It can be seen that the relative displacement for the initial design along the lateral side was lower than 40 μm. This magnitude was expected to enhance the bone growth. However, for the second and third designs, the values of relative displacement were higher than 40 μm. However, the value was within the area of critical threshold value. Although the magnitude did not exceed the maximum limit value, it was estimated to decrease the amount of bone growth to the implant surface. Similar to the initial design, the magnitudes of relative displacement for the second and third designs were larger on the proximal region due to bending loading.
Lateral side
(a)
(b)
(c)
Fig. 6 Selection of node on the lateral side: a cylindrical design (initial design), b cylindrical with larger shoulder (second design) and c rectangular tapered (third design)
Study of Primary Stability of Hip Implant for Semi Hip Replacement …
141
140 Location of node on the implant (mm)
Higher than 40μm at proximal region
120 100 Higher than 40 at middle region
80 60 40
Initial Design
20
Second Design Third Design
0 0
30
60
90
120
150
180
Relative Displacement (μm)
Fig. 7 Comparison of relative displacement for initial design along lateral sides
Generally, the value displacement for the second design along the lateral side was between 2 and 62 μm. This magnitude did not lead to loosening of prosthesis. Due to the larger proximal area at the shoulder, it decreased the relative displacement from 33 to 24 μm. This indicated that the larger surface area of prosthesis on the proximal region produced proximal fixation of prosthesis and increased the primary stability of prosthesis. For the third design, the magnitude of relative displacement along the lateral side was between 1 and 49 μm. Similar to the relative displacement of the second design, this magnitude did not lead to loosening of prosthesis. Overall, the values of relative displacement for all designs were sufficient to induce the bone growth to prosthesis surface. Figure 8 shows the selection of node on the medial side. A graph has been plotted in Fig. 9 along the medial side to investigate the stability on the medial side; it can be observed that the highest relative displacements for all designs of prosthesis along medial side also occurred at the proximal region. For the second design, the maximum value of relative displacement was 149 μm, and for the third design, the highest value of relative displacement was 102 μm. Table 3 presents the summary of relative displacement for initial, second and third designs on the lateral and medial sides.
142
H. Abdullah et al.
Medial side
Fig. 8 Selection of node on the lateral side: a cylindrical design (initial design), b cylindirical with larger shoulder (second design) and c rectangular tapered (third design)
Location of node on the implant (mm)
140 120 100 80 60 40 Initial Design Second Design Third Design
20 0 0
30
60
90
120
150
Relative Displacement (μm) Fig. 9 Comparison of relative displacement for initial design along medial sides
180
Study of Primary Stability of Hip Implant for Semi Hip Replacement …
143
Table 3 Summary of relative displacement on lateral and medial side Prosthesis
Lateral
Medial
Highest value (μm)
Lowest value (μm)
Highest value (μm)
Lowest value (μm)
Initial design
40
2
146
2
Second design
62
2
149
3
Third design
59
1
102
2
4 Conclusion To achieve the hip implant’s a better stability, the dimensions of the implant are referring to size of the bone from a patient. This study focused on the implant design based on the collarless and cementless types. The stability of the hip implant for hip surgery was evaluated by determining the relative displacement of node on the surface of the implant between the node on the bone surface by using the finite element analysis. The relative displacement was measured along the medial and lateral sides. Then, the value was compared with the threshold value, which is 40– 150 μm. For the first design, it was found that the relative displacement on the lateral side was between 5 and 3 μm. These amounts were within the threshold limit of between 40 and 150 μm. However, along the medial side, the relative displacement was between 5 and 146 μm. The maximum relative displacement was close to the maximum threshold value and was expected that it could lead to the loosening of the implant if a higher loading was applied. Therefore, several other geometries need to be created in order to produce a hip implant with a better stability. Acknowledgements This research was supported by University Tun Hussein Onn Malaysia (UTHM) through Tier 1 (vot Q384). The authors also would convey their appreciation to the University Teknologi Malaysia.
References 1. Sahai N, Saxena KK, Gupta N, Garg S, Bora U, Baishya UJ, Borah V (2021) Designing and simulation of a lightweight hip implant stem: an FEM based approach. Adv Mater Process Technol 00(00):1–9 2. Chethan KN, Shyamasunder Bhat N, Zuber M, Satish Shenoy B (2019) Finite element analysis of different hip implant designs along with femur under static loading conditions. J Biomed Phys Eng 9(5):507–516 3. Bhawe AK, Shah KM, Somani S, Shenoy BS, Bhat NS, Zuber M, Chethan KN (2022) Static structural analysis of the effect of change in femoral head sizes used in total hip arthroplasty using finite element method. Cogent Eng 9(1) 4. Faris M, Manap A, Shuib S, Ismail H, Zainudin M, Shaari NS (2022) Comparative analysis of TiNbHA with existing materials for hip implant using finite element method. In: Proceedings of mechanical engineering research day, pp 109–110
144
H. Abdullah et al.
5. Rosli INAM, Shuib S, Shokri AA, Ghani ARA, Hamizan NS, Arrif IM (2021) Development of hip implant: gait study and finite element analysis. In: 1st national biomedical engineering conference, NBEC 2021, pp 30–35 6. Putra AD, Andoko, Wulandari R, Kurniawan GA (2020) Simulation of hip joint implants using finite element method with time and load variations. Key Eng Mater 851:111–121 7. Annanto GP, Saputra E, Jamari J, Bayuseno AP, Ismail R, Tauviqirrahman M, Anwar IB (2018) Numerical analysis of stress distribution on artificial hip joint due to jump activity. E3S Web Conf 73:1–5 8. Panwar K, Cutter B, Holmboe M, Card R, Pistel W, Law JI (2016) Instability in total hip arthroplasty. INTECH 11:13 9. Fontalis A, Epinette JA, Thaler M, Zagra L, Khanduja V, Haddad FS (2021) Advances and innovations in total hip arthroplasty. Sicot-J 7 10. Malfroy Camine V, Rüdiger HA, Pioletti DP, Terrier A (2018) Effect of a collar on subsidence and local micromotion of cementless femoral stems: in vitro comparative study based on micro-computerised tomography. Int Orthop 42(1):49–57 11. Bieger R, Freitag T, Ignatius A, Reichel H, Dürselen L (2016) Primary stability of a shoulderless Zweymüller hip stem: a comparative in vitro micromotion study. J Orthop Surg Res 11(1):10–15 12. Ismail NF, Shuib S, Yahaya MA, Romli AZ, Shokri AA (2018) Finite element analysis of uncemented total hip replacement: the effect of bone-implant interface. Int J Eng Technol (UAE) 7(4):230–234 13. Kanaizumi A, Suzuki D, Nagoya S, Teramoto A, Yamashita T (2022) Patient-specific threedimensional evaluation of interface micromotion in two different short stem designs in cementless total hip arthroplasty: a finite element analysis. J Orthop Surg Res 17(1):1–9 14. Hosny HAH, Srinivasan SCM, Hall MJ, Keenan J, Fekry H (2017) Achievement of primary stability using 3D-CT guided custom design femoral stems in patients with proximal femoral deformity: EBRA-FCA analysis. Acta Orthop Belg 83(4):617–623 15. Albini Lomami H, Damour C, Rosi G, Poudrel AS, Dubory A, Flouzat-Lachaniette CH, Haiat G (2020) Ex vivo estimation of cementless femoral stem stability using an instrumented hammer. Clin Biomech 76(April):105006 16. Vollmer A, Saravi B, Lang G, Adolphs N, Hazard D, Giers V, Stoll P (2020) Factors influencing primary and secondary implant stability—a retrospective cohort study with 582 implants in 272 patients. Appl Sci (Switzerland) 10(22):1–14 17. Singh S, Harsha AP (2016) Analysis of femoral components of cemented total hip arthroplasty. J Inst Eng (India) Ser D 97(2):113–120 18. Rafiq M, Kadir A, Hansen UN (2007) The effect of physiological load configuration on interface micromotion in cementless femoral stems. Jurnal Mekanikal 23:50–61 19. Chanda S, Mukherjee K, Gupta S, Pratihar DK (2020) A comparative assessment of two designs of hip stem using rule-based simulation of combined osseointegration and remodelling. Proc Inst Mech Eng, Part H J Eng Med 234(1):118–128
Investigation of Mental Health Condition Among Factory Worker During Covid pandemic–A Cross-Sectional Study Irna Syahira Hassan, Nur Fazhilah Abdul Razak, Junaidah Zakaria , and Ezrin Hani Sukadarin
Abstract The coronavirus disease 2019 (COVID-19) was initially reported in December 2019 in Wuhan, China, after a cluster of unusual pneumonia cases. On March 11, 2020, the World Health Organization (WHO) declared the outbreak a pandemic. Many workplaces are affected by work-related psychosocial risks and stress, as well as the adverse health and economic implications. Workers have been challenged, stretched, and tested in ways they have never been before, as have pharmaceutical manufacturing workers’ mental health, as they are responsible for providing optimal medication manufacturing that aids in therapy, treatment, and patient life during COVID-19. This study is to assess the level of depression, anxiety, stress, and associated factors among factory workers during COVID-19. A crosssectional study was conducted using online surveys to assess workers’ mental health and related factors. The study involves 201 manufacturing workers from various departments. The data were collected using a questionnaire distributed via an online platform. The data collection instrument consists of three parts: (1) sociodemographic, (2) related factors associated with mental health conditions, and (3) the DASS-21 survey. This study found that most workers have normal mental health conditions. Depression and anxiety have a significant correlation with sociodemographic characteristics and organizational factors. Meanwhile, stress has a significant correlation with socioeconomic and organizational aspects. All three mental health conditions have a significant correlation with organizational factors such as working experiences (>6 years) and preparation for workflow management during COVID19, with p-values of 0.028 and 0.023, respectively. The study’s findings may assist authorities in establishing ways to diagnose mental distress early, thereby reducing mental or psychological disease among factory workers. I. S. Hassan · N. F. A. Razak (B) Faculty of Industrial Sciences and Technology, University Malaysia Pahang Al-Sultan Abdullah, Pekan, Malaysia e-mail: [email protected] J. Zakaria · E. H. Sukadarin Department of Chemical Engineering Technology, Faculty of Engineering Technology, Bachelor of Technology in Occupational Safety & Health (with Hons.), Universiti Tun Hussein Onn Malaysia (UTHM), Johor, Malaysia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_12
145
146
I. S. Hassan et al.
Keywords COVID-19 pandemic · Mental health · DASS-21 · Psychosocial risk factors
1 Introduction Throughout history, there have been several battles against epidemic diseases. Countries have faced complex social and economic challenges during cholera, plague, malaria, and tuberculosis outbreaks. However, following the discoveries of SARS, MERS, H1N1, and EBOLA at the beginning of the twenty-first century, a new type of coronavirus emerged in Wuhan, the capital city of Hubei Province in China [1]. COVID-19, caused by an unknown severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), began in the Chinese city of Wuhan in December 2020 [2]. Since then, the virus has rapidly spread throughout China and the rest of the world. As a result, on March 11, 2020, the World Health Organization (WHO) declared COVID-19 a global pandemic. COVID-19 was first transmitted to the local population in Malaysia by travellers between January and February 2020, resulting in widespread infection in early March when the Tablighi Jamaat cluster of conditions emerged. The government enforced a movement control order (MCO) from March 18, 2020, to June 9, 2020, prohibiting citizens from organizing social activities or gatherings, including cultural, religious, athletic, work-related, and educational activities. Psychological discomfort in the population has been observed as a result of protracted isolation during movement lockdowns during outbreak epidemics; this appears as a variety of symptoms such as poor mood, insomnia, tension, anxiety, despair, frustration, irritability, and emotional tiredness [3]. The pandemic has a profound effect on the world’s economies in several ways, including company closures, an increase in unemployment, a decrease in exports, a reduction in oil prices, an increase in hunger, an increase in the global death rate, and population growth [4]. The deadly virus appears to significantly impact people, causing terror, anger, tension, and anxiety. Previous research has shown that crises significantly impact individuals’ work and psychological well-being. A crisis is a stressful and upsetting event in a person’s life [5]. According to a previous study, the major job stressors include a severe workload with unrealistic deadlines, a work-family imbalance, and job uncertainty [6]. Depression, anxiety, and stress may make people more susceptible and vulnerable to the COVID-19 virus infection [6]. Workplace mental health disorders may be caused by excessively tight deadlines, repetitive work, an insufficient work climate, and dissatisfaction with peers and immediate superiors. Workers struggled with the strains of the COVID-19 pandemic, the emotional challenges of social isolation, and achieving work–life balance while working remotely. The same situation is not evocative or frustrating for everyone, and everyone does not experience the same negative thoughts and emotions when they are depressed or stressed. It must be tackled as a matter of severe occupational health. Therefore, any mental health
Investigation of Mental Health Condition Among Factory Worker …
147
disorders such as depression, anxiety, and stress should be handled cautiously by the individual and their organizations, which should provide proper and practical support. Modern life, particularly during the COVID-19 outbreak, is full of mental health problems such as depression, anxiety, and stress. Even though depression is the leading cause of disability worldwide, it frequently goes undiagnosed and untreated. Meanwhile, anxiety is regarded as one of the three most harmful emotional factors, causing numerous incurable problems and disorders in a person’s life [7]. Mental health issues are among the most expensive burdens that developing-world organizations and governments face. Psychological distress, identified as psychological and physical symptoms associated with an emotional state of distress, is a growing public health problem in Malaysia, with comparable social and economic effects and consequences [8–10]. In addition, the correlation between mental health and factory workers has been identified by several similar studies. They found that mental health problems were significantly higher with high job demands [11–13]. According to a Malaysian survey, 70% of respondents registered elevated anxiety levels during the early stages of the COVID-19 outbreak. Psychological distress is often associated with poor physical health and increased healthcare use, which harms employees and employers through reduced work participation, increased sick leave, and higher absenteeism and presentism [14]. The lack of effective treatment leaves people alone with these issues, leading to a downward spiral of despair [15]. The COVID-19 pandemic has intensified the global mental health crisis, with nearly onethird of Asia Pacific’s remote workers admitting that the pandemic has worsened workplace burnout. Since 2020, many studies have analysed the psychological impact of the COVID19 pandemic on various populations, for example, studies among medical workers [16, 17], Chinese residents [18], older adults, children, and adolescents, and college students [19–21], educators such as teachers and lecturers [22]; however, so far there is still inadequate research on pharmaceutical manufacturing workers during the COVID-19 pandemic. The main objective is to assess depression, anxiety, and stress among factory workers from different departments working during the COVID-19 pandemic. The association between sociodemographic characteristics, socioeconomic, organizational factors, work environmental factors, and special health conditions with depression, anxiety, and stress among factory workers working during the COVID-19 pandemic will also be explored and discussed in this paper.
2 Methodology A cross-sectional study is conducted through an online survey to determine the worker’s depression level, anxiety, and stress. A walkthrough observation is also performed to evaluate the significant factors of mental health disorders. The inclusion criteria were both male and female respondents work in different departments at the factory during the COVID-19 pandemic with provided informed consent to
148
I. S. Hassan et al.
participate in the study. Meanwhile, respondents who are severely ill and not in a condition to answer the questions were excluded from the study.
2.1 Participants Pharmaceutical manufacturing is the selected company based in Sungai Petani, Kedah, Malaysia, ultimately controlled by the Malaysian government. The population’s sample size is determined using the purposive sampling technique. This study focused on male and female workers in the factory. About 201 workers were involved in this study.
2.2 Instrumentation The respondents self-administer the structured self-report questionnaire through an online survey that consists of three parts. Part I–Information related to sociodemographic characteristics of manufacturing workers. Part II–Questionnaire related to factors associated with depression, anxiety, and stress such as sociodemographic characteristics, socioeconomic status, organizational factors, environmental factors, and particular health conditions. Part III–Based on the depression, anxiety, and stress scale–21 (DASS-21) [23]. Each of the three DASS-21 scales have seven items broken down into subscales of the same material. Since it is freely available on the DASS-21 official website, the translation has been widely used in Malaysia [24]. In addition, the DASS-21 is a validated tool used in different Malaysian people to identify depression, anxiety, and stress symptoms.
2.3 Data Analysis Statistical Package for Social Sciences (SPSS) software version 2.0 was used. Descriptive analysis is calculated, including the frequency, percentage mean, and standard deviation. Pearson’s chi-square test and multivariate analysis are used in this study. In this study, the Pearson’s chi-square test and multivariate analysis will be used. The significance level will be set at 0.05 for the analyses. This test was selected because the analysed data were categorical and involved the associations for two variables. Data normality test was conducted by using the Kolmogorov–Smirnov test with p > 0.05 taken as the normal distribution.
Investigation of Mental Health Condition Among Factory Worker …
149
3 Result 3.1 Demographic Analysis Based on Table 1, most respondents are between 31 and 40 years old, which is 103 (51.2%). Respondents aged between 21 and 30, and 41 and 50 years old are 39 (19.4%) and 30 (14.9%), respectively, while the rest, only 29 (14.4%) people, are age 50 years old. Most of the respondents are female, that is 119 (59.2%) people, while the rest, only 82 (40.8%) respondents, are male. Most of the respondents’ religion is Islam, which is 180 (89.6%), followed by the Hindu, which is 19 (9.5%), while the rest, only 1 (0.5%), belong to others. 79.1% are married, 15.9% are unmarried, and 5.0% of the respondents are divorced/separated. The type of family, socioeconomic factors, and other essential elements that are required in this study are also provided in the table.
3.2 Depression, Anxiety, and Stress Figure 1 shows the descriptive finding of depression, anxiety, and stress level among workers. Based on Fig. 1, 10.95% of respondents had mild levels of depression, 6.47% had a moderate level of depression while only 1.00% and 0.50% of the respondents had a severe and extremely severe levels of depression, respectively. For the anxiety levels among respondents where the majority of the respondents, i.e. 78.11%, had a normal level of anxiety, 5.97% had mild level of anxiety, and 10.45% had a moderate level of anxiety whereas 2.99% and 2.49% of the respondents had a severe and extremely severe level of anxiety respectively. On the stress scale findings of DASS21, the majority of the respondents (92.54%) had a normal level of stress, 4.48% of the respondents had mild level of stress, whereas the moderate and severe levels of stress both had 1.49% and there were no respondents that had an extremely severe level of stress during COVID-19 pandemic.
3.3 Depression, Anxiety, Stress and Demographic Factors The multivariate analysis for the depression, anxiety, and stress toward demographic factors in Table 2 revealed that females are almost twofold more likely to have anxiety compared to males (AOR (adjusted odd ratio) = 1.944, CI (confidence interval) = 0.795–4.753). Respondents aged between 31 and 40 years old were significantly threefold more likely to have depression (AOR = 3.463, CI = 1.132– 10.596) compared to those with the chance of developing anxiety and stress. Surprisingly, married people were found to have a higher chance of developing depression significantly (AOR = 5.942, CI = 2.008–7.588) compared to anxiety and stress.
150
I. S. Hassan et al.
Table 1 Demographic analysis of the study respondents Variables
Category
Number
Percentage (%)
Age
20–30 years’ old 31–40 years’ old 41–50 years’ old > 50 years old
39 103 30 29
19.4 51.2 14.9 14.4
Gender
Female Male
119 82
59.2 40.8
Religion
Islam Christianity Hindu Buddhism
180 1 19 1
89.6 0.5 9.5 0.5
Marital status
Unmarried Married Divorced/separated
32 159 10
15.9 79.1 5.0
Race group
Malay Indian Chinese
179 21 1
89.1 10.4 0.5
Family type
Nuclear Extended Joint Blended
173 6 10 12
86.1 3.0 5.0 6.0
Socio-economic Status
Bottom 40% (B40) Middle 40% (M40) Top 20% (T20)
147 49 5
73.1 24.4 2.5
Total family income
RM10000
147 40 9 5
73.1 19.9 4.5 2.5
Training/orientation of COVID-19
Yes No
22 179
10.9 89.1
Working hours per day
6–10 h 11–15 h >15 h
196 5 0
97.5 2.5 0
Current work condition
Work from home Day shift work (morning-evening) Swing shift work (afternoon-midnight) Night shift work (midnight-morning)
15 179 4 3
7.5 89.1 2.0 1.5
Working experience (years)
6 years
3 28 15 155
1.5 13.9 7.5 77.1
Investigation of Mental Health Condition Among Factory Worker …
151
Fig. 1 Depression, anxiety, and stress severity level of the respondents (N = 201)
3.4 Depression, Anxiety, Stress and Socioeconomic Factors Table 3 shows the factor analysis for socioeconomic status; those who live at Semeling were found to have threefold higher and more likely to have stress symptoms (AOR = 3.366, CI = 1.366–5.904) compared to those who live in Gurun (AOR = 2.899, CI = 0.899–2.899) and Sungai Petani (AOR = 1.859, CI = 0.270–2.307); however, it is not significant. Compared to other levels of education like degree, SPM (Malaysian secondary school certificate), and SKM (Malaysian technical certificate), those with Diploma were found to have a higher chance of developing Stress (AOR = 6.090, CI = 1.926–4.680) significantly. But, for the family income, no significant result was found to relate to the perceived depression, anxiety or stress among workers.
4.570
0.126
0.001*
>50 years old
Married (ref: unmarried)
Note Odd Ratio [95% Confidence Intervals] Significant if p < 0.05
5.942
1.559
41–50 years old 0.539
Marital status
3.463
31–40 years old 0.029*
0.815
Age (ref: 20–30 years old)
0.582
[2.008,7.588]
[0.653,1.969]
[0.379,6.420]
[1.132,10.596]
[0.393,1.689]
0.719
0.204
0.390
0.735
0.145
p-value
Female (ref: male)
Anxiety 95% CI
p-value
Odd ratio
Depression
Gender
Variables
Table 2 Depression, anxiety, stress and demographic
1.231
4.290
1.999
0.828
1.944
Odd ratio
[0.397,3.813]
[0.454,4.536]
[0.412,9.691]
[0.276,2.480]
[0.795,4.753]
95% CI
0.314
0.158
0.690
0.649
0.668
p-value
Stress
0.458
0.121
0.636
0.692
1.345
Odd ratio
[0.100,2.091]
[0.006,2.268]
[0.069,5.888]
[0.143,3.363]
[0.347,5.210]
95% CI
152 I. S. Hassan et al.
Investigation of Mental Health Condition Among Factory Worker …
153
Table 3 Depression, anxiety, stress and socioeconomic factors Variables
Current residence Sungai Petani (ref: others)
Level education (ref: others)
Family income (ref: > RM10000)
Depression
Anxiety
Stress
p-value Odd ratio[95% CI]
Sig
p-value Odd ratio[95% CI]
Odd ratio [95% CI]
0.850
1.260 0.868 0.816 0.994 [0.114,3.890] [0.075, 8.920]
1.859 [0.270, 2.307]
Gurun
0.694
0.477 0.694 0.477 0.630 [0.012,1.883] [0.012, 1.883]
2.899 [0.899, 2.899]
Semeling
0.315
0.177 0.995 1.191 1.000 [0.006,5.176] [0.000, 1.294]
3.366 [1.366, 5.904]
SPM
0.938
1.095 0.611 0.557 0.000* [0.111,1.808] [0.058, 5.334]
2.351 [1.051, 4.094]
Diploma
0.912
1.156 0.809 0.732 0.000* [0.088,5.253] [0.058, 9.160]
6.090 [1.926, 4.680]
Degree
0.502
0.450 0.550 0.491 0.336 [0.044,4.617] [0.047, 5.073]
1.056 [3.973, 6.391]
6 years
6–10 h (ref: 11–15 h)
0.248 0.985
0.943
0.272
0.099
0.956
0.264
1.000
0.000*
0.000*
0.000*
0.952
0.949
0.530
p-value
1–3 years
Anxiety
p-value
Odd ratio [95% CI]
Depression
4–6 years
Working hours/day
Working experiences (ref: X > 0.2)
Skewness (X < 0.85)
Aspect ratio (X < 5.0)
0.50
1,107,096
258,085
0.76236
0.83635
1.9684
0.35
1,421,267
322,065
0.76068
0.84902
1.9506
0.25
2,178,835
473,489
0.75979
0.81491
1.9251
0.20
3,263,762
688,495
0.75863
0.80627
1.9106
Table 4 Mesh quality analysis for stenosed model between different sizes of elements Element size (mm)
Number of elements
Number of nodes
Mesh quality Orthogonal quality (1.0 > X > 0.2)
Skewness (X < 0.85)
Aspect ratio (X < 5.0)
0.50
1,119,573
260,487
0.76197
0.8414
1.9683
0.35
1,439,808
325,413
0.76076
0.84844
1.9495
0.25
2,198,826
477,112
0.75983
0.86395
1.9243
0.20
3,258,710
687,571
0.75889
0.82018
1.9098
3.2 Prediction of Potential Region for Atherosclerosis to Occur with Laminar Flow Condition Healthy peripheral artery model under straight leg condition (180°) was analyzed by using laminar flow conditions to predict high-risk regions to develop atherosclerosis. In the sub-section below, detailed findings on the velocity and WSS profile are observed to investigate flow behavior and potential region to occur atherosclerosis.
3.2.1
Velocity Profile of Healthy Peripheral Artery
Figure 4 illustrates the velocity distribution in healthy peripheral arteries with laminar flow conditions. The inlet velocity of 0.467 m/s was defined at the start of the femoral artery. The flow velocity increases along the femoral artery towards the branching of SFA and Profunda. Due to the difference in artery diameter (larger to smaller) at the profunda and femoral artery, the velocity distribution was the highest at the initial point of branching while the velocity resumed being lower at the SFA. The increase in velocity was found at the popliteal artery because the popliteal artery gets narrower towards the downstream and the structure is curved in nature. In addition, a similar occurrence as branching (Profunda and SFA) was observed at the branching of the Posterior and Anterior Tibial Artery as illustrated in Fig. 5.
230
U. Z. Shahrulakmar et al.
Fig. 4 Velocity contour of a healthy peripheral artery
As overall, the velocity increases as it moves along the downstream of the artery due to the diameter variation throughout the peripheral arterial structure. The highest velocity recorded was 1.706 m/s at the branching of the profunda. This happens due to the presence of bifurcation and significant differences in the diameter of SFA and Profunda. In addition, the velocity profile also can be influenced by the angle of the bifurcation.
3.2.2
Wall Shear Stress (WSS) Profile of Healthy Peripheral Artery
Wall shear stress is another dominating parameter that influences blood flow behavior in the artery. It is known that wall shear stress fluctuates throughout the flow cycle and results in arterial wall thickening. Figure 6 shows the WSS contour for a healthy peripheral artery. High WSS elongates the endothelial cells and forces them to align in the direction of the flow. Low WSS had a negligible effect on the cell but increases intercellular permeability and consequently increases plaque formation in the region.
Prediction of Atherosclerosis in Peripheral Arterial Disease Using …
231
Fig. 5 Detail velocity contours of a healthy peripheral artery at the branching of profunda and SFA. The highest velocity is recorded in the red circle Fig. 6 WSS contour of a healthy peripheral artery
232
U. Z. Shahrulakmar et al.
A few critical locations such as the branching of profunda and SFA, popliteal artery, and branching of posterior and anterior tibial artery exhibit low wall shear stress conditions. A larger region of low wall shear stress is found at the beginning of SFA after the branching. This may occur due to abrupt velocity change from upstream to downstream due to reducing the size of the artery. Consequently, this region has a higher tendency for atherosclerosis to occur where similar findings were reported by Gogineni and Ravigururajan [5].
3.3 Comparison of High Potential Region to Develop Atherosclerosis with Laminar Flow Condition The prediction of the critical region to develop stenosis in the peripheral artery was identified through the findings from velocity and WSS profile of healthy peripheral artery. The upstream of SFA illustrates low wall shear stress and subsequently has a high potential for atherosclerosis to occur. Hence, a stenosed model was constructed by using trapezoidal stenosis at the low-stress region in SFA. The outcome of the stenosed model is described in the below section accordingly.
3.3.1
Velocity Profile of Stenosed Peripheral Artery
Figure 7a, b demonstrate the comparison in flow specifically at the branching of profunda and SFA. There are significant changes in the velocity distribution with the presence of stenosis at SFA. When there is the presence of stenosis at the branching of the profunda and SFA, the velocity at the profunda increases and lowers at SFA compared to the laminar flow in a healthy artery. The maximum velocity (A1) in a healthy artery was recorded 1.038 m/s while the maximum velocity (A2) for a stenosed artery is 2.132 m/s. This happens when the size of the artery is smaller, the pressure reduces, and this forces a high-speed flow and sudden increase in the artery after the stenosed location will increase the pressure and subsequently reduce the flow velocity of the blood abruptly. Thus, lower velocity flow also will induce for development of stenosis at the downstream region.
3.3.2
Wall Shear Stress (WSS) Profile of Stenosed Peripheral Artery
Figure 8 describes the influence of stenosis at the predicted location on WSS findings. It is seen that there are significant changes in WSS at the downstream region after the affected area. This is because WSS is closely related to the velocity and viscosity of the blood in the artery. When there is a decrease in the velocity flow after constricted location, the WSS at the respected location also lowers. Correlatedly, this
Prediction of Atherosclerosis in Peripheral Arterial Disease Using …
233
Fig. 7 Flow comparison at branching of SFA and profunda: a healthy condition and b stenosed condition
will enhance the development of stenosis at the lower region. As shown in Fig. 9, at the branching of anterior and posterior tibial artery as well as anterior tibial artery, the minimum WSS of a stenosed model (B2) is 0.641 Pa and the minimum WSS of a healthy model (B1) is 1.194 Pa where stenosed model obtained lower WSS value compared to a healthy artery. This proves that stenosis at one location can worsen the condition and impact lesions, especially at the downstream region. Besides that, stenosis also may induce transitional flow which can risk plaque to rupture. This can occur when impacted or weak plaque could not withstand the flow force and tend to break. This gets riskier to the patient where ruptured plaque will flow towards the narrow artery at the downstream region. When there is no sufficient space to go through, this plaque may fully block the blood flow at downstream. Besides, ruptured plaque also will flow to the downstream region (narrow artery) and may fully block the flow at the lower region.
234 Fig. 8 WSS contour of a stenosed peripheral artery
U. Z. Shahrulakmar et al.
Prediction of Atherosclerosis in Peripheral Arterial Disease Using …
235
Fig. 9 Comparison of WSS contours of a stenosed peripheral artery at ATA: a healthy condition and b stenosed condition
4 Conclusion Using computational fluid dynamics (CFD) for predicting atherosclerosis is a relatively new approach to predicting and assessing the progression of this disease. This technique has the potential to provide valuable insight into the disease process, as well as aid in the development of new treatments. Compared to traditional methods, CFD offers greater accuracy and flexibility to capture the complex flow patterns in peripheral arteries. Furthermore, CFD simulations can be used to provide a better understanding of the underlying fluid mechanics and the role of fluid dynamics in the disease process. With further development and refinement, CFD can become a valuable tool for predicting and managing atherosclerosis.
References 1. Corti A, Chiastra C, Colombo M, Garbey M, Migliavacca F, Casarin S (2020) A fully coupled computational fluid dynamics—agent-based model of atherosclerotic plaque development: Multiscale modeling framework and parameter sensitivity analysis. Comput Biol Med 118(January):103623. https://doi.org/10.1016/j.compbiomed.2020.103623 2. Colombo M et al (2020) Computing patient-specific hemodynamics in stented femoral artery models obtained from computed tomography using a validated 3D reconstruction method. Med Eng Phys 75:23–35. https://doi.org/10.1016/j.medengphy.2019.10.005
236
U. Z. Shahrulakmar et al.
3. Fowkes FGR, Aboyans V, Fowkes FJI, McDermott MM, Sampson UKA, Criqui MH (2017) Peripheral artery disease: epidemiology and global perspectives. Nat Rev Cardiol 14(3):156– 170. https://doi.org/10.1038/nrcardio.2016.179 4. Morley RL, Sharma A, Horsch AD, Hinchliffe RJ (2018) Peripheral artery disease. BMJ 360(February):1–8. https://doi.org/10.1136/bmj.j5842 5. Gogineni A, Ravigururajan TS (2017) Flow dynamics and wall shear stresses in a bifurcated femoral artery. J Biomed Eng Med Devices 02(02). https://doi.org/10.4172/2475-7586.100 0130 6. Sun N, Torii R, Wood NB, Hughes AD, Thom SAM, Xu XY (Feb 2009) Computational modeling of LDL and albumin transport in an in vivo CT image-based human right coronary artery. J Biomech Eng 131(2). https://doi.org/10.1115/1.3005161 7. Hewlin RL, Kizito JP (2018) Development of an experimental and digital cardiovascular arterial model for transient hemodynamic and postural change studies: ‘A Preliminary Framework Analysis,’. Cardiovasc Eng Technol 9(1). https://doi.org/10.1007/s13239-017-0332-z 8. Johari NH et al (2019) Disturbed flow in a stenosed carotid artery bifurcation: comparison of RANS-based transitional model and LES with experimental measurements. Int J Appl Mech 11(4):1–21. https://doi.org/10.1142/S1758825119500327 9. Miller AJ, Takahashi EA, Harmsen WS, Mara KC, Misra S (2017) Treatment of superficial femoral artery restenosis. J Vasc Interv Radiol 28(12):1681–1686. https://doi.org/10.1016/j. jvir.2017.07.032 10. Kawarada O et al (2020) Peak systolic velocity ratio derived from quantitative vessel analysis for restenosis after femoropopliteal intervention: a multidisciplinary review from Endovascular Asia. Cardiovasc Interv Ther 35(1):52–61. https://doi.org/10.1007/s12928-019-00602-z 11. Razavi MK, Flanigan DPT, White SM, Rice TB (2021) A real-time blood flow measurement device for patients with peripheral artery disease. J Vasc Interv Radiol 32(3):453–458. https:// doi.org/10.1016/j.jvir.2020.09.006 12. Salman HE, Yazicioglu Y (2019) Computational analysis for non-invasive detection of stenosis in peripheral arteries. Med Eng Phys 70:39–50. https://doi.org/10.1016/j.medengphy.2019. 06.007 13. Avolio AP (1980) Multi-branched model of the human arterial system. Med Biol Eng Comput 18(6):709–718. https://doi.org/10.1007/BF02441895 14. Li X et al (2016) Why is ABI effective in detecting vascular stenosis? Investigation based on multibranch hemodynamic model. Sci World J 2013(March):2013. https://doi.org/10.1155/ 2013/185691 15. Wolf YG, Kobzantsev Z, Zelmanovich L (2006) Size of normal and aneurysmal popliteal arteries: a duplex ultrasound study. J Vasc Surg 43(3):488–492. https://doi.org/10.1016/j.jvs. 2005.11.026 16. Sandgren T, Sonesson B, Ahlgren AR, Lanne T (1999) The diameter of the common femoral artery in healthy human: influence of sex, age, and body size. J Vasc Surg 29(3):503–510. https://doi.org/10.1016/S0741-5214(99)70279-X 17. Shahrulakmar UZ, Omar MN, Johari NH (2021) Brief review on recent advancement of computational analysis on hemodynamics in peripheral artery disease 18. Berglund H et al (1997) Highly localized arterial remodeling in patients with coronary atherosclerosis: an intravascular ultrasound study. Circulation 96(5):1470–1476. https://doi. org/10.1161/01.CIR.96.5.1470 19. Lorenzini G, Casalena E (2008) CFD analysis of pulsatile blood flow in an atherosclerotic human artery with eccentric plaques. J Biomech 41(9):1862–1870. https://doi.org/10.1016/j. jbiomech.2008.04.009 20. Moser KW, Kutter EC, Georgiadis JG, Buckius RO, Morris HD, Torczynski JR (2000) Velocity measurements of flow through a step stenosis using Magnetic Resonance Imaging. Exp Fluids 29(5):438–447. https://doi.org/10.1007/s003480000110 21. Tang D, Yang C, Kobayashi S, Zheng J, Vito RP (2003) Effect of stenosis asymmetry on blood flow and artery compression: a three-dimensional fluid-structure interaction model. Ann Biomed Eng 31(10):1182–1193. https://doi.org/10.1114/1.1615577
Prediction of Atherosclerosis in Peripheral Arterial Disease Using …
237
22. Crawford DW (2017) Wall Shear Stress Estimates in 114(November 1992):515–520 23. Wilson RF et al (1988) Pathophysiology and natural history the effect of coronary angioplasty on coronary 77(4):873–886 24. Sinha Roy A, Back LH, Banerjee RK (2006) Guidewire flow obstruction effect on pressure drop-flow relationship in moderate coronary artery stenosis. J Biomech 39(5):853–864. https:// doi.org/10.1016/j.jbiomech.2005.01.020 25. Kamangar S et al (2017) The influence of geometrical shapes of stenosis on the blood flow in stenosed artery. Sains Malays 46(10):1923–1933. https://doi.org/10.17576/jsm-2017-4610-31 26. Totorean AF et al (2022) Patient-specific image-based computational fluid dynamics analysis of abdominal aorta and branches. J Pers Med 12(9). https://doi.org/10.3390/jpm12091502 27. Liu H et al (2021) Comparison of Newtonian and non-Newtonian fluid models in blood flow simulation in patients with intracranial arterial stenosis. Front Physiol 12(September):1–11. https://doi.org/10.3389/fphys.2021.718540 28. Kadhim SK, Al-Azawy MG, Ali SAG, Kadhim MQ (2021) The influence of non-Newtonian model on properties of blood flow through a left coronary artery with presence of different double stenosis. Int J Heat Technol 39(3):895–905. https://doi.org/10.18280/ijht.390324 29. Quemada D (1978) Rheology of concentrated disperse systems II. A model for non-Newtonian shear viscosity in steady flows. Rheol Acta 17(6):632–642. https://doi.org/10.1007/BF0152 2036 30. Marcinkowska-Gapi´nska A, Gapinski J, Elikowski W, Jaroszyk F, Kubisz L (2007) Comparison of three rheological models of shear flow behavior studied on blood samples from postinfarction patients. Med Biol Eng Comput 45(9):837–844. https://doi.org/10.1007/s11517-0070236-4
Impact Analysis of Motorcycle Helmet: Finite Element Modeling N. Aimi Huda and M. S. Salwani
Abstract The motorcycle helmet is necessary for riders to wear it while riding a motorcycle. The principal reason for a motorcycle helmet is to soak up the energy of the collision and minimizes the risk of head injury to the rider’s head in the event of an impact and improves safety. This research aims to model different types of face helmets and analyze the impact performance of motorcycle safety helmets. The three most significant designs that are widely used are full-face, open-face and half-face helmets. The designs will then be simulated in Inventor Nastran. The FEA will be conducted to investigate the maximum von Misses stress and total deformation of the helmet. It can be concluded that half-face design helmets yield the highest von Misses stress and total deformation value of 5.7 x 105 MPa and 3.9 mm, respectively. The results provide a potential solution for ensuring better motorcycle helmet for protection. Keywords Motorcycle helmet · Finite Element Analysis · Impact
1 Introduction The use of helmets reduces the risk of mild to severe brain injuries among crash injury sufferers [1]. A helmet is a type of head protection that is worn on the head. A helmet, in particular, works in conjunction with the skull to protect the human brain. In the event of an accident, wearing a bicycle helmet can significantly reduce injuries. Helmets reduce severe lacerations or fractures of the midface, nose, and eyes, as well as brain damage. When compared to non-helmet users, wearing a helmet was related to a lower hospital or ICU days, and these individuals were less likely to die [2] Another important function of motorcycle helmets is to prevent brain injuries, which can be quite serious and result in permanent impairment or even death. As a result, a N. Aimi Huda · M. S. Salwani (B) Faculty of Mechanical and Automotive Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, 26600 Pekan, Pahang, Malaysia e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_19
239
240
N. Aimi Huda and M. S. Salwani
protective helmet reduces the amount of impact energy reaching the head, decreasing the severity or chance of injury. Helmets keep the head comfortable by reducing wind noise and functioning as a shield from wind blasts, bad weather, and any form of item, in addition to protecting the head in motorcycle accidents. They cushion the head and absorb the impact in the event of an accident, allowing the impact to last longer [3]. A modern motorcycle helmet has six fundamental components: a very thin and tough outer shell, a thick and soft impact-absorbing inner liner, comfort padding, a retention system, a visor, and a ventilation system [4, 5] From my literature review, the hard outer shell is typically made of thermoplastic materials like polycarbonate (PC) or acrylonitrile–butadiene–styrene (ABS), or even composite materials like fibre reinforced plastics (FRP) like glass reinforced plastic (GRP) or carbon reinforced plastic (CRP), or even just carbon fibre or Kevlar® [6]. Impact analysis is carried out in the composite helmet using FRC and glass fibre in a study. Fibre reinforced composite materials, which are made with glass fibre as reinforcements and a matrix, have recently gained the interest of researchers due to their low density, high specific mechanical strengths, ease of use and renewability. The objectives for this study are to model full-face, open-face and half-face helmet subjected to impact load and to analyze impact performance of motorcycle safety helmet.
2 Methodology 2.1 Design The motorcycle helmet model is divided into three models, which are full-face, openface and half-face helmets. In this project, we focus on one main component: the outer shell of each helmet. Each of the helmet models used the same material, which is glass reinforced plastic (GRP). All the helmet models are drawn in SOLIDWORKS and transferred into Inventor Nastran. Finite element analysis for the helmets was carried out in Inventor Nastran (see Table 1). To overcome the helmet design issue, a well-structured design method has been developed based on the readily available helmet in the market that complied with Malaysian Standard MS1 and Regulation UNR 22. For example, the systematic approach framework and the mechanical design process are well-known design approaches that can be applied. Table 1 Material properties of glass reinforced plastic (GRP) [6] Material
Young’s modulus (GPa)
Poisson ratio
Yield strength (MPa)
Density (kg/m3 )
Glass reinforced plastic
8
0.1
75
1830
Impact Analysis of Motorcycle Helmet: Finite Element Modeling
241
Fig. 1 Open-face helmet
This study discusses the three types of helmets, as shown in Figs. 1, 2 and 3. The helmet protects the ears and eyes from dust, air flow and covers the entire head from the back head base.
2.2 Simulation During the simulation, boundary conditions are essential to ensure that the simulated scope does not differ from the objective. A 5 km/h velocity is applied to the motorcycle helmet in −x direction. Frontal collision has been simulated since face is considered as the most vulnerable area of the head when using helmet. In this project, a rigid wall is fully fixed to evaluate the helmet designs’ impact response. Each helmet collided with the wall in the front direction. Figure 4 is the simulation setup for all helmet models.
242
N. Aimi Huda and M. S. Salwani
Fig. 2 Full-face helmet
3 Results and Discussion 3.1 Total Deformation Figures 5, 6 and 7 shows the total deformation of the motorcycle helmet. Total deformations were obtained as the squared root of the square of directional deformation. Written deformation in Figs. 5, 6 and 7 must deduct the distance between the helmet and the rigid wall to obtain the actual total deformation. The maximum value of the total displacement of half-face helmets is 3.9 mm, for the open-face helmet is 2.5 mm while the total displacement of the full-face helmet is 3.0 mm. In conclusion, the value of the total displacement for the half-face helmet is higher compared to other helmets.
Impact Analysis of Motorcycle Helmet: Finite Element Modeling
243
Fig. 3 Half-face helmet
Fig. 4 Simulation setup for all helmet models
3.2 Von Mises Stress Figure 8 shows the output of von Mises stress. It is used to predict the yielding of materials under complex loading. From the analysis, von Misses stress in Fig. 8 had obtained the maximum stress of 3.7 × 105 MPa, respectively, while in Fig. 9, obtained the maximum stress 2.1 × 105 MPa. Lastly, for Fig. 10, the maximum stress is 5.7 × 105 MPa. From the von Mises stress results, we could see that the area around the side of the outer shell experienced the maximum stress.
244
N. Aimi Huda and M. S. Salwani
Fig. 5 Deformation for full-face helmet
Fig. 6 Deformation for open-face helmet
The von Misses stresses decrease from a full-face helmet to an open-face helmet and then increase to the half-face helmet. In this analysis, we would be able to identify. stress-concentrated areas. As we can see from the analysis, the place that has the maximum value of von Misses stress is on the side of the outer shell is a bit thinner than the other part of the helmet to compensate for ears.
Impact Analysis of Motorcycle Helmet: Finite Element Modeling
Fig. 7 Deformation for half-face helmet
Fig. 8 von Misses stress for full-face helmet
245
246
N. Aimi Huda and M. S. Salwani
Fig. 9 von Misses stress for open-face helmet
Fig. 10 von Misses stress for half-face helmet
4 Conclusion In this study, three (open-face, half-face and full-face) helmet designs subjected to impact have been carried out based on predefined conditions. It can be concluded that half-face design helmets yield the highest von Misses stress and total deformation value of 5.7 x 105 MPa and 3.9 mm, respectively.
Impact Analysis of Motorcycle Helmet: Finite Element Modeling
247
Acknowledgements The authors would like to thank University Malaysia Pahang for providing the financial support under the University Internal Fundamental Research Grant No. RDU220365 and Postgraduate Research Grant Scheme (PGRS190337).
References 1. Sisimwo PK, Onchiri GM (2018) Epidemiology of head injuries and helmet use among motorcycle crash injury: a quantitative analysis from a local hospital in Western Kenya. Pan Afr Med J 31:1–6 2. Jones MM, Bayer R (2007) Paternalism and its discontents: motorcycle helmet laws, libertarian values, and public health. Am J Public Health 97(2):208–217 3. Liu DS, Chen YT (2017) A finite element investigation into the impact performance of an open-face motorcycle helmet with ventilation slots. Appl Sci 7(3):279 4. Scott LR et al (2019) Helmet use and bicycle-related trauma injury outcomes. Brain Inj 33(13– 14):1597–1601 5. Fernandes FAO, Alves De Sousa RJ (2013) Motorcycle helmets—a state of the art review. Accid Anal Prev 56:1–21 6. Thomas A, Thilak JAJ (2017) Impact analysis on composite helmet by using FRC and glass fiber by using Ansys. Int Res J Eng Technol 4(3):1629–1634
The Protective Performance of Different Types of Motorcycle Helmets in Terms of HIC and BrIC N. Q. Radzuan, M. H. A. Hassan , M. N. Omar, and K. A. Abu Kassim
Abstract This research evaluates the protective performance of full-face, open-face, and half-coverage motorcycle helmets by taking the head injury criterion (HIC) and brain injury criterion (BrIC) as performance indicators. A pendulum test rig was developed to produce a 5.58 ± 0.29 m/s impact speed. The researchers gave a head impact with or without a helmet to an anthropomorphic test device (ATD) head and neck called Hybrid III at its frontal, rear, and side areas. The Hybrid III had an installed Shimmer 200 g IMU sensor in its skull. The raw data output was linear and rotational velocities recorded using ConsensysPro software version 1.6.0. The linear velocity data are then processed by MATLAB® 2016b software because its raw data are uncalibrated. Calibrated linear and rotational velocities were then used to calculate HIC and BrIC. The research can determine no definite best helmet type through the crash impact experiment; as a result it shows the inconsistency of HIC score among three different types of helmets at each impact location. Furthermore, the research found that the helmet type did not provide significant protection towards rotational impact. It is worth mentioning that side impact may cause the highest injury severity due to rotational motion. Keywords Motorcycle helmet · Pendulum test rig · head impact · Head injury criterion · Brain injury criterion
N. Q. Radzuan · M. H. A. Hassan (B) · M. N. Omar Faculty of Mechanical and Automotive Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, 26600 Pekan, Pahang, Malaysia e-mail: [email protected] K. A. Abu Kassim Malaysian Institute of Road Safety Research, 43000 Kajang, Malaysia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_20
249
250
N. Q. Radzuan et al.
1 Introduction Traumatic brain injury (TBI) can be categorized into primary or secondary brain injury and open or closed brain injury. It might also depend on its severity degree. The three categories of injury severity are mild, moderate, and severe. Open brain injury includes a penetrating object towards the skull resulting in an open wound and breakage. Meanwhile, closed brain injury occurs when the head is not broken or cracked, but the fragile brain tissue is damaged due to being shaken or rotated. Some examples of TBI include concussions, contusions, brain haemorrhages, intracranial hematomas, coup-contrecoup brain injury, diffuse axonal injury (DAI), and penetrating brain injury. The TBI occurred may continually lead to secondary brain injury, which is called neurodegenerative disease, such as epilepsy disease [1, 2], Parkinson’s disease [3, 4], amyotrophic lateral sclerosis disease [3, 5, 6], Alzheimer’s disease [7, 8], and spinal muscular atrophy disease [9]. A worldwide TBI-related published statistics data were recorded in The Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) report throughout the years. The data resource is helpful for researchers and stakeholders to assess the current status of any disease cases. The diseases could burden the nation in the twenty-first century; therefore, stakeholders could take any countermeasure action to reverse or lower the statistics of the disease burden. The GBD report covers over 204 countries and territories and focuses on fatal and non-fatal injuries that lead to disabilities in life quality. James et al. (2019) mentioned that unintentional accidents such as falls are the leading factors causing TBI, followed by road traffic injuries [10]. The Centers for Disease Control and Prevention reported about 60,565 and 60,611 death cases related to TBI in the United States in 2018 and 2019, respectively [11]. Alas, no available published TBI statistics for Malaysia cases of recent years by the Ministry of Health supported service called the National Trauma Database, Malaysia (NTrD). In 2011, NTrD published major trauma statistics based on data collected back in January 2009 until December 2009 through 8 emergency and trauma departments of hospitals around Malaysia, namely Hospital Selayang, Hospital Sungai Buloh, Hospital Ampang, Hospital Kuala Lumpur and Hospital Tengku Ampuan Rahimah in Klang Valley area, Hospital Sultanah Bahiyah in Kedah, Hospital Sultanah Aminah in Johor, and Hospital Pulau Pinang. The emergency and trauma departments recorded 166,786 admissions in that particular year regardless of major or minor trauma at any body region of the patients. 78.35% of 4453 primary trauma patients had head and neck injuries with Abbreviated Injury Score (AIS) ≥ 3. The score indicates serious, severe, critical, and maximal injury, which implies currently untreatable. It is also reported that road traffic injuries contributed to nearly 80% of Malaysia’s trauma cases in 2009, predominantly male and younger people aged 15–24 [12, 13]. Motorcyclists and their pillion riders had the highest statistics surpassing other vehicles by 72.36%. In middle- and low-income countries such as India, Indonesia, Vietnam, Cambodia, and Malaysia, under-bone– and scooter–type motorcycles are easy to
The Protective Performance of Different Types of Motorcycle Helmets …
251
commute to destinations, especially during traffic jams in urban areas [14, 15]. Henceforth, half of the fatal accidents in Malaysia each year involve motorcycles [16–19]. It has been an obligation since 1973 to wear a motorcycle helmet as protection equipment to avoid head injury occurrences in Malaysia [20–22]. An observational study was conducted from May 2011 to February 2015 towards a traumatic motorcyclist injury admitted to Hospital Sultanah Aminah, Johor. The authors acknowledged 90% of the study population wore motorcycle helmets during the initial assessment, yet, head injury still became a significant factor in fatality [19]. The effectiveness of motorcycle helmets in protecting the head region, especially its implication on brain injury, has been debated worldwide. Full face helmet was the best protection in reducing the risk of head injury [23–25]. Nevertheless, Hitosugi et al. (2004), Yu et al. (2011), Ramli et al. (2014), and Ramli and Oxley (2016) stated there was no significant difference between full-face helmets and open-face helmets or other types of helmets in head injury severity except improper helmet wearing [26–29]. Servadei et al. (2003) and Olsen et al. (2016) admitted the regulation of universal helmet law at the country level helped in reducing head injury [30, 31] contradicting the Malaysian case scenario when the universal helmet law was first legislated where the rate of motorcyclist fatality was unusually spiked [20]. The lack of public awareness and law enforcement back then was the reason for no significant increase in motorcycle helmet percentage usage. Motorcycle helmet certification standard was established to ensure their quality before being marketed; Federal Motor Vehicle Safety Standards (FMVSS) No. 218, Snell M2015/M2020, British Standard Institution (BS 6658), ECE 22.05, JIS T 8133: 2000, depending on country. In Malaysia, the certification was provided by the Department of Standards Malaysia (STANDARDS MALAYSIA) or SIRIM Berhad. The test that was performed for the certification standard was evaluating head injury criterion (HIC) only. Yet, Meng (2019) pointed out that other than penetrating brain injury, head injury during road traffic accidents also included closed brain injury, which is mainly caused by rotational acceleration such as subdural haematomas or diffuse brain injuries [31–34]. Henceforth, there is a need to study the association between rotational motion and TBI [35]. The primary objective of the research is to evaluate the protective performance of three different types of motorcycle helmets in preventing head injury. HIC and BrIC scores would be considered indicators of the evaluation. The research may investigate which helmet performs best in the highest possibility of sustaining TBI at different impact locations among frontal, rear, and side. Previous studies and international standards completed crash impact experimental work by helmet dropping in monorail or guided fall. Still, the researchers used a novel testing method, pendulum impact, for the data collection.
252
N. Q. Radzuan et al.
2 Methodology The research has used a novel experimental method and apparatus to evaluate motorcycle helmet effectiveness. The testing methodology can produce linear and rotational motion to measure the head injury severity.
2.1 Pendulum Test Rig A pendulum test rig was developed by imitating Thorne et al. in striking NOCSAE head form [36]. However, the test rig is customized, which best suits the materials available in the research lab. The primary material used is aluminium profiles, which made it solid and durable, measuring 2.00 m × 0.62 m × 2.06 m, as shown in Fig. 1. The test rig was designed in pendulum style for the easiness of repetitive impact. The 1.50-m pendulum arm can be released from the 90° to the lower angles depending on the approximate desired speed. The impact of this research occurred at the lowest gravity, which resulted in the highest possible speed. The theory of impact speed was calculated using the conversation energy theorem as presented in Eq. 1: mgh =
1 2 mv 2
(1)
where m is the mass of the impacted object in kilogram, g is the constant related to Earth’s gravitational force accelerating objects towards its centre, h is the height of the pendulum head from a released position in metres, and v2 is the speed of the impact. The impact speed by calculation is 5.42 m/s; however, after testing the five Fig. 1 Pendulum test rig
The Protective Performance of Different Types of Motorcycle Helmets …
253
trials using a mobile app called SpeedClock, the average speed recorded was 5.58 ± 0.29 m/s. The research employed one speed only throughout the study with the thought that accidents may occur at various conditions so do its impact speed. Therefore, the helmet’s performance was evaluated based on the 5.58 m/s. The international standards for motorcycle helmet certification had implemented impact speed between 4.3–7.8 m/s, for instances where FMVSS No. 218 used 5.2 m/s (hemi anvil) and 6.0 m/ s (flat anvil), BS 6658 used 4.6–7.5 m/s (flat anvil) and 4.3–7.0 m/s (hemi anvil), ECE 22.05 used 7.5 m/s, and Snell M2000 and M2005 used 7.8 m/s (flat anvil), 6.6 m/s (hemi anvil), and 7.8 m/s (edge anvil). The motorcycle helmet in many experiments has been tested from 3.05 to 10.0 m/s [32, 37–42]. Thus, this research has chosen a speed in between that range.
2.2 The Hybrid III Anthropomorphic Test Device An anthropomorphic test device (ATD) called a crash test dummy was used to replace actual humans in the crash test experiment. The usage of ATD copies the impact absorption that might happen to actual humans in road traffic accidents. This research used a head and neck Hybrid III ATD attached to a platform on the ground. Humanetics developed Hybrid III with the association of the Society of Automotive Engineers (SAE) Biomechanics Committees and the National Highway Transport and Safety Administration (NHTSA). It is recognized worldwide and used in many automotive applications and road safety testing. The Hybrid III was designed as biofidelic, with size, weight, stiffness, impact absorption and dissipation, mimicking the actual human. Usually, ATD is categorized into age, impact direction, size, and sex [43]. The Hybrid III head and neck are made of cast aluminium parts weighted 6.08 kg. Meanwhile, its skin is removable vinyl skin. Its neck is part of segmented rubber and aluminium and can simulate rotation flexion and extension response to the impact test.
2.3 Sensor Shimmer 200 g IMU sensor was placed inside the Hybrid III skull at the head’s centre of gravity. It was used to collect the raw data of linear and rotational velocities during the crash impact experiment by three accelerometers and a gyroscope. The data collected by the Shimmer were sent directly to the laptop through a Bluetooth connection. Simultaneously, ConsensysPro software version 1.6.0 recorded and screened the data on a computer. The raw data of linear velocities were in uncalibrated mode. Therefore, it must be coded into MATLAB® 2016b software for producing calibrated data. The final output is linear acceleration for every 0.001 s in axes x, y, and z. On the contrary,
254
N. Q. Radzuan et al.
ConsensysPro quickly converted the raw data of rotational velocities into .csv format, which was easily viewed on Microsoft Excel for further analysis. The calibrated linear and rotational motion data was prepared for the following process. The data were considered through a line graph in Microsoft Excel to shorten it approximately at the impact area. This needs to be done to ensure the data is easy to manage.
2.4 Signal Processing The gyroscope in the sensor may measure the rotational velocities in three-axis X, Y, and Z. The rotational motion data in all three axes may have gone through numerical differentiation once to get rotational velocities. However, this data should be filtered first through a fourth-order Butterworth filter with a cut-off frequency of 167 Hz before eliminating error accumulation. The process can be achieved by installing the Microsoft Excel add-in [19]. The resultant linear acceleration was calculated for data from axes X, Y, and Z. Still, the highest value of rotational velocities in axes X, Y, and Z was determined for the rotational motion output calculation.
2.5 Experiment Procedure Three types of motorcycle helmets were used in the crash impact experiment, as listed in Table 1. The impact locations were set to be at the frontal, rear, and side areas of the motorcycle helmets. At first, the pendulum was released in a free fall motion in the direction of the non-helmeted Hybrid III. Next, the experiment proceeded to impact Hybrid III with the helmet. This action was performed repeatedly towards different types of motorcycle helmets. Nonetheless, the impact at each impact location was once only to prevent data inconsistency due to the degradation of helmet performance. Both full-face and open-face helmets are SIRIM certified. SIRIM Berhad is a certification body in the country that inspects and approves product distributors before being marketed to the user. Meanwhile, the half-coverage helmet has no certification approval. Half-coverage helmet is the lightest among the helmets. This also concluded that more materials were included in the full-face and open-face helmets. Figure 2 depicts a series of photographs capturing the moment of impact when a pendulum cylinder strikes an open-face helmet. A nylon cap was placed over the Table 1 Details of each motorcycle helmet No
Type of motorcycle helmet
Brand
Weight (g)
Certification
1
Full-face helmet
Zeus ZS-813
1664
Certified
2
Open-face helmet
X-DOT G618-N
1069
3
Half-coverage helmet
GTmotor Harley Retro
610.67
Not Certified
The Protective Performance of Different Types of Motorcycle Helmets …
255
Fig. 2 Photographs of the pendulum impact taken before, during, and after the pendulum cylinder struck the open-face helmet
metal portion of the cylinder to prevent it from unintentionally damaging the skin of the ATD. The photographs depict the helmet before, during, and after the impact. Hybrid III’s neck was also not locked to the test platform because the human neck may not be strengthened during impact. Hence, the Hybrid III may slide after being impacted by the pendulum weight. The decision is because the victim does not expect the impact in any actual accident. The situation could differ for head impact experimental procedures for sports such as soccer heading. The player may purposely strengthen his neck before commencing soccer heading.
2.6 Output Calculation The evaluation measures the head injury criterion (HIC) and brain injury criterion (BrIC) as indicators of helmet protective performance. Peak linear acceleration (PLA) is the first data extracted from the linear velocities to compute HIC. PLA is the maximum value of the resultant linear acceleration calculated using Eq. 2: |R| =
√
x 2 + y2 + z2.
(2)
Next, the dataset nearest the PLA was tabulated for HIC calculation. PLA and HIC are linear motion indicators commonly used in international helmet standards. HIC score is the summation of the biggest area under the curve (AUC) within 0.015 s that contains PLA. The formula of HIC is given in Eq. 3: [ H I C = (t2 − t1 )
1 t2 − t1
] 2.5
t2
a(t)dt
max
(3)
t1
where a(t)dt is the acceleration, and t2 and t1 are the final and initial times HIC is calculated. The rotational motion of this research focused on BrIC based on the extensive research by Takhounts et al. (2013), Gabler (2017), and Craig et al. (2020) [44–46]. BrIC has started being considered in most helmet testing lately. There are many head injury predictors discussed in TBI research arenas. However, this research concentrated on HIC and BrIC only. BrIC was calculated using the formula given in Eq. 4:
256
N. Q. Radzuan et al.
┌ ( |( ) ) )2 ( ) | max(|ωx |) )2 max |ω y | max(|ωz |) 2 | Br I C = + + ωxC ω yC ωzC
(4)
where ω[x,y,z] is the angular velocity in the direction of the −X, −Y, and −Z axes in rad/s. The value of critical angular velocities, ωxC , ω yC , ωzC , was provided by Takhounts et al. (2013) [44].
3 Results and Discussion The risk of head injury can be reduced by wearing a helmet, as shown in Fig. 3. The risk varies between 37 and 58% at the side impact, 34 and 56% at the frontal impact, and 21 and 41% at the rear impact. The risk varies according to the type of motorcycle helmet. The result finds no definite helmet that may best protect the head injury at every impact location. This can be seen that an open-face helmet performs better at protecting against side impact, and a full-face helmet performs better at protecting against frontal impact. Surprisingly, the uncertified half-coverage helmet has the lowest HIC score in rear impact protection. The graph in Fig. 4 depicts the helmet slightly helps in reducing the risk of injury severity in the case of rotational motion by 4–8% at side-impact, 13–18% at frontal impact, and 5–12% at rear impact. Therefore, the current helmet in the market heavily focuses on reducing injury risk in terms of linear motion but not rotational motion. In addition, side-impact notably may give a higher risk of injury severity in terms of rotational motion because the BrIC of side impact is the highest compared to frontal and rear impact. Tsai et al. (1995), Sung et al. (2016), and Erhardt et al. (2016) mentioned that a full-face helmet provides better protective performance based on their case study.
Fig. 3 The comparison of head injury criterion
The Protective Performance of Different Types of Motorcycle Helmets …
257
Fig. 4 The comparison of brain injury criterion
How best would the full-face helmet be in a scientific impact crash experiment? This research offers mechanical data through both HIC and BrIC in Figs. 3 and 4 to support the findings from several clinical studies that claimed no significant difference between full-face helmets and open-face helmets or other types in head injury severity [26–29]. The risk of head injury due to linear motion could be solved by adding more padding to the rear side of the helmet. This is due to the highest HIC score at the rear impact amongst all helmets. The case is different for the rotational motion. In an accident, the head could be rotated, for example, at 30 rad/s, despite wearing any ordinary marketed helmet. Henceforth, the risk of injury due to rotational motion may still occur, although more padding was to be added. Pertaining to the issue, adding more padding to the helmet might not be an effective solution to rotational injury. Both Aare (2003) and Finan et al. (2008) agreed coefficient of friction has a significant influence on rotational acceleration [47, 48]. Hence, reducing friction between the helmet and head or between the helmet and impact surface may improve head injury due to rotational motion. Fernandes et al. stated two possible solutions to the issue are Multi-directional Impact Protection System (MIPS) and Phillips Head Protection System (PHPS) technologies [49]. MIPS technology was developed and founded by Peter Halldin, Hans vons Holst, and Svein Klein, focusing on rotational-induced brain injury. The mechanism of MIPS covers the former statement of reducing friction mentioned earlier between the helmet and head, as shown in Fig. 5. A low-friction layer is placed between the helmet’s hard shell and comfort liner to ensure the head may rotate less in an accident, unlike the commonly marketed helmet, which has padding foam liner glued to its shell. MIPS may reduce rotational velocities and acceleration by approximately 25% and 40%, respectively, compared to commonly marketed helmets [50], further reducing the BrIC score. On the other hand, Phillips Helmet takes the latter part of reducing friction between helmet and impact surface. It is based on patented research work by Kenneth David Phillips [51], which was then commercialized by Lazer Helmets company [52] as
258
N. Q. Radzuan et al.
Fig. 5 Mechanism of MIPS in protecting brain injury [49]
PHPS or SuperSkin technology in 2009. Figure 6 shows the outer shell of the Phillips helmet has a layer of a synthetic lubricant-based membrane mimicking the human scalp. The skin of the outer shell may slide or be pushed to the side in an oblique impacted accident. This action may decrease friction between the helmet and the impact surface, thus further reducing rotational velocity and acceleration. Over 50% rotational acceleration can be reduced, declining 67.5% of intracerebral shearing risk [49]. Several things could be improved with the methodology of the experiment described. First, the use of a pendulum impact test raises concerns about the validity of the results. International helmet certification evaluations typically use a monorail impact drop test with a sensor of 20,000 g, which may be a more realistic simulation of a motorcycle crash. It would be beneficial for future studies to compare the validity of pendulum impact tests in quantifying the effectiveness of motorcycle helmets. Fig. 6 Low friction membrane layer at the outer shell of Phillips helmet [53]
The Protective Performance of Different Types of Motorcycle Helmets …
259
Second, the neck of the Hybrid III test dummy was not locked to the test platform, allowing the torso to move after impact, which may lead to data inconsistency. In a real accident, the neck promptly may be strengthened in anticipation of impact. On the other hand, the unlocked neck in this experiment and the moving torso of the Hybrid III may result in different outcomes than what would be seen in a real-life crash. This is because low friction due to the sliding movement could resulting data inconsistency, especially in rotational motion.
4 Conclusion The crash impact experiment is accomplished by developing a novel pendulum test rig that was used to perform a head impact experiment. The experiment was performed on three types of motorcycle helmets: full-face, open-face, and half-coverage. A 5.58 ± 0.29 m/s speed impacted the helmet’s frontal, rear, and side areas. The result shows mixed HIC and BrIC scores among the three helmets; hence, no definite conclusion can be made as the best protective performance helmet for protecting at every impact location. Helmets provide reasonably good protection at linear motion as measured by international helmet standards. However, the currently marketed helmet does not offer significant protection for rotational movement. The side location of the motorcycle helmet is critical to receiving greater rotational injury severity due to its high BrIC score among all helmets. Henceforth, manufacturers should consider MIPS or PHPS technologies in helmet design. Overall, while the study provides some valuable insights into the performance of different types of motorcycle helmets in terms of linear and rotational motion, the limitations in the methodology raise concerns about the generalizability of the results. Future research should address these limitations to provide more reliable conclusions about the effectiveness of motorcycle helmets in reducing the risk of head injury. Acknowledgements The authors want to acknowledge ASEAN NCAP, FIA Foundation, Global NCAP, OEMs and the Society of Automotive Engineers Malaysia (SAE Malaysia) for funding this research under the ASEAN NCAP Holistic Collaborative Research (ANCHOR III) grant. Also, the authors are thankful to the Universiti Malaysia Pahang for providing the research facilities.
References 1. Ding K, Gupta PK, Diaz-Arrastia R (2016) Epilepsy after traumatic brain injury. In: Grant G, Laskowitz D (ed) Translational research in traumatic brain injury. CRC Press/Taylor and Francis Group
260
N. Q. Radzuan et al.
2. Burke J et al (2021) Association of posttraumatic epilepsy with 1-year outcomes after traumatic brain injury. JAMA Netw Open 4(12):1–14. https://doi.org/10.1001/jamanetworkopen.2021. 40191 3. Sreepad B et al (2020) Spinal cord and brain injury protection: testing concept for a protective device. Accid Anal Prev 22(4):1–14. https://doi.org/10.2217/nmt-2016-0017 4. Marin C, Fuentes M, Alobid I, Tubita V, Rojas-Lechuga MJ, Mullol J (2022) Olfactory bulb excitotoxicity as a gap-filling mechanism underlying the link between traumatic brain injuryinduced secondary neuronal degeneration and Parkinson’s disease-like pathology. Neurochem Res 47(4):1025–1036. https://doi.org/10.1007/s11064-021-03503-x 5. Liu G et al (2021) Head injury and amyotrophic lateral sclerosis: a meta-analysis. Neuroepidemiology 55(1):11–19. https://doi.org/10.1159/000510987 6. Gu D, Ou S, Tang M, Yin Z, Wang Z, Liu G (2021) Trauma and amyotrophic lateral sclerosis: a systematic review and meta-analysis. Amyotroph Lateral Scler Frontotemporal Degener 22(3– 4):170–185. https://doi.org/10.1080/21678421.2020.1861024 7. Dams-O’Connor K, Guetta G, Hahn-Ketter AE, Fedor A (2016) Traumatic brain injury as a risk factor for Alzheimer’s disease: current knowledge and future directions. Neurodegener Dis Manag 6(5):417–429. https://doi.org/10.2217/nmt-2016-0017 8. Li L, Liang J, Fu H (2021) An update on the association between traumatic brain injury and Alzheimer’s disease: focus on Tau pathology and synaptic dysfunction. Neurosci Biobehav Rev 120:372–386. https://doi.org/10.1016/j.neubiorev.2020.10.020 9. Johannsen J, Weiss D, Daubmann A, Schmitz L, Denecke J (2021) Evaluation of putative CSF biomarkers in paediatric spinal muscular atrophy (SMA) patients before and during treatment with nusinersen. J Cell Mol Med 25(17):8419–8431. https://doi.org/10.1111/jcmm.16802 10. James SL et al (2018) Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: a systematic analysis for the global burden of disease study 2017. The Lancet 392(10159):1789–1858. https://doi.org/10.1016/S0140-6736(18)32279-7 11. Centers for Disease Control and Prevention (2022) Surveillance report of traumatic brain injuryrelated hospitalizations and deaths by age group, sex, and mechanism of injury—United States, 2018 and 2019. United States of America 12. Sabariah Faizah J, Abd Wahab M, Abdul Wahab MY, Tze Ming Y, Mohd Saiboon I (2011) National trauma database January 2009 to December 2009-fourth report. Sungai Buloh, Selangor 13. Arulsamy A, Shaikh MF (2020) Current status of traumatic brain injury research in Malaysia: a systematic review. Neurosci Res Notes 3(4):1–21. https://doi.org/10.31117/neuroscirn.v3i4.52 14. Radzuan NQ, Hassan MHA, Khairil Anwar AK, Ahmad Azad Ab R, Intan Suhana MR, Nur Aqilah O (2021) The association of socio-demographic characteristics towards driver behaviour and traffic fatality in Selangor, Malaysia. In: Human-centered technology for a better tomorrow, pp 394–405 15. Fauzi WMSBW, Bin Omar AR, Bin Rashid H (2017). Enhancement of postura motergoTM : from an ergonomic motorcycle test rig to a full-scale simulator. J Mech Eng SI 4(4):43–59 16. Sultan Z, Ngadiman NI, Dela F, Kadir A, Roslan NF, Moeinaddini M (2016) Plan Malays J Malays Inst Plan (Special Issue IV):135–146 17. Abdul Manan MM, Jonsson T, Várhelyi A (2013) Development of a safety performance function for motorcycle accident fatalities on Malaysian primary roads. Saf Sci 60:13–20. https://doi. org/10.1016/j.ssci.2013.06.005 18. Malaysia Road Safety Department and JKJR (2018) Buku Statistik Keselamatan Jalan Raya, Kuala Lumpur, Malaysia 19. Henry Tan CL, Tan JH, Mohamad Y, Ariffin AC, Imran R, Tuan Mat TNA (2019) Clinical characteristics of 1653 injured motorcyclists and factors that predict mortality from motorcycle crashes in Malaysia. Chin J Traumatol English Edn 22(2):69–74. https://doi.org/10.1016/j.cjtee. 2018.11.001 20. Darma Y (2017) A time series analysis of road traffic fatalities in Malaysia. University of Malaya
The Protective Performance of Different Types of Motorcycle Helmets …
261
21. Law TH, Radin Sohadi RU (2004) The effect of economic crisis and targeted motorcycle safety program on traffic deaths in Malaysia. J Inst Eng Malays 65(3/4):28–32 22. Oxley J, O’Hern S, Jamaludin A (2018) An observational study of restraint and helmet wearing behaviour in Malaysia. Transp Res F: Traffic Psychol Behav 56:176–184. https://doi.org/10. 1016/j.trf.2018.03.028 23. Tsai Y, Wang J, Huang W (1995) Case-control study of the effectiveness of different types of helmets for the prevention of head injuries among motorcycle riders in Taipei, Taiwan. 142(9):974–981 24. Sung KM et al. (2016) The preventive effect of head injury by helmet type in motorcycle crashes: a rural Korean single-center observational study. BioMed Res Int 2016. https://doi. org/10.1155/2016/1849134 25. Erhardt T, Rice T, Troszak L, Zhu M (2016) Motorcycle helmet type and the risk of head injury and neck injury during motorcycle collisions in California. Accid Anal Prev 86:23–28. https:// doi.org/10.1016/j.aap.2015.10.004 26. Hitosugi M, Shigeta A, Takatsu A, Yokoyama T, Tokudome S (2004) Analysis of fatal injuries to motorcyclists by helmet type. Am J Forensic Med Pathol 25(2):125–128. https://doi.org/10. 1097/01.paf.0000127397.67081.f5 27. Yu WY, Chen CY, Chiu WT, Lin MR (2011) Effectiveness of different types of motorcycle helmets and effects of their improper use on head injuries. Int J Epidemiol 40(3):794–803. https://doi.org/10.1093/ije/dyr040 28. Ramli R, Oxley J, Hillard P, Mohd Sadullah AF, McClure R (2014) The effect of motorcycle helmet type, components and fixation status on facial injury in Klang Valley, Malaysia: a case control study. BMC Emerg Med 14(1):1–11. https://doi.org/10.1186/1471-227X-14-17 29. Ramli R, Oxley J (2016) Motorcycle helmet fixation status is more crucial than helmet type in providing protection to the head. Injury 47(11):2442–2449. https://doi.org/10.1016/j.injury. 2016.09.022 30. Olsen CS et al. (2016) Motorcycle helmet effectiveness in reducing head, face and brain injuries by state and helmet law. Inj Epidemiol 3(1). https://doi.org/10.1186/s40621-016-0072-9 31. Servadei F, Begliomini C, Gardini E, Giustini M, Taggi F, Kraus J (2003) Effect of Italy’s motorcycle helmet law on traumatic brain injuries. Inj Prev 9:257–260 32. Meng S (2019) Towards improved motorcycle helmet test methods for head impact protection: using experimental and numerical methods. KTH Royal Institute of Technology 33. Singleton MD (2017) Differential protective effects of motorcycle helmets against head injury differential protective effects of motorcycle helmets against head injury. Traffic Inj Prev ISSN 9588:387–392. https://doi.org/10.1080/15389588.2016.1211271 34. Mallory A, Duffy S, Rhule H (2013) Head injuries to helmeted and unhelmeted motorcyclists in US Trauma data. 44(0):424–430 35. Xiao Z, Wang L, Mo F, Lv X, Yang C (2020) Influences of impact scenarios and vehicle frontend design on head injury risk of motorcyclist. Acc Anal Prev 145(June). https://doi.org/10. 1016/j.aap.2020.105697 36. Thorne B (2016) Pendulum based impact testing of athletic helmets using the NOCSAE headform, University of Nevada. http://digitalscholarship.unlv.edu/thesesdissertations/2907 37. Engsberg JR, Standeven JW, Shurtleff TL, Tricamo JM, Landau WM (2009) Spinal cord and brain injury protection: testing concept for a protective device. Spinal Cord 47(8):634–639. https://doi.org/10.1038/sc.2009.1 38. DeMarco AL, Chimich DD, Gardiner JC, Nightingale RW, Siegmunda GP (2010) The impact response of motorcycle helmets at different impact severities. Accid Anal Prev 42:1778–1784. https://doi.org/10.2217/nmt-2016-0017 39. Whyte T, Gibson T, Eager D, Milthorpe B (2017) Full-face motorcycle helmet protection from facial impacts: an investigation using THOR dummy impacts and SIMon finite element head model. Inj Prev 23(3):205–210. https://doi.org/10.1136/injuryprev-2015-041925 40. Ebrahimi I, Golnaraghi F, Wang GG (2015) Factors influencing the oblique impact test of motorcycle helmets. Traffic Inj Prev 16(4):404–408. https://doi.org/10.1080/15389588.2014. 937804
262
N. Q. Radzuan et al.
41. Lloyd JD (2017) Biomechanical evaluation of motorcycle helmets: protection against head and brain injuries. J Forensic Biomech 08(03). https://doi.org/10.4172/2090-2697.1000137 42. Mertz HJ (2002) Anthropomorphic test devices. Acc Inj 66–84. https://doi.org/10.1007/9781-4757-2264-2_4 43. Hasnun M, Hassan A (2016) Modelling and analysis of soccer heading and protective headgear to understand and prevent mild traumatic brain injury. Universiti Malaysia Pahang 44. Takhounts EG, Craig MJ, Moorhouse K, McFadden J, Hasija V (2013) Development of brain injury criteria (BrIC). SAE Tech Pap 2013(November):243–266. https://doi.org/10.4271/201322-0010 45. Gabler LF (2017) Development of improved metrics for predicting brain strain in diverse impacts. University of Virginia 46. Craig M, Parent D, Lee E, Rudd R, Takhounts E (2020) Injury criteria for the THOR 50th male ATD. D.C., United States, Washington 47. Aare M (2003) Prevention of head injuries focusing specifically on oblique impacts. Royal Institute of Technology (KTH) 48. Finan JD, Nightingale RW, Myers BS (2008) The influence of reduced friction on head injury metrics in helmeted head impacts. Traffic Inj Prev 9(5):483–488. https://doi.org/10.1080/153 89580802272427 49. Fernandes FAO, Alves De Sousa RJ (2013) Motorcycle helmets—a state of the art review. Accid Anal Prev 56:1–21. https://doi.org/10.1016/j.aap.2013.03.011 50. Halldin P, Aare M, Kleiven S, von Holst H (2003) Improved helmet design and test methods to reduce rotational induced brain injuries. In: RTO specialist meeting, the NATO’s Research and Technology Organization (RTO) 51. Philips KD (2004) Protective headgear and protective armour and a method of modifying protective headgear and protective armour. US 2004/0168246 A1 52. Lazer Helmets (2009) Lazer helmets: a history of innovations. https://www.lazerhelmets.com/ brand.php. Accessed 27 June 2022 53. Costley-White J (2010) New helmet could protect against rotational injuries. BikeRadar. https:// www.bikeradar.com/news/new-helmet-could-protect-against-rotational-injuries/. Accessed 27 June 2022
Measuring Running Performance Through Technology: A Brief Review Siti Rabiatull Aisha Idris
Abstract Running has recently gained popularity, and many runners use wearable technology to enhance their race preparation. This paper provides a brief review of measuring running performance through technology usage. The focus is on the effectiveness of those related technologies in providing accurate data, which at the same time could improve runners’ performance as well as their technique. The finding from this study showed that existing technologies could provide good data (at least for individual reference) since other aspects of running techniques, such as stride length, cadence, and power, were integrated into the technology. Nonetheless, it is hoped that future technology will encourage the development of quick interfaces that could enhance running techniques and training by considering injury-related features in human–computer interaction and running. Keywords Running · Running technology · Sports
1 Introduction Today’s runners have easier access than ever to sporting innovation and technology. The predictions of a tech-rich future have mostly come true, from running watches (smart watches), GPS trackers, and heart rate monitors to creative approaches to laces, scales, and clothes. Performance-oriented measures, such as pace, travelled distance, and elapsed time, are prevalent in technology-supported running. New wearable assistive devices are currently being produced that provide information on a runner’s technique and hence can help to increase running economy as well as lower the risk of injury [1, 2]. For complex tasks like gait retraining, which is anticipated to gain from the development of wearable devices, concurrent visual motor S. R. A. Idris (B) Faculty of Mechanical & Automotive Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, 26600 Pekan, Pahang, Malaysia e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_21
263
264
S. R. A. Idris
input has frequently been shown to be an effective strategy [3, 4]. Compared to traditional motion-capture technologies, wearable technologies’ continued development and growing use present a chance to gather quantitative data “in the field” less intrusively, for longer periods, and with fewer spatial restrictions [5–8]. New-generation sensors can provide real-time feedback and enable prospective investigations on large cohorts; they are compact, portable, barely noticeable, inexpensive, and simple to use [9]. Measuring running performance through technology has become increasingly popular and accessible in recent years. Technological advancements have provided runners with a variety of tools and devices to track their performance, monitor their progress, and analyze their data. The most common ways technology can be used to measure running performance are GPS watches and running apps [10, 11]. The GPS watches and smartphone apps equipped with GPS capabilities allow runners to track their distance, pace, and route in real-time. These devices provide valuable data on speed, distance covered, elevation, and even heart rate. They often offer features such as interval training, goal setting, and personalized coaching. Other than that, heart rate monitors are also often being used since they can provide insights into a runner’s cardiovascular effort and help optimize training intensity [12, 13]. Some devices use chest straps to measure heart rate accurately, while others utilize wrist-based optical sensors. By monitoring heart rate during workouts, runners can determine their training zones and ensure they are training at the appropriate intensity. Meanwhile, power meters, like those used in cycling, are becoming more prevalent in the running world [14, 15]. These devices measure the power output of a runner, which considers factors such as pace, incline, and body weight. Running power can provide valuable information on the overall effort and efficiency of a runner, helping them optimize their training and pacing strategies. In addition, running stride sensors were also used to measure running performance [16, 17]. They were typically attached to the shoe or placed on a specific part of the body, providing detailed information about a runner’s cadence, stride length, vertical oscillation, and ground contact time. These metrics can help identify areas for improvement in running form and efficiency. Then, video analysis [18, 19]. Even though not a direct technologybased measurement, video analysis has become more accessible with smartphones and wearable cameras. Recording and analyzing running form can provide valuable insights into biomechanics, allowing runners to identify and correct any flaws that may be affecting performance or increasing the risk of injury. Finally, online platforms and social communities [20, 21]. Numerous online platforms and social communities exist where runners can store and analyze their training data, set goals, and compare their performance with others. These platforms often provide detailed statistics, training plans, and the ability to connect with coaches or fellow runners for support and motivation. While there have been significant advancements in technologies for measuring running performance, there are still several improvements that can be carried out. Instead of fully exploiting the potential of new technologies to shed light on the connections between biomechanics (such as movement methods) and injury-related characteristics, studies frequently describe the potential of those technologies. This is
Measuring Running Performance Through Technology: A Brief Review
265
because there is presently little information regarding the connections between some biomechanical assessments, their resultant values, and injuries or injury risks. (e.g. recovery status, inciting factors) [9, 22]. Therefore, this paper summarizes information related to the characteristics of the technologies, types of measures used, and the effectiveness of those technologies in providing accurate data, which at the same time could improve runners’ performance and technique and prevent injury.
2 Running Technologies 2.1 Methods and Model There were several methods and models for measuring running performances used among athletes or runners [23–25]. The most common method is basic measurements which include height, weight, body mass index, skinfolds, and muscle measurements. Secondly is timing. This is the simplest and most used method of measuring out a known distance and timing the athlete sprinting from one side to the other. Thirdly is the minimal power model. This is a brand-new, simplistic, and all-encompassing model for human running performance that makes use of the relative metabolic cost of movement. Fourthly is the universal running model. This model uses two indices that gauge stamina and the rate at which the body absorbs the most oxygen to describe how well runners perform. The next method was mathematical models. There were several mathematical models available in the literature to estimate performance in longdistance running events such as the 5, 10 km, half-marathon, and marathon. After that, wearable devices. An athlete’s performance indices can be derived from their regular exercise performance using models that can make use of data from wearable devices. Finally, the other method was performance measurement versus evaluation. Performance measurement is ongoing and part of a continuous improvement process, while evaluation is discrete. These methods and models can provide insights into different aspects of running performance, such as endurance, velocity, the metabolic cost of transport, and individual determinants of fitness. Nonetheless, it is noted that in running performance models, basic measurements and performance indices are two different types of variables used to predict performance. Basic measurements include height, weight, body mass index, skinfolds, and muscle measurements [23, 25]. These variables are individual-dependent and can be used to estimate the relationship between energy expenditure and performance. On the other hand, performance indices quantify different aspects of performance and provide a unique insight into basic determinants of fitness in a large population of runners over a wide range of exercise capacities and long-time scales [25]. Performance indices are required to predict race performance and contain no additional individual-dependent quantities. For example, the universal running model characterizes a runner’s performance with two indices that measure endurance (endurance index) and the velocity of maximum oxygen
266
S. R. A. Idris
uptake [25]. These indices can be obtained from regular exercise performance data collected from wearable devices. In summary, basic measurements are individualdependent variables that can estimate the relationship between energy expenditure and performance, while performance indices are population-based variables that provide insights into different aspects of running performance. In addition, those running performance models can be used to inform training programs for athletes by identifying strengths and weaknesses, measuring progress, designing custom training programs, and providing feedback but still lacking information or feedback on injury prevention. It is noted that those methods and models were generated based on selected users. Hence, data presented to the individual may not be accurate and unbalanced. The task of training individualized models has received most of the published work in the literature (i.e., phase two) [26]. To this aim, techniques based on transfer learning were proposed to address the issue of an unbalanced dataset, where labelled data from one type of user can also aid in optimizing the models for another type of user [27]. Additionally, progressive learning has been suggested to expedite online training while simultaneously giving learners the freedom to acquire new categories of user activities [28–30]. When a user’s behaviour evolves or a new type of user activity demands to be recognized, incremental learning techniques have proven to be effective at adapting [31]. In contrast to the customized training model’s issue (i.e., phase two), the other model that received less attention was the baseline training model (i.e., phase one). The baseline model is very important and should be considered as a priority; otherwise, improvement on the accuracy of human activity recognition (HAR) in the second phase cannot be carried out [32]. Developing a strong baseline model is quite challenging, and there can be two main challenges: i. Inadequate training data—the baseline model is complicated and needs to consider a variety of user categories (e.g., age, gender, body height), type of activity, and methods for embedding sensors (e.g., helmet, limbs, hand-held). Such a model needs a huge volume of labelled data to process. For training an effective industrial-class DL model, an initial training dataset produced in experimental conditions is inadequate. ii. Biased training data—Normal practice to increase training dataset is through data user feedback. Nonetheless, the data was generated based on selected users. This biased data would affect the baseline model, which cannot fit most of the users. Several methods have been proposed to overcome those problems, and they can be divided into two categories which are data-oriented and model-oriented. Dataoriented is used to improve training dataset quality. For instance, modifying the size of the training dataset through resampling technique by decreasing or increasing a certain class of data [33]. Meanwhile, model-oriented methods focus on designing learning algorithms that could tolerate certain dataset flaws. For instance, a vast dataset from a different field can be added to a small dataset via transfer learning. [34].
Measuring Running Performance Through Technology: A Brief Review
267
2.2 Measurement for Running Performance and Technique In general, running performance can be measured through running economy, which is the volume of oxygen consumed per kilogram of body weight per kilometre. Running economy can be measured in a lab-based test where the person runs on a treadmill at a variety of speeds for long enough that they achieve a physiological steady state while having their VO2 measured. To improve running economy and ultimately enhance running performance, several strategies can be employed, such as consistent training, strength training, improving form, gradually increasing mileage, sprint training, avoiding downhill running, and plyometric exercises or running drills. These are known as real-time feedback. Knowing which running technique biomechanical elements to employ for realtime feedback is crucial because wearables can accurately assess an increasing number of them. Items that are appropriate for real-time feedback include (i) are closely related to injuries or running economy, (ii) can be monitored precisely under varied conditions, and (iii) are changeable [35]. Real-time input can be given to lower injuries and boost running efficiency on such components: stride/step frequency, stride/step length, contact time, kinematic (trunk flexion, hip flexion–extension, vertical displacement of the pelvis during stance, leg extension, knee flexion, ankle eversion, foot strike, ankle dorsiflexion), kinetic (braking force, vertical impact peak, vertical loading rate). Because of the limitations of laboratory-based studies, for example, due to their low sample sizes and limited ability to measure the multifactorial nature of running injuries, as well as the fact that they typically only determine technique once before the follow-up, even though the technique may alter during the monitoring phase, there may be inconsistent relationships in prospective studies on injuries [36]. Since data collected in-field are not subject to these restrictions, it can be utilized to discover novel connections between running performance, technique, and injuries [37]. As training becomes more serious, accurate and precise data are important among wearable users [38–40]. Hence, validated and trustworthy variables for real-time feedback should be used in wearables. The validity and dependability of biomechanical items generated from sensors like accelerometers and pressure insoles have been the subject of several research, and these items are also progressively being tested in environments that more closely resemble real-world situations. Even though many (mostly spatiotemporal) variables can be reliably measured, this is not always the case for every variable. This is owing, for instance, to the sampling frequency, operating range, or sensor placements [41, 42]. Therefore, clinicians, design engineers, and researchers must study whether variables have been proven, ideally in settings that represent real-world use. Lastly, the end-user should be able to modify the variable needed, to exist in the wearable. Almost all variables in running can be changed, although some are probably simpler and easier to change directly. For instance, switching to a forefoot striking pattern is simpler when the stepping pace can be increased at once rather than doing so while maintaining the baseline step rate [43].
268
S. R. A. Idris
2.3 Quantifying the Workload Intensity, frequency, and duration/distance influence a running program’s workload. Running injuries have been linked to sudden increases in workload [44]. The intensity of running can be assessed in a variety of ways because running duration and frequency are easy to quantify, and hence it is crucial to understand which measurements are pertinent for real-time feedback. The suitable variables for real-time feedback should include highly correlated with the tangible metabolic and/or mechanical intensity, they can be measured with accuracy, and they are also adjustable by the athlete. The relative amount of energy expended per minute (kJ kg 1 min) or the caloric expenditure per kilometre (kcal kg 1 km) is commonly used to describe the metabolic intensity of exercise [45, 46]. The term “mechanical intensity” is used to describe tissue strain and stress since these are crucial elements in (mal) adaptation. Wearables typically rely on indirect measurements because they are unable to detect energy consumption or tissue stresses and strains directly. There is no one approach that is best for providing real-time feedback because each intensity metric has its benefits and drawbacks. For new runners who are still wondering about intensity, a combination of different measurements may provide the best indication of the real intensity, depending on how the perceived intensity relates to an objective intensity measure like heart rate [47]. However, it was shown that highly trained athletes frequently have low training during high-intensity sessions and too much training during lowintensity sessions, which makes it beneficial to have real-time feedback, especially on running intensity [48]. Lastly, heart rate’s real-time feedback data have aided users in keeping up their speed, indicating that they can (simply) change this variable [40]. However, not all runners are aware of the appropriate running intensity. Therefore, they would welcome guidance [40, 49]. Therefore, rather than just providing numbers, real-time feedback should indicate whether the runner should aim to decrease or raise based on the intended outcome of the session.
3 The Effectiveness of Running Technologies in Improving Performance and Technique 3.1 Technology that Detects Technique The existing technologies have made huge progress in measuring parameters involved to improve running technique [41, 42, 50]. Researchers have demonstrated that it is feasible to define technical parameters such as vertical oscillation (VO), knee flexion– extension angles, cadence, foot strike type (FST), and ground contact time (GCT) using wearable sensors and camera-based motion detection [51–53]. This enables the profiling of running styles, the identification of improvement opportunities, and the
Measuring Running Performance Through Technology: A Brief Review
269
detection of fatigue indicators. Nonetheless, the runners primarily have access to the technique-related parameters measured by the sensors after the run, frequently in the form of kinematic analysis. Due to the lack of feedback during the actual motions, changing their running style/technique is difficult. Although certain technologies can detect technique-related factors in real-time, little research has been done on how to use the information to construct helpful interfaces, such as systems that can help runners as they are doing running movements. It was found that information such as VO2 max, running power, and cadence, which was offered by current smartwatches, was a number to improve their running technique. For some people, it is hard to understand the use of such numbers/data to enhance their running technique. Moreover, it was discovered that it is inefficient and eventually dangerous to repeatedly address the watch while running to observe the results of movement adjustments. This is due to a lack of injury-related information embedded in the smartwatch. There was a study on kinematic analysis in running using wearable sensors that reported that the device will notify athletes when their arms move imprecisely, which can reduce their running economy [4]. As a result, the system serves as a constant running coach, providing real-time feedback on a specific aspect of running technique. Meanwhile, the other method also exists to notify the user, which is through auditory feedback, in case they go beyond their training limit [54]. Whereas Sensoria Fitness Socks (SFS) use socks with built-in pressure sensors in them, can measure cadence and foot strike type, and can directly transfer the information to the application on a smartphone. When runners deviate from user-defined run characteristics, the application employs audible feedback to notify them. Those systems were used to notify the users/runners when the movement is imprudent. Nonetheless, based on their distinct feedback, it might be challenging to determine how to correct a mistaken movement and to what extent it needs correction [55]. Therefore, run training interface designs should take advantage of the fact that technique-related metrics are becoming detectable in real-time to help runners with their movements rather than portraying the motions as visual information on a screen after a run.
4 Conclusion This paper summarizes information related to characteristics of the technologies, types of measures used, and the effectiveness of those technologies in providing accurate data, which at the same time could improve runners’ performance and technique and prevent injury. Based on existing information and data obtained from current running technologies, it is evident that real-time input on running forms is primarily pertinent for people who have an existing or recurrent injury or display a form that raises their injury risk.
270
S. R. A. Idris
References 1. Anderson T (1996) Biomechanics and running economy. Sports Med 22(2):76–89 2. Moore IS (2016) Is there an economical running technique? A review of modifiable biomechanical factors affecting running economy. Sport Med (Auckland, N.z.) 46 793–807 3. Napier C, Esculier JF, Hunt MA (2017) Gait retraining: out of the lab and onto the streets with the benefit of wearables. Br J Sports Med 51(23):1642–1643 4. Strohrmann C et al (2011) Out of the lab and into the woods: kinematic analysis in running using wearable sensors. In: Ubiquitous computing 5. Paquette MR et al (2020) Moving beyond weekly “distance”: optimizing quantification of training load in runners. J Orthop Sports Phys Ther 50(10):564–569 6. Naughton M et al (2020) Correction to quantifying the collision dose in rugby league: a systematic review, meta-analysis, and critical analysis. Sports Med Open 6(1):50 7. Macadam P et al (2019) Quantification of the validity and reliability of sprint performance metrics computed using inertial sensors: a systematic review. Gait Posture 73:26–38 8. Vanwanseele B et al (2020) Accelerometer-based data can provide a better estimate of cumulative load during running compared to GPS-based parameters. Front Sport Act Living 2 9. Van Hooren B et al (2020) Real-time feedback by wearables in running: current approaches, challenges and suggestions for improvements. J Sports Sci 38(2):214–230 10. Salazar N (2022) The paradox of mobility technology usage: how GPS sports watches keep “active lifestylers” (im)mobile. Mobil Humanit 1(1):62–75 11. Johansson RE et al (2020) Accuracy of GPS sport watches in measuring distance in an ultramarathon running race. Int J Sports Sci Coach 15(2):212–219 12. Pasadyn SR et al (2019) Accuracy of commercially available heart rate monitors in athletes: a prospective study. Cardiovasc Diagn Ther 9(4):379–385 13. Flatt AA, Howells D (2019) Effects of varying training load on heart rate variability and running performance among an Olympic rugby sevens team. J Sci Med Sport 22(2):222–226 14. Cerezuela-Espejo V et al (2021) Are we ready to measure running power? Repeatability and concurrent validity of five commercial technologies. Eur J Sport Sci 21(3):341–350 15. Leo P et al (2022) Power profiling and the power-duration relationship in cycling: a narrative review. Eur J Appl Physiol 122(2):301–316 16. Zhang H, Guo Y, Zanotto D (2020) Accurate ambulatory gait analysis in walking and running using machine learning models. IEEE Trans Neural Syst Rehabil Eng 28(1):191–202 17. Rana M, Mittal V (2020) Wearable sensors for real-time kinematics analysis in sports: a review. IEEE Sens J 21(2):1187–1207 18. Palucci Vieira LH et al (2019) Match running performance in young soccer players: a systematic review. Sports Med 49(2):289–318 19. Pons E et al (2019) A comparison of a GPS device and a multi-camera video technology during official soccer matches agreement between systems. PLoS ONE 14(8):e0220729 20. Tsai T-H et al (2021) Running on a social exercise platform: applying self-determination theory to increase motivation to participate in a sporting event. Comput Hum Behav 114:106523 21. Couture J (2021) Reflections from the ‘Strava-sphere’: Kudos, community, and (self-) surveillance on a social network for athletes. Qual Res Sport, Exerc Health 13(1):184–200 22. Ceyssens L et al (2019) Biomechanical risk factors associated with running-related injuries: a systematic review. Sports Med 49(7):1095–1115 23. Alvero-Cruz JR et al (2020) Predictive performance models in long-distance runners: a narrative review. Int J Environ Res Public Health 17(21) 24. Mulligan M, Adam G, Emig T (2018) A minimal power model for human running performance. PLoS ONE 13(11):e0206645 25. Emig T, Peltonen J (2020) Human running performance from real-world big data. Nat Commun 11(1):4936 26. Wang J et al (2017) Deep learning for sensor-based activity recognition: a survey. Pattern Recognit Lett 119:3–11
Measuring Running Performance Through Technology: A Brief Review
271
27. Ding R et al (2019) Empirical study and improvement on deep transfer learning for human activity recognition. Sensors 19(1):57 28. Hu C et al (2018) A novel random forests based class incremental learning method for activity recognition. Pattern Recogn 78:277–290 29. Ntalampiras S, Roveri M (2016) An incremental learning mechanism for human activity recognition, pp 1–6 30. Wang Z et al (2012) An incremental learning method based on probabilistic neural networks and adjustable fuzzy clustering for human activity recognition by using wearable sensors. IEEE Trans Inf Technol Biomed 16(4):691–699 31. Yang Q, Gu Y, Wu D (2019) Survey of incremental learning. In: 2019 Chinese control and decision conference (CCDC), pp 399–404 32. Siirtola P, Koskimäki H, Röning J (2019) Personalizing human activity recognition models using incremental learning. abs/1905.12628 33. Shorten C, Khoshgoftaar TM (2019) A survey on image data augmentation for deep learning. J Big Data 6(1):60 34. Ramentol E et al (2012) SMOTE-RSB*: a hybrid preprocessing approach based on oversampling and undersampling for high imbalanced data sets using SMOTE and rough sets theory. Knowl Inf Syst 33(2):245–265 35. Phillips E et al (2013) Harnessing and understanding feedback technology in applied settings. Sports Med 43(10):919–925 36. Shen P et al (2021) Effects of running biomechanics on the occurrence of iliotibial band syndrome in male runners during an eight-week running program—a prospective study. Sport Biomech 20(5):560–570 37. Kiernan D et al (2018) Accelerometer-based prediction of running injury in national collegiate athletic association track athletes. J Biomech 73:201–209 38. Clermont CA et al (2020) Runners’ perspectives on ‘Smart’ wearable technology and its use for preventing injury. Int J Hum-Comput Interact 36(1):31–40 39. Rupp MA et al (2016) The impact of technological trust and self-determined motivation on intentions to use wearable fitness technology. Proc Human Factors Ergon Soc Ann Meet 60(1):1434–1438 40. Kuru A (2016) Exploring experience of runners with sports tracking technology. Int J HumComput Interact 32(11):847–860 41. Mitschke C, Kiesewetter P, Milani TL (2018) The effect of the accelerometer operating range on biomechanical parameters: stride length, velocity, and peak tibial acceleration during running. Sensors 18(1):130 42. Mitschke C, Zaumseil F, Milani TL (2017) The influence of inertial sensor sampling frequency on the accuracy of measurement parameters in rearfoot running. Comput Methods Biomech Biomed Engin 20(14):1502–1511 43. Huang Y et al (2019) Foot strike pattern, step rate, and trunk posture combined gait modifications to reduce impact loading during running. J Biomech 86:102–109 44. Damsted C et al (2018) Is there evidence for an association between changes in training load and running-related injuries? A systematic review. Int J Sports Phys Ther 13(6):931–942 45. Achten J, Jeukendrup AE (2003) Heart rate monitoring. Sports Med 33(7):517–538 46. Fletcher JR, Esau SP, MacIntosh BR (2009) Economy of running: beyond the measurement of oxygen uptake. J Appl Physiol 107(6):1918–1922 47. Tholander J, Nylander S (2015) Snot, sweat, pain, mud, and snow: performance and experience in the use of sports watches. In: Proceedings of the 33rd annual ACM conference on human factors in computing systems 48. Foster C et al (2001) Differences in perceptions of training by coaches and athletes. S Afr J Med Sci 8:3–7 49. Lazar A et al (2015) Why we use and abandon smart devices. In: Proceedings of the 2015 ACM international joint conference on pervasive and ubiquitous computing 50. Benson LC et al (2022) Is this the real life, or is this just laboratory? A scoping review of IMU-based running gait analysis. Sensors 22(5):1722
272
S. R. A. Idris
51. Altman AR, Davis IS (2012) A kinematic method for footstrike pattern detection in barefoot and shod runners. Gait Posture 35(2):298–300 52. Eskofier B, Musho E, Schlarb H (2013) Pattern classification of foot strike type using bodyworn accelerometers. In: 2013 IEEE international conference on body sensor networks, pp 1–4 53. Uddin MZ et al (2021) Estimation of mechanical power output employing deep learning on inertial measurement data in roller ski skating. Sensors 21(19):6500 54. Eriksson V, Bresin R (2010) Improving running mechanics by use of interactive sonification. In: Proceedings of the interaction sonification workshop (ISon) 2010 55. Sigrist R et al (2013) Augmented visual, auditory, haptic, and multimodal feedback in motor learning: a review. Psychon Bull Rev 20(1):21–53
Experimental Study of Gait Monitoring on Wearable Shoes Insole and Analysis: A Review Nur Wahida Saadion
and Mohd Azrul Hisham Mohd Adib
Abstract Gait analysis can be as straightforward as observational screening to identify any abnormalities that are visible to the unaided eye. By focusing on a typical gesture that arises in the gait cycle or walking pattern, customized wearable insoles help to improve the walking pattern by raising the arch to the ideal level, which makes the foot more balanced. This paper focuses on the review of the experimental research study regarding stance time, swing phase, step length, and stride length of the gait analysis towards gait abnormalities. A few methods were tested to develop an ideal wearable shoe insole for future users. More than half of the authors are concerned to analyze gait analysis via the stance phase compare to other parameters. Nevertheless, the crucial part of gait analysis is the analysis of the stance phase and swing phase, which most of the authors did not highlight as the issue in the research study. Hence, this paper also concluded the research studies similarities and limitations for future improvement. Keywords Gait analysis · Wearable · Insole · Sensor · Body balance
N. W. Saadion (B) · M. A. H. M. Adib Human Engineering Group (HEG), Faculty of Mechanical & Automotive Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah (UMPSA), 26600 Pekan, Pahang, Malaysia e-mail: [email protected] M. A. H. M. Adib e-mail: [email protected] M. A. H. M. Adib Medical Engineering and Health Intervention Team (MedEHiT), Centre for Advanced Industrial Technology, Universiti Malaysia Pahang Al-Sultan Abdullah (UMPSA), 26600 Pekan, Pahang, Malaysia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_22
273
274
N. W. Saadion and M. A. H. M. Adib
1 Introduction Wearable shoe insole is aimed to measure the physiological indicators without interfering with daily activities and provides indicative data of a person’s health. By concentrating on gestures that arise in the gait cycle or walking pattern, customized wearable insoles help to align feet and ankles, ease pain, and eventually improve walking form by raising the arch to the ideal level, which makes the foot more balanced. Gait analysis is a clinical approach to evaluate, record, and make any required improvement in terms of dynamic posture and coordination during movement. During the evaluation process, the therapist needs to monitor the minor shifts in movement such as rotations and tilts or knee movement and foot placement [1]. A crucial indicator of the general health of an elder is their preferred walking speed. Spastic gait, musculoskeletal hip pathology, and antalgic gait can affect gait analysis in terms of foot-dragging and knee pain. Lower limbs are usually associated with spasticity as a factor of plantarflexed ankles displayed by “tip-toe” walking and toe dragging. Gradually, reduced weight bearing on the affected side characterizes antalgic gait caused by knee discomfort. During the stance phase, the knee will still bend and probable toe weight-bearing occurs. Thus, this approach may help to detect and treat diseases early, as well as to cut back on frequent trips to or lengthy stays in pricey healthcare institutions. This study focuses on reviewing the trend of the experimental research study on the gait analysis towards gait abnormalities concerning the method, similarities, and limitations on the parameters in terms of stance time, swing phase, step length, and stride length.
2 Methodology 2.1 Development of In-Shoe Wearable Pressure Sensor Using an Android Application [2] Vertical Ground Reaction Force (VGRF) [2] which complies with spatial and temporal characteristics and the preferred transition pace was the aim of the studies. By utilizing a 2 by 2 inches printed circuit board (PCB) mounted beside the shoe to monitor, VGRF can be explored. Compromising to minimize cost, the TekScan FlexiForce sensor was used on the heel, 1st and 5th metatarsals to compare peak pressures of the subject’s walking pattern toward 30 subjects. The researcher’s goal in measuring VGRF by measuring the subject’s walking in normal mode and slow mode can be shown in Fig. 1. The paper concludes peak pressure can vary due to knee flexion by body weight.
Experimental Study of Gait Monitoring on Wearable Shoes Insole …
275
Fig. 1 a Insole sensor position and b the average body weight and peak pressure of one subject [2]
2.2 In-Sole Shoe Foot Pressure Monitoring for Gait Analysis [3] The [3] depicts resistive pressure sensors in contact with the person’s knee and toe developed on 15 regions as shown in Fig. 2. To supply secure power to the device, Bluetooth technology was utilized as an ideal device for indoor or outdoor purposes. Microcontroller was employed to translate the data from the sensor. Static position experiments were performed with different subjects to reach the goal of this paper which is to evaluate pressure exerted on the foot standing on one leg and both legs for 60 s as presented below. To compare graphs (b) left and right, the pressure applied on both legs is less than one’s that applied to one leg. This study of foot pressure brings benefits in the analysis of gait for Parkinson’s patients, diabetics, and people who have had orthopedic surgery.
Fig. 2 a In-shoe plantar measurement design system and b pressure versus time left; person standing on both legs right; person standing on a leg [3]
276
N. W. Saadion and M. A. H. M. Adib
Fig. 3 SmartStep block diagram and the developed design [4]
2.3 SmartStep: A Fully Integrated, Low-Power Insole Monitor [4] The author in [4] reflects on resolving the limitation of previous work regarding shoe monitoring systems. The current device was improved to an always-on system powered by a rechargeable battery in terms of an FSR sensor attached to PCB, IMU sensor, and flash storage as shown in Fig. 3. Based on the reliability outcome, the device successfully demonstrated 99% of data can be retrieved in a one-hour experiment through multiple disconnection or reconnection events. In terms of battery life, it can run for more than 24 h on a 40 mAh battery and can last for two days with a single charge on 12 h maximum daily wear.
2.4 SmartStep 2.0: A Completely Wireless, Versatile Insole Monitoring System [5] In [5], the researcher suggested a solution to improvise on wireless charging method due to battery charging issues. SmartStep 2.0 electronics insole had been improved by a 100 mAh battery encapsulated in epoxy resin and cushioned by orthotic foam as shown in Fig. 4. To validate the objective of this study, the author improved the heels of the shoes to be less than 4.5 cm to charge the battery with sufficient voltage. The device was then tested on a male subject in a free-living environment. The procedure was conducted three times to find the average in three days. Within 3 days, the functionality of realtime data logging with offline transmission (RTOL) was proven successful in sensor data logging and transfer to the base station. The insole system is also actively used with a single charge for nearly 5 days.
Experimental Study of Gait Monitoring on Wearable Shoes Insole …
277
Fig. 4 SmartStep 2.0 insole monitor [5]
2.5 Piezoelectric Pressure Sensing Shoe Insole [6] This study is focusing on collecting quantitative foot pressure distribution data [6] by evaluating foot or posture pathologies. The researcher is concerned with early detection and preventative amputation of high-risk patients. The pressure sensor of the device was placed as shown in Fig. 5. On the other hand, the main interface of the device to the user is under the responsibility of the microcontroller and mobile application. This device will rely on the user’s foot physiology in conjunction with their walking style influencing the amount of power that the piezo generators can supply, as well as the pressure distribution. To have an accurate depiction, tolerance of the pressure sensor range must be maintained for all parts in contact with the insole. The tolerance analysis outcome depicts that the voltage reading for pressure will be depending on the arch type.
Fig. 5 a Sensor position on the insole, and b foot pressure distribution by arch [6]
278
N. W. Saadion and M. A. H. M. Adib
Fig. 6 Hardware components of RT-GAIT [7]
2.6 Development of the RT-GAIT, a Real-Time Feedback Device to Improve the Gait of Individuals with Stroke [7] In [7], the author offered an evaluation of real-time feedback in a clinical study. The author made a hypothesis at the start of the study: targeted the device should be reliable to receive real-time feedback and improve temporal gait asymmetry and walking ability in stroke patients. Two pressure sensors were located in a flexible plastic insole as shown in Fig. 6, to measure the stance times of both lower legs during walking. The device was tested on the left hemiparesis post-stroke individual for three days walk with and without the RT-GAIT. The support time in the affected lower extremity’s single limb improved by 80%, while the percentage of the gait cycle of the same single limb support improved by 83%. Unfortunately, the device system has a constraint limit to utilize sensors, which provide less information compared to combining it with IMU sensors like accelerometers and gyroscopes.
2.7 Smart Insole: Wearable Sensor Device for Unobtrusive Gait Monitoring in Daily Life [8] Smart Insole [8] differs from other development research since it aimed to solve the aforementioned problems regarding gait laboratories, video, and IMU, and real-life gait monitoring. There are a few advantages listed by this device as shown in Fig. 7 including step count, step pace, swing time, and center of pressure (COP) shifting velocity as well as dynamic balance status and potential fall risk in real-life. The author offered gait monitoring and analyzing features for their future use. To test the reliability of the device, this research includes 7 activity segments related to sitting, stair climbing, and walking. Overall, Smart Insole displays the same result as the pedometer and human observation. The evaluation resulted that Smart Insole is robust
Experimental Study of Gait Monitoring on Wearable Shoes Insole …
279
Fig. 7 The mechanical design of Smart Insole [8]
and reliable, with proven accuracy of more than 99% in step count. Nevertheless, the limitation of this device highlighted on lack of wireless charging features.
2.8 Design and Development of Integrated Insole System for Gait Analysis [9] The developmental research [9] presents a monitoring device by using inertia measurement sensors. The initial development of this device required to be worn on both feet by strapping it to the subject’s lower legs with an insole inserted inside a sandal or shoes as shown in Fig. 8. Impressively, the device can stand alone in freeliving conditions without any complication. To validate the device’s performance, an experiment on a healthy subject was chosen for the initial preliminary design toward average stride length, step length, velocity, and step count. The reliability of the device can be guaranteed based on the result recorded in Table 1 based on the acquired signal towards the gait parameters measured.
280
N. W. Saadion and M. A. H. M. Adib
Fig. 8 Subject with wearable module [9]
Table 1 Gait parameters measured [9] Subject
Distance traveled (m)
Stride length (m)
Step length (m)
Avg velocity (m/ s)
Stride count
Trial 1
10
1.0965
0.52
0.6202
10
Trial 2
12
1.3333
0.665
0.6821
9
Trial 3
14
1.2507
0.635
0.6895
11
Trial 4
16
1.1429
0.596
0.6966
14
2.9 A Reliable Gait Phase Detection System [10] Closed-loop FES walking applications [10] using a portable gait phase detection system (GPDS) were investigated in this paper. The GPDS presents an innovative latest sensor configuration utilizing force-sensitive resistors and a miniature gyroscope chip as shown in Fig. 9 to measure the force load on a shoe insole and the rotational velocity of the foot. This study targeted a few parameters: stance, heel-off, swing, and heel-strike phase. The researcher validates the device’s reliability by experimenting to prove the success rate which represents by two groups: healthy individuals and individuals with gait pathologies by walking on few obstacles and climbing and descending stairs. Overall, the experiments showed the detection success rate for both group subjects was greater than 99%, and only in the stair climbing and descending tasks, subjects with impaired gait dominated up to 96%.
Experimental Study of Gait Monitoring on Wearable Shoes Insole …
281
Fig. 9 Placement of the sensors used by the GPDS [10]
2.10 A Shoe-Integrated Sensor System for Long-Term Center of Pressure Evaluation [11] In [11], a flexible insole device was introduced to unbothered the user’s normal gait. The device is designated to recognize human activity to automate the CoP monitoring process. Unfortunately, some limitations can create a larger bias between the two systems by altering the insole shape. Nevertheless, the experiment results indicate sufficient accuracy of region consistency in plantar pressure sampling of the two systems. The device is considered qualified for clinical purposes since the ICC values are also above 0.95 in the ML direction and above 0.9 in the AP direction as in Fig. 10. Overall findings demonstrated that this study system can provide almost the required accuracy as a commercial device but with more affordable price and self-analyze as an added advantage.
Fig. 10 a The dynamic CoP traces detected during walking and b the display of the DAT module, the insole, and the overall hardware in a shoe [11]
282
N. W. Saadion and M. A. H. M. Adib
3 Result and Discussion 3.1 Finding Wearable shoe insole can be divided into 3 criteria that are vital to the gait analysis research which are the device sensor, the parameter, and features that are owned by the device as shown in Table 2. The minimal parameter that a shoe insole should measure in gait analysis is by taking into consideration of step length, stride length, number of steps, and swing phase. Nevertheless, most likely the sensor used in the research were conducted towards healthy people more than people with disorders which are not aligned with the original purpose of the device which was developed for people who have neurological disorders. The features of the research that had been layout in this paper are also in continuous development since it still did not achieve the same standard as commercial equipment, but with much lower cost, consumers may experience their own proven self-based analysis device without the need for a professional therapist to utilize. The sensor is the most essential element of a wearable shoe insole device since it recognizes the majority of the user’s motion and movement. All of the authors agreed to position the main sensor on the heel and first metatarsal, but only 60% of them located it on the fifth metatarsal. The position of the sensor varies among 40% of the writers simply because it is determined by the parameter to be evaluated. Since the decisive criterion should be minimally measured are stride length or cadence, step length, step count, and swing phase, parameters are being restricted in the device, unfortunately out of ten papers that have been reviewed, only half of the research which also represent the highest percentage among 4 parameters measured stance phase as their main parameter to have research on. Others have been less than 20%, and the majority of them employed pressure on the foot and normal walking activities as the primary parameter to evaluate gait analysis. In terms of the features, most authors propose data storage and wireless connectivity because the parameters must be recorded for the user to self-analyze based on the performance results, and more than half of them mentioned designing a lightweight and thin design. Almost 60% of the author did not discuss on how the device had been proven to be prolonged due to the device’s lack of a rechargeable battery. To recap, when compared to current technology and product, the device cannot serve as the next successor to measure gait analysis efficiently because most of the parameters are not provided.
3.2 Limitation Gait analysis consists of several stages but the most important of which are the stance phase and swing phase, which most researchers fail to emphasize the importance of assessing those two for older people. A standard gait evaluation usually necessitates at least one complete gait cycle, which includes a foot strike with the observed
x
x
In-Sole shoe foot x pressure monitoring for gait analysis [3]
x
x
x
SmartStep: a fully integrated, low-power insole monitor [4]
SmartStep 2.0—a completely wireless, versatile insole monitoring system [5]
Piezoelectric pressure sensing shoe insole [6]
x
x
o
Step length
Development of x in-shoe wearable pressure sensor using an android application [2]
List of the paper/ Parameter gait monitoring Cadence/ device stride length
x
x
x
x
x
Stance phase
Table 2 Summaries of interest issues studied
x
x
x
x
x
Swing phase
o
o
o
o
o
Wireless connectivity
Features
o
o
o
o
o
Data storage
x
o
o
x
x
Lightweight/ thin design
o
o
o
x
x
Rechargeable battery
o
o
o
o
o
Heel
Sensor
o
o
o
o
o
1st metatarsal
o
x
x
o
o
(continued)
5th metatarsal
Experimental Study of Gait Monitoring on Wearable Shoes Insole … 283
x
o
x
Smart Insole: wearable sensor device for unobtrusive gait monitoring in daily life [8]
o Design and development of an integrated insole system for gait analysis [9]
x
x
Step length
Development of the RT-GAIT, a real-time feedback device to improve the gait of individuals with stroke [7]
List of the paper/ Parameter gait monitoring Cadence/ device stride length
Table 2 (continued)
o
o
o
Stance phase
o
o
x
Swing phase
o
o
x
Wireless connectivity
Features
o
o
o
Data storage
o
o
x
Lightweight/ thin design
x
o
x
Rechargeable battery
o
o
o
Heel
Sensor
o
o
o
1st metatarsal
o
o
x
(continued)
5th metatarsal
284 N. W. Saadion and M. A. H. M. Adib
x
x
1
10
A shoe-integrated sensor system for long-term center of pressure evaluation [11]
Total
Percentage % (out of 10 journals)
50
5
o
o
Stance phase
o—An issue covered in the experiment x—Issues that were not covered in the experiment
20
2
x
x
Step length
A reliable gait phase detection system [10]
List of the paper/ Parameter gait monitoring Cadence/ device stride length
Table 2 (continued)
30
3
x
o
Swing phase
90
9
o
o
Wireless connectivity
Features
100
10
o
o
Data storage
60
6
o
o
Lightweight/ thin design
40
4
x
x
Rechargeable battery
100
10
o
o
Heel
Sensor
100
10
o
o
1st metatarsal
60
6
o
x
5th metatarsal
Experimental Study of Gait Monitoring on Wearable Shoes Insole … 285
286
N. W. Saadion and M. A. H. M. Adib
leg, a contralateral foot strike with the other leg, and another foot strike with the observed leg, rather than a static position that only measures pressure on both feet. Nevertheless, sensor selection is also critical in determining what parameters can be measured besides sensor position also will help a lot in gathering a lot of data that will be needed to analyze pathologies and injuries. To make sure the device has a reliable result, several criteria should be considered, such as the use of an inertial measurement unit (IMU) that can measure the x, y, and z axes, which will be more accurate, the array position of a pressure sensor, a rechargeable battery that will extend the device’s use, a user-friendly application, a thin design of the insole, and the option to add a cushion for comfort, as its main goal is for those who are older and have any injuries or neurological disorders.
3.3 Similarities Despite the differences in brand and model, the features used in the gait analysis are remarkably similar to all ten journal articles, steering the wearable shoe insole to a reliable analysis. The activity of evaluating a subject’s gait analysis makes a significant contribution to determining whether the gait is normal or abnormal. The majority of the author will present concise gait analysis using the same activities such as walking, ascending and descending stairs, and assessing the pressure, stand, and swing phases. This evaluation may start and end with your feet, but it involves your entire body. The gait cycle is the sequence of events or periods witnessed by the researcher as the subject moves from one foot to the other. The analyst will analyze your movement for inefficiencies, foot rolls, and other biomechanical aspects. Their evaluation will reveal the underlying causes of malfunctions in your foot mechanism.
4 Conclusion To summarize, wearable shoe insoles are a broad field of study that is still evolving. The primary goal of each researcher differs depending on the parameter chosen to be measured. However, more than half of the authors were concerned about measuring the stance and swing phase in the study. Others were more concerned with the device’s reliability and how to improve it rather than upgrading the parameters that the user could access for analysis. After all, the main parameter to study gait analysis has been insufficiently revealed to the user. Acknowledgements This research work was strongly supported by the Ministry of Higher Education (MOHE) under FRGS/1/2021/TK0/UMP/02/25 grant, RDU210129 and RDU210332 grant from Universiti Malaysia Pahang Al-Sultan Abdullah (UMPSA), which provided the research materials and equipment. The authors have no conflicts of interest that are relevant to the content of this review.
Experimental Study of Gait Monitoring on Wearable Shoes Insole …
287
References 1. Simancek JA (2013) Deep tissue massage treatment. https://www.sciencedirect.com/topics/ neuroscience/gait-analysis 2. Mostfa AA, Xiao W, Sharkawy AN (2022) Development of in-shoe wearable pressure sensor using an android application. AIP Conf Proc 2386(January). https://doi.org/10.1063/5.0066831 3. Malvade PS, Joshi AK, Madhe SP (2017) In-sole shoe foot pressure monitoring for gait analysis. In: International conference on computing, communication, control and automation, ICCUBEA 2017, May 2018. https://doi.org/10.1109/ICCUBEA.2017.8463769 4. Hegde N, Sazonov E (2014) Smartstep: a fully integrated, low-power insole monitor. Electron 3(2):381–397. https://doi.org/10.3390/electronics3020381 5. Hegde N, Sazonov ES (2015) SmartStep 2.0—a completely wireless, versatile insole monitoring system. In: Proceedings—2015 IEEE international conference on bioinformatics and biomedicine, BIBM 2015, April 2018, pp 746–749. https://doi.org/10.1109/BIBM.2015.735 9779 6. Kozel TG, Lee A, Michal TAK (2019) Piezoelectric pressure sensing shoe insole 7. Hegde N, Fulk GD, Sazonov ES (2015) Development of the RT-GAIT, a real-time feedback device to improve gait of individuals with stroke. In: Proceedings annual international conference of the IEEE engineering medicine and biology society EMBS, vol 2015-Novem, September, pp 5724–5727. https://doi.org/10.1109/EMBC.2015.7319692 8. Lin F, Wang A, Zhuang Y, Tomita MR, Xu W (2016) Smart insole: a wearable sensor device for unobtrusive gait monitoring in daily life. IEEE Trans Ind Inform 12(6):2281–2291. https:// doi.org/10.1109/TII.2016.2585643 9. Aggarwal A, Gupta R, Agarwal R (2018) Design and development of integrated insole system for gait analysis. In: 2018 11th international conference on contemporary computing IC3 2018, August, pp 1–5. https://doi.org/10.1109/IC3.2018.8530543 10. Pappas IPI, Keller T, Mangold S, Popovic MR, Dietz V, Morari M (2004) A reliable gyroscopebased gait-phase detection sensor embedded in a shoe insole. IEEE Sens J 4(2):268–274. https:// doi.org/10.1109/JSEN.2004.823671 11. Guo R et al (2021) A shoe-integrated sensor system for long-term center of pressure evaluation. IEEE Sens J 21(23):27037–27044. https://doi.org/10.1109/JSEN.2021.3116249
Preliminary Ergonomics Analysis of Sit-Stand (STS) Desk on the Patient with Lower Back Pain Problem: A Case Study Muhammad Rafli Salim Hasan Raza, Mohd Azrul Hisham Mohd Adib, and Nurul Shahida Mohd Shalahim
Abstract Pain in the thoracic spine is referred to as thoracic back pain that is situated mainly between the shoulder blades at the rear of the chest (the thorax). It stretches about at the waist level from the base of the neck to the beginning of the lumbar spine. The long hours of working in a static position, especially at the desk, significantly affect the back and the patient feels pain, which can lead to musculoskeletal disorders (MSDs). Therefore, this study focuses to investigate the posture impact of the adult with back pain using a Sit-to-Stand (STS) desk product through ergonomics simulation by the Rapid Upper Limb Assessment (RULA) analysis. The development of the virtual environment for ergonomics analysis is created. The study is conducted in three conditions: work-standing, work-sitting on the bedside, and work-sitting on the bed. As a result, all three postures show the same score, and modification of the environment on the posture may be needed. Also, the ergonomic effect of these postures shows the good design and comfort of the STS desk. Keywords Digital human modeling · Musculoskeletal disorders · Ergonomics · Sit-Stand desk
M. R. S. H. Raza (B) · M. A. H. M. Adib · N. S. M. Shalahim Human Engineering Group (HEG), Faculty of Mechanical & Automotive Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah (UMPSA), 26600 Pekan, Pahang, Malaysia e-mail: [email protected] M. A. H. M. Adib e-mail: [email protected] N. S. M. Shalahim e-mail: [email protected] M. A. H. M. Adib · N. S. M. Shalahim Medical Engineering and Health Intervention Team (MedEHiT), Centre for Advanced Industrial Technology, Universiti Malaysia Pahang Al-Sultan Abdullah (UMPSA), 26600 Pekan, Pahang, Malaysia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_23
289
290
M. R. S. H. Raza et al.
1 Introduction Sit-Stand (STS) desk is erected on the idea that reducing sitting time by standing to work throughout the computer routine has a required consequence on health outcomes [1]. Adapting the work-related environment through the overview of the STS desk is one possible method for reducing prolonged sitting during the workday among office employees. Sit-stand workstations allow the user to substitute between sitting and standing postures at their desk, giving them the chance to reduce their sitting time at work. Evidence about STS desks from the occupational health and ergonomics literature has focused on musculoskeletal health and work performance [2]. Musculoskeletal disorders (MSDs) are very frequent in patients seeking physiotherapy [3]. Ergonomics should ideally be considered across all domains of life, including in the clinic, office, and at home [4]. The effectiveness of the treatment and the quality of the goods in the rehabilitation can be improved by emphasizing ergonomics [5]. MSDs are injuries or disorders of the muscles, nerves, tendons, joints, cartilage, and spinal discs that affect the human body’s movement or musculoskeletal system [6]. Two categories of risk factors may cause MSD, which are ergonomic risk factors and individual risk factors. Some of the ergonomic risk factors are awkward postures, repetition, vibration, force, and contact stress [7]. Physiotherapy or physical therapy is a medical specialty focused on the rehabilitation of disabilities and the promotion of mobility, the capacity to work, quality of life, and future movement through tests, assessments, diagnoses, and physical intervention [8]. Several principles of importance to the ergonomics of posture and movement are derived from a range of specialist fields, namely biomechanics, physiology, and anthropometrics [9]. Ergonomics is the relationship between the worker and the job, or the tasks performed, and focuses on designing work areas or tasks in a way that supports and improves performance [10]. Rapid Upper Limb Assessment (RULA) is a simple and quick way to assess the risk of MSDs in a range of working positions. RULA splits the body into segments that can be oblique autonomously, conferring to movement planes, and offers a scoring system for muscle activity across the whole body, whether it is static, dynamic, fastchanging, or unsteady, and where manual treatment may happen, which is stated to as a connection score because it is significant in load management but may not continuously be completed with the hands. The human body is designed to be in motion, and optimum posture plays a crucial role in maintaining a healthy musculoskeletal system. Sitting and standing workstations can have a significant impact on posture, and the correct positioning can help prevent back pain and other musculoskeletal disorders. This study intends to explore the impact of posture on back pain and the benefits of using a Sit-Stand desk to achieve optimal posture while working. The study will use ergonomic simulation and the RULA analysis to evaluate the effectiveness of the Sit-Stand desk in promoting good posture.
Preliminary Ergonomics Analysis of Sit-Stand (STS) Desk …
291
2 Methodology 2.1 Observational Method and Case Study Data gathering began with interviews and observations of the adult who use the STS desk. The next step includes collecting anthropometric data of the adult, images, and videos of an adult using the STS desk to demonstrate their job postures for continued study. This study was conducted at Block D, Kolej Kediaman Dua (KK2), Universiti Malaysia Pahang. The STS desk was specially developed to help an adult with back pain. The adjustable height of the STS desk ranges from 25.4 to 76.2 cm at a few different angles which operate on three different actuators independently. The table also operates simultaneously with the height adjustment that can be controlled using mobile apps. The participant information sheet (PIS) and the consent form (CF) were given to the participant before the start of the assessment. A 28-year-old male subject who has back pain problems willingly took part in the experiment. The participant is informed to should not engage in any strenuous physical activity on the day of the measurement. The subject was given a comprehensive description of the experimental procedure before the experiment started. The subjects’ statistical information was gathered using a meter scale, measuring tape, and weight machine. Table 2 displays the anthropometric details of the participant.
2.2 Digital Human Modelling (DHM) Digital human modeling (DHM) which incorporates ergonomic requirements on human subjects without direct measurement during the early design process is also beneficial for designers or ergonomists [11]. Numerous studies revealed that the posture analysis techniques used by DHM were largely accurate [12]. The RULA assessment approach is used in this study through ergonomic design analysis (human builder) in CATIA software to examine the task’s posture. Table 1 represents the Rapid Upper Limb Assessment (RULA) levels of musculoskeletal disorder (MSD) risk. The static posture and arm supported/person leaning are used as RULA parameters in the analysis. It describes the different levels of risk based on a scoring system used in the ergonomic analysis. Table 1 shows ranges from negligible risk level (score 1–2) to very high-risk level (score 7). The scores are used to guide ergonomic interventions and to prevent and manage MSDs in the workplace. As seen in Fig. 1, the process flow diagram for the development of the virtual environment for ergonomics analysis begins with design in SolidWorks 2022 to construct the CAD models for the STS desk. Then, ergonomics analysis is performed to determine whether the STS desk component can deform in an undesirable way or if a critical stress state develops in any shape, then exported to CATIA in STEP file format. Next, the CATIA V5 Human Builder module is where the digital human is
292
M. R. S. H. Raza et al.
Table 1 The basic RULA level of MSD risk descriptions Final score
Level of risk
Requirements of action
1–2
Negligible
No action required
3–4
Low
Change may be needed
5–6
Medium
Further investigation and change soon
7
Very high
Investigate and change immediately
created. To change the human model’s posture for RULA analysis, lifelike characteristics are then added. Table 1 indicated the risk level of every position is determined using RULA. This allows for an efficient assessment of the adults. The posture of the upper arm, forearm, wrist, wrist twist, neck, trunk, and leg are used by CATIA to determine the total RULA score. The lowest RULA score is 1, and the highest RULA score is 7 [13].
3 Result and Discussion 3.1 Anthropometric Measurement Table 2 shows the detailed description of the participant involved in the study.
3.2 Analysis of Body Posture of the Subject During Work-Standing In this section, the RULA analysis was observed based on three main body posture positions; work-standing, work-sitting on the bedside, and work-sitting on the bed. The results for body posture during standing while working are shown in Fig. 2. Figure 3 shows the RULA analysis for standing while working on the STS desk with approximately 110° angle on the lower arm position and a 15° bending neck position. RULA analysis shows that the posture level final score is 3 and in yellow color. This means that further investigation and change may be needed. The neck, trunk, and leg show a scoring of 3 alongside with forearm scoring 2, which contributed to the RULA analysis final score for standing while working is 3.
Preliminary Ergonomics Analysis of Sit-Stand (STS) Desk …
293
Fig. 1 Process flow diagram for the development of the virtual environment for ergonomics analysis
3.3 Analysis of Body Posture of the Subject During Work-Sitting on the Bedside Figures 4 and 5 shows the RULA analysis for posture sitting on the bedside with approximately 20° angle on the upper arm position and 12° bending neck position and 110° on the lower arm angle. The neck, trunk, and leg show a scoring of 3 alongside the wrist and arm scoring 3 which contributed to the RULA analysis final
294 Table 2 Description of participant’s anthropometric measurement
M. R. S. H. Raza et al.
Characteristics
Value
Gender
Male
Age (years)
28
Height (cm)
167
Weight (Kg)
55
BMI (kg/m2 )
19.7
Waist height (cm)
95
Chest height (cm)
135
Fig. 2 Body posture of the subject during standing while working on the STS desk
Fig. 3 The RULA analysis for posture standing while working
Preliminary Ergonomics Analysis of Sit-Stand (STS) Desk …
295
score for standing while working 3. RULA analysis demonstrates the posture level final score is 3 and in yellow color. This means that further study and change may be needed. Subject legs are properly reinforced to the ground and supported by STS desk base frame as a footrest which eliminates the ergonomic risk factor of static and sustained posture.
Fig. 4 Body posture of the subject during working on the STS desk while sitting on the bedside
Fig. 5 The RULA analysis for posture sitting on the bedside
296
M. R. S. H. Raza et al.
Fig. 6 Body posture of the subject during working on the STS desk while sitting on the bed
3.4 Analysis of the Body Posture of the Subject During Work-Sitting on the Bed Figures 6 and 7 shows the RULA analysis for posture sitting on the bed with approximately 20° angle on the lower arm position and 12° bending neck position and 125° on the lower arm angle. The neck, trunk, and leg show a scoring of 3 alongside with forearm scoring 2 with, the wrist and arm adding 2 more scores. The RULA analysis shows the posture level final score is 3 and in yellow color. This indicates further investigation and change may be desirable. From the results, we know the study of ergonomics has grown and its importance in product design. Therefore, it is critical to establish reliable methods for assessing a new product’s ergonomics from the start of its conceptual definition. The presence of ergonomic risk factor external contact stress which is a condition observed when part of the participant’s body rubs against the corner of the STS desk immediately intervenes after the day of observation by adding padding to the corner of the STS desk. The link between working postures with the risk of MSDs has been convincingly proven. Many existing studies have shown the decrease of risk with the aid of ergonomically designed products [12], redesigned the working layout [13], embracing diversity in the workers using effective design [14], variation in biomedical exposure, and a few other important aids. The method for determining the task risk of the participant using the STS desk product by applying an analytical tool has been described in this study. RULA is one of the analytic instruments used to assess the subjects and determine the condition that needs to be ergonomically modified and designed to avoid all ergonomic risks that could contribute to musculoskeletal disorders (MSD). Future work will require the use of more advanced hardware for
Preliminary Ergonomics Analysis of Sit-Stand (STS) Desk …
297
Fig. 7 The RULA analysis for posture sitting on the bed
complex ergonomics evaluation, such as total human body monitoring and virtual environment rendering. Besides, future research in anthropometric studies can be carried out and STS desk design can be considered based on physical and comfort considerations. Finally, based on the results of the RULA analysis, we can accomplish that using STS desk product when working while standing, working while sitting on the bedside, and working while sitting on the bed share the same result, all three postures score 3 on RULA scoring analysis which indicated that further investigation and change may be needed. Further analysis should be carried out to ensure the STS desk user will not encounter any major risk factors in MSDs when working.
4 Conclusion This study has a real effect on adult performance during working at a desk. The incorrect posture induces muscle fatigue leading to an increase in muscle activity. In addition, the adult would also be injured if it happened continuously. The ergonomic effect of the posture was evaluated to enhance the comfort and design of the STS desk. This knowledge can benefit the designer when designing a product without any hazards. The evaluation also shows the individual has been exposed to all physical risk factors including the neck, the trunk, and the upper extremities. Acknowledgements This research work was strongly supported by the Ministry of Higher Education (MOHE) under FRGS/1/2021/TK0/UMP/02/25 grant, RDU210129 and RDU210332 grant from Universiti Malaysia Pahang Al-Sultan Abdullah (UMPSA) which provided the research materials and equipment. The authors have no conflicts of interest that are relevant to the content of this paper.
298
M. R. S. H. Raza et al.
References 1. Chambers AJ, Robertson MM, Baker NA (2019) The effect of sit-stand desks on office worker behavioral and health outcomes: a scoping review. App Ergo 78:37–53 2. Chau JY, Daley M, Dunn S, Srinivasan A, Do A, Bauman AE, Van Der Ploeg HP (2014) The effectiveness of sit-stand workstation for changing office works’ sitting time: results from the stand@work randomized controlled trial pilot. 11:1–10 3. Integrating ergonomics tools in physical therapy for musculoskeletal risk assessment and rehabilitation-a review. Integrating ergonomics tools in physical therapy for musculoskeletal ergonomics risk assessment and rehabilitation–a review. https://www.researchgate.net/public ation/268078947. Accessed 15 November 2022 4. Musculoskeletal disorder risk factors in manual material handling. https://ergo-plus.com/mus culoskeletal-disorder-msd-risk-factors-manual-material-handling/. Accessed 15 November 2022 5. Work-related musculoskeletal disorders & ergonomics I workplace health strategies by condition I workplace health promotion I CDC. https://www.cdc.gov/workplacehealthpromotion/ health-strategies/musculoskeletal-disorders/index.html. Accessed 15 November 2022 6. The causes of musculoskeletal disorders (MSDs) I ErgoPlus. https://ergo-plus.com/causes-mus culoskeletal-disorders-msds/. Accessed 15 November 2022 7. Jiandani M, Mhatre B (2018) Physical therapy diagnosis: how is it different? J Postgrad Med 64(2):69. https://doi.org/10.4103/jpgmjpgm_691_17 8. Hutting N et al (2017) Physical therapists and importance of work participation in patients with musculoskeletal disorders: a focus group study. BMC Musculoskelet Disord 18(1):196. https://doi.org/10.1186/s12891-017-1546-9 9. Azrul M, Mohd H, Arifin RA, Abdul MH Development of physiotherapy-treadmill (PhyMill) as rehabilitation technology tools for kid with Cerebral Palsy, pp 1–10 10. Ariffin RA, Adib MAHM, Shalahim NSM, Daud N, Hasni NHM (2020) An ergonomic perspective of user need on physio-treadmill (PhyMill) criteria: knowledge and awareness of cerebral palsy among future parents. J Phys Conf Ser 1529(5). https://doi.org/10.1088/1742-6596/1529/ 5/052071 11. De Magistris G et al (2013) Dynamic control of DHM for ergonomic assessments. Int J Ind Ergon 43(2):170–180. https://doi.org/10.1016/).Ergon.2013.01.003 12. Ahmed S, Irshad L, Demirel HO, Tumer IY (2019) A comparison between virtual reality and digital human modeling for proactive ergonomic design. Lecture notes in computer science (including subseries lecture notes in artificial intelligence and lecture notes in bioinformatics), LNCS, vol 11581, no February, pp 3–21. https://doi.org/10.1007/978-3-030-22216-1_1 13. RULA: Postural loading assessment tools for Malaysia mining industry. https://www.res earchgate.net/publication/303143232_RULA_Postural_loading_assessment_tools_for_Mal aysia_mining_industry. Accessed 15 November 2022 14. McAtamney L, Hignett S (2004) Rapid entire body assessment. Handbook of human factors and ergonomics methods, vol 31, pp 8-1–8-11. https://doi.org/10.1201/9780203489925.ch8
Developing a Survey Tool to Measure Human Factors Constructs for Personal Hearing Protector (PHP) Use Among Industrial Workers—First Phase Nur Syafiqah Fauzan , Mirta Widia , and Ezrin Hani Sukadarin
Abstract The survey tool (questionnaire) is one of the most widely used tools to collect data. This paper aims to develop a survey tool of human factors and personal hearing protector (PHP) use among industrial workers. The survey tool is developed based on the combination of the Health Promotion Model (HPM) and Health Belief Model (HBM). Development of the eight main constructs in the survey included narrative literature and qualitative review by the two expert panels of researchers in the related field. Overall, this tool produced good comments from the experts. Some of the items were removed due to poor match in terms of content. This current research is crucial to investigate the factors and PHP usage among targeted industrial workers. This study can serve as the primary instrument for determining the human factors and personal hearing protectors used for industrial workers in various sectors. This survey tool can contribute to an improved understanding of the human factors that may influence the consistent use of PHP in an excessive noise work area. Keywords Survey tool · Factors · Personal hearing protector · Industrial workers
N. S. Fauzan · M. Widia (B) Faculty of Industrial Sciences and Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, Lebuh Persiaran Tun Khalil Yaakob, 26300 Kuantan, Pahang, Malaysia e-mail: [email protected] M. Widia Centre for Advanced Industrial Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, 26600 Pekan, Pahang, Malaysia E. H. Sukadarin Department of Chemical Engineering Technology, Faculty of Engineering Technology, Universiti Tun Hussien Onn Malaysia (UTHM), Johor, Malaysia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_24
299
300
N. S. Fauzan et al.
1 Introduction The World Health Organization (WHO) estimated that around 466 million people have hearing loss globally [1]. Besides, Occupational Noise Related Hearing disorders were reported as the highest among other occupational poisoning and disease cases from 2016 until 2019 [2]. The manufacturing sector reported the highest percentage of confirmed occupational poisoning and disease cases in 2019 [3]. A study by Rasasoran et al. [4] shows that a high prevalence of hearing loss was reported among workers in the noise-exposed palm oil industries. On the other hand, the automotive industry in China reported about 62.53% of them exceeded 85 dB(A) of the personal noise level. Most of the excessive noise comes from various automotive industry jobs, such as surface treatment, metal cutting, stamping, grinding, welding, forging, assembly and plastic moulding [5]. Besides the manufacturing industries, other industries, such as the construction industry, were found to have personal noise exposure level issue among both machine and non-machine workers [6]. Therefore, understanding factors that facilitate or hinder specific safety behaviours’ performance is crucial [7]. A study by Reddy et al. [8] found that both personal and environmental factors for intrapersonal, interpersonal, organisational, community and policy influenced the use of hearing protectors. A study by Acharya [9] found that using personal protective equipment (PPE) among workers was significantly associated with the gender of the respondents and encouragement to use PPE. Several factors lead to using hearing protectors among workers, such as exposure level, individual risk perceptions, a company’s safety climate [10] and social modelling [11]. Nath et al. [12] reported that the significant challenge of issuing personal protective equipment is comfort, and it can be adequate if PPE is worn correctly. Thus, determining factors that play an important role in the usage of PPE is the first step before planning and implementing an intervention to increase PPE use [13]. Thus, this paper aims to present the process of developing a survey tool for measuring human factors and PHP use among industrial workers. The dimensional construct of human factors influencing PHP use among industrial workers was determined. Then, the finalised items within the study construct via qualitative assessment were determined.
2 Materials and Methods 2.1 Survey Tool Development The survey tool was developed in several stages. First, a survey tool is developed based on the guiding validated framework from the combination of the Health Belief Model (HBM) [14] and the Health Promotion Model (HPM) [15].
Developing a Survey Tool to Measure Human Factors Constructs …
301
Second, a narrative literature review was carried out to identify relevant questions closely related to the main objective of the survey tool development. The questions were adapted and modified from the previous studies. The researchers developed the items of each construct from past research by combining two selected frameworks: The Health Belief Model and the Health Promotion Model. The two (2) appointed experts reviewed the constructs used in the survey tools. Another previous study used two expert panels to review the instrument before it could be used by the end users [16]. According to Presser et al. [17], the review method assesses any worries associated with the questionnaire in advance of annoying ideas or unsuitable wording of questions. Each expert commented on the modified items and indicated their decision to remove, keep, or modify them [18].
2.2 Conceptual Framework on the Human Factors and Personal Hearing Protector (PHP) Use Among Industrial Workers The framework had eight constructs. A theoretical framework on the relationship between human factors and PHP usage among industrial workers was used to develop the survey tool. The independent constructs are the interpersonal influences, perceived susceptibility, perceived severity, perceived benefit, perceived barriers, perceived self-efficacy and cues to action. The dependent construct is the use of PHP (see Fig. 1).
3 Results and Discussion The developed survey tool consisted of two parts. The first part focuses on sociodemographics. The second part comprised eight study constructs: interpersonal influence, perceived susceptibility, perceived severity, perceived benefit, perceived barrier, perceived self-efficacy, cues to action and use of PHP. The survey tool applied a 5-point Likert scale. All seven (7) constructs were used scale from 1 (strongly disagree) to 5 (strongly agree). This scale is ranged from 1 (strongly disagree) to 5 (strongly agree), which indicates the respondent’s agreement with each item [20]. Besides, the last construct for Personal Hearing Protector use was to use a scale from 1 as ‘Never’ to a scale of 5 as ‘Always’. The Likert scale is one of the most basic and widely used psychometric tools in sociology, psychology, information systems, politics, economics and other fields [21]. 5-point rating scales are less confusing and may boost the response rate.
302
N. S. Fauzan et al.
Fig. 1 Conceptual framework. Adapted from Health Belief Model and Health Promotion Model [14, 15, 19]
3.1 Sociodemographic The questions for sociodemographic section consists of eighteen (18) questions (gender, age, nationality, marital status, educational level, working experience, type of personal hearing protector, working type, worker’s experience with the ear problem and illnesses, family history of hearing disorder/loss, worker’s audiometric test and hearing re-examination experience, worker’s satisfaction on current PHP and the worker’s position). Table 1 shows the first phase of sociodemographic information is finalised after the qualitative review by the appointed two (2) experts. Overall, items inside this section remain. However, certain items need improvements due to incorrect or overlapping items, and wrong chosen words.
3.2 Analyses of Human Factors Constructs for PHP Use In the beginning, the total items for this tool are 97 items which cover all the study construct. During the qualitative assessment process, certain items considered perfect matches were maintained as it is, while the items considered moderate matches were maintained after refining some of the sentences. Besides, the items that were considered poor matches by the experts were removed. Items that were consistently deemed unnecessary were removed, and the modified items were altered [18]. After
Developing a Survey Tool to Measure Human Factors Constructs …
303
Table 1 Finalise the first phase of Sociodemographic information Sociodemographic
Items
S1
Gender
Male
S2
Age
Female ≤24 year 25–34 years 35–44 years ≥ 45 years S3
Nationality
S4
Marital status
Malaysian Non-Malaysian Single Married Others (widowed, divorced)
S5
Educational level
Primary Secondary Certificate Diploma Bachelor degree Master Ph.D.
S6
Working experience
≤1 year 2–5 years 6–10 years ≥11 years
S7
Type of personal hearing protector (PHP)
Earplug Ear muff Combination
S8
Working type
S9
Do you experience ringing in the ears or sound heard differently in Yes each ear? No
S10
Have you suffered any illness that has affected your hearing (e.g., infection, tinnitus, discharge, etc.)?
Yes
S11
Have you ever had an ear operation or any other major operation that affected your hearing?
Yes
S12
Any family history of hearing loss/disorder?
Yes
Regularly Shift
No No No
S13
Have you had an audiometric test before?
Yes (continued)
304
N. S. Fauzan et al.
Table 1 (continued) Sociodemographic
Items No
S14
Have you suffered any illness listed here?
S15
If YES, please tick (/) ONE or more illnesses listed here
Yes No Vision impairment Alzheimer’s disease Diabetes mellitus Cancer Vertigo Dizziness Psychosocial health Cardiovascular disease Stroke
S16
Experience of hearing re-examination
S17
Are you satisfied with your current personal hearing protector?
Yes No Yes No Not applicable
S18
Position
Manager Engineers Executive Supervisor Technician Operator General worker
the qualitative assessment, the total items retained for this current study are 79 (see Table 2). Table 3 shows the finalised first phase of Human Factors Construct for PHP use after considering all the comments by the appointed experts. According to this table, the construct for interpersonal influence has a total of eight (8) items, perceived susceptibility has seven (7) items, perceived severity has eight (8) items, perceived benefit has eight (8) items, perceived barrier eighteen (18) items, perceived selfefficacy have fifteen (15) items, cues to action have ten (10) items and use of personal hearing protector (PHP) have five (5) items.
Developing a Survey Tool to Measure Human Factors Constructs …
305
Table 2 Changes the number of items in the study construct after qualitative content validation No.
Human factors (construct)
Items (prior expert review)
Source
No of items
Items retained for the current study
Source
No of items
1
Interpersonal influence (II)
10
[13, 22, 23]
8
[13, 22, 23]
2
Perceived susceptibility (PS)
9
[23, 24]
7
[23, 24]
3
Perceived severity (PV)
10
[23–25]
8
[23–25]
4
Perceived benefit (PB)
14
[22–30]
8
[23, 28, 31]
5
Perceived barrier (PR)
20
[13, 22, 23, 25–28]
18
[13, 22, 23, 28, 31]
6
Perceived self-efficacy (PSE)
17
[13, 22, 23, 26, 28, 32]
15
[13, 22, 23, 28, 31, 32]
7
Cues to action (CA)
12
[23, 29]
10
[23, 29]
8
Use of personal hearing protector (PHP)
5
[33, 34]
5
[33, 34]
Total of items
97
79
4 Conclusion The first phase of survey tools for measuring the human factors constructs for PHP use among industrial workers was successfully developed via qualitative content validation by the two (2) experts. However, other studies conducted content validity using a quantitative approach [35]. This study manages to develop eight human factor constructs consisting of 79 items, including interpersonal influence, perceived susceptibility, perceived severity, perceived benefit, perceived barrier, perceived self-efficacy, cues to action and use of personal hearing protector (PHP). The sociodemographic parts are successfully developed, consisting of eighteen (18) items. However, there are some highlights to be noted every time the content validation is planned; (1) the survey tool/questionnaire has to be translated and used in a dual language in English and Malay before it will be given to the targeted group, (2) provide a guide to the respondents on how to fill the survey questionnaire form and (3) involvement of experienced industrial practitioner such as safety officer as
306
N. S. Fauzan et al.
Table 3 Finalise first phase research construct of Human Factors for PHP use Response options
Question (1) Interpersonal Influence II1 II2 II3
My team leader often uses a personal hearing protector when exposed to a noisy workplace
1. Strongly Disagree 2. Disagree My co-workers often use personal hearing protectors when 3. Neither Agree not Disagree exposed to a noisy workplace 4. Agree My co-workers expect me to wear a personal hearing 5. Strongly Agree protector when I am in a noisy work environment
II4
My family members encourage me to use a personal hearing protector when I am in a noisy work environment
II5
My supervisor expects me to wear a personal hearing protector when I am in a noisy work environment regularly
II6
Everyone in this company expects me to wear a personal hearing protector regularly
II7
My co-worker expects me to wear a personal hearing protector every day
II8
My company management encourages me to wear a personal hearing protector every day
(2) Perceived Susceptibility PS1 PS2 PS3 PS4
I believe my chances of developing a hearing loss problem 1. Strongly Disagree are high 2. Disagree 3. Neither Agree nor I worry about getting a hearing loss problem Disagree I know people in this career field who got a hearing loss 4. Agree problem 5. Strongly Agree Small exposures to noise hazards won’t me to a hearing loss problem
PS5
Everybody can get hearing loss problems, including office workers
PS6
I am at risk of a hearing loss problem
PS7
I can have a hearing loss problem even without experiencing any signs or symptoms
(3) Perceived Severity PV1
The thought of getting a hearing loss problem deeply concerns me
PV2
If I developed a hearing loss problem, my career would be in jeopardy
PV3
Problems I would experience from the hearing loss problem would last a long time
PV4
A hearing loss problem will lead to permanent changes in my health
PV5
My financial security would be endangered if I developed a hearing loss problem
1. Strongly Disagree 2. Disagree 3. Neither Agree not Disagree 4. Agree 5. Strongly Agree
(continued)
Developing a Survey Tool to Measure Human Factors Constructs …
307
Table 3 (continued) Response options
Question PV6
I am afraid to even think about getting a hearing loss problem
PV7
There are no drugs to manage hearing loss problems
PV8
Hearing loss complications would endanger my problem
(4) Perceived Benefit PB1
Feeling safe while wearing a personal hearing protector
PB2
Feeling useful while wearing a personal hearing protector
PB3 PB4
1. Strongly Disagree 2. Disagree 3. Neither Agree nor Wearing a personal hearing protector will prevent future Disagree hearing problems for me 4. Agree A personal hearing protector prevents exposure to the noise 5. Strongly Agree hazards I am around on the job
PB5
I benefit from wearing a personal hearing protector
PB6
I think wearing a personal hearing protector every time I am in loud environments is important
PB7
I am convinced I can prevent hearing loss by wearing hearing protectors whenever I am in loud environments
PB8
If I wear a personal hearing protector, I can protect my hearing
(5) Perceived Barrier PR1
Wearing a personal hearing protector is uncomfortable
PR2
I think using a personal hearing protector will slow my speed
PR3 PR4
1. Strongly Disagree 2. Disagree 3. Neither Agree not Disagree A personal hearing protector limits my ability to hear what 4. Agree I want to hear 5. Strongly Agree I think it will be hard to hear warning signals (like backup beeps) if I am wearing hearing protectors
PR5
I don’t feel like wearing a personal hearing protector at the workplace
PR6
I think earmuffs make my head sweat too much
PR7
Personal hearing protectors are uncomfortable to wear
PR8
A personal hearing protector limits my ability to communicate with others
PR9
Wearing a personal hearing protector is annoying
PR10
The size of the personal hearing protector is not fit for me, so I don’t wear it
PR11
I don’t like to wear anything on my ears while performing a job task
PR12
I think a personal hearing protector puts too much pressure on my ears (continued)
308
N. S. Fauzan et al.
Table 3 (continued) Response options
Question PR13
A personal hearing protector interferes with my ability to do my job
PR14
A personal hearing protector is not always available to me
PR15
My co-workers would make fun of me for wearing a personal hearing protector
PR16
I would need to develop a new habit of wearing a personal hearing protector, and that is difficult
PR17
A personal hearing protector is expensive
PR18
There are disadvantages to wearing a personal hearing protector
(6) Perceived Self-efficacy PSE1
If using a personal hearing protector was comfortable, I would definitely use it
PSE2
If a personal hearing protector was easy to obtain, I would definitely use it
PSE3
I know when I should use a hearing protector
PSE4
I can wear a personal hearing protector regularly in a noisy workplace
PSE5
I wear a personal hearing protector regularly, even though my colleagues around me are not in the habit of wearing a personal hearing protector
PSE6
When my personal hearing protector is not functioning, I will inform my supervisor to get a new one for me
PSE7
I can inspect or check for any defects in the personal hearing protector before wearing it
PSE8
I am sure how to tell when a personal hearing protector needs to be replaced
PSE9
I can wear a personal hearing protector properly
PSE10
I can wear a personal hearing protector even if I have to wear other personal protective equipment (PPE)
PSE11
I am confident the usage of a personal hearing protector can reduce the noise exposure to me
PSE12
I am confident that I will remember to use a personal hearing protector when I am exposed to noise hazards
PSE13
I am confident I can obtain the proper personal hearing protector when I am exposed to noisy hazards at work
PSE14
I am confident that my job performance will not be adversely impacted by wearing a personal hearing protector
PSE15
I am confident that after wearing the proper PHP throughout my career will help prevent me from getting a hearing loss issue
1. Strongly Disagree 2. Disagree 3. Neither Agree not Disagree 4. Agree 5. Strongly Agree
(7) Cues to Action (continued)
Developing a Survey Tool to Measure Human Factors Constructs …
309
Table 3 (continued) Response options
Question CA1 CA2 CA3
A reminder from my supervisor every day would be important to wear of personal hearing protector
1. Strongly Disagree 2. Disagree Inspection from my supervisor would improve my wear of 3. Neither Agree not Disagree personal hearing protectors 4. Agree The fact that OSHA fines me or my employer for not 5. Strongly Agree wearing a personal hearing protector is important
CA4
Posters in my workplace would serve as important reminders to wear personal hearing protectors
CA5
The threat of disciplinary action is an important factor in ensuring I wear a personal hearing protector
CA6
Having a personal hearing protector at the location of the hazard is critical to ensure that I wear it
CA7
If I see others wearing personal hearing protectors in my area, then it reminds me to use them
CA8
Regular and frequent education on the importance of personal hearing protectors improves how often I wear them
CA9
My supervisor sets the example of wearing a personal hearing protector when being exposed to hazards
CA10
Training provided by my supervisor about PHP and the importance of personal hearing protectors was helpful
Use of Personal Hearing Protector (PHP) UP1 UP2 UP3
How often do you wear personal hearing protectors during 1. Never the past week when in high-noise areas? 2. Rarely How often do you wear personal hearing protectors during 3. Sometimes 4. Very Often the past month when in high-noise areas? 5. Always How often do you wear personal hearing protectors during the past three months when in high-noise areas?
UP4
How often are you aware of the compliance of wearing a personal hearing protector when working in an excessive noise area?
UP5
How often do you make sure that your personal hearing protector is well-fitted?
an appointed expert panel in reviewing the content of the questionnaire. Before distributing the finalised survey tool (questionnaire), a briefing must be conducted on the targeted respondents to ensure workers’ understanding in answering the questions. The consent form must be given to the respondent to ensure the participants understand and know the study’s purpose before carrying out the sampling process. Therefore, this survey tool can contribute toward an improved understanding of the human factors that may influence the consistent use of PHP among workers working in an excessively noisy work area.
310
N. S. Fauzan et al.
Acknowledgements The authors would like to thank the Universiti Malaysia Pahang, Malaysia, for providing financial support under Internal Research grant RDU190388.
References 1. 2. 3. 4.
5.
6.
7.
8.
9. 10. 11.
12.
13.
14. 15. 16.
17.
18.
WHO (2018) Addressing the rising prevalence of hearing loss, no 02 DOSH (2019) Annual report 2019. Department of Occupational Safety and Health, Putrajaya DOSH (2019) Occupational poisoning and disease statistics 2019, Putrajaya Rasasoran KA, Atil A, Jeffree MS, Saupin S, Lukman KA (2021) Hearing loss and associated factors among noise-exposed workers in palm oil mills. Risk Manag Healthc Policy 14:3653– 3658. https://doi.org/10.2147/RMHP.S319858 Chen Y et al (2019) Prevalence and determinants of noise-induced hearing loss among workers in the automotive industry in China: a pilot study. J Occup Health 61(5):387–397. https://doi. org/10.1002/1348-9585.12066 Fauzan NS, Mirta W, Kee HY (2019) Psychological impact of noise exposure among machine and non- machine operators in construction industry. In: IOP conference series: materials science and engineering, vol 702, no 012055, pp 1–7. https://doi.org/10.1088/1757-899X/702/ 1/012055 Morgan JI, Curcuruto M, Steer M, Bazzoli A (2020) Implementing the theoretical domains framework in occupational safety: development of the safety behaviour change questionnaire, no December. https://doi.org/10.1016/j.ssci.2020.105135 Reddy RK, Welch D, Thorne P, Ameratunga S (2012) Hearing protection use in manufacturing workers: a qualitative study. Noise Heal 14(59):202–209. https://doi.org/10.4103/1463-1741. 99896 Acharya S (2014) Journal of public health open access utilization pattern of personal protective equipment. Heal Prospect J Public Heal 13(Ii):24–28 Arezes PM, Miguel AS (2005) Hearing protection use in industry: the role of risk perception. Saf Sci 43(4):253–267. https://doi.org/10.1016/j.ssci.2005.07.002 Ronis DL, Hong O, Lusk SL (2006) Comparison of the original and revised structures of the health promotion model in predicting construction workers’ use of hearing protection, pp 3–17. https://doi.org/10.1002/nur Nath ND, Behzadan AH, Paal SG (2020) Automation in construction deep learning for site safety: real-time detection of personal protective equipment. Autom Constr 112(July 2019):103085. https://doi.org/10.1016/j.autcon.2020.103085 Lu L, Shi L, Han L, Ling L (2015) Individual and organisational factors associated with the use of personal protective equipment by Chinese migrant workers exposed to organic solvents. Saf Sci 76:168–174. https://doi.org/10.1016/j.ssci.2014.11.025 Rosenstock IM (1974) The health belief model and preventive health behavior. Health Educ Monogr 2(4):354–386. https://doi.org/10.1177/109019817400200405 Pender NJ (1987) Health promotion in nursing practice, 2nd ed. Appleton & Lange, Norwalk, CT Makhilan NI, Fauzan NS, Shariffudin N, Mohd Zin A (2021) Development of safety assessment system for food premises. Int J Ind Manag 11(1):274–286. https://doi.org/10.15282/ijim.11.1. 2021.6117 Presser S et al (2004) Methods for testing and evaluating survey questions university of Maryland university of Michigan U.S. census bureau office for national statistics university of Michigan 68(1):109–130 Bai Y, Li J, Bai Y, Ma W, Yang X, Ma F (2018) Development and validation of a questionnaire to evaluate the factors influencing training transfer among nursing professionals. BMC Health Serv Res 18(1):1–9. https://doi.org/10.1186/s12913-018-2910-7
Developing a Survey Tool to Measure Human Factors Constructs …
311
19. Rosenstock IM (1974) Historical origins of the health belief model. Heal Educ Behav 2(4):328– 335. https://doi.org/10.1177/109019817400200403 20. Vagias WM (2006) Likert-type scale response anchors 21. Taherdoost H (2019) What is the best response scale for survey and questionnaire design; Review of different lengths of Rating Scale/Attitude Scale/Likert Scale. Int J Acad Res Manag 8(1):2296–1747. https://hal.archives-ouvertes.fr/hal-02557308 22. Dehdari T, Rahimi T, Aryaeian N (2014) Developing and testing a measurement tool for assessing predictors of breakfast consumption based on a health promotion model. J Nutr Educ Behav.https://doi.org/10.1016/j.jneb.2013.12.007 23. Wright T, Adhikari A, Yin J, Vogel R, Smallwood S (2019) Issue of compliance with use of personal protective equipment among wastewater workers across the southeast region of the United States, pp. 1–18 24. Gregorio ER, Kobayashi J, Robert J, Medina C, Simbulan NP (2016) Knowledge, attitudes, and related practices of Filipino seafarers regarding cardiovascular diseases, no January. https:// doi.org/10.5603/IMH.2016.0039 25. Hwang JH et al (2017) A survey of parental perception and pattern of action in response to influenza-like illness in their children: including healthcare use and vaccination in Korea, pp 204–211 26. Wu T, Pender N (2002) Determinants of physical activity among Taiwanese adolescents: an application of the health promotion model, 25–36. https://doi.org/10.1002/nur.10021 27. Lu L, Shi L, Han L, Ling L (2015) Individual and organisational factors associated with the use of personal protective equipment by Chinese migrant workers exposed to organic. Saf Sci 76:168–174. https://doi.org/10.1016/j.ssci.2014.11.025 28. Haghi R, Ashouri A, Karimy M, Rouhani-Tonekaboni N, Kasmaei P (2021) The role of correlated factors based on Pender health promotion model in brushing behavior in the 13–16 years old students of Guilan, Iran, pp 1–10 29. Barakat AM (2020) Preventive health behaviours during coronavirus disease 2019 pandemic based on health belief model among Egyptians 30. Polovich M (2010) ScholarWorks @ Georgia State University Nurses’ use of hazardous drug safe handling precautions 31. Svensson EB, Morata TC, Nylén P, Krieg EF, Johnson AC (2004) Beliefs and attitudes among Swedish workers regarding the risk of hearing loss. Int J Audiol 43(10):585–593. https://doi. org/10.1080/14992020400050075 32. Johnson JL, Ratner PA, Bottorff JL (1993) An exploration of Pender’s health promotion model using LISREL. Nurs Res 42(3):132–138 33. Mccullagh M, Lusk MSL, Ronis MDL (2002) A test of the Pender health promotion model 51(1):33–39 34. Nichol K, Bigelow P, O’Brien-Pallas L, McGeer A, Manno M, Holness DL (2008) The individual, environmental, and organisational factors that influence nurses’ use of facial protection to prevent occupational transmission of communicable respiratory illness in acute care hospitals. Am J Infect Control 36(7):481–487. https://doi.org/10.1016/j.ajic.2007.12.004 35. Ishanuddin NM, Sukadarin EH, Nawi NSM, Widia M, Rashid AAA, Aziz HA, Zakaria J, Fauzan NS, Osman H, Roslin EN, Jawi ZM (2021) Design and implementation content validity study: development of an instrument for measuring consumers’ perception of automatic emergency braking (AEB). In: International conference on applied human factors and ergonomics, pp 382–392
A Review on the Pedal Error Cases Among Car Drivers in Malaysia Nursya Mimie Ayuny Ismail, Mohamad Zairi Baharom, Zulkifli Ahmad, Mohd Hasnun Arif Hassan, Juffrizal Karjanto, Zulhaidi Mohd Jawi, and Khairil Anwar Abu Kassim
Abstract Human error has caused about 90% of car accidents. The problem of sudden acceleration while driving is the main focus of this analysis. Every year, it has caused people injuries and deaths on Malaysian roads. This paper aims to observe the mistakes in pedal use reported in Malaysian news. The aim is to analyse why these mistakes happen and what causes them. Words like “tertekan pedal”, “pedal minyak”, pedal error, pedal misapplication, and sudden unintended acceleration are used to find related news articles. The study will analyse 35 pedal error cases between 2016 and 2022. The aspects discussed in the analysis include the age and gender of the drivers, where the accidents happened, and what the drivers were doing before the crashes, injuries, and fatalities. The results show an increasing trend of pedal errors, especially among older drivers. The findings show that 46% of the 35 pedal error crashes occurred in commercial parking lots. Almost half of the 35 crashes, 57%, occurred during parking manoeuvres. Most pedal misapplication accidents (54%) did not hurt the driver. Only 20% of the pedestrians involved in the accidents were not injured, and three died. However, the small number of pedal error cases and the subcategories have limited the analysis. More pedal error situations need to analyse to gain a more in-depth analysis. Keywords Pedal mistake · Pedal error · Unexpected acceleration · Pedal application error · Sudden unintended acceleration N. M. A. Ismail (B) · M. Z. Baharom · Z. Ahmad · M. H. A. Hassan Faculty of Mechanical and Automotive Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, 26600 Pekan, Pahang, Malaysia e-mail: [email protected] M. Z. Baharom e-mail: [email protected] J. Karjanto Faculty of Mechanical Engineering, Universiti Teknikal Malaysia Melaka, Hang Tuah Jaya, Durian Tunggal, 76100 Melaka, Malaysia Z. Mohd Jawi · K. A. Abu Kassim Malaysian Institute of Road Safety Research (MIROS), 43000 Kajang, Selangor, Malaysia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_25
313
314
N. M. A. Ismail et al.
1 Introduction There has been an increase in road traffic accidents worldwide despite improved vehicle safety features [1]. Road traffic accidents result in the deaths of over one million people each year globally, making them the eleventh most common cause of death [2]. According to a previous study, human error is the leading cause of approximately 90% of road traffic accidents [3]. Among the South-East Asian (ASEAN) countries, Malaysia has the highest risk of road fatalities per 100,000 population [4]. Police statistics show that there were 402,626 road accidents in Malaysia between January and September 2022 that killed 4,379 people. The data marks a significant rise compared to the 255,532 road accidents and 3,324 deaths reported during the same period last year [5]. Many factors cause road accidents. Sudden unintended acceleration (SUA) is one of the factors that cause road accidents which leads to injuries and deaths on the road. Several authors suggest that pedal error is crucial in these accidents [6, 7]. Driver rehabilitation specialists (DRSs) explained that a pedal application error is when a driver accidentally presses the accelerator instead of the brake pedal, no matter how fast they are going or if they try to fix it [8]. In a panel discussion, they explained about three types of drivers who misuse their pedals. People in the first group often move their foot between the accelerator and brake pedals because they are unsure which one to press or where to put it. The second type comprises people who make a pedal application mistake but then fix it. The third type includes drivers who cannot correct their errors, no matter how fast or slow they are going [8]. According to Baharom et al.’s [9] research, out of the 321 respondents surveyed, 118 reported experiencing pedal errors during their time as drivers. The most commonly mentioned situations where pedal errors occurred were parking, driving, and hard braking. The respondents reported the highest occurrence of pedal errors while parking (37%), driving (24%), and during hard braking (24%). Furthermore, 9% of those who made pedal errors admitted to causing an accident due to their mistake [9]. The government’s implementation of the Malaysian Road Safety Plan 2014– 2020 (PKJRM 2014–2020) led to an annual decrease of 4.8% in road accidents. The Malaysia Road Safety Plan 2022–2030 (PKJRM 2022–2030) was established to ensure continued progress. This new plan aims to reduce traffic accident fatalities by 50% by the end of 2030 [10]. The study of the literature showed that Malaysia has many pedal error cases. Words such as “tertekan pedal”, “pedal minyak”, “pedal error”, “pedal mistake”, and “sudden unintended acceleration” were some of the search terms used to find linked news stories. Also, the words were used to look for any recent pedal-related incidents in the news. The review looked at different aspects of these accidents, such as the driver’s age, gender, the location of the crash, what the driver was doing right before the collision, injuries, and deaths caused by the accidents.
A Review on the Pedal Error Cases Among Car Drivers in Malaysia
315
2 Aim and Methods This paper looks at pedal mistake cases in Malaysia’s news to determine how often they happened and what caused them. It also lays the groundwork for future research on how to stop and reduce pedal mistakes while driving. The number and trend of pedal error cases in Malaysia are looked at based on the driver’s age, gender, the location of the crash, the driver’s actions before the collision, injuries, and deaths. Figure 1 shows how the screening process works and how the reports are chosen. Malaysian news sites like Utusan Malaysia, Sinar Harian, Berita Harian, Kosmo, and Harian Metro were used to get the information. We thoroughly searched news and media websites for pedal misapplication accidents to start our study. We used words like “pedal error”, “pedal misapplication”, “wrong pedal”, “tertekan pedal”, and “pedal minyak” to find the related articles. We used Malay words as keywords because most news sites in Malaysia are in Malay. Through this process, we found 45 news reports from April 2016 to October 2022 [11–55], of which 40 included quotes from The Royal Malaysia Police (PDRM). The reports were reviewed and grouped based on the driver’s ages and genders. It was noticed that five of the reports did not state how old or what gender the drivers were who made the pedal error. As a result, this paper looks at 35 news reports about pedal mistakes. The media reports were looked at and put into groups based on the age of the driver: teens (20 and under), young adults (21–35), middle-aged adults (36–55), young-old (56–75), and old-old (76 and above). The reports were also sorted by gender, crash location, actions taken before the crash (like driving, reversing, turning), injuries, and deaths. The way the age groups are put together is in line with a study done in the past [8].
3 Analyses 3.1 Pedal Error Cases in Malaysia Reported by News Media from 2016 to 2022 Even though only some people are fully aware, pedal mistakes still need to be solved in Malaysia. Figure 2 shows that the number of pedal mistake reports increases yearly. For example, 11 cases were reported by news sources like Utusan Malaysia, Harian Metro, Sinar Harian, Kosmo, and Berita Harian in 2020, and 12 cases in 2022. Figure 2 shows that pedal misapplication is still a significant and severe problem in Malaysia. Nevertheless, the cases are lower in 2021 than in 2020 and 2022. The Malaysian Movement Control Order (MCO), which was made to stop the spread of COVID-19, could be the reason for this drop. During the MCO, there were probably fewer transport actions because of the rules that were put in place.
316
N. M. A. Ismail et al.
START Pedal error identification through news media searching (n=45) • Berita Harian (n=16) [11-26] . Harian Metro (n=7) [27-33] . Sinar Harian (n=8) [34-41] . Utusan Malaysia (n=4) [42-45] . Kosmo (n=3) [46-48] . Others (n=7) [49-55]
Screening 1: Are the pedal error reports reliable and ` valid? (reports/news confirmed by Malaysia Royal Police (PDRM))
NO
Reports excluded due to no confirmation of police stated (n=5)
YES
Screening 2: Do the pedal error reports `include the driver’s age and gender?
YES Reports will be further investigated and analyzed (n=35) Detail reports analysis Summarize analysis
END Fig. 1 Screening and reports selection process
Reports excluded due NO to no driver’s age and gender stated (n=5) . No age (n=1) . No gender (n=4)
A Review on the Pedal Error Cases Among Car Drivers in Malaysia
317
14
Number of case(s)
12
12
11
10 8 6
6 4 2
1
1
2016
2017
2
2
0 2018
2019
2020
2021
2022
Year Fig. 2 Number of pedal error cases in Malaysia reported in news media from 2016 to 2022 (n = 35)
3.2 Driver Age Figure 3 shows the number of pedal misapplication accidents mentioned in the news, broken down by the drivers’ age. Drivers between the ages of 56 and 75, in the “young-old” age group, have the most pedal misapplication accidents. Based on the 35 cases, it is clear that the number of pedal errors goes up as the driver gets older. This means a driver’s chance of making a pedal mistake increases as they age. A previous study has shown that older drivers are more likely to make mistakes with their feet [56]. Therefore, it seems possible that older drivers are more likely to make mistakes when using the pedals of their cars.
Number of driver(s)
12
10
10 8
7
6
4
4 2
1
1
5 3
2
1
1
0 20 or below
21 to 35
36 to 55
56 to 75
76 and older
Age group Male
Female
Fig. 3 Number of drivers in pedal error crashes reported by the news media (n = 35)
318 Table 1 Male to female driver proportion in pedal application error crashes by age group. (news media, 2016–2022, n = 35)
N. M. A. Ismail et al.
Driver age group
Driver gender Male (%)
Female (%)
190 mm Hg in women” [20]. A significant study on Italian Olympic competitors found that
Study of Anxiety Parameters and Sensors Related to Monitoring …
331
the upper normal range for blood pressure during tests of maximum effort is “220/ 85 mmHg for male athletes and 200/80 mmHg for female athletes” [21].
2.4 Electromyography (EMG) Anxiety disorders (ADs), the most common type of mental illness in the world, affect 264 million people globally [22]. Although clinical symptoms may now be used by psychiatrists to determine if a patient has anxiety, these may be inaccurate or incorrectly stated. While other research using machine learning techniques demonstrates that it is possible to recognize anxiety through physiological analysis [13]. Anxiety levels and physiological parameters are not significantly correlated, according to old psychophysiology studies [23]. Researchers have developed novel technologies to enhance well-being while reducing morbidity, mortality, and healthcare costs because of these difficulties and inconsistencies [24]. Most of the earlier research has been on detecting stress through physiological signs. Numerous studies have employed a variety of physiological markers, including the “electrocardiogram (ECG), electroencephalogram (EEG), galvanic skin response (GSR), electromyogram (EMG), and arterial blood pressure (ABP),” to identify various levels of stress. Some of the earlier research relied on just one signal to determine stress levels [10] “ECG or Heart Rate Variability (HRV)” signals have typically been emphasized as the most accurate stress indicators [25]. Even though it doesn’t get as much attention as the ECG and GSR, the EMG signal is a reliable sign of physiological stress [17]. Since the upper trapezius muscle’s EMG signal has been used as a stress indicator for some time, Most earlier studies have mainly focused on how the signal’s characteristics change under binary stress [26]. The upper trapezius muscle’s EMG signal has been employed as a stress indicator, and most earlier studies have concentrated on examining the variation in each EMG signal’s distinct properties for binary stress detection [9]. With 100%, 97.6%, and 96.2% accuracy, respectively, EMG and ECG signals can accurately diagnose stress levels for two, three, and four degrees of stress. The right trapezius muscle’s EMG signal has also been shown to be more sensitive to stress than other muscles. Together, these findings imply that the EMG signal was just as effective in detecting stress as the well-regarded ECG signal [27].
332
N. K. Kamarudin et al.
3 Result and Discussion 3.1 Research Finding The heart’s parasympathetic activity decreases and its sympathetic activity increases when a person is under mental stress, pushing energy to the muscles. An EMG thus shows increased muscle activity in comparison to a resting state. It has been demonstrated that EMG signals are more reliable and can detect stress just as well as ECG. Several signal recording and processing factors must be considered to accomplish this. By bringing the EMG signal back to normal and eliminating ECG interference, the accuracy of stress detection can be improved. An intriguing finding for archery athletes suggests that less experienced archers have lower accuracy levels and higher heart rates than their more experienced counterparts. This finding suggests that experienced archers may have better control over their arousal levels, resulting in a better balance of the sympathetic and parasympathetic nervous systems. The fact that experienced archers have lower heart rates is evidence in favor of this assertion. Statistical analysis revealed significant heart rate differences between expert and inexperienced archers. Higher sympathetic activity, which is associated with overstimulation, may be harmful to a sport like archery, which requires precise control and moderate levels of arousal. However, it is unclear whether the lower performance and elevated heart rates observed in inexperienced archers are the result of over-arousal or simply a lack of experience. This research highlights the ongoing development of features and methods for monitoring anxiety and arousal levels in athletes. Notably, according to Table 1, approximately 76.7% of the authors believe that heart rate is the primary parameter for detecting anxiety or nervousness in athletes, outweighing the importance of emotional assessment. Heart rate measurements are less difficult to obtain than blood pressure readings because blood pressure detection requires specialized equipment. Finally, the relationship between anxiety, performance, and physiological responses in archery athletes is being studied. According to the findings, experienced archers have better performance, lower heart rates, and better arousal control. Heart rate measurements have surpassed emotional assessment as a prominent tool for anxiety detection. As research advances, the inclusion of advanced monitoring devices, such as Heart Rate and Blood Pressure Readers and EMG sensors, holds promise for improving anxiety monitoring and optimizing performance outcomes in archery athletes.
3.2 Limitation When a person is under mental stress, their parasympathetic activity will decline and their sympathetic activity will rise, which will cause the energy to be transferred to their muscles. According to the results of a thorough study on the detection of
x
o o o o o
3. Associations Between Psychological Inflexibility and Mental Health Problems During the Covid-19 Pandemic: A Three-Level Meta-Analytic Review [3]
4. Exploration of Anxiety Factors Among Students of Distance Learning: A Case Study of Alama Iqbal Open University Exploration of Anxiety Factors Among Students of Dl: A Case Study [4]
5. Anxiety Status of Junior Archers in Covid-19 During Training Isolation Period Towards the Shooting Performance [5]
6. Experiential Avoidance in Depression, Anxiety, Obsessive–Compulsive Related, And Posttraumatic Stress Disorders: A Comprehensive Systematic Review and Meta-Analysis [6]
7. Effects of Psychological Interventions on Competitive Anxiety in Sport: A Meta-Analysis [7]
x o x
x o o o
10. A Study of The Relationship Between Anxiety, Cognitive Emotion Regulation, and Heart Rate Variability in Athletes [10]
11. Part of The Biological and Physical Anthropology [11]
12. Effect of Heart Rate on Shooting Performance in Elite Archers [12]
13. Test Anxiety and Physiological Arousal: A Systematic Review and Meta-Analysis [13]
o o
x
o o
8. Determinants of Anxiety in Elite Athletes: A Systematic Review and Meta-Analysis [8]
9. Trained Athletes and Cognitive Function: A Systematic Review and Meta-Analysis [9]
x
o
o
o
x x
o
1. Effect of Competition Anxiety on Athletes’ Sports Performance: Implication for Coach [1]
Blood pressure (BP)
Parameter
2. Article Anxiety Level in Students of Public Speaking: Causes and Remedies Journal of Education o and Educational Development [2]
Heart
List of paper
o
o
x
x
x
o
x
x
x
x
o
x
x
Electromyography (EMG)
Table 1 Related thirty research studies that used main parameters: heart rate, blood pressure, and electromyography in the detection of anxiety
(continued)
o
x
o
x
x
x
x
x
x
x
x
x
x
Sensor involved
Study of Anxiety Parameters and Sensors Related to Monitoring … 333
o x
o x x o
15. Blood Pressure and Hypertension [15]
16. Global Epidemiology, Health Burden and Effective Interventions for Elevated Blood Pressure and Hypertension [16]
17. Exercise Prescription in The Treatment of Hypertension [17]
o o o o o o
o o o o x
20. Sports Pre-Competitive Anxiety Levels Among Good and Poor Performing Intercollegiate Athletes [20]
21. To What Extent Can We Shorten HRV Analysis in Wearable Sensing? A Case Study on Mental Stress Detection [21]
22. Keep the Stress Away with Soda: Stress Detection and Alleviation System [22]
23. Test Anxiety and Physiological Arousal: A Systematic Review and Meta-Analysis
24. Acceptance and Commitment Therapy Informed Behavioral Health Interventions Delivered by Non-Mental Health Professionals: A Systematic Review [23]
25. Enhanced the Anxiety Monitoring System Among Athletes with IOT for Sports Performance: A x Review [24]
o o
o x
18. Playing Under Pressure: EEG Monitoring of Activation in Professional Tennis Players [18]
19. Prevalence of Masked Hypertension in Untreated and Treated Patients with Office Blood Pressure Below 130/80 Mm Hg [19]
o
o
Blood pressure (BP)
14. Blood Pressure Response and Pulse Arrival Time During Exercise Testing in Well-Trained Individuals [14]
Parameter
Heart
List of paper
Table 1 (continued)
o
x
o
o
o
o
o
o
x
o
x
x
Electromyography (EMG)
(continued)
o
x
o
o
o
o
o
o
x
o
o
x
Sensor involved
334 N. K. Kamarudin et al.
o o x
o x
28. Smart Devices and Wearable Technologies to Detect and Monitor Mental Health Conditions and o Stress: A Systematic Review [26] o o 76.7
27. Exercise Prescription in The Treatment of Hypertension
29. Trained Athletes and Cognitive Function: A Systematic Review and Meta-Analysis
30. Anxiety Performance Among Athletes in Response to Theories and Standard Instruments: A Systematic Review [27]
Percentage
66.7
o
x
Blood pressure (BP)
26. Heart Rate Variability and Pre-Competitive Anxiety According to The Demanding Level of The Match in Female Soccer Athletes [25]
Parameter
Heart
List of paper
Table 1 (continued)
43.0
o
x
x
o
x
Electromyography (EMG)
40.0
x
x
o
x
x
Sensor involved
Study of Anxiety Parameters and Sensors Related to Monitoring … 335
336
N. K. Kamarudin et al.
stress using EMG and ECG signals, the carefully chosen EMG signal was the most accurate stress indicator in earlier studies and could detect stress just as well as ECG. Several signal recording and processing factors must be considered to accomplish this. By removing ECG interference from the EMG signal and normalizing the EMG features using the median Absolute Deviation, stress detection is made more effective (MAD). More research should be done on the use of the stress detection framework using a combination of EMG and ECG signals, though. Papers that used the ECG and EMG signals alone, together, or in combination with other physiological signals should be done to better compare the results of this work with those of previous studies.
3.3 Similarities Despite differing opinions, most papers use similar parameters to identify anxiety or stress using the parameter of heart rate (HR). Blood pressure (BP) and electromyography are two methods (EMG).
4 Conclusion This review article has outlined some of the ways that strength and conditioning specialists may be better able to understand the various types of stressors that athletes experience, the effects these stressors may have on athletic performance, and recommendations for encouraging athletes to develop efficient coping mechanisms to lessen the potential detrimental physiological and psychological effects of stress. It has been said that a device is being created to help athletes identify the areas of their bodies where they feel more stressed or anxious so that they can apply what they learn in training to different situations, like competition. Acknowledgements This research work is strongly supported by the internal grant RDU190399 from Universiti Malaysia Pahang Al-Sultan Abdullah (UMPSA) which provided the research materials and equipment. The authors have no conflicts of interest that are relevant to the content of this review.
References 1. Marwat N, Syed UI, Luqman MS, Manzoor M (Jun 2021) Effect of competition anxiety on athletes sports performance: implication for coach. Humanit Soc Sci Rev 9(3):1460–1464. https://doi.org/10.18510/hssr.2021.93146 2. Raja F (2017) Article anxiety level in students of public speaking: causes and remedies. J Educ Educ Dev
Study of Anxiety Parameters and Sensors Related to Monitoring …
337
3. Yao X, Xu X, Chan KL, Chen S, Assink M, Gao S (01 Jan 2023) Associations between psychological inflexibility and mental health problems during the COVID-19 pandemic: a three-level meta-analytic review. J Affect Disord 320:148–160. https://doi.org/10.1016/j.jad. 2022.09.116 4. Ajmal M, Ahmad S, Scholar P (2019) Exploration of anxiety factors among students of distance learning: a case study of Allama Iqbal open university exploration of anxiety factors among students of DL: a case study of AIOU 68 5. Ahmad WNW, Adib MAHBM, Nashir IM, Txi MRS, Salleh FNM, Tang JR (2023) Anxiety status of junior archers in COVID-19 during training isolation period towards the shooting performance. Int J Publ Health Sci 12(1):361–370. https://doi.org/10.11591/ijphs.v12i1.21982 6. Akbari M, Seydavi M, Hosseini ZS, Krafft J, Levin ME (01 Apr 2022) Experiential avoidance in depression, anxiety, obsessive-compulsive related, and posttraumatic stress disorders: a comprehensive systematic review and meta-analysis. J Context Behav Sci 24:65–78. https:// doi.org/10.1016/j.jcbs.2022.03.007 7. Ong NCH, Chua JHE (01 Jan 2021) Effects of psychological interventions on competitive anxiety in sport: a meta-analysis. Psychol Sport Exerc 52. https://doi.org/10.1016/j.psychs port.2020.101836 8. Rice SM et al (2019) Determinants of anxiety in elite athletes: a systematic review and metaanalysis. Br J Sports Med 53(11):722–730. https://doi.org/10.1136/bjsports-2019-100620 9. Logan NE, Henry DA, Hillman CH, Kramer AF (2022) Trained athletes and cognitive function: a systematic review and meta-analysis. Int J Sport Exerc Psychol. https://doi.org/10.1080/161 2197X.2022.2084764 10. Horvath E, Kovacs MT, Toth D, Toth L (2022) A study of the relationship between anxiety, cognitive emotion regulation and heart rate variability in athletes. J Phys Educ Sport 22(2):528– 534. https://doi.org/10.7752/jpes.2022.02066 11. Theses Dissertations M, Dorshorst T (2019) Part of the biological and physical anthropology commons recommended citation recommended citation Dorshorst, Tabitha. https://doi.org/10. 7275/15119161 12. Açıkada C, Hazır T, Asçı A, Hazır Aytar S, Tınazcı C (2019) Effect of heart rate on shooting performance in elite archers. https://doi.org/10.1016/j.heliyon.2019 13. Roos AL et al (01 Jun 2021) Test anxiety and physiological arousal: a systematic review and meta-analysis. Educ Psychol Rev 33(2):579–618. https://doi.org/10.1007/s10648-020-09543-z 14. Heimark S et al (Jul 2022) Blood pressure response and pulse arrival time during exercise testing in well-trained individuals. Front Physiol 13. https://doi.org/10.3389/fphys.2022.863855 15. DeGuire J, Clarke J, Rouleau K, Roy J, Bushnik T (2019) Blood pressure and hypertension. Health Rep 30(2):14–21. https://doi.org/10.25318/82-003-x201900200002 16. Zhou B, Perel P, Mensah GA, Ezzati M (01 Nov 2021) Global epidemiology, health burden and effective interventions for elevated blood pressure and hypertension. Nat Rev Cardiol 18(11):785–802. https://doi.org/10.1038/s41569-021-00559-8 17. Ichiro Miura S (01 Feb 2023) Exercise prescription in the treatment of hypertension. Hypertens Res 46(2):521–522. https://doi.org/10.1038/s41440-022-01083-z 18. Pineda H (Apr 2022) Playing under pressure: EEG monitoring of activation in professional tennis players. Physiol Behav 247. https://doi.org/10.1016/j.physbeh.2022.113723 19. De La Sierra A et al (12 Jun 2018) Prevalence of masked hypertension in untreated and treated patients with office blood pressure below 130/80 mm Hg. Circulation 137(24):2651–2653. https://doi.org/10.1161/CIRCULATIONAHA.118.034619 20. Hussain F, Shah M, Ali A (Mar 2021) Sports pre-competitive anxiety levels among good and poor performing intercollegiate athletes. sjesr 4(1):515–519. https://doi.org/10.36902/sjesrvol4-iss1-2021(515-519) 21. Castaldo R, Montesinos L, Melillo P, Massaro S, Pecchia L (2017) To what extent can we shorten HRV analysis in wearable sensing? A case study on mental stress detection. In: IFMBE. Springer Verlag, pp 643–646. https://doi.org/10.1007/978-981-10-5122-7_161 22. Akmandor AO, Jha NK (2017) Keep the stress away with SoDA: stress detection and alleviation system. IEEE Trans Multi-Scale Comput Syst 3(4):269–282. https://doi.org/10.1109/TMSCS. 2017.2703613
338
N. K. Kamarudin et al.
23. Arnold T et al (01 Apr 2022) Acceptance and commitment therapy informed behavioral health interventions delivered by non-mental health professionals: a systematic review. J Context Behav Sci 24:185–196. https://doi.org/10.1016/j.jcbs.2022.05.005 24. Wan Ahmad WN, Hisham Mohd Adib MA, Sut Txi MR (Nov 2022) Enhanced the anxiety monitoring system among athletes with IoT for sports performance: a review. J Phys Educ Sport 22(11):2700–2707. https://doi.org/10.7752/jpes.2022.11344 25. Ayuso-Moreno R, Fuentes-García JP, Collado-Mateo D, Villafaina S (Aug 2020) Heart rate variability and pre-competitive anxiety according to the demanding level of the match in female soccer athletes. Physiol Behav 222. https://doi.org/10.1016/j.physbeh.2020.112926 26. Hickey BA et al (02 May 2021) Smart devices and wearable technologies to detect and monitor mental health conditions and stress: a systematic review. Sensors 21(10). https://doi.org/10. 3390/s21103461 27. Wan Ahmad WN, Ghazalli Z, Mohd Adib MAH (2022) Anxiety performance among athletes in response to theories and standard instruments: a systematic review, pp 379–394. https://doi. org/10.1007/978-981-16-4115-2_30
EEG and EMG-Based Multimodal Driver Drowsiness Detection: A CWT and Improved VGG-16 Pipeline Mamunur Rashid, Mahfuzah Mustafa, Norizam Sulaiman, and Md Nahidul Islam
Abstract Monitoring distracted driving is crucial for ensuring road safety and avoiding the economic and social consequences of loss. Most past research has focused solely on EEG or EMG; hence, they are liable for delivering inaccurate results. Integrated methods utilize the advantages of EEG and EMG while eliminating their disadvantages. Therefore, it is preferable to merge as many options as possible to improve the accuracy of sleepiness identification. This article presents a novel electroencephalogram (EEG) and electromyogram (EMG)-based multimodal drowsiness detection system. Initially, an experiment simulating driving was undertaken to gather EEG and EMG data in alert and drowsy states. Continuous wavelet transformation (CWT) has been utilized for EEG and EMG signals, converting timedomain representations to time–frequency representations. A pre-trained VGG16 model has been used to classify the discriminating features. In the architecture of the enhanced VGG16, several convolutional layers have been eliminated, and new layers have been incorporated into the fully connected unit. In addition, the annotated time– frequency images are utilized in the process of fine-tuning the higher levels of the neural network design. The proposed multimodal system detected driver drowsiness with a validation accuracy of 92.50%. Therefore, the proposed EEG and EMG-based multimodal system with CWT and an enhanced VGG-16 pipeline will result in a more reliable method for detecting driver drowsiness. Keywords Driver drowsiness · EEG · EMG · Deep learning (DL) · CNN · VGG 16
M. Rashid (B) · M. Mustafa · N. Sulaiman · M. N. Islam Faculty of Electrical and Electronics Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, 26600 Pekan, Pahang, Malaysia e-mail: [email protected] M. Mustafa e-mail: [email protected] N. Sulaiman e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_27
339
340
M. Rashid et al.
1 Introduction In the past few years, there has been a dramatic surge in road collisions, resulting in catastrophic human life and property losses. Numerous studies indicate that driving when drowsy (fatigued) is a leading contributor to road crashes [1]. Many types of research [1, 2] suggest a significant association between physiological signals and the cognitive state of drivers and that these signals can be utilized to identify driver drowsiness precisely. Among these physiological signals, electroencephalography (EEG), electrooculogram (EOG), Electromyogram (EMG) and Electrocardiogram (ECG) or their combination are widely utilized [2]. Electroencephalography (EEG) is a method of recording the electrical potentials emitted by nerve cells in the cerebral cortex; it has a large amount of sample data with a high temporal precision, which can be interpreted as either physiological or psychological information [3]. Due to EEG’s ability to directly represent the actions of the brain’s nerve cells has been hailed as one of the most effective technologies for revealing the swings in drivers’ mental states while on the road [1]. Multiple computational methods using EEG data have been designed to monitor and analyse driver fatigue [4–6]. An approach for classifying driver tiredness based on electroencephalogram data was presented by Chai et al. [7], which combined sparse-deep belief networks with autoregressive modelling. Zeng et al. [2] created two EEGConv models for classifying mental states to detect driver drowsiness. Several physiological aspects of muscular movements alter and relax throughout the shift from wakefulness to sleep. Consequently, the activity of one’s muscles can be monitored for signs of sleepiness. Electrical muscle stimulation (EMG) is the gold standard for movement detection since it monitors electrical signals related to muscle activation using surface electrodes on the driver’s skin [8]. Electricity is produced along the membrane of a muscle, with each depolarization caused by a contraction. Electrodes can detect and record the action potential’s effects on the electrical potential of the muscular membrane [9]. Mahmoodi et al. [9] used surface EMG signal classification to induce driver fatigue. Five statistical features and a k-nearest neighbour classifier were utilized in their investigation. Wang et al. [10] demonstrated a sleepiness detection method based on electromyography (EMG) and electrical muscle stimulation where the peak factors of the EMG were the determinant of the drowsiness. Previous research may have increased the recognition rate of driver tiredness for a particular modality, but because of individual variances leading to differences in the physiological parameters of different individuals, the drowsiness assessment model’s capacity still faces some challenges. Physiological techniques that combine two or more modalities have the potential to capitalize on the advantages of each while minimizing the drawbacks. In the case of data analysis, recent studies have combined traditional machine learning methods with deep learning to categorize driver fatigue. Higher accuracy with conventional machine learning approaches requires robust feature learning techniques. Due to the automatically efficient feature learning capacity, deep learning algorithms can offer a viable alternative. In light of
EEG and EMG-Based Multimodal Driver Drowsiness Detection …
341
the preceding, the authors of this research combined features extracted from EEG and EMG signals and employed transfer learning to identify the drowsy condition in subjects accurately. Some studies [11, 12] used EEG and EMG-based integrated frameworks to detect drowsiness but didn’t achieve the required classification accuracy for real-life applications. In this article, the time–frequency image of the EEG and EMG signal were extracted and concatenated. Secondly, a pre-trained transfer learning method (i.e., VGG-16) has been designed to classify the drowsiness and alert states. Lower-level weights for the target neural network are provided by the proposed model, which has been pre-trained on large datasets of raw images; higher-level weights are explicitly customized for the driver drowsiness detection job. Finally, the validity of the proposed method was verified by a self-captured dataset. The core contributions of this article are listed below: • EEG and EMG data have been collected from twelve individuals to propose a multimodal drowsiness detection system. • Time–frequency image from EEG and EMG has been extracted using CWT. • A pre-trained transfer learning method (i.e., VGG-16) has been proposed to classify the drowsiness and alert states. The paper is organized as follows: Sect. 2 describes the complete methodology, including the experimental set-up for the driver drowsiness, CWT-based time–frequency imaging process and transfer learning-based classification technique. Section 3 shows the related experimental results, discusses these with other existing models, and demonstrates the efficiency of the proposed method. Finally, Sect. 4 concludes our study.
2 Method and Materials This study has proposed a multimodal driver drowsiness detection approach based on time–frequency representation and a transfer learning-based classification model pipeline. Figure 1 illustrates the complete architecture of the proposed method. The raw signals (EEG and EMG) were initially collected and preprocessed. The clean data has been transformed into time–frequency images using CWT. Then, the time– frequency images of all channels of a single trial were concatenated. Since the EEG and EMG data have been collected from 14 and 2 channels, the concatenated image consists of sixteen images for a single trial. Then, the total trials were split into training and validation datasets. Finally, the CWT images have been employed in the proposed transfer learning-based improved VGG-16 model to identify driver drowsiness.
342
M. Rashid et al.
Fig. 1 Complete methodology of proposed driver drowsiness detection system
2.1 Continuous Wavelet Transform Finding relevant patterns, trends, anomalies and temporal or spatial dependencies could be beneficial before applying any machine-learning technique [13]. The CWT is a method for extracting time–frequency features that provide multi-scale signal refining via scaling and translating operations. Following the data preprocessing stage, the segmented dataset undergoes a CWT transformation from the time domain to the time–frequency domain. The CWT can automatically adapt the time–frequency signal analysis criteria and lucidly describe the signal frequency shift over time. The raw data is transformed into 2-D time–frequency pictures from 1-D time-domain signals in this work using CWT as a feature extraction technique. To construct the wavelet set, the mother wavelet is scaled and translated, which is a family of wavelets ψ(t), shown as ) ( 1 t −τ ψ S,τ (t) = √ ψ S S
(1)
Here, S represents the scale parameter inversely related to frequency, and τ represents the translation parameter. The signal x(t) can be achieved by a complex conjugate convolution operation, mathematically defined as follows [14]: 1 (s, τ ) = = √ s
x(t)ψ
∗
(
) t −τ dt S
(2)
where ψ ∗ (·) denotes the complex conjugate of the above function ψ ∗ (·) and this operation decomposes the signal x(t) in a series of wavelet coefficients, in which the base function is the wavelet family. The S and τ are the parameters in the family wavelets. The signal x(t) is transformed and projected to the time and scale dimensions of the family wavelets.
EEG and EMG-Based Multimodal Driver Drowsiness Detection …
343
2.2 Improved-VGG16 Model One of the best approaches to pattern recognition was introduced by LeCun et al. [15]: convolutional neural networks (CNN). This approach extracts the visual characteristics from the input picture using locally learned filters. The convolution, pooling and fully connected layers make up the internal layer structure of CNN. Figure 2 illustrates typical CNN architecture. The convolutional layers of a CNN are responsible for extracting features from the dataset, while the fully connected layers are responsible for classifying these features. Adding more layers to the network or making it deeper is the most direct way to improve the deep neural network’s feature learning capabilities. Two potential problems may arise from this. The first problem is that the extended network is more susceptible to over-fitting since a more profound or broader model has more parameters. Secondly, there’s the problem of the ever-increasing demand for computing resources. To overcome these concerns, researchers have widely used transfer learning (TL) algorithms [16–18]. But most of the methods are used for natural image classification. Besides, the TL algorithms employ the ‘ImageNet’ weight, which is trained with the raw images, and the provided time images are not comparable; more layers where the weight is updated with the ‘ImageNet’ weight need to be fine-tuned. This procedure facilitates the integration of time–frequency images into the TL architecture. Islam et al. [19] proposed an improved VGG16 model based on the VGG16 transfer learning model to address the concern. Since the ‘ImageNet’ weight and the time–frequency images differ, they used fine-tuning techniques to improve the overall performance. Their study froze some convolutional layers using the ‘ImageNet’ weights. Then, convolutional layers were removed, and new layers were added to the fully connected block. The dropout layer was added after each dense layer to prevent the over-fitting problem. Finally, they used fine-tuning techniques to retrain the non-frozen layers. This technique helps in achieving significant improvement in EEG-based AEP signal classification. For driver drowsiness detection using EEG and EMG signals, this study utilized the improved VGG16 classifier. Figure 3 shows the total parameters used to build the improved VGG16 model.
Fig. 2 Architecture of convolutional neural network
344
M. Rashid et al.
Fig. 3 Parameter of proposed VGG-16 architecture
2.3 Experimental Data Twelve individuals aged 24 and 35 were selected for this study. Participants didn’t mention any neural or mental issues. Everyone said having normal vision and hearing. All contestants have held their driving licenses for a minimum of two years. They were instructed to rest well and refrain from consuming coffee, alcohol and tea 48 h before the trial. Each subject was made aware of the experimental protocol. In this work, EMG data was obtained with the Shimmer Consensys EMG Development Kit, which monitors two channels of non-invasive surface EMG and depicts muscle movement at the monitoring site. An Android app known as Shimmer capture was utilized to record the EMG data. Among two pairs of electrodes, one pair was placed on the lower forearm and another on the upper forearm. The data was collected at the sampling rate of 250 Hz. An Emotiv Epoc with fourteen channels was used to record the EEG data. According to the 10–20 electrode positioning system, the fourteen electrodes of this device were positioned at the locations of AF3, F7, F3, FC5, T7, P7, O1, O2, P8, T8, FC6, F4, F8 and AF4. Figure 4 illustrates the distribution of
EEG and EMG-Based Multimodal Driver Drowsiness Detection …
345
Fig. 4 Electrode positioning map of Emotiv Epoc
electrodes of Emotiv Epoc. The Emotiv Pro software was used to capture the dataset. The sampling frequency of this device is 128 Hz. This study considers the EEG and EMG signals for two different states (alert and drowsy). Figure 5 illustrates the complete data acquisition protocol for EEG and EMG signals. In this protocol, the data of the first 20 min have been considered for alert condition, and the last 20 min have been considered for drowsy state. The first five minutes have been considered for adaptation time in both situations. The interval between alert and drowsy varied depending on the subject. As a result, when a particular subject felt sleepy, they reported it, and that portion of the data was considered for the drowsy state. After data processing, the total number of trials for EEG and EMG is 1200, where 600 trials are for the drowsy state and another 600 for the alert condition. Each trial consists of three seconds. Among 1200 trials, 60% were used for training, and the rest were used to validate the model.
Fig. 5 Data acquisition protocol for alert and drowsiness state
346
M. Rashid et al.
2.4 Performance Evaluation To test the performance of the model, this study has calculated classification accuracy (CA), Precision, Recall, F1_Score and Cohen Kappa score (K) [20–22]. Accuracy =
TP + TN × 100 TP + TN + FP + FN
(3)
TP × 100 TP + FP
(4)
Precision = Recall =
TP × 100 TP + FN
(5)
where TP = true positive, TN = true negative, FP = false positive and FN = false negative. F1 =
1 × Precision × Recall × 100 Precision × Recall
(6)
The formula to calculate Cohen’s kappa for two raters is: K =
Po − Pe 1 − Pe
(7)
where, Po = the relative observed agreement among raters. Pe = the hypothetical probability of chance agreement.
3 Results and Discussions We evaluated the proposed method using five evaluation metrics: classification accuracy, precision, recall, F1-Score and Cohen Kappa score. To compute these metrics, we have extracted the confusion matrix. Figure 6 illustrates the confusion matrix of the testing dataset. We experimented using EEG and EMG data separately, and then the integrated EEG and EMG data were also assessed to check the effectiveness of the proposed multimodal system. The confusion metrics of Fig. 5 illustrate all these three for three modalities. The classification of a single EEG-based drowsiness detection system is 91%, whereas the EMG system achieved 90% accuracy. The multimodal EEG and EMG-based system obtained an accuracy of 92.50%. From this accuracy, it is clear that the multimodal EEG and EMG system outperforms the single EEG or EMG systems. Another point we may interpret from the confusion matrix is that the drossy state of EEG has been recognized more precisely than the
EEG and EMG-Based Multimodal Driver Drowsiness Detection …
347
Fig. 6 Confusion Metrix of the proposed system
alert state. In the case of EEG, only 12 drowsy trials out of 235 drowsy trials are misclassified. On the other hand, the alert state of EMG has been identified more accurately than the drowsy state. Here, only 16 alert trials out of 245 alert trials are wrongly identified. However, the number of misclassified alert and drowsy state trials are significantly lower in the case of EEG and EMG-based multimodal systems. This observation indicates that the multimodal system consisting of EEG and EMG is far better than the single modality-based system. The precision, recall, F1-Score, and Cohen Kappa score of the VGG-16 model for the multimodal system are 0.9270, 0.9191, 0.9231 and 0.8499, respectively. We have also analysed the overall accuracy and loss curve of the proposed VGG16 model for training and validation datasets. Figure 7 illustrates the accuracy curve. From the curve, it is clear that the training accuracy reaches the maximum stable point at a point only in around 25 epochs. Here, the training accuracy is 99.97%. However, the validation accuracy is slightly fluctuating, but it also goes to the highest point in only around 25 epochs. Similarly, Fig. 8 shows the loss curve of the proposed VGG-16 model. Finally, we have compared the outcome of the proposed study with state-of-the-art methods. Table 1 shows the comparison summary of the related studies. According to this table, the proposed multimodal system achieves the highest classification accuracy. The authors in [7] used the autoregressive model to extract the feature Fig. 7 The accuracy curve of the proposed improved VGG-16 model
348
M. Rashid et al.
Fig. 8 The loss curve of the proposed improved VGG-16 model
from EEG and then classified the extracted features using a deep belief network. They achieved a classification accuracy of 90.6%. A single modality (i.e., EMG) based study was conducted by [9], where authors used different statistical features including variance, range, spectral power, shape factor and kurtosis. They classified these features using six machine learning techniques where k-NN achieved the best accuracy of 90%. Very few studies have been conducted to justify the effectiveness of EEG-EMG-based multimodal systems. In [11], authors proposed a hidden Markov model (HMM) to classify the features from EEG and EMG modalities. The average accuracy from 12 subjects was 83.8% which is far lower than our proposed method. Another study based on EEG and EMG in [23] proposed a large margin projected transudative support vector machine (LMPROJ) model to classify the power spectral density feature. The classification accuracy of the method was below 85%. Table 1 Comparison table of related studies References
Modality
Feature
Classifier
Accuracy (%)
[7]
EEG
AR
DBN
90.6
[2]
EEG
Raw data
CNN
91.788
[24]
EEG
PSD
SVM
90.7
[9]
EMG
SF
k-NN
90
[25]
EEG
Entropy
SVM
90.7
[23]
EEG and EMG
PSD
LMPROJ
> Total number of articles (n=236) >>Remove *Duplicates *Other language *Not full article *Follow up, Project report, Book Chapter *Not abide by inclusion criteria *Non-ergonomic workplace
Eligibility
(Core) (n=27) Springer (n=27) WilliePub (n=5) ResearchGate (n=51) SemanticScholar (n=15) ScienceDirect (n=28) Other (n=12) Elsevier (n=71)
Screening
Identification
The prevalent reviews on the effects of occupational ergonomic risk factors on musculoskeletal diseases of the neck and elbow [5]; shoulder [13]; hip and knee [14], and psychological stress [15] have identified the eight generic categories of ergonomic risk factors which are of interest in the occupational spectrum: (i) force exertion and heavy lifting; (ii) demanding posture; (iii) repetitiveness or prolonged activities; (iv) Articles included in review (n=148)
440
K. M. Ruzairi et al.
Table 1 Ergonomic risk factors evaluation method for working conditions Risk factor
Method
Sources
Lifting of heavy loads
National Institute for Occupational Safety and Health lifting equation (NIOSH-Eq)
Waters et al. 1993 [6]
Awkward postures
Ovako Working Analysis System (OWAS)
Karhu et al. 1977 [7]
Rapid Upper Limb Assessment (RULA)
McAtamney and Nigel Corlett 1993 [8]
Rapid Entire Body Assessment (REBA)
Hignett and McAtamney 2000 [9]
Repetitive work
Noise assessment
The Occupational Repetitive Action Occhipinti 1998 [10] Tool (OCRA)) European Assembly Worksheet (EAWS)
Schaub et al. 2013 [11]
Daily Noise Dosage (DND)
Aryanezhad et al. 2009 [12]
whole-body or segmental vibration; (v) temperature extremes; (vi) localised contact stresses; (vi) Hand tools and equipment; and (vii) Personnel Relations. The ergonomic issue arises when workers are exposed to a risk factor or a combination of risk factors. MSD results from a combination of biomechanical and psychosocial risk factors at the workplace [3, 16]. Workers will be affected by emotional disturbance (Prodanovska-Stojcevska et al. 2016), physical strain and sprains, and muscle stress. These will affect mechanical joint stability and muscle activity (Romaiguere et al. 1993), as well as have some auditory effects [17]. Meanwhile, a previous study also discovered that MSD is a symptom of physical and psychological risk factors at the workplace. Biomechanical risk factors occur when a person accommodates psychosocial demands, which leads to a stress response, which can produce muscle stiffness or static muscle loading or build other biological responses [3]. According to Baek et al. [18], there is a correlation between psychosocial risk factors and MSD, as it can refer to various models of psychosocial factors among those from Bongers [19]. It was mentioned in Bongers et al. [19] that when there is a factor of biomechanical exposure to the human body, such as localised contact pressures, this will directly influence problems by changing posture through stress when there is a psychosocial disorder. Furthermore, Moon and Sauter [20] also identified an uninterrupted pathway between work methods, including ergonomics, organisational systems, and physical work surroundings. Psychosocial factors are more concerned with the mental response of job dissatisfaction to a working condition that can lead to physical strain. Psychosocial risk factors may affect workers’ psychological response to their work and influence the risk of low back disorders. For example, mental workload is associated with the risk of low back pain symptoms. At the workplace, low social support, hostile behaviour, conflicting demands, and the addition of family problems are among the causes of the negative psychosocial
A Review of Biomechanical and Psychosocial Risk Factors Among …
441
impact of organisations. This situation may be getting worsen with poor management systems. When the problem with work arises it also will affect worker’s lifestyles such as taking unhealthy substances, job satisfaction, and organisational attribute. The important part is that the psychosocial components have been denoted to be significantly associated with MSD. According to Ariëns et al. [21], psychosocial risk factors also seem to play a significant role in the development of neck pain it has emerged that it has a reciprocal influence. Low decision latitude, low social support, and job dissatisfaction are among the significant predictors of neck pain [21]. Bongers et al. [19] considered two paths of action concerning psychosocial factors on the appearance of MSD: (i) the direct effect of these factors on the biomechanical load of the individual as shown in previous research, among which are roofers working with unique work environments (kneeling and awkward postures) and extreme temperature and other cogent arguments. The exposure to vibration can cause a reduction in heart rate variability (ii) the workers evaluate psychosocial factors as being potential threats for which a solution must be found. It has also been proven by Aptel and Cnockaert [22] that when pressure is felt, it will increase in the muscular tonus following the central nervous system (CNS) activation of the reticulate, which increases the biomechanical load, thus resulting in inflammation of the tendons and tissue deformation. Hales and Bernard [23] found that psychosocial demands may be highly correlated with physical demands in certain situations. This suggests that any association between psychosocial risk factors and MSDs may reflect the relationship between physical risk factors and MSDs.
3.3 Musculoskeletal Disorders (MSD) and Its Effects Based on the literature review, the symptoms of MSD also vary according to the level of risk industrial activity and sociodemographic variables such as age, marital status, level of education, and personal lifestyle [15]. Prolonged exposure may lead to signs and symptoms of the disease until the more severe illness is called a disorder that makes people more vulnerable to injury. Uncontrolled ergonomic risk factors without proper solutions will affect employees and the environment. In addition, the substantial changes in work organisation along with the organisation’s need to incur costs for workers’ compensation claims will also add to the substantial economic burden. Emphasis on the development of MSD in this study, the authors indicate various large groups of MSD components. Pramitasari et al. [24] stated that MSD is the single largest category of workplace injuries and disorders that affect the human body’s movement or musculoskeletal system with different levels of severity (i.e. muscles, tendons, ligaments, nerves, discs, blood vessels). Predominantly, MSD has been shown to cause drowsiness, economic burden monotonous work [13], and muscle tiredness. These results suggest that lack of recognition and work delays contribute to the stresses that facilitate the emergence of MSDs [25].
442
K. M. Ruzairi et al.
Researchers also began to explore psychosocial risk exposure according to age and gender. Collins et al. [26] found notable different effects for neck and shoulder disorders, where males are found to be less affected than females. Another case study over the past year found that prolonged exposure to monotonous, repetitive work with arms outstretched in front of the body or twisted arms was associated with elbow pain in women, but not in men [27]. Elbow pain is also related to biomechanical factors among women subjects because of low support at work and mental resilience [27]. This signifies that women are more likely to suffer from MSD symptoms [26]. However, the emphasis on health aspects such as physical exercise, healthy food intake, and leisure activities proved to contribute to disease prevention. Concerning work related by age, the findings by Collins et al. [26] indicated that with increasing age and prolonged exposure, it is not difficult to understand why this group of elderly people shows significant MSD symptoms.
4 Conclusion Biomechanical risk factors contribute to MSD. With that, it unsurprisingly arises, when combined with psychosocial factors, that results may worsen. Finally, workers‘ experience of biomechanical factors and exposure to psychological factors were significantly associated with MSDs and susceptible to their occurrence. Many wellknown syndromes from negative biomechanical experience and psychosocial risk factors contributing to MSD have been scientifically shown. High job demands and vibration are the most exposure in most of the studies reviewed. Most of these studies were experimental, but there were also self-reported symptoms from observational studies. As discussed in previous studies, there are concerns about which combinations of risk factors would be most affected by MSD variability outcomes. Previous studies also have been done to prove the biomechanical relationship between neck, shoulder, and upper limb symptoms. Studying this relationship with the symptoms of knee disorder (lower limb) is also necessary. There is a high potential for combining risk factors in longitudinal research studies. Acknowledgements The authors would like to thank Universiti Malaysia Pahang (UMP) (www. ump.edu.my) for providing financial support under the Internal Research Grant RDU210334 and the UMP Postgraduate Grant Research Scheme PGRS210353
References 1. European Commission (2006) Directive 2006/42/EC of the european parliament and of the council of 17 May 2006 on machinery, and amending directive 95/16/EC (recast), official journal of the european union, L 157, 9 June 2006
A Review of Biomechanical and Psychosocial Risk Factors Among …
443
2. National Institute for Occupational Safety and Health (NIOSH) (1997) Musculoskeletal disorders and workplace factors: a critical review of epidemiologic evidence for work-related musculoskeletal disorders of the neck, upper extremity, and low back. DHHS (NIOSH) Publication No. 97B141, Cincinnati, OH 3. Jaffar N, Abdul-Tharim AH, Mohd-Kamar IF, Lop NS (2011) A literature review of ergonomics risk factors in construction industry. Procedia Eng 20(2011):89–97. https://doi.org/10.1016/j. proeng.2011.11.142 4. Lop NS, Salleh NM, Zain FMY, Saidin MT (2019) Ergonomic risk factors (ERF) and their association with musculoskeletal disorders (MSDs) among Malaysian construction trade workers: concreters. Int J Acad Res Bus Soc Sci 9(9):1269–1282. https://doi.org/10.6007/IJARBSS/v9i9/6420 5. Caffaro F, Cremasco MM, Preti C, Cavallo E (2016) Ergonomic analysis of the effects of a telehandler’s active suspended cab on whole body vibration level and operator comfort. Int J Indus Ergon 53:19–26. https://doi.org/10.1016/j.ergon.2015.10.009 6. Waters TR, Putz-Anderson V, Garg A, Fine LJ (1993) Revised NIOSH equation for the design and evaluation of manual lifting tasks. Ergonomics 36(7):749–776. https://doi.org/10.1080/ 00140139308967940 7. Karhu O, Kansi P, Kuorinka I (1997) Correcting working postures in industry: a practical method for analysis. Appl Ergon 8(4):199–201. https://doi.org/10.1016/0003-6870(77)901 64-8 8. McAtamney L, Nigel CE (1993) RULA: a survey method for the investigation of work-related upper limb disorders. Appl Ergon 24(2):91–99. https://doi.org/10.1016/0003-6870(93)90080-s 9. Hignett S, Mcatamney L (2000) Rapid entire body assessment (REBA). Appl Ergon 31(2):201– 205 10. Occhipinti E (1998) OCRA: a concise index for the assessment of exposure to repetitive movements of the upper limbs. Ergonomics 41(9):1290–1311. https://doi.org/10.1080/001401398 186315 11. Schaub K, Caragnano G, Britzke B, Bruder R (2013) The European assembly worksheet. Theor Issues Ergon Sci 14(6):616–639. https://doi.org/10.1080/1463922X.2012.678283 12. Aryanezhad MB, Kheirkhah AS, Deljoo V et al (2009) Designing safe job rotation schedules based upon workers’ skills. Int J Adv Manuf Technol 41:193–199. https://doi.org/10.1007/s00 170-008-1446-0 13. Athirah Diyana MY, Karmegam K, Shamsul BMT, Irniza R et al (2019) Risk factors analysis: work-related musculoskeletal disorders among male traffic policemen using high-powered motorcycles. Int J Ind Ergon 74:102863. https://doi.org/10.1016/j.ergon.2019.102863 14. Asadi H, Yu D, Mott JH (2019) Risk factors for musculoskeletal injuries in airline maintenance, repair & overhaul. Int J Ind Ergon 70:107–115. https://doi.org/10.1016/j.ergon.2019.01.008 15. Sekkaya F, Imbeau D, Chinniah Y, Dube P, Marcellis Warin ND et al (2018). J Appl Ergon 72:69–87. https://doi.org/10.1016/j.apergo.2018.05.005 16. Hoogendoorn WE, van Poppel MNM, Bongers PM, Koes BW, Bouter LM (2000) Systematic review of psychosocial factors at work and private life as risk factors for back pain. Spine J 25(16):2114–2125. https://doi.org/10.1097/00007632-200008150-00017 17. Tirabasso A, Botti T, Cerini L, Di Giovanni R, Lunghi A, Marchetti E et al (2015) Human hearing effects by noise and hand-arm vibration on distortion products otoacoustic emissions. In: Proceedings of the 22nd international congress on sound and vibration. International Institute of Acoustics and Vibrations 18. Baek K, Yang S, Lee M, Chung I (2018) The association of workplace psychosocial factors and musculoskeletal pain among Korean emotional laborers. Saf Health Work 9(2):216–223. https://doi.org/10.1016/j.shaw.2017.09.004 19. Bongers PM, De Winter CR, Kompier MA, Hildebrandt VH (1993) Psychosocial factor at work and musculoskeletal disease. Scand J Work Environ Health 19:297–312. https://doi.org/ 10.5271/sjweh.1470 20. Moon SD, Sauter SL (1996) Psychosocial aspects of musculoskeletal disorders in office work. Beyond biomechanics. Taylor and Francis
444
K. M. Ruzairi et al.
21. Ariëns GA, van Mechelen W, Bongers PM, Bouter LM, van der Wal G (2001) Psychosocial risk factors for neck pain: a systematic review. Am J Ind Med 39(2):180–193. https://doi.org/ 10.1002/1097-0274(200102)39:2%3c180::aid-ajim1005%3e3.0.co;2-# 22. Aptel and Cnockaert (2002) Stress and work-related musculoskeletal disorders of upper extremities. TUTB Newsletter N0 19–20 23. Hales TR, Bernard BP (1996) Epidemiology of work-related musculoskeletal disorders. Orthop Clin North Am 27(4):679–709 24. Pramitasari R, Pitaksanurat S, Phajan T, Laohasiriwong W (2015) Association between ergonomics risk factors and work-related musculoskeletal disorders in beverage factory workers, Indonesia. Proceeding international seminar and workshop on public health action. “Building Healthy Community” 25. Lanfranchi J, Duveau A (2008) Explicative models of musculoskeletal disorders (MSD): from biomechanical and psychosocial factors to clinical analysis of ergonomics. Revue Européenne de Psychologie Appliquée 58(4):201–213 26. Collins JD, O’Sullivan LW (2015) Musculoskeletal disorder prevalence and psychosocial risk exposures by age and gender in a cohort of office based employees in two academic institutions. Int J Indus Ergon 46:85–97. https://doi.org/10.1016/j.ergon.2014.12.013 27. Haahr JP, Andersen JH (2003) Physical and psychosocial risk factors for lateral epicondylitis: a population-based case-referent study. Occup Environ Med 2003(60):322–329
Knowledge and Awareness of Road Safety Among University Students Nur Nadhirah Najwa Musni, Wan Norlinda Roshana Mohd Nawi , and Mirta Widia
Abstract The growing field of road safety research is a response to the higher number of traffic accidents. Due to the importance of road safety legislation, knowledge, and awareness in reducing road accidents, especially among young drivers, this study aimed to determine the relationship between university students’ road safety knowledge and awareness. To achieve the purpose of these findings, a study was prepared and delivered to university students using an online questionnaire as the instrument of research. The questionnaire is divided into three sections: demographic information, understanding of road signs and road safety rules, and awareness of road safety. This survey seeks to engage 313 students between the ages of 18 and 26 who are enrolled at the University Malaysia Pahang. The anticipated outcome for section B is that the majority of participants will be able to interpret road signs and will have a greater understanding of road safety. For section C, university students were supposed to know road safety awareness and poor driving behaviors. The result indicated that the relationship between knowledge of road signs and road safety awareness among university students is significant, as indicated (p = 0.012), and that there is also a significant relationship between knowledge of road law regulations and road safety awareness among university students (p = 0.000). This study is anticipated to meet its intended aims of determining the relationship between university students’ road safety knowledge and awareness. Keywords Road Safety · Knowledge · Awareness
N. N. N. Musni · W. N. R. Mohd Nawi (B) · M. Widia Faculty of Industrial Sciences and Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, Kuantan, Malaysia e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_36
445
446
N. N. N. Musni et al.
1 Introduction Road safety refers to the strategies and techniques taken to enhance the safety of road users and prevent deaths. Road accidents were among the most important societal issues in Malaysia. Given that road accidents can occur spontaneously or unintentionally, it appears necessary to comprehend the elements that lead to their occurrence in order to prevent them from becoming an everyday occurrence [1]. Globally, road safety is a serious problem due to the rising number of deaths and injuries caused by vehicle accidents. Personality qualities and views of road users may influence how drivers, particularly adolescents, respond to various traffic scenarios. Inexperienced drivers are among the most dangerous motorists, especially during their first few months of driving alone [2]. The increasing frequency of traffic accidents has increased the need of studying road safety. The Malaysian Institute of Road Safety Research (MIROS) has developed a mortality rates pattern derived from a linear regression average (ARIMA) guide for predicting the number of deaths in Malaysia until 2020 and aiding the outcomes of the launched plan against a goal of reducing vehicle accidents fatalities by 50%. Even though numerous road safety bodies have been identified in government organisations, particular organisations, and civic agencies, and many road safety campaigns have been launched recently, road accidents continue to occur in all places, on regular days or during holiday periods [3]. As a result, there is a need for research into how institutional factors can influence road safety, and this study intends to give a more perspective by analysing the knowledge and awareness of road safety among university students at the University Malaysia Pahang.
2 Methodology This study employed a descriptive research approach by creating an online questionnaire using Google Forms and distributing it across multiple social media networks. About 313 participants between the ages of 18 and 26 years old who are students at the University of Malaysia Pahang participated in this study. All participants are driving either their automobiles or those of another. Participants were asked to complete a series of questionnaires that will serve as the main data source for an online survey built on the Google Forms platform. The questionnaire was adapted from a previous study and adjusted according to the latest driving license book. The questionnaire is divided into three sections to achieve this study’s objectives. The participant’s history is examined in Part A, and Part B tests the participant’s knowledge of traffic laws, regulations, and road signage. Part C intends to evaluate the participant’s attitudes and knowledge regarding road safety. The evaluation of the experience and understanding claims uses closed-ended true or false, yes-or-no, and Likert scale questions, with each subject requiring participants to choose an answer.
Knowledge and Awareness of Road Safety Among University Students
447
Table 1 Demographic results of participants Background of respondents Gender Age
Type driving license
Item
Percentage (%)
Male
129
41.2
Female
184
58.8
18–22
189
60.4
23–24
92
29.4
25 and above
32
10.2
B2 (motorcycle)
34
10.9
D/DA (car)
139
44.4
Full (motorcycle and car)
121
38.7
None Did any of the accident cases occur increasing awareness about road safety on you?
Frequency (N = 313)
19
6.1
Yes
303
96.8
No
10
3.2
3 Results and Discussion 3.1 Demographic Analysis Table 1, shows that there were 184 more female participants than male participants in the study (129 students). Most participants (60.4%) were between the ages of 18 and 22. Second, 29.4% of those between the ages of 23 and 24. However, only a small proportion of the population is over the age of 25. Participants between the ages of 20 and 24 are university students who are not first-year students and are allowed to drive their vehicles there. Out of the 313 participants, 34 had B2 licenses (10.9%), 139 had D/DA licenses (44.4%), 121 had both vehicle and motorcycle licenses (38.7%), and 19 had no licenses (6.1%). The 305 students who took part agreed that the accident raised their awareness of the need for road safety, which must be performed regularly to recognise the value of life.
3.2 Knowledge of Road Safety Road Sign Figure 1 demonstrates the findings of the analysis in road sign knowledge. Following the direction and speed limit ends here were the only two road signs that participants were unable to correctly recognise, with only 58.1 and 83.7% of participants knowing about these signals, respectively. Parking prohibited and stopping prohibited are two road signs that 93 and 92% were able to properly recognise. Although both
448
N. N. N. Musni et al.
Fig. 1 Response to the road sign
are signs with colour and images that do not exactly correspond to the meaning or warning they depict. This is additionally true for the speed limit end here sign, without experience or education, identifying the definition may be challenging. Apart from top-scoring road signs such as “no entry,” (99.4%), road intersection (99%), highway speed limit (96.5%), and obstruction marker (94.9%) where in which the illustration and colour represent the signage indicating it is designed to notify which might conveniently identify and comprehend. Except for top-scoring road signs such as “no entry” (99.4%), “road intersection” (99%), “highway speed limit” (96.5%), and “obstacle marker” (94.9%), where the illustration and colour indicate, the signage is intended to inform those who can readily recognise and comprehend it. Road Law and Regulations The majority of participants have a fundamental awareness of road safety since each question is answered correctly by the majority. For Question 1 “The motorist may turn on the danger lights if he or she is in a hurry and needs to accelerate.” This statement has been properly answered by 90.4% of the participants. Within a double-line road marking, the driver is permitted to pass the vehicle in front of him, as stated in Question 2. 96.8% are aware that passing another vehicle within a double-lined road marking is prohibited. The purpose of road surface markings is to guide and inform both cars and pedestrians. 96.2% are aware that rest stops are not authorised at highway emergency stops. The emergency lane exists solely for emergency purposes, allowing ambulances, fire and rescue, and police to reduce traffic and reach their destination as rapidly as possible, especially during peak hours such as business hours [4]. Helmets are vital for protecting motorcyclists from the direct impact of crashes. 98.1% of respondents concurred with this statement. 97.4% concurred that road signs should be easy to read and effective. 95.5% of responders correctly identify the placement of the letter ‘P’ sticker. Incorrect installation of a kid safety seat may result in fatalities or major events that could have been prevented if the seat had been used and installed properly [5]. Only 69.6% are aware that child safety seats must be placed in the back seat. The number of right responses to this question is low because
Knowledge and Awareness of Road Safety Among University Students
449
the vast majority of college students are single and childless. Therefore, it may be difficult to provide adequate responses to inquiries regarding car seat installation. About 94.9% of participants agree that a yellow traffic signal means the driver is about to stop. The same three procedures were utilised before for traffic-light recognition: colour classification, filtering, and identification. Using a basic data structure, colour categorisation verifies image parts for the red, yellow, and green traffic light colours. Moreover, according to the results, some participants did not recognise that certain behaviours constitute traffic offenses that put their lives and the lives of others at risk, such as participants who questioned the necessity of wearing a helmet. (1.9%), passing in a double lane (3.2%), halting in an emergency lane (3.8%), and speeding on yellow lights to beat red lights (3%). (5.1%). Figure 2 illustrates the information about the percentage of road laws and regulations knowledge among participants. Awareness of Road Safety The parametric test would be selected if the test was normal. If the test was not normal, a non-parametric test would be applied. The alpha level (α) in SPSS was set to 0.05 because it was the most common and was usually used as a reference for conducting normality testing. Because the data set was less than 2000, the sample questionnaires in this study were examined using Shapiro–Wilk’s normality test. In a normal distribution, the significant value must be greater than 0.05 (p > 0.05) which indicates that the hypothesis is accepted. Meantime, if the achieved p-value was less than 0.05 (p < α), the hypothesis would be rejected if the data was not normally distributed. The significant value for the normality test in this result was
Fig. 2 Response to road law and regulations
450
N. N. N. Musni et al.
0.000 as shown in Table 2. As a result, the significant value was less than 0.05 (p < 0.05), and it was classified as a non-normal distribution with an indeed significance. Table 3 provides descriptive statistics on road safety awareness and practices. Almost all the participants chose a scale of 4 or 5 which agreed and strongly agreed for all the questions for awareness excluding questions of “There is a language barrier in the road safety awareness campaign” and “Often reply to text messages while on the road”. The World Health Organization (WHO) also proposes reinforcing the management of institutional road safety and appointing the sense of ownership of governmental organisation to assist state initiatives for road safety. This organisation ensures the efficacy of traffic and transportation policy, such as well-executed strategic road safety planning [6]. For the question “There is a language barrier in the road safety awareness campaign”, the participant had chosen scales 3 or 4 which is either agree or disagree and agree. The communication barriers and diverse cultures are examples of several issues that the organisation is facing in having made its awareness campaign successful [7]. For the question “Often reply to text messages while on the road”, the participant had chosen a scale of 1 or 2 which strongly disagrees and disagrees. Using a smartphone while driving is one of the biggest aspects of this day that does have a significant influence on distracting drivers. While drivers use their smartphones for Facebook or to send messages, they probably have spent between 40 and 60% of their time gazing away from the road, compared to maybe 10% of the time looking down ordinarily [8]. Relationship Between Knowledge and Awareness of Road Safety Table 4 shows the Chi-Square test between mean road awareness and mean knowledge of road signs. In the column on the right of this output table, there is the Asymptotic Significance, also known as the p-value, of the chi-square from running data in SPSS. This value determines the statistical value of the relationship between knowledge of road signs and road awareness among university students. In the tests of significance, if p < 0.05, it could be assumed that these two variables have a statistically significant correlation. The p-value in the Chi-Square output is p = 0.012. This indicates that the relationship between knowledge of road signs and awareness of road safety among university students is significant and the research hypothesis is accepted. Teenage road traffic accidents are increasing for a variety of well-known reasons, including curiosity, reckless driving, and peer pressure [9]. Table 5 shows the Chi-Square test between mean road awareness and mean knowledge of road law regulations. In the column on the right of this output table, there is the Table 2 Test of normality for awareness of road safety Kolmogorov-Smirnova MEANRA a Lilliefors
Shapiro–Wilk
Statistic
df
Sig
Statistic
Df
Sig
0.149
313
0.000
0.829
313
0.000
significance correction
Knowledge and Awareness of Road Safety Among University Students
451
Table 3 Descriptive statistics on road safety awareness and practices N
Mean
Std deviation Skewness
Kurtosis
Statistic Statistic Statistic
Statistic Std. Error Statistic Std. Error
313
4.88
0.398
−4.684
0.138
32.069
0.275
The internet is 313 one of the most suitable media to spread awareness of road safety
4.65
0.563
−1.809
0.138
5.291
0.275
Road safety campaigns were able to increase road safety awareness
313
4.56
0.623
−1.520
0.138
3.465
0.275
There is a language barrier in the road safety awareness campaign
313
3.85
0.998
−0.935
0.138
0.772
0.275
Road safety should be taught as early as primary school
313
4.75
0.556
−2.828
0.138
10.490
0.275
Is easier to look 313 at the screen than voice instructions on map navigation
4.07
0.854
−1.247
0.138
2.421
0.275
Often reply to text messages while on the road
313
1.36
0.821
2.815
0.138
8.088
0.275
Putting on a seatbelt while driving
313
4.75
0.777
−3.654
0.138
13.413
0.275
More awareness of safety with the presence of authority
313
4.55
0.678
−2.060
0.138
6.655
0.275
Road safety awareness is important for road users
(continued)
452
N. N. N. Musni et al.
Table 3 (continued) N
Mean
Std deviation Skewness
Statistic Statistic Statistic The quantity of 313 restriction penalty and fines is enough to lower traffic violations Valid N (listwise)
4.18
0.899
Kurtosis
Statistic Std. Error Statistic Std. Error −1.107
0.138
1.216
0.275
313
Table 4 Road awareness and knowledge of road sign relationship Value
df
Asymp. sig. (2-sided)
Pearson chi-square
106.675a
76
0.012*
Likelihood ratio
91.124
76
0.114
Linear-by-linear association
9.994
1
0.002
N of valid cases
313
a 80
cells (80.0%) have an expected count of less than 5. The minimum expected count is 0.02
Asymptotic Significance, also known as the p-value, of the chi-square from running data in SPSS. This value determines the statistical significance of the relationship between knowledge of road law regulations and road awareness among university students. In the tests of significance, if p < 0.05, it could be concluded that there is a statistically significant relationship between these two variables. The p-value in chisquare output is p = 0.000. This indicates that the relationship between knowledge of road law regulations and awareness of road safety among university students is significant and the research hypothesis is accepted. Table 5 Road awareness and knowledge of road law regulation relationship Value
df
Asymp. sig. (2- sided)
Pearson chi-square
407.118a
76
0.000
Likelihood ratio
82.367
76
0.289
Linear-by-linear association
11.151
1
0.001
N of valid cases
313
a 81
cells (81.0%) have an expected count of less than 5. The minimum expected count is 0.00
Knowledge and Awareness of Road Safety Among University Students
453
4 Conclusion The research concluded that there is a significant relationship between knowledge and awareness of road safety among university students at the University Malaysia Pahang. Road safety must be given much attention, and road safety awareness should be reinforced, because ensuring appropriate road users could indeed save many lives, particularly among young people and inexperienced drivers [1]. As a result, they may practice road safety measures based on their knowledge while driving. Regardless, road traffic accident prevention should be included in the education curriculum, allowing students to learn advanced skills and knowledge, while also preventing road traffic accidents. Acknowledgements The authors would like to thank Universiti Malaysia Pahang Al-Sultan Abdullah (UMPSA), Faculty of Industrial Science & Technology (FIST), for the financial support (RDU210363).
References 1. Abd Rahman R, Mazle HA, Lim WM, Mohd Masirin MI, Hassan MF (2021) Road safety awareness among university students: case study at Malaysia Universiti Tun Hussein Onn Malaysia, Johor. J Civil Eng 2. Ghasemi N, Acerra E, Vignali V, Lantieri C, Simone A, Imine H (2020) Road safety review update by using innovative technologies to investigate driver behaviour. Transp Res Procedia 45(2019):368–375. https://doi.org/10.1016/j.trpro.2020.03.028 3. Kamarudin MKA, Abd Wahab N, Umar R, Shakir Mohd Saudi A, Hafiz Md Saad M, Rosdi NR, Sarah Alisa Abdul Razak N, Murtadha Merzuki M, Salam Abdullah A, Amirah S, Mohd Ridzuan A (2018) Road traffic accident in Malaysia: trends, selected underlying, determinants and status intervention. Int J Eng Technol 7(4.34):112. https://doi.org/10.14419/ijet.v7i4.34.23839 4. Of A, Rules T (2020) View of knowledge and awareness of traffic rules and the impact of traffic exposure on college students. 17(7). https://mail.palarch.nl/index.php/jae/article/view/ 1325/1363 5. Schwebel DC, Johnston A, Rouse J (2017) Teaching infant car seat installation via interactive visual presence: an experimental trial. Traffic Inj Prev 18(2):188–192. https://doi.org/10.1080/ 15389588.2016.1225204 6. Eusofe Z, Evdorides H (2017) Assessment of road safety management at institutional level in Malaysia: a case study. IATSS Res 41(4):172–181. https://doi.org/10.1016/j.iatssr.2017.03.002 7. Choocharukul K, Sriroongvikrai K (2017) Road safety awareness and comprehension of road signs from international tourist’s perspectives: a case study of Thailand. Transp Res Procedia 25:4518–4528. https://doi.org/10.1016/j.trpro.2017.05.348 8. Abd Rahman R, Sakim N, Lim WM, Mohd Masirin MI, Hassan MF (2021) Road safety and traffic injuries due to distracted driving of smartphone usage among university students. J Civil Eng Sci Technol 12(1):46–55. https://doi.org/10.33736/jcest.3343.2021 9. Kulkarni V, Kanchan T, Palanivel C, Papanna MK, Kumar N, Unnikrishnan B (2013) Awareness and practice of road safety measures among undergraduate medical students in a South Indian state. J Forensic Leg Med 20(4):226–229. https://doi.org/10.1016/j.jflm.2012.09.022
Riding Towards Safety: Examining the Patterns of Motorcycle Accidents in Malaysia N. Q. Radzuan, M. H. A. Hassan , M. N. Omar, N. A. Othman, and K. A. Abu Kassim
Abstract The present study aimed to analyse the pattern of motorcycle accidents in Malaysia within five years, from 2015 to 2019. The research objective was to understand the characteristics and trends of motorcycle accidents and injuries to identify areas for improvement in motorcycle safety. Data was collected using traffic accident reports provided by the Royal Malaysian Police. The data was analysed to determine patterns and trends in the number of accidents, types of injuries, victims’ body injury location, the usage of protective headgear, and the location of opponent vehicle damage. The study results showed that motorcycle accidents remain a significant issue in Malaysia, with a relatively consistent number of accidents yearly. Head injuries were the major contributor towards motorcycle fatalities. The analysis also revealed that many fatal victims who suffered from head injuries were wearing helmets which raised doubts about the effectiveness of the protective headgear and the possibility of counterfeit helmets or improper usage. The study suggests establishing a motorcycle safety rating program similar to ASEAN NCAP for passenger cars. This program would raise awareness of motorcycle safety and aid consumers in making informed purchase decisions. Additionally, the study recommends rating the effectiveness of protective gear such as helmets in reducing head injury risk in accidents. The study provides a comprehensive analysis of motorcycle accidents in Malaysia. It highlights the need for continued efforts to improve motorcycle safety through programs and better protective headgear to decrease motorcycle accidents and injuries.
N. Q. Radzuan · M. H. A. Hassan (B) · M. N. Omar Faculty of Mechanical and Automotive Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, 26600 Pekan, Pahang, Malaysia e-mail: [email protected] N. A. Othman Faculty of Electrical and Electronics Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, 26600 Pekan, Pahang, Malaysia K. A. A. Kassim Malaysian Institute of Road Safety Research, 43000 Kajang, Malaysia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_37
455
456
N. Q. Radzuan et al.
Keywords Motorcycle helmet · ASEAN NCAP · Safety rating · Types of injuries · Location of body injuries · Location of opponent vehicle damage
1 Introduction The Global Status Report on Road Safety 2018 by the World Health Organization has highlighted the dire reality that traffic fatalities have emerged as the eminent cause of death, particularly among children and young male adults aged 5–29 years [1]. It calls for attention that among injury-related causes of death, traffic accidents are the only ones included in the list, with health-related issues leading to the list of causes of death. Furthermore, the incidence of transport injuries, specifically traffic injuries, accounted for a staggering 84.8% of all transport injuries in 2017 [2]. Middle to low-income countries continue to struggle with this issue despite the commendable efforts of high-income countries such as the United States of America (USA) and Europe, which have successfully reduced the rate of traffic fatalities to 15.6 and 9.3 deaths per 100,000 people, respectively. Africa and Southeast Asia, in particular, have been grappling with a high rate of traffic fatalities. They currently stand at 26.6 and 20.7 deaths per 100,000 people, respectively. Policymakers are presently engaged in a rigorous process of deliberation to devise solutions aimed at reducing traffic fatalities.
1.1 Malaysian Traffic Fatalities Preliminary data analysis indicates that Malaysia holds a disturbingly high position on the global scale, ranking 19th among the top 182 nations in terms of the staggering number of traffic fatalities per 100,000 population [3]. The nation has recorded an average of 6,652 traffic fatalities per annum between 2005 and 2019, as shown in Fig. 1. Malaysia’s population surged to over 32 million in 2017, with a corresponding increase in registered vehicles. This also has led to an alarming rise in the risk of vehicular accidents [4]. The Department of Statistics Malaysia (2018) has highlighted transport accidents as one of Malaysia’s leading causes of death, ranking among the top five in 2017 [5]. As depicted in Fig. 2, a cursory examination of the data reveals that the trajectory of Malaysia’s traffic fatality index has exhibited a positive trend, with a gradual decrease beginning in the year 2015. However, it is a matter of grave concern for the nation as it currently holds the unfortunate distinction of ranking third among its Southeast Asian counterparts, which is preceded only by Thailand and Vietnam. The alarming rate of traffic fatalities in Southeast Asia has led to the development of the ASEAN New Car Assessment Program (ASEAN NCAP), which is a collaborative initiative of the region’s nations aimed at mitigating the staggering number of vehicular accidents and fatalities through an extensive car safety rating program. The program endeavours
Riding Towards Safety: Examining the Patterns of Motorcycle …
457
Fig. 1 Traffic fatality number in Malaysia from 2005 to 2019
to raise public awareness about the paramount importance of vehicle safety and to incentivise passenger car manufacturers to incorporate advanced safety features in their vehicles. The ultimate goal is to reduce the frequency of accidents and fatalities on the road, therefore making the roads in Southeast Asia safer for everyone. The ASEAN NCAP conducts rigorous crash tests on various passenger car models and evaluates them based on their performance in multiple areas, including adult and child occupant protection, pedestrian protection, and safety assist technologies. The results of these tests are then made available to the public hence providing consumers with valuable information that they can use to make more informed decisions when
Fig. 2 Traffic fatality index among Southeast Asian Countries
458
N. Q. Radzuan et al.
purchasing a safe car [6]. Car safety features such as airbags, seat belts, and antilock brakes are specifically designed to protect occupants in the event of a crash. They can significantly reduce the risk of injury or fatality. The program also tests the vehicle’s ability to protect pedestrians and motorcyclists, which also helps minimise pedestrian and motorcyclist fatalities in accidents [7].
1.2 Motorcyclist Issue in Malaysia The populace of Malaysia has developed an inclination towards the use of standardtype or scooter-type motorcycles with engine capacities ranging between 50 and 1000 cc as a convenient means of transportation to various destinations. This is owing to the affordability of purchasing motorcycles and their low fuel consumption. This mode of transportation is also an optimal choice during traffic congestion, particularly in urban areas. This trend resembles other heavily populated Southeast Asian nations such as Thailand, Vietnam, Indonesia, and the Philippines. Conversely, the riding culture in high-income countries in the USA and Europe is primarily for leisure. The rise of motorcycles on the roads of Malaysia has preceded a dominant representation of motorcycle accidents in the nation’s traffic fatality data, accounting for nearly half of all fatal accidents each year. The issue of motorcycling in Malaysia is distinctive with the term “Mat Rempit” or daredevil riders. These daredevil riders are often involved in illegal street racing and dangerous riding stunts. Most daredevil riders in Malaysia are young Malay males, some as young as 12 years old [8–11]. Furthermore, these riders are believed to possess only learner driving licenses or no licenses [12]. Additionally, young adults aged 16–25 with low income and education levels are likelier to violate traffic rules. This statement is based on research by Borhan and colleagues that focused on the socio-demographics and risk-taking behaviour of motorcyclists at signalised junctions in Sungai Buloh, Selangor [13]. It is relevant to note that the results of this study do not represent the entire population of Malaysian motorcyclists. An observational study conducted from 2011 to 2015 on traumatic motorcyclist injuries admitted to Hospital Sultanah Aminah, Johor, revealed that 90% of the study population wore motorcycle helmets during the initial assessment. Nevertheless, head injuries remained a significant factor in fatality [14].
2 Objective and Methodology The present study aims to understand the characteristics and trends of motorcycle accidents and injuries within five years, from 2015 to 2019. The data utilised in this study was sourced from the Traffic Division of the Royal Malaysia Police (RMP), Bukit Aman, which has been decoded into accident reports under the amendment to traffic laws and regulations known as POLIS 27 [Pindaan 1/09] (POL27). Collecting
Riding Towards Safety: Examining the Patterns of Motorcycle … Table 1 Preliminary data on traffic accidents
459
No
Year
Acquired accident data
1
2015
23,034
2
2016
23,874
3
2017
19,931
4
2018
9,350
5
2019
15,474
Total
91, 663
POL27 accident reports is paramount in understanding Malaysia’s current state of accident patterns. It can be utilised by the Department of Statistics Malaysia to predict future accident data. The data analysed in this study includes information such as the location, date, day, and month of the accident, the total and type of vehicles involved, the total number of drivers, riders, pillion passengers, and pedestrians who were either fatal or injured, the type of accidents, and more. However, this study focuses on a specific subset of data, including the vehicle type, injury type, victims’ body injury location, status of protective headgear, and location of opponent vehicle damage. The number of initial accident data per year is listed in Table 1.
2.1 The Screening Method The data screening process was performed in a sequential manner beginning with the removal of redundant vehicle plate numbers to eliminate any potential bias, as shown in Fig. 3. Subsequently, a statistical analysis of motorcycle accidents was conducted compared to other types of vehicles, including buses, four-wheel-drive vehicles (4 × 4 s), lorries and trailers, passenger cars, taxis, rickshaws, vans, and bicycles. Only vehicles tested in NCAP programs, specifically motorcycles, 4 × 4 s, passenger cars, taxis, and vans, were considered to refine the data further. Also, any single report numbers were removed to focus solely on the interactions between motorcycle accidents and other types of vehicles. Once the data was deemed ready, three types of analyses were performed. The first analysis pertained to the motorcyclist injury type with parameters considered fatal, severe, slight, and no injury. The second analysis focused on the location of injuries sustained by motorcyclists, including the back region, chest, hands, head, hips, legs, multiple locations, and neck. The third analysis examined the site of damage sustained by the opponent vehicle, including the frontal, left side, multiple locations, no damage, rear, right side, and upper side. A further analysis was performed to filter only head injuries to assess whether the victim utilised the correct retention status on protective headgear, particularly for the section of motorcyclist injury location. This procedure was repeated each
460
N. Q. Radzuan et al.
Fig. 3 Research flowchart
year from 2015 until 2019 to obtain a comprehensive understanding of the pattern of motorcycle accidents within the specified time frame.
Riding Towards Safety: Examining the Patterns of Motorcycle …
461
3 Results and Discussion Figure 4 illustrates the percentages of motorcycle involvement in accidents according to year, with a steady increase from 50.01% in 2015 to nearly 52% in 2019. This trend suggests that the number of motorcycle accidents is on the rise. This highlights the need for intervention to address this issue. The percentages of motorcyclist injury type according to the year in Fig. 5 illustrate a consistent trend of fatal and slight injuries accounting for a higher share compared to no harm and severe injury. The data shows that fatal injuries among motorcyclists ranged from 27% in 2015 to 36% in 2018. Furthermore, the percentage of severe injuries among motorcyclists indicated a decrease over the five years, ranging from 22% in 2015 to 17% in 2018. The graph implies a complex pattern of motorcycle accidents: the percentage of motorcyclist fatalities rose from 2015 through 2018 while severe injuries decreased, and percentages of slight injuries remained static within four years. One possible explanation for this trend could be an increase in motorcycle accidents overall, as depicted in Fig. 4, leading to a higher number of fatalities and static percentages of slight injuries, even as the severity of injuries decreases. This could happen because of an increase in the number of motorcyclists on the road. Another possible explanation could be an increased focus on reducing severe injuries among motorcyclists, which decreased the percentage of severe injuries and static percentage of slight injuries. It also happens because the implementation of interventions and safety measures has been improved regarding emergency response protocols, advanced medical treatments, and increased use of protective gear. However, these measures might not be effective in preventing fatalities. The graph in Fig. 6 through Fig. 8 shows a five-year trend of specific focus on victims’ body injury locations and protective headgear usage among motorcyclists.
Fig. 4 Percentages of motorcycles involved in accidents among other vehicle types
462
N. Q. Radzuan et al.
Fig. 5 Percentages of motorcyclist injury type
Figure 6a through Fig. 8a illustrate the highest and second highest data of percentages of injury locations among fatal, severely injured, and slightly injured motorcyclists, respectively. Figure 6a shows that most fatal injuries among motorcyclists occur in the head, with percentages ranging from 53% in 2015 to 33% in 2019. Notably, there was a significant decrease in the percentage of head injuries in 2019. In addition, multiple locations accounted for a considerable percentage of fatal injuries among motorcyclists ranging from 35% in 2015 to 30% in 2019. It is significant to note that ‘multiple locations’ in the data may also include head injuries. Figures 7a and 8a illustrate the percentages of injury locations among severely and slightly injured motorcyclists, respectively. Both figures show that most injuries among these groups occur in the legs, with percentages ranging from 34 to 42% for severely injured motorcyclists and 39% to 41% for slightly injured motorcyclists. The percentage of multiple locations among severely injured motorcyclists ranges from 24 to 29%, and among slightly injured motorcyclists ranges from 25 to 29%. Figures 6b through Fig. 8b illustrate the percentages of protective headgear usage among head injuries of fatal, severely injured, and slightly injured motorcyclists, respectively. Overall, the data shows that most motorcyclists involved in accidents were wearing helmets, with percentages ranging from 67 to 85% across the five years. The percentages of motorcyclists who unbuckled their helmets or no helmets at the time of the accident were relatively low, ranging from 1 to 7% for unbuckled helmets and from 15 to 28% for no helmets. However, it is worth mentioning that the data also shows a small percentage of motorcyclists wearing religious turbans at the time of the accident. Malaysian traffic law permits Sikhs and Muslims to wear religious head coverings for religious reasons replacing helmets, although it was believed it does not protect from head injury [15, 16]. These findings suggest that while protective headgear among motorcyclists is relatively high, there is still room for improvement in ensuring proper helmet usage. The
Riding Towards Safety: Examining the Patterns of Motorcycle …
463
Fig. 6 a Percentages of injury locations among fatal motorcyclists. b Percentages of protective headgear usage among head injuries of the fatal motorcyclists.
Fig. 7 a Percentages of injury locations among severely injured motorcyclists. b Percentages of protective headgear usage among head injuries of the severely injured motorcyclists
high percentage of fatal and severe injuries in the head and legs suggests that there is a need for targeted intervention in these areas, such as education and enforcement of
464
N. Q. Radzuan et al.
Fig. 8 a Percentages of injury locations among slightly injured motorcyclists. b Percentages of protective headgear usage among head injuries of the slightly injured motorcyclists
helmet laws. Furthermore, religious turbans in the data highlight the need for accommodations for religious head coverings that provide equivalent head protection. However, it may be challenging to provide such changes. The data in Fig. 9 shows that most opponent vehicle damage occurred in the frontal area, with percentages from 42% in 2015 to 37% in 2019. This trend suggests that most motorcycle accidents involve a collision with the front of the opponent vehicle. Additionally, the percentage of opponent vehicle damage on the right side ranges from 19% in 2015 to 16% in 2018. In contrast, the percentage of opponent vehicle damage on the left side was 17% in 2017. The data also shows a slight decrease in the percentage of the frontal area from 2015 to 2019, while the percentages of the right side remain relatively consistent. This suggests that there may be a decrease in the frequency of head-on collisions between motorcycles and other vehicles. It is vital to mention that motorcycle accidents remain a significant issue among other vehicles, with fatal injuries and head injuries being the major contributors. Furthermore, the analysis revealed that many fatal victims who suffered from head injuries were wearing helmets. The findings seem in line with the discovery by Tan and colleagues in a 2011 to 2015 case study in Johor that 90% of fatal victims due to head injuries wore helmets [14]. This reserve doubts the effectiveness of the protective headgear and the possibility of counterfeit helmets or improper usage. To address these concerns, it is essential to address concerns about motorcycle safety by implementing a rating program similar to the ASEAN NCAP for passenger cars. The current ASEAN NCAP has positively impacted passenger car manufacturers by encouraging them to include motorcycle safety as part of their rating pillars
Riding Towards Safety: Examining the Patterns of Motorcycle …
465
Fig. 9 Percentages of the location of opponent vehicle damage
[7]. However, a program specifically tailored to motorcycles would raise awareness of the importance of motorcycle safety and provide consumers with valuable information to make informed purchasing decisions. Such a program would help improve motorcycles’ overall safety on the road and reduce the number of accidents and injuries caused by motorcycle use. An ASEAN NCAP for motorcycles could establish a rating system based on several protective parameters. These parameters could include the effectiveness of the motorcycle’s braking system, the level of protection provided by the motorcycle’s frame and bodywork, and the point of the motorcycle’s safety features, such as airbags and anti-lock brakes [17–20]. Moreover, ASEAN NCAP for motorcycles could also rate the effectiveness of the motorcycle’s protective headgear, such as helmets. This is crucial in reducing the risk of head injury in an accident. A rating system for the helmet’s safety performance, durability, and features like quick-release buckle, ventilation, and impact resistance could be included.
4 Conclusion In conclusion, examining motorcycle accidents in Malaysia from 2015 to 2019 reveals several significant findings. Firstly, the percentage of motorcycle accidents experienced a slight increase during this period. Additionally, fatal cases witnessed a substantial rise of 27–36% between 2015 and 2018, with a slight decrease in 2019. Furthermore, head injuries emerged as the leading cause of motorcyclist fatalities, despite most riders wearing helmets. Moreover, leg injuries consistently dominated the statistics for severe and slight injuries. However, when explicitly focusing on head injuries, most motorcyclists wore helmets. Lastly, the investigation highlighted
466
N. Q. Radzuan et al.
frontal vehicle damage as the most common pattern for opponent vehicles crashing into motorcycles, indicating head-on collisions. The data presented in this study highlights the need for continued efforts to improve motorcycle safety in Malaysia, including implementing a safety rating program for motorcycles. This rating system would provide consumers with critical information on various motorcycle models’ safety features and performance. This, too, may enable them to make informed purchasing decisions, ultimately reducing motorcycle accidents and injuries. Acknowledgements The authors would like to acknowledge ASEAN NCAP, FIA Foundation, Global NCAP, OEMs and the Society of Automotive Engineers Malaysia (SAE Malaysia) for funding this research under the ASEAN NCAP Holistic Collaborative Research (ANCHOR IV) grant (UIC221510). Also, the authors are thankful to the Universiti Malaysia Pahang Al-Sultan Abdullah for providing the facilities to conduct the research.
References 1. WHO (2018) Global status report on road safety 2018. Geneva 2. James SL et al (2018) Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: a systematic analysis for the global burden of disease study 2017. Lancet 392(10159):1789–1858 3. Manan M (2014) Factors associated with motorcyclists’ safety at access points along primary roads in Malaysia 4. JKJR (2020) Buku Statistik Keselamatan Jalan Raya 5. Department of Statistics Malaysia (2018) Press release statistics on causes of death, Malaysia 6. Jamaludin AS et al (2021) Malaysian road traffic crash data: where do we stand now. J Mod Manuf Syst Technol 5(2):88–94 7. Abdul Khalid MS et al (2021) A review of motorcycle safety technologies from the motorcycle and passenger car perspectives. J Soc Automot Eng Malaysia 5(3):417–429 8. Wong LP (2011) Socio-demographic and behavioural characteristics of illegal motorcycle street racers in Malaysia. BMC Public Health 11 9. Nurullah AS, Makol-Abdul PR, Rahman SA (2012) Gender and motivations for street racing in Malaysia. J Sociol Res 3(1):67–79 10. Ismail R, Din NC, Lee OL, Ibrahim N, Sukimi F (2015) Role of sensation seeking and aggression on risk riding behaviors among motorcycle street racers in Malaysia. J Soc Sci Human 1:169–179 11. Amit N, Ismail R, Ibrahim N, Said Z, Ghazali SE (2016) Sensation seeking and self-esteem differences among illegal street racers in Malaysia. Mediterr J Soc Sci 7(1):96–102 12. Ismail R, Ibrahim N (2007) Faktor-faktorMempengaruhi Keterlibatan Remaja Dalam Perlumbaan Motorsikal Haram dan Hubungannya Dengan Jenis Personaliti. Sokongan Sosial dan Coping Skill’, Bangi, Selangor 13. Borhan MN, Ibrahim ANH, Aziz A, Yazid MRM (2018) The relationship between the demographic, personal, and social factors of Malaysian motorcyclists and risk taking behavior at signalised intersections. Accid Anal Prev 121:94–100 14. Tan HCL, Tan JH, Mohamad Y, Ariffin AC, Imran AC, Tuan Mat TNA (2019) Clinical characteristics of 1653 injured motorcyclists and factors that predict mortality from motorcycle crashes in Malaysia. Chin J Traumatol English Ed 22(2):69–74 15. UNECE (2016) The United Nations motorcycle helmet study; “How it works and how to join it” series, Geneva
Riding Towards Safety: Examining the Patterns of Motorcycle …
467
16. Oxley J, O’Hern S, Jamaludin A (2018) An observational study of restraint and helmet wearing behaviour in Malaysia. Transp Res Part F Traffic Psychol Behav 56:176–184 17. Ariffin AH, Hamzah A, Solah MS, Paiman NF, Mohd Jawi Z, Md Isa MH (2017) Comparative analysis of motorcycle braking performance in emergency situation. J Soc Automot Eng Malaysia 1(2):137–145 18. Kumaresh G, Lich T, Skiera A, Moennich J (2017) Benefit mapping of anti-lock braking system for motorcycles from India to Indonesia. J Soc Automot Eng Malaysia 1(2):166–178 19. Dinges J, Hoover T (2018) A comparison of motorcycle braking performance with and without anti-lock braking on dry surfaces 20. Teoh ER (2018) Motorcycle crashes potentially preventable by three crash avoidance technologies on passenger vehicles. Traffic Inj Prev 19(5):513–517
Development of Automatic Cervical Brace for Neck Pain Rehabilitation M. Z. Ahmad Fazril, Nur Haizal Mat Yaacob, Norsuhaily Abu Bakar, Mohamad Shaban AlSmadi, and Nasrul Hadi Johari
Abstract Conventional cervical braces are mainly in solid and rigid positioning, manually adjusted to a required neck position. Hence, the braces cannot provide disentangling changes for the home patient. This project attempts to develop an automatic self-balancing cervical brace with a combination of a rigid brace structure and the assistance of an Arduino controller. The prototype new cervical brace was tested for user experience recommendations, and several mechanical design and electronic parts recommendations have been addressed for further improvement. The test also indicates that a proper capacity servo motor and accelerometer are required to maintain the 0˚angle with minimum battery usage for an extended period. This study has contributed to the exploration of new theoretical and practical understanding of cervical brace development, i.e., the integration of automatic control usage parallel with an innovative solution in replacing rigid braces to increase satisfaction among home patients during rehabilitation. Keywords Cervical brace · Self-balancing · Rehabilitation
M. Z. A. Fazril · N. H. M. Yaacob Faculty of Innovative Design and Technology, Universiti Sultan Zainal Abidin, 21300 Terengganu, Malaysia N. A. Bakar (B) Faculty of Applied Social Sciences, Universiti Sultan Zainal Abidin, 21300 Terengganu, Malaysia e-mail: [email protected] M. S. AlSmadi Faculty Education Sciences, Irbid National University, Irbid, Jordan N. H. Johari Centre for Advanced Industrial Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, 26600 Pekan, Pahang, Malaysia Faculty of Mechanical and Automotive Engineering Technology, Universiti Malaysia Pahang Al-Sultan Abdullah, 26600 Pekan, Pahang, Malaysia © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024 M. H. A. Hassan et al. (eds.), Proceedings of the 2nd Human Engineering Symposium, Lecture Notes in Mechanical Engineering, https://doi.org/10.1007/978-981-99-6890-9_38
469
470
M. Z. A. Fazril et al.
1 Introduction The cervical brace is generally made from a metal type brace and attaches to a patient with neck trauma to the skull used to strengthen the neck structures and stabilize the spine. First used in 1967 during the Vietnam War, the cervical collar was used to recover spinal-wound soldiers [1]. The medical staff at that time used the cervical brace as an immobilizer, they used sandbags and placed them on the side of the head to transport patients with cervical spine fractures, but these bags did not cover the column. Pins were screwed into the skin above the eyebrows to hold the brace in place. The halo is attached by metal rods to a plastic vest worn above the chest and back (Fig. 1). A halo vest is the most rigid external immobilizer, especially in the upper cervical spine. Lauweryns (2010) [3] mentioned that there are limits to 75% flexion–extension at C1–C2. After a neck fracture or dislocation, halo braces are widely used. Middendorp and Hosman (2008) [4] reported that occipital condyle fractures, occipitocervical dislocation, C1 (most common), and C2 fractures, with an estimated average healing time of 3–4 months, indicate definitive care Jonas and Anders (2016) [5]. For 8–12 weeks, most individuals wear a halo brace. Albeit numerous cervical collars are industrially accessible, there is no agreement on which offers the best insurance, with contemplates indicating significant varieties in their capacity to limit the cervical scope of movement. Fig. 1 Cervical vest to restrict motion of the upper cervical spine. Figure adapted from Randall and Nicholas (2017) [2]
Development of Automatic Cervical Brace for Neck Pain Rehabilitation
471
Alexandra et al. (2018) [6] reported that the pathophysiological base of pain in the neck is complex. Inflammation of the tissues may be caused by infection, degradation of the disk or joint, immobility, trauma, or psychological stress, and nociception triggered by irritation. Cheng et al. (2013) [7] reported that neck pain musculoskeletal could be initially harmful or nontraumatic. In a clinical trial done by Obelieniene et al. (1999) [8], the traumatic neck pain associated with hyperextension syndrome is most commonly associated with soft tissue injury around the neck. As many as 40% of whiplash injuries are estimated to have long-term consequences. Patients’ physical therapy cervical pain treatments have recently improved thanks to various designs of assistive devices. However, the devices can only offer fixed position procedures of care and exclude dynamic changes and orientation of the head and neck [9]. A brand-new dynamic neck brace that combines the benefits of multiple assistive devices may adjust traction and support in symmetrical or asymmetrical postures. Additionally, it can provide an evaluation of the biomechanical factors relating to head and neck stiffness to diagnose head and neck problems. These contributions offer vital information for the future design of dynamic neck braces and the rehabilitation of neck discomfort patients [9, 10]. This project designed and developed a tool mechanism to help the affected cervical spine home patient overcome the common musculoskeletal condition and stabilize the symptom using Arduino as an open-source arrangement for structure device adventures. The Arduino will assist the system of self-balancing apparatus in an unstable dynamic system and maintain the equilibrium state of cervical spine structures.
2 Methodology The study began by compiling and collecting all relevant information about the cervical brace from multiple sources. A questionnaire for quantitative sampling to identify customer needs about the existing cervical brace was developed and distributed. The expectation of the new design is based on the inclusion criteria, i.e., (1) patients in the care of spine condition, (2) treated cervical spinal column injury patients, (3) patients with collar removal, (4) medical personnel and practitioners, (5) caretakers, ages, and genders. A mockup of a full-size concept model, considered an Alpha prototype for a design review of mechanical functionality, allows a concept to be tested, visualizing how the ideas will be materialized (Fig. 2). The 3D prototype was generated through 3D printing to accurately represent the size, shape, design, and color. The existing cervical brace in the market was identified to be the benchmark model and technically blended with the questionnaire survey result. In designing any physical object, material selection is an essential step to ensure the quality and minimize the costs of the product. All 3D files were converted into STL (stereolithography) file type for fabrication using Autodesk® Fusion 360 CAD software. Then the file was transferred into 3D
472
M. Z. A. Fazril et al.
Head support Chin support
Chest support
Controller parts
Fig. 2 An improvised halo brace for integration test
open-source slicing application printing software known as Ultimaker Cura® 4.8 before being printed using a 3D printer. After that, all 3D printed parts and standard components were assembled into a complete unit of the new cervical brace design to perform support neck function. The self-correcting brace requires a proper electronic platform. Hence, the brace chin support that consists of a connected arm below the jaw level position was set in a moveable position guided by a stepper motor and MPU6050 circuit board equipped with a gyroscope and accelerometer. The accelerometer sends the acceleration forces of X, Y, and Z axes, and will calculate gravitational acceleration along the axes. This will provide precise information about the orientation of the sensor. A micro-step A4988 is a micro-step driver with an integrated translator to control bipolar stepper motors for fast operation. Lithium polymer (Li-Po) batteries with DC supply of 11.7 V and AA batteries were the power source for the battery.
3 Results and Discussion A quantitative survey was conducted among the 57 respondents of targeted medical staff, rehabilitation personnel, and medical students from around Malaysia on the Halo brace innovation requirements. The respondents answered the survey forms and tested the prototype of the new cervical brace. 100% of respondents have agreed that an innovated and upgraded cervical brace should replace the conventional halo brace, with 88% recommending the autonomous control of the brace. The automatic and autonomous control brace is presumed to help patients during recovery instead of manual precautions that are prone to unexpected injury.
Development of Automatic Cervical Brace for Neck Pain Rehabilitation
473
The new cervical brace in Fig. 2 aims to strengthen the high cervical spine’s external and self-balancing control. Protecting the cervical back and neck after surgery or trauma is essential. The invention also intends to keep both adult and young patients with cranium in their heads. The functional features that arise in this new cervical brace are described in Table 1. The new cervical brace was then tested on a full scale with an Arduino controller and motor. The findings of the testing and action taken are listed in Table 2. The tests also recorded design weakness in the placement of the proper Gyroscope MPU-6050 position and failed to maintain a proper 0° base or reference point. It also affects the ability of the chin support to have an appropriate tilt angle for individual settings. The problem was solved by adding MPU-6050 in an adequately secured fitting position. The tilting angle of the mounting in the control of effect on the X-axis orientation is controlled within the testing limit angle of −30° min and max 30º; the gyroscope is effectively utilized to detect and quantify an object’s deviation from its ideal alignment, as well as to maintain that orientation and angular velocity. This alignment angle value can be set and altered depending on the individual’s level of severity control needed by the physician. The set value of angle time response Table 1 Main functional features Description
Part Head support
To support immobilization during the healing period
Chin support
The primary support is to retain the head above the spinal cord instead of driving into the wrong position
Chest support
The chest is the primary support for overall brace functionality
Smooth gearing movement
As user individual setting capability, smooth gearing movement, and single axis tilt control, the function is mainly added features for the easiness of the user
Precise positioning control
Precise positioning control and firm exterior stability of a cervical spine, ease of use, reduced complicationrate, minimized patient conditions, and early patient mobilization are some of the advantages of the cervical brace
Table 2 The full scale new cervical brace testing result Part
Test result/findings
Improvement
Chin support
Inappropriate design: the support is blocking the mouth opening and the jaw interferes when tilting downwards
New design has been proposed. An anthropometric size of chin support platform was included. The chin support is directly connected to the servo motor for automated adjustment based on the jaw position
Head support
Poor neck posture and angle positioning, not flexible. Discomfort surface due to material choice
Flexible and comfort design was proposed together with TPU material
Chest support
Thick and heavy
Reduce the support size and thickness
474
M. Z. A. Fazril et al.
Fig. 3 Final design of cervical brace using automatic self-balancing concept
is set to 50 ms in the mean of the activation signal from MPU-6050 to the response of the servo motor. Servo motor activation angle actuation is open from 0º to 180º, which allows a considerable angle to be set for the servo motor’s allowable angle for this testing purpose. After all the improvements have been made, the final product is ready for the user experience stage and expects further recommendations for the design reliability before the pre-commercial scale (Fig. 3).
4 Conclusion and Recommendations As for the conclusion, the study has proposed an innovation of an automatic selfbalancing cervical support for patients with neck trauma. The design and development process has discovered the importance of user experience in addressing issues on the cervical brace, which is not only on the mechanical part but also the electronic controller functionality. Further recommendations for design reliability are expected before the product can be on the production line.
References 1. Arishita GI, Vayer JJ, Bellamy RF (1989) Cervical spine immobilization of penetrating neck wounds in a hostile environment. J Trauma 29(3):332–337 2. Hlubek RJ, Theodore N (2017) Halo vest immobilization. In: Holly L, Anderson P (eds) Essentials of spinal stabilization. Springer, Cham 3. Lauweryns P (2010) Role of conservative treatment of cervical spine injuries. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deformity Soc Eur Sect Cervical Spine Res Soc 19(1):S23–S26
Development of Automatic Cervical Brace for Neck Pain Rehabilitation
475
4. van Middendorp JJ, Hosman AJ (2008) Evaluation of morbidity, mortality and outcome following cervical spine injuries in elderly patients. Eur Spine J 17(11):1538–1539 5. Walbom J, Kruse A (2016) Traumatic atlantoaxial rotatory dislocation and displaced ossiculum terminale epiphysiolysis treated with a halo device: a case report. Childs Nerv Syst 32:1321– 1325 6. Alexandra I, Taryn Lawler J, Adam R (2018) Chapter 68—neck pain. In: Rakel D (ed) Integrative medicine (4th edn). Elsevier, pp 676–688 7. Cheng JS, McGirt MJ, Devin C (2013) Neck pain.: Kelley’s textbook of rheumatology, 9th edn. In: Firestein GS et al (eds) Part 6-differential diagnosis of regional and diffuse musculoskeletal pain. WB Saunders Company, Philadelphia, pp 625–238 8. Obelieniene D, Schrader H, Bovim G, Miseviciene I, Sand T (1999) Pain after whiplash: a prospective controlled inception cohort study. J Neurol Neurosurg Psychiatry 66(3):279–283 9. Ibrahem ME, El-Wakad MT, El-Mohandes MS, Sami SA (2022) Implementation and evaluation of a dynamic neck brace rehabilitation device prototype. J Healthc Eng. 27:1–12 10. Zhang H, Agrawal SK (2017) Kinematic design of a dynamic brace for measurement of head/ neck motion. IEEE Robot Autom Lett 2(3):1428–1435