Approach to Temporary Structures Monitoring

Home , Hinder

The Pennsylvania State University

The Graduate School

College of Engineering

A CYBER-PHYSICAL SYSTEMS (CPS) APPROACH

TO TEMPORARY STRUCTURES MONITORING

A Dissertation in

Architectural Engineering

Xiao Yuan

Submitted in Partial Fulfillment

of the Requirements

for the Degree of

Doctor of Philosophy

December, 2016

i The dissertation of Xiao Yuan was reviewed and approved* by the following:

Chimay J. Anumba Dean & Professor, College of Design and Construction, University of Florida Adjunct Professor, Department of Architectural Engineering, Pennsylvania State University Dissertation Co-Advisor Co-Chair of Committee

M. Kevin Parfitt Professor, Department of Architectural Engineering, Pennsylvania State University Head of the Department of Architectural Engineering Dissertation Co-Advisor Co-Chair of Committee

Dinghao Wu Assistant Professor, College of Information Science and Technology Pennsylvania State University

Stephen Treado Associate Professor, Department of Architectural Engineering Pennsylvania State University

* Signatures are on file in the Graduate School

ABSTRACT

The construction industry has had a high record of structural failures and safety problems for decades. Some of these problems, such as occupational injuries, quality of structure, and speed of construction project are affected by temporary structures (i.e. scaffolding, temporary support system, and formwork system). In many cases, these temporary structures are regarded as static structures without appropriate monitoring of temporal changes in their stability. To be specific, it is estimated that three quarters of the construction workers in the United States work on or near temporary structures. The improper management of temporary structures results in 100 deaths, 4500 injuries, and costs $90 million every year. While significant deployments are being made in the structural health monitoring of constructed facilities, such as bridges, dams and other civil infrastructure, inadequate attention has been paid to temporary structures.

Developments in information and communications technologies, notably the advent of Cyber-Physical Systems (CPS), are changing the way in which structures are monitored. With the bidirectional coordination possible between physical artefacts and their virtual representations, CPS offers an approach that can facilitate the monitoring and active control of temporary structures in such a way as to prevent structural failures and safety hazards on the job site. This research focused on the development and validation of a Cyber-Physical Systems approach to safety monitoring of the temporary structures widely used on the construction job site. The objectives of the research include investigating CPS implementation in construction industry, assessing CPS applicability to temporary structures, reviewing safety regulations of temporary structures, developing a CPS-based prototype system for temporary structures monitoring, and evaluating the performance of the prototype CPS system.

In order to conduct the research, several research methods were utilized, including literature reviews, rapid prototyping, laboratory experiments, and an evaluation workshop. In particular, research problems were clarified through literature review. In reviewing the key features of CPS, enabling technologies and current applications in various industries were summarized to identify the opportunities for CPS use in tackling important problems in the construction industry. To establish the applicability of CPS in temporary structures, the safety issues of temporary

iii structures were investigated to identify the failure patterns of some commonly used temporary structures. The potential benefits and barriers of CPS application to temporary structures were also explored. Based on the CPS applicability analysis, a CPS prototype system for temporary structures monitoring was designed and developed by linking the virtual model (based on the Navisworks Management platform) and the physical structures. This prototype system was tested through laboratory experiments under 5 failure scenarios and evaluated for performance by industry experts.

The evaluation of the developed CPS prototype system involved key industry experts and confirmed that CPS offers an effective approach to advanced monitoring of temporary structures; the developed CPS prototype system works effectively in the monitoring of temporary structures; and there are a number of benefits and barriers associated with CPS application to temporary structures. In addition to the safety monitoring of scaffolding systems, which was demonstrated as an example of temporary structures, various potential application areas have been identified for CPS application to temporary structures monitoring. While the proposed TSM method has provided an example of how to closely integrate the virtual model and the physical components of temporary structures for structural monitoring, more applications and extensible benefits can be realized based on it.

In general, this research explored and validated CPS application for the monitoring of temporary structures. It identified CPS benefits and applicability to the area of temporary structures monitoring, developed a CPS prototype for the monitoring of temporary structures, conducted experimental tests to check and refine the prototype system, and conducted workshop evaluation of the prototype system for performance analysis. By taking advantage of CPS, this research provides a new approach to assisting safety superintendents in the monitoring of temporary structures through bi-directional communication between physical structures and their virtual models. This research has contributed to a deeper understanding of the potential of Cyber-Physical Systems (CPS) to address critical problems in the construction industry, using temporary structures monitoring as an effective example application.

TABLE OF CONTENTS

LIST OF FIGURES ...... viii LIST OF TABLES ...... x ACKNOWLEDGEMENTS ...... xi

Chapter 1: Introduction ...... 1 1.1 Background ...... 1 1.2 Research Overview ...... 3 1.2.1 Research Aim and Objectives ...... 3 1.2.2 Expected Contributions ...... 3 1.3 Thesis Structure ...... 4 1.4 Summary ...... 6 Chapter 2: Research Methodologies ...... 7 2.1 Overview of Research Methodologies ...... 7 2.1.1 Qualitative Research Methodologies ...... 7 2.1.2 Quantitative Research Methodologies ...... 8 2.1.3 Hybrid Research Methodology ...... 9 2.2 Methodologies in Information Systems Research ...... 10 2.3 Research Methods Adopted and Justification ...... 11 2.3.1 Literature Review ...... 11 2.3.2 Rapid Prototyping ...... 12 2.3.3 Laboratory and Field Experiment ...... 12 2.3.4 Evaluation Workshop ...... 13 2.4 Summary ...... 13 Chapter 3: Applicability of CPS to Safety Monitoring of ...... 14 Temporary Structures ...... 14 3.1 Introduction to CPS ...... 14 3.1.1 Key definitions of CPS ...... 14 3.1.2 Key features of CPS ...... 15 3.2 CPS Applications ...... 16 3.2.1 CPS Applications in Other Industries ...... 16 3.2.2 CPS Applications in the Built Environment ...... 20 3.3 Overview of Temporary Structures ...... 22 3.3.1 Common Types of Temporary Structures ...... 22 3.3.1.1 Earthwork Shoring/ Sheeting System ...... 23 3.3.1.2 Temporary Bracing System ...... 25 3.3.1.3 Underpinning of Foundations ...... 27 3.3.1.4 Scaffolding System ...... 29 3.3.1.5 Formwork ...... 31

v 3.3.1.6 Temporary Performance Stages ...... 33 3.4 Prevention of Temporary Structural Failures ...... 34 3.4.1 Regulations & Standards ...... 35 3.4.2 Recommended Practices ...... 36 3.4.3 Education ...... 37 3.4.4 Limitations of Conventional Methods ...... 37 3.5 CPS applicability to temporary structures monitoring ...... 38 3.5.1 Use of BIM for temporary structures management ...... 38 3.5.2 Use of Data Acquisition System (DAQ) for temporary structures management ...... 38 3.6 Potential Areas for CPS Applications in Temporary Structures ...... 39 3.6.1 Temporary performance stages ...... 39 3.6.2 Temporary support systems ...... 40 3.7 Potential Benefits and Barriers ...... 41 3.7.1 Potential benefits ...... 41 3.7.2 Potential barriers to CPS implementation ...... 42 3.8 Summary ...... 43 Chapter 4 Prototype System Development ...... 44 4.1 System Development Framework ...... 44 4.2 Framework of CPS-based Scaffold Systems Monitoring...... 47 4.2.1 Identify the Key Information to Check ...... 47 4.2.2 Establish Warning Threshold ...... 48 4.2.3 Tie Physical Scaffold to its Corresponding Virtual Model ...... 48 4.2.4 Transfer Information from Physical Scaffold to Virtual Model ...... 49 4.2.5 Potential Structural Hazards Prediction in the Virtual Model...... 49 4.2.6 Visualization of Structural Deficiency ...... 50 4.2.7 Remote Monitoring of Scaffold Systems through Virtual Model ...... 50 4.3 User Needs Identification ...... 51 4.3.1 End User Requirements Identification ...... 51 4.3.2 System Requirements ...... 52 4.4 Choice of Development Environment ...... 54 4.4.1 Hardware Environment ...... 54 4.4.2 Software Environment ...... 55 4.5 Overview of System Architecture ...... 57 4.6 Development of TSM ...... 58 4.6.1 3D Model and Plug-in Development...... 58 4.6.2 Database Development ...... 59 4.6.3 Mobile App for Human-Machine Interaction ...... 60 4.7 Summary ...... 61 Chapter 5 Physical Experiments with Temporary Structures...... 62 5.1 Experimental Set-up ...... 62 5.1.1 Set-up of Scaffolding System ...... 62 5.1.2 Set-up of Sensors ...... 63 5.1.3 Set-up of DAQ System ...... 70 5.2 Experimental Scenarios Design ...... 74 5.2.1 Failure Pattern of Scaffolding System ...... 75 5.2.2 Experimental Design ...... 77 5.3 Experimental Tests and Results ...... 81

vi 5.3.1 Experiment 1: Overloading of Scaffolding System at Post 3 ...... 81 5.3.2 Experiment 2: Base Settlement of Scaffolding System at Post 1 ...... 83 5.3.3 Experiment 3: Lateral Load of Scaffolding System at Post 1 ...... 84 5.3.4 Experiment 4: Displacement of Scaffolding System at Plank 3 ...... 85 5.3.5 Experiment 5: Lack of Diagonal Brace at Post 3 ...... 86 5.4 Summary ...... 86 Chapter 6 Prototype System Operation ...... 87 6.1 Linking Physical Structures to Prototype System ...... 87 6.2 System Work Flow ...... 90 6.3 User-interface of Prototype System ...... 91 6.4 Integration of Temporary Structure Performance with Prototype System ...... 95 6.4.1 Integration of Physical Structures and Their Virtual Models ...... 95 6.4.2 Experimental Test of TSM System ...... 97 6.5 Summary ...... 113 Chapter 7 Prototype System Evaluation ...... 114 7.1 Evaluation Approach ...... 114 7.1.1 Evaluation Objectives and Approach ...... 114 7.1.2 Selection of Participants ...... 115 7.1.3 Questionnaire Design ...... 115 7.2 Conduct of Evaluations and Outcomes ...... 116 7.2.1 Evaluation Process ...... 116 7.2.2 Evaluation Outcomes ...... 118 7.3 Benefits of TSM Prototype System ...... 125 7.4 Limitations of TSM Prototype System ...... 126 7.5 Summary ...... 127 Chapter 8 Discussions and Conclusions ...... 129 8.1 Summary ...... 129 8.1.1 Summary ...... 129 8.1.2 Tasks and Activities ...... 129 8.2 Conclusions ...... 132 8.3 Contributions ...... 134 8.4 Research Limitations ...... 135 8.5 Practical Deployment Considerations...... 136 8.5.1 Limitations of Laboratory Experiments ...... 136 8.5.2 Potential Opportunities for Field Applications of TSM System ...... 138 8.6 Recommendations for future Research ...... 139 8.7 Concluding Remarks ...... 141 REFERENCES ...... 143 APPENDIX A ...... 153

vii LIST OF FIGURES

Figure 4-1: Possible System Architecture of CPS-based Temporary Structures Monitoring...... 44 Figure 4-2: Framework of CPS-based Scaffold Systems Monitoring ...... 47 Figure 4-3: Virtual Model and Physical Set-up of Frame Scaffold ...... 49 Figure 4-4: System architecture of CPS-based TSM ...... 58 Figure 4-5: Revit Model of Frame Scaffold ...... 59 Figure 4-6: 3D Model of Scaffold in Navisworks...... 59 Figure 4-7: Mobile App of TSM ...... 60 Figure 5-1: Set-up of Frame Scaffold System ...... 62 Figure 5-2: Load Cell ...... 63 Figure 5-3: Switch Sensor ...... 64 Figure 5-4: Displacement Sensor ...... 64 Figure 5-5: Bi-directional Inclinometer ...... 65 Figure 5-6: Calibration of Load Cells ...... 66 Figure 5-7: Calibration of Displacement Sensors ...... 67 Figure 5-8: Calibration of Inclinometers ...... 68 Figure 5-9: Sensor Layout for Experimental Test ...... 69 Figure 5-10: Supporter of Displacement Sensor under Scaffolding Plank ...... 70 Figure 5-11: Set-up of DAQ hardware ...... 71 Figure 5-12: Virtual Channel for Switch Sensors at DAQ System ...... 72 Figure 5-13: Virtual Channel for Analog Signals at DAQ System ...... 73 Figure 5-14: Experimental Scenario of Overloading at Post 3 ...... 78 Figure 5-15: Experimental Scenario of Base Settlement at Post 1 ...... 79 Figure 5-16: Experimental Scenario of Lateral Load at Post 1 ...... 80 Figure 5-17: Experimental Scenario of Displacement of Scaffold Plank 3 ...... 80 Figure 5-18: Experimental Scenario of Lack of Diagonal Brace at Post 3 ...... 81 Figure 5-19: Initial Test of Overloading at Post 3 ...... 82 Figure 5-20: Initial Test of Base Settlement at Post 1 ...... 83 Figure 5-21: Initial Test of Lateral Load at Post 1 ...... 84 Figure 5-22: Initial Test of Displacement of Plank 3 ...... 85 Figure 6-1: Structure of On-cloud Database ...... 89 Figure 6-2: Database for Base Settlement at Initial Stage Viewed from Heidi SQL ...... 90 Figure 6-3: System Workflow ...... 91 Figure 6-4: User Interface of CPS-based TSM ...... 93 Figure 6-5: User Interface of TSM at Initial Stage of Base Settlement Test ...... 94 Figure 6-6: Warning Message on Portable TSM Interface ...... 94 Figure 6-7: User Interface of TSM at Second Stage of Base Settlement Test...... 95 Figure 6-8: Physical-to-Cyber Bridge ...... 96 Figure 6-9: Cyber-to-Physical Bridge ...... 96 Figure 6-10: Process of Experimental Test 1 ...... 99 Figure 6-11: Layout of Experimental 1 ...... 100 Figure 6-12: Warning of Potential Hazard – Experiment 1 ...... 100 Figure 6-13: Warning Information of Experiment 1 at Portable Device ...... 101 Figure 6-14: Process of Experimental Test 2 ...... 102 Figure 6-15: Layout of Experimental Test 2 ...... 102 Figure 6-16: In progress - Experimental Test 2...... 103 Figure 6-17: Warning of Potential Hazard – Experiment 2 ...... 103 Figure 6-18: Warning Information of Experiment 2 at Portable Device ...... 104

viii Figure 6-19: Process of Experimental Test 3 ...... 105 Figure 6-20: Layout of Experimental Test 3 ...... 106 Figure 6-21: Warning of Potential Hazard – Experiment 3 ...... 106 Figure 6-22: Warning Information of Experiment 3 at Portable Device ...... 107 Figure 6-23: Process of Experimental Test 4 ...... 108 Figure 6-24: Layout of Experimental Test 4 ...... 109 Figure 6-25: Warning of Potential Hazard – Experiment 4 ...... 109 Figure 6-26: Warning Information of Experiment 4 at Portable Device ...... 110 Figure 6-27: Process of Experimental Test 4 ...... 111 Figure 6-28: Layout of Experimental Test 5 ...... 111 Figure 6-29: Warning of Potential Hazard – Experiment 5 ...... 112 Figure 6-30: Warning Information of Experiment 5 at Portable Device ...... 112 Figure 7-1: Process of Evaluation Work Shop ...... 118 Figure 7-2: Usefulness of TSM System ...... 118 Figure 7-3: Likeness to Embrace TSM ...... 119 Figure 7-4: Reality of Five Failure Scenarios ...... 120 Figure 7-5: Accuracy of TSM System ...... 120 Figure 7-6: Performance of Linkage between Physical Structures and Virtual Model ...... 121 Figure 7-7: Importance of Linkage between Physical Structures and Virtual Model ...... 121 Figure 7-8: Adequacy of Warning Alarm ...... 122 Figure 7-9: TSM Performance Evaluation ...... 123 Figure 7-10: Importance of Warning Notification ...... 123 Figure 7-11: Trust Level of TSM by Engineering Experts ...... 124 Figure 7-12: Ease of Learning TSM ...... 125

ix LIST OF TABLES

Table 3-1: Key Features of CPS ...... 16 Table 3-2: Fatal Occupational Injuries Due to Cave-in, From 1992 to 2010 (BLS, 2007, 2012) ...... 23 Table 5-1: Key Components of Frame Scaffold System ...... 63 Table 5-2: Fatal Occupational Injuries Related to Scaffolding, of Year 2003-2011 ...... 74

x ACKNOWLEDGEMENTS

I would love to take this opportunity to thank my advisor, Dr. Chimay J. Anumba, for all his support and encouragement to me for the past four and half years. This research would be impossible without his instruction and guidance. The countless discussion with him equipped me with a solid foundation to the research project, which also guided me as a light during the struggling moments. I cherish the experience working at his supervision as he showed me the importance of being curious in exploring, being persistent in goals, and being considerable to others.

I would like to express my gratitude to my co-advisor, Prof. M. Kevin Parfitt. His patience and support provided me with strong motivation in tackling the obstacles in learning the required knowledge. The discussion with him enlightened me on various aspects of the research, including the exploration of building failures and the design of experimental tests. I would also like to thank the other two members of the committee, Dr. Dinghao Wu and Dr. Stephen Treado, for their valuable suggestions in keeping the research conducted in the appropriate direction.

I am thankful for Mr. Paul Kremer and Mr. Corey Wilkinson for their help in dealing with the technical issues for the experimental set up and test. I would like to thank Mr. John Bechtel, Mr. Don Fronk, and the other staffs at the Office of Physical Plant for offering the prototype system evaluation. This research also got help from my friends, Zhen Xie, Dengfeng Li, Chong Zhou, and Zhuoqi Yang, who offered me valuable discussion in improving the system design. Throughout my doctoral program, I was lucky to receive the warm support from my friends, including Bo Gu, Zhiwang Zhang, Yifan Liu, Fuju Wu, Lee Jiang, Amanda Webb, Yewande Abraham, Yan Chen, Fangxiao Liu, Miaomiao Niu, Bo Lin, and others.

Most importantly, I would love to thank my family for their unreserved support to my study and my life goal. I cherish the warm love and support from my husband, Weizhuo Sun, who gave me the happiness and the strength to work towards my goal. Thank my parents, Ms. Li Chun and Mr. Xiaoming Yuan for their love and support to me. They taught me the meaning of love, brought me up in the happy family with countless sweet memories and always stand by myside without hesitation. All their love is the essential foundation to my research, as well as my life.

xi Chapter 1: Introduction

1.1 Background

Cyber-Physical Systems, commonly known as CPS, can be termed as the effective bidirectional integration of computational resources with physical processes, components and systems. Embedded computers and networks monitor and control the physical processes with feedback loops, while physical processes affect computations and vice versa (Derler et al. 2012). By definition, a CPS involves a high degree of integration between computing (virtual) and physical systems (Wu and Li 2011), which is supported by the networked implementation of CPS (Anumba et al. 2010). Distributed applications are also common which involve distributed management and/or distributed operations such as a power grid. Other features of CPS include the ability to provide timely service in the face of real-time constraints (Wan and Alagar 2012), to adapt to changing situations through dynamic reorganizing/reconfiguration (Shi et al. 2011), to automatically control a physical system based on continuous tracking (Wan et al. 2011), and to integrate several different communication systems and devices (Wan et al. 2011).

As indicated by the key features, CPS offers a potential solution to addressing emerging problems, and has been implemented in several industry sectors. In the manufacturing industry, CPS has been deployed to help manage dynamic changes in production (Kaihara and Yao 2012). Relative to the power grid, smart grid technology is being developed using CPS applications (Krogh et al. 2008). CPS have also been implemented in the transportation industry to promote the development of intelligent traffic systems (Gong and Li 2013). The healthcare industry is increasingly relying on CPS for networked medical systems and health information networks (Shi et al. 2011). These initial attempts of CPS applications in the industry sectors mentioned above have given rise to the recognition of the importance of CPS to the construction industry. As a result, CPS applicability and potential benefits have been explored in various areas of the construction industry, including project delivery process (Anumba et al. 2010), light fixture monitoring and control (Akanmu et al. 2012), structural health monitoring (Hackmann et al. 2014), and temporary structures monitoring (Yuan et al. 2014). Based on these investigations, CPS has been identified as having considerable potential in the construction industry, particularly

1 in addressing those problems that require bidirectional coordination between physical systems and their virtual representations.

The term ‘temporary structures’ refers to systems and assemblies used for temporary support or bracing of permanent work during construction, and structures built for temporary use. The former are defined as the elements of civil engineering work, which support or enable the permanent works (Grant and Pallett 2012). Included are temporary support systems such as earthwork sheeting & shoring, temporary bracing, soil backfill for underground walls, formwork systems, scaffolding, and underpinning of foundations. The second category includes temporary or emergency shelters, public art projects, lateral earth retaining structures in construction zones, construction access barriers, temporary grandstands and bleachers, and indoor and outdoor theatrical stages (Parfitt 2009).

In the construction industry, which accounts for more than one third (36%) of all U.S. workplace fatalities (Zhang et al. 2015), the safety problem relative to temporary structures remains serious (Parfitt 2009). The last four decades have seen numerous collapses related to improper erection and monitoring of temporary structures. In 1973, the improper removal of forms triggered a progressive collapse of the Skyline Plaza (Bailey’s Crossroads, VA), killing 14 construction workers and injuring 34 others (Feld and Carper 1997). Another example was the collapse of a section of the University of Washington football stadium expansion in 1987 due to premature removal of temporary guy wires (Feld and Carper 1997). A major scaffold system on a 49-story building on 43rd street in New York’s Time Square collapsed in 1998 as a result of bracing removal, resulting in the death of one individual, several injuries and hundreds displaced from their residences (Stewart 2010). In general, it is estimated that three quarters of construction workers work on or near temporary structures (OSHA 2014). The improper management of temporary structures results in 100 deaths, 4500 injuries, and costs $90 million every year (OSHA 2014). Thus the improvement of temporary structures monitoring is urgently needed (Parfitt 2009).

Recent advances in information technologies have resulted in the emergence of cyber-physical systems (CPS), which offer a promising and more effective approach to temporary structures

2 monitoring (Yuan et al. 2014). The use of CPS provides an opportunity for changes in the physical structure to be captured and reflected in a virtual model. Conversely, changes in the virtual model can be communicated to sensors embedded in or attached to the physical components. This bi-directional coordination between physical and virtual systems enables the temporary structures to be continuously monitored and assessed for performance in order that potential hazards can be identified and addressed prior to an accident, irrespective of causation.

1.2 Research Overview

This section covers the research objectives and expected contributions. In general, six specific objectives are identified. In review of the objectives, several contributions to the academic research and the construction industry can be obtained and is discussed in detail in this section. 1.2.1 Research Aim and Objectives The proposed project aims to investigate the potential for improved monitoring of temporary structures to improve the safety of construction workers and the serviceability of temporary structures through the implementation of Cyber-Physical Systems (CPS). The specific objectives of this project are to: 1) Investigate CPS applications in other industry sectors, such as manufacturing, healthcare, transportation, and power grid. 2) Assess the applicability of CPS in the monitoring of temporary structures, including a review of the enabling technologies. 3) Review the safety regulations and requirements for a specific type of temporary structure. 4) Develop and test a CPS-based monitoring system for the specific type of temporary structure identified earlier. 5) Evaluate (and refine) the prototype CPS monitoring system based on feedback from laboratory experiments and industry experts. 1.2.2 Expected Contributions This study seeks to evaluate the applicability of CPS in safety monitoring of temporary structures. The outcome of this research is intended to provide opportunities for intelligent monitoring of temporary structures in the construction industry via mutual communication between a physical structure and its corresponding virtual model. Generally, the expected contributions of this research project include:

3  Identify CPS applicability to temporary structures monitoring, including the identification of the need of CPS for temporary structures monitoring, benefits and implementation barriers of CPS to temporary structures, as well as the enabling technologies of CPS that support CSP applicability to the monitoring of temporary structure.  A prototype CPS-based system for safety monitoring of temporary structures, which demonstrates in construction industry can be obtained through CPS.  The developed system benefits the safety management of construction industry through the following ways: o Real time inspection: in lieu of inspecting specific influential factors, CPS examine the performance of temporary structure components every few seconds, and takes into consideration all loads imposed on the components, thus ensuring structural stability. o Quick problem identification: Implementation of a dynamic CPS environment has the potential to shorten the time interval between the onset of an initial hazard. o Automated sensing and control: CPS enable bi-directional communication between physical components and their virtual representations. From the job site to the virtual model, movement of physical components will be detected by sensors and sent to the virtual model, where the difference between designed and actual structure will be highlighted on the model. From the virtual model to the job site, once potential hazards are detected, safety alert will be sent from the virtual model to the workers or automatically stabilize temporary structures.  Provide a new approach in assisting the monitoring of temporary structures though the close linkage between the physical structures and their virtual models with the help of data acquisition system, virtual model platform, on-cloud database, and others.  Contribute to the knowledge of the benefits and implementation barriers of CPS to the construction industry, particularly to the area of temporary structure monitoring.

1.3 Thesis Structure

This thesis has the following 8 major chapters, which are briefly described below: Chapter 1: Introduction Chapter 1 gives an introduction of the research project, such as the research background and research objectives to be accomplished. It provides an overview of the research project by stating

4 the overall aim and specific research objectives. It also summarizes the methods adopted and expected contributions Chapter 2: Research methodologies Chapter 2 examines several research methods in general, including qualitative research methodologies, quantitative research methodologies, and hybrid methodologies. In particular, it presents an overview of the research methods adopted in the research of information systems, with a particular focus on information systems in the construction industry. Based on the overview of research methodologies, it identified the ones that were appropriate to this project. Chapter 3: Applicability of CPS to safety monitoring of temporary structures This chapter investigates CPS applicability to safety monitoring of temporary structures. It starts with the identification of an urgent need for safety management of temporary structures. This is followed by a comprehensive review of the conventional approaches to temporary structures monitoring and existing CPS applications was conducted. Several potential areas of temporary structures for CPS applications were explored; based on the previous reviews, it made the case for CPS deployment in temporary structures monitoring with potential benefits and limitations explored. Chapter 4: Prototype system development In Chapter 4, the development of a prototype system for temporary structures monitoring is presented. This is based on the system development framework for temporary structures monitoring and the identification of end user requirements. A discussion of the system development environment, including software environment and hardware environment, as well as system architecture design is also presented for a clear guidance on the prototype system development. Chapter 5: Physical experiments with temporary structures This chapter demonstrates the set-up of the hardware for experimental work, including a temporary structure and sensors. It presents five experimental scenarios that were designed for evaluation of the performance of temporary structures. It also describes the initial experimental tests conducted under each experimental scenario to see if the performance of the temporary structures can be captured as accurately as expected. Chapter 6: Prototype system operation Chapter 6 describes how the virtual model and the physical structures were integrated into a

5 comprehensive CPS for temporary structures monitoring. It demonstrates how the prototype system works through system workflow and the user-interfaces of the developed prototype system were presented through initial experimental tests. It also describes how the prototype system was further tested through a comprehensive experiment based on the five experimental scenarios identified in chapter 5. Chapter 7: Prototype system evaluation and refinement This chapter is concerned with the evaluation of the prototype system through an evaluation workshop. It discusses the selection of the evaluation method and the design of the evaluation questionnaire. It also describes the feedback from the evaluation workshop, and the prototype system’s benefits and limitations. Chapter 8: Discussion and conclusions In chapter 8, the research project is summarized, and its key findings and contributions to knowledge outlined. It also discusses the research limitations and practical considerations in full-scale adoption of the prototype on site. Finally, it also presents recommendations for practical implementation and future research.

1.4 Summary

This chapter has presented the research background by reviewing the current problems of temporary structures in the construction industry and the potential solution based on taking advantage of advanced information systems such as CPS. It outlined the aims and objectives of the research, along with the expected scientific contributions. In particular, it is expected that this research will demonstrate CPS applicability to temporary structures monitoring, develop a prototype system for CPS applicability validation, conduct experimental tests for system validation, as well as end-user evaluation of the system’s performance. For clear understanding of the thesis, a summary of the thesis structure with a brief introduction to each chapter was also provided.

6 Chapter 2: Research Methodologies

2.1 Overview of Research Methodologies

This chapter examines a variety of research methodologies generally used in academic research. They include qualitative research methodologies, quantitative research methodologies, and hybrid methodologies. 2.1.1 Qualitative Research Methodologies Qualitative research is a method of inquiry which is used for an in-depth understanding of human behaviors and the reason behind it. It investigates how and why specific outcome occurs (Yin, 1984), and interpret the meaning, metaphors, and symbols of the social world (Asher and Miller, 2011). The qualitative research is conducted in natural settings for descriptive results, through the analysis of data that primarily from observations, interviews, and documents (Yin, 1984). This research method is frequently employed in the social sciences, as well as market research and others (Denzin and Lincoln, 2005). There are several qualitative approaches developed for different research objective. Among which, the most frequently used qualitative research methodologies include grounded theory, ethnography method, phenomenology method, and case study.

 Grounded theory: grounded theory is a research methodology applied in social sciences. It is a systematic generation of theory based on analysis of collected data (Martin and Turner, 1986) to understand the process when there are inadequate theories for explanation (Creswell, 2008). The process of conducting grounded theory research is to first of all identify research problem and collect data for analysis. A broad theory can be developed by coding data and identify relationship between variables. After that, the theory should be evaluated before use. The data comes from interviews, observations, documents, historical records, and videotapes (Bugday, 2012). While the developed theory might not work in every situation, it fits at least the analyzed data set.  Enthnography: enthnography method is originally applied in the anthropology. It aims at understanding the culture of a group or organization through the observation of social interaction, behaviors and perceptions (Reeves et al. 2008). Most of the data collected in enthnography approach comes from participant observation, which requires researchers become participant of

7 the organization while observing it. Interviews and documents can also serve as data sources.  Phenomenology: phenomenological research describes the meaning for several individuals of their lived experience of a concept or a phenomenon by listening to the different stories of participants. It aims to interpret phenomenon and how this world comes to be (Waters). As is stated by Willis (2007), phenomenology looks into the subjectivity of reality to reveal the importance of understanding how human view themselves and the world.  Case study: Thomas (2011) defines case studies as the analysis of people, events, decisions, periods, projects, policies, institutions, or other systems that are studied holistically by one or more methods. The case adopted for analysis can be one or multiple, which provides data and examples for researchers to understand the underlying principles or what the case explicates. Instead of random pick up, a case is selected based on the research objectives. Generally, a case rich in information is preferred. To be specific, there are three types of cases that are usually adopted, including key cases (identified due to the inherent interest of the case), outlier cases (extreme, deviant, or atypical case), and local knowledge cases (identified based on the in-depth knowledge of researcher). Through detailed analysis of limited amounts of cases, researchers seek for extended experience and in-depth understanding of the world. 2.1.2 Quantitative Research Methodologies Quantitative research method explains the phenomena by collecting numerical data that are analyzed with mathematically based method (Aliaga and Gunderson, 2005). The measurements for qualitative research includes populations and sampling, random assignment, and generalizability. The populations and sampling represents the subject that is studied. The selection and identification of appropriate sampling is critical to the research result. The random assignment differs from random sampling, and is mostly used for experiments. The generalizability looks into the evaluation of a general conclusion. Generally, there are five research approaches in quantitative research, such as experiments, behavioral measures, surveys, social network analysis, and archival & meta-analysis.  Experiments: an experiment is conducted to verify, establish or reject a hypothesis. It is highly used in a variety of disciplines, such as engineering science and medical science. Generally, there are two types of variables in an experiment, namely the independent variables and dependent variables. The researcher manipulates the independent variables and measures the change of dependent variables to test the hypothesis. There are three main kinds of experiments, the

8 controlled experiment, the quasi-experiments, and the field experiments. The controlled experiment usually has two data sets, one is identical and another one is tested with an independent variable being manipulated by researcher. The results of the two data sets will then be compared for further investigation. It works well in identifying relationship between one or multiple factors and outcomes. When it is difficult to conduct controlled experiment, researchers may turn to the quasi-experiment by observing variables of the system under study without manipulation. In this way, the effects of one or some factors are identified when the influence of other variables remain constant. On contrary to laboratory experiment, field experiment is conducted in the natural situation without control of the variables. It tests a hypothesis or an approach in the real world, which makes the result to be more convincing.  Surveys: survey research is often used to assess thoughts, opinions, and feelings (Shaughnessy, et al. 2011). It is adopted by researchers in various areas for analysis of human behaviors in social and phycology area, for marketing and advertising in business, for improvement and revolution in politic area. The sample selected for survey can be limited and specific, as well as broad and worldwide, which is generally determined based on the research purpose. Besides, the survey can be conducted via several ways, such as mail survey, personal interview, telephone interview, and internet surveys. The mail surveys are sent to selected people with instruction on how to fill and return it; the personal interview is usually a face-to-face interview, conducted by the researcher or his assistant at a public place; the telephone interview helps to reduce the cost of time spent on survey, and provide access to wider sample of people around the world. However, a telephone interview during people’s work may not be welcome. People intend to block the call or not respond seriously. This limitation greatly affects the validity of the survey results. To make it more flexible while keeping the advantage of telephone interview, the internet interview is commonly used recently. Generally, the survey method can have a large sample with little cost and time. Besides, it can efficiently ask lots of questions for further analysis. However, the bias due to response should be taken into consideration. 2.1.3 Hybrid Research Methodology While there has been a distinct line between qualitative and quantitative methodology, researchers have come to realize the importance of mixing quantitative and qualitative research method, namely hybrid research methodology, also known as triangulation research methodology. The hybrid research methodology is initially justified by Denzin in 1970. Instead of replacing

9 qualitative or quantitative research method, the use of hybrid research methodology is to draw strengths and minimize weakness of the two methodologies for higher evaluation and deep understanding of the research problem. Therefore, it is recommended to adopt hybrid research methodology either for the entire or part of study Denzin (1984).

2.2 Methodologies in Information Systems Research

Information system is an emerging scientific discipline with more knowledge to be discovered and created (Caplinskas and Vasilecas, 2004). Various kinds of methodologies have been adopted in the study on information systems. The mostly common approaches, including laboratory and field experiment, case study, surveys, and hybrid research method, are discussed in detail below.  Laboratory and field experiment: the laboratory and field experiment serves as the dominant approach to information technology studies for years. It is conducted to examine the effects on one or multiple variables. Usually, the information technology is set as the independent variable for evaluation of particular outcomes (Kaplan and Duchon, 1988).  Case study: as an investigation of a contemporary phenomenon, the case study has been advanced for research on information systems. The case study is usually used for better understanding of the nature and complexity of the process of information systems taking places (Benbasat, et al., 1987).  Survey: there has been a great reliance on survey method in information system studies (Goldstein, et al., 1986). Owning to the simple and easing scoring and effective analysis of relationships among variables, Survey method is commonly used by the information systems research community (Newsted , et al., 1998).  Hybrid research: The hybrid research methodology has been reviewed for the information system study by researchers (Kaplan and Duchon, 1988). As a pragmatic area, the information system has its origin in a variety of reference disciplines. While the study on information system traditionally relies on quantitative methods for data analysis, the importance of a better organization and interpretation has been recognized (Kaplan and Duchon, 1988). The application of information technologies in the built environment is a relatively new area. The information technologies are usually adopted to facilitate the information process involved throughout the life cycle of buildings (Björk, 1999). Based on different research objectives, a variety of methodologies are utilized for the research on information technologies in the built

10 environment. They include ethnographic research, surveys, case study, action research and experiments (Fellows and Liu, 1997). As is discussed in previous sections, the ethnographic research examines the society through the observation of social interaction, behaviors and perceptions. The case study supports the identification of needs and research gaps of information technologies in this area. Due to the problem-driven feature of information technology application in the built environment, the action research is commonly adopted to solve an immediate problem. As the most commonly used research method in information system research, the use of experiments helps the evaluation and refinement of information technology applications in the built environment. Besides, the research on information technologies in the built environment involves multiple disciplines, which then calls for the adoption of hybrid research methodologies.

2.3 Research Methods Adopted and Justification

The research program aims to investigate the potential of improving safety monitoring of temporary structures with the aid of CPS. In doing this, six major objectives are to be achieved. They include (1) the investigation of CPS applications in other industry sectors and construction industry; (2) the identification of CPS applicability in the design and monitoring of temporary structures; (3) review of the safety regulations and requirements for a specific type of temporary structure; (4) exploration of the available technologies for CPS implementation in temporary structure monitoring; (5) prototype system development for temporary structures; (6) experimental test of prototype system; (7) evaluation (and refinement) of the prototype CPS monitoring system. Therefore, a variety of research methods are adopted based on research purpose, including literature review, prototype development, laboratory and field experiment, and system refinement. 2.3.1 Literature Review A literature review is an objective, thorough summary and critical analysis of the relevant available research and non-research literature on the topic being studied. It is required that the sources to be reviewed, and the structure and flow in literature review should be carefully examined. A good literature review goes over the written papers for a comprehensive understanding of a state of the art of knowledge in related areas. Besides, it helps to identify the gap for further research direction and examine appropriate solutions.

Literature review is conducted throughout the research program, including initial literature review, exploratory literature review, focused literature review, and refined literature review. Firstly, an initial literature review is performed for an overview of the research in construction industry and CPS. With interests, observations and initial questions arising, an exploratory literature review is conducted to examine the appropriate temporary structures for safety monitoring with aid of CPS, and clarify the applicability of CPS in this area. Besides, research methodologies will be reviewed and selected in accordance to the research purpose. Based on previous identifications, a focused literature review helps to format the system design of the safety monitoring CPS of scaffold system. Finally, refined literature review will focus on the evaluation of research questions, research methods, research results, contributions, as well as research limitations. Throughout the research process, resources for literature review come from journals, conferences, key researchers, institutions, etc. These resources evaluate the research questions and contributions, for they have been reviewed and developed by leading academic. The initial and exploratory literature review is presented in detail in chapter 2 and 3. 2.3.2 Rapid Prototyping Rapid prototyping is generally defined as a group of techniques used to quickly fabricate a scale model of a physical part or assembly using three-dimensional computer aided design (CAD) data (Efunda, 2013). It benefits the research by increasing effective communication and decreasing development time. Basically, the rapid prototyping involves the development of an innovation system, which is then used in the analysis, design, evaluation and development of the innovation work (Jones and Richey, 2000). The rapid prototyping method is adopted in this research to develop a safety monitoring CPS of scaffold system, which includes system analysis, system design, and the development of the modules for information exchange and control. 2.3.3 Laboratory and Field Experiment In an attempt to refine and evaluate the proposed approach of improving safety monitoring of scaffold systems with the aid of CPS, the developed prototype system will be tested through laboratory and field experiment. A real world situation (such as change of loads, foundation settlement of scaffolds, and movement of scaffolding components), which usually causes structural failures of scaffolding, will be simulated with the implementation of developed prototype system. Data collected from the laboratory and field experiment can be analyzed for

12 system refinement. 2.3.4 Evaluation Workshop To evaluate the proposed approach and research contributions, as well as capture system limitations, a system evaluation is conducted among a selected amount of engineering experts and construction workers through a workshop. In carrying out this evaluation, a demonstration of the laboratory and field experiment will be presented to the engineering experts for questionnaires on effectiveness and potential uses of the developed prototype system. To be specific, the system can be evaluated for functionality, system reliability, measurement accuracy, operation and maintenance easy, technology maturity, portability, adaptability, and cost. The focus group discussion right after the questionnaire can help to clarify feedbacks from engineering expert, as well as in depth discussion for potential applications and improvement.

2.4 Summary

This chapter reviews the research methodologies generally used in academic research, including qualitative research methods, quantitative research methods, and hybrid research methods. By examining the research purpose and the research methods adopted in information systems research, four major research methods are identified to be used for this research project. In particular, the literature review will be used for CPS applicability analysis for temporary structures monitoring; the CPS prototype system will be developed for CSP application validation through rapid prototyping; the developed prototype system will be further tested through laboratory experiment and evaluated through evaluation workshop.

13 Chapter 3: Applicability of CPS to Safety Monitoring of Temporary Structures

3.1 Introduction to CPS

Due to the different research scopes, researchers view CPS in different ways. A better understanding of its definition and key features is helpful to the identification of potential benefits and barriers to CPS application in temporary structures. 3.1.1 Key definitions of CPS Several researchers have defined CPS as follows: • Cyber-Physical Systems are smart systems that have cyber technologies, both hardware and software, deeply embedded in and interacting with physical components (Smart System 2012). This definition regards CPS as an integration of both software and hardware, where the physical and virtual components can interact with each other through embedded instruments. • Cyber-Physical Systems are physical and engineered systems whose operations are monitored, coordinated, controlled and integrated by a computing and communication core (Rajkumar et al. 2010). From this perspective, the focus of CPS is more about the control and communication of the physical system through the computing system. It highlights the function of remote control and monitoring of the real world by CPS. • Cyber-Physical Systems are integrations of computation with physical processes, wherein networked embedded computers monitor and/or control physical processes based upon local (i.e. in-network) and remote (i.e. back-end) computational models (Krogh et al. 2008). Based on this definition, it can be concluded that CPS enables the computing system to control the physical system through a network. • Cyber-Physical Systems are large-scale interconnected systems of heterogeneous components that are envisioned to provide integration of computation with physical processes (Lee 2007). Therefore, CPS provide a solution to integrate several heterogeneous components, so that the physical world can be well controlled by the computing system with different functions based on the users’ purpose. In general, CPS are used as an interaction system where the physical world and virtual world can communicate and interact with each other seamlessly. Considering for the purposes of this paper,

14 a CPS is simply considered as the effective bidirectional integration of computation with physical processes. Embedded computers and networks monitor and control the physical processes with feedback loops, where physical processes affect computations and vice versa (Derler et al. 2012). 3.1.2 Key features of CPS By definition identified above, a CPS involves a high degree of integration between computing (virtual) and physical systems (Greenwood et al. 2015). Distributed applications are also common which involve distributed management and/ or distributed operations such as a power grid. Another feature of CPS is the ability to provide timely service in the face of real-time constraints (Wan and Alagar 2012).

In general, the key features of CPS can be summarized as integration, distributed system, real-time system, virtualization, adaptability, automation, heterogeneity, and uncertainty (Table 1). By integration, the CPS combines the physical and computing system through mutual information exchange and control, instead of only tracking the physical world with a computing system or vice versa (Wu and Li 2011). Besides, CPS provides distributed management for it can be a large system consisting of several distributed systems. In this way, by using CPS, multiple parties can remotely obtain access to the project from different places for project management or cooperation. Meanwhile, CPS is capable of real-time communication, which enables continuous integration and up-to-date information exchange between the virtual and physical systems. The integration between virtual and physical systems can be updated in real time. In terms of virtualization, CPS creates a virtualized interface for users to remotely analyze and track the physical system. It also enables all the physical components to be identified and recognized accurately in the virtual model (Penna et al. 2010). The feature of adaptability means that CPS is capable of adapting to changing situations through dynamic reorganizing /reconfiguration (Khaitan and McCalley, 2014). The automation feature enables the automatic control of physical system according to continuous tracking. During the automatic control process, actuators are triggered based on system analysis and command. From the perspective of heterogeneity, CPS integrate several different systems together with standard communication and information exchange. It also integrates various devices, including sensors, mobile devices, high-end workstations and servers (Wan et al. 2011).

Overall, CPS integrates the physical structure and its corresponding cyber systems seamlessly at all scales and levels. Automatic and resilient human-machine interaction is supported by CPS for improved control, efficiency and reliability of the physical systems (Lee et al. 2015). Table 3-1: Key Features of CPS

Researcher Integration Distributed Real-time Virtual Adaptability Automation Heterogeneity Wu and li Combination;

(2011) networked cyber Shi et al. Integrated capability highly adaptive complex (2011) networked in physic automation world Anumba tagging and interaction virtual et al. (2010) tracking physics distributed Wan and sharing system; Alagar real-time heterogeneity resource Geographically (2012) distributed Wan et al. couple with heterogeneity (2011) physics

partial distributed Geisberger autonomy; networked; control; systems adaptive et al. (2011) human-machine of system cooperation

3.2 CPS Applications

In exploring CPS applicability in temporary structures, it is important to understand its applications in other industry sectors where CPS has been adopted and developed for advanced management. In addition, a review of current CPS applications in the realm of the built environment could help to identify the current research gaps, examine CPS applicability, as well as to identify potential application areas in temporary structures monitoring. 3.2.1 CPS Applications in Other Industries CPS was initially developed in other industries such as manufacturing industry, power grid, transportation industry, and healthcare industry. Some of the applications in these industry sectors are briefly reviewed in this section.

3.2.1.1 Manufacturing Industry

Smart manufacturing refers to the manufacturing industry which, with the application of CPS for

16 the tracking and monitoring of a product throughout its life cycle, can optimize its performance and efficiency (Coalition 2011). Current manufacturing resource aggregation (MRA) cannot adapt to dynamic factors, which greatly impact the quality and efficiency of the manufacturing industry. In viewing this problem, Liu and Su (2011) pointed out that CPS can provide real-time information on physical resources and thus support the realization of dynamic MRA. Liu and Su’s study gives a new solution for dynamic MRA with the aid of CPS. Taking advantage of the real-time feature of CPS, Kaihara and Yao (2012) developed a Real-Virtual Integrated Scheduling System based on the concept that CPS is the combination of the computational and physical worlds (Lee 2007). Their developed system helps to manage dynamic changes at production sites through the dynamic scheduling of the real system and simulation of the virtual system.

The main challenges for CPS the implementation in manufacturing industry include network integration, affordability, and the interoperability of engineering systems (Aderson 2011). The use of CPS in manufacturing industry requires seamless connection between each node. While the manufacturing industry’s network has enabled the life-cycle management of products, the implementation of CPS requires high-speed, stable, and seamless communication for real-time inspection and control. Therefore, more integrated network should be set up before CPS can be fully implemented in this area. Due to the high cost of both retrofitting existing control system and the initial set up of CPS, there is no motivation for small and medium-sized companies to investigate CPS. Besides, most of the manufacturers use different kinds of engineering systems for production management, and even customize their own software and system. The lack of an information exchange standard regulating the data type of the manufacturing files from various manufacturers prohibits the quick promotion of CPS, for CPS has to be adjusted and re-designed for each manufacturing company due to the lack of interoperability of engineering systems (Smart Systems 2012).

3.2.1.2 Power Grid

The smart grid is a complex ecosystem of heterogeneous (cooperating) entities that interact to provide the envisioned functionality (Karnouskos 2011). In view of the fact that the infrastructure for both the electric grid and water distribution systems is aging with technical and

17 reliability problems, Krogh et al. (2008) proposed that the electric grid and other utilities can use CPS technologies for a smarter and more efficient system. CPS is valued as an integral part of smart grid (Karnouskos 2011). To verify the correctness of cyber-physical composition, Sun et al. (2007) introduced a model to avoid interference among components of a system. Furthermore, since CPS imposes increasing uncertainties on controlled systems, Zhu and Basa (2011) proposed a holistic theoretical framework and applied it to power systems. This framework helps to maintain an appropriate level of CPS operation even during unexpected system interruptions.

With CPS implementations in smart power grids, challenges to be dealt with include system security, data analysis, and the integration of new technologies. The system security requires that the system should not easily crash even when there is a high consumption demand, or when the system is attacked. Besides, while the use of CPS enables multiparty access to the power grid management system to get or provide energy to the system, such distributed accesses to the smart power grid impose the risk of hacker attacks. Because the power grid produces large amounts of data for analysis, an appropriate data fusion method is critical to CPS for automatic control of the power grid system. Besides, the integration of CPS with current power grid management system calls for great efforts in analyzing existing system, a proper way of adding CPS to existing power grid management system, as well as the use of new technologies for enhanced functions with use of CPS, such as energy storage technologies (Baheti and Gill 2011).

3.2.1.3 Transportation Industry

Recent traffic problems, such as traffic jams and accidents, call for the improvement of transportation system (Gong and Li 2013). CPS offers the potential to an intelligent traffic system through real time detection of system status, optimal route plan, and adjustment of control strategies (Gong and Li 2013); and is valued as the promising approach to aerospace education, research, training, and accelerated workforce development (Noor 2011), and the transformation of the national airspace system (NITRD 2012). Four research areas, including (1) new functionality, such as higher capacity, greater safety, more efficiency; (2) integrated flight deck system for (semi)autonomous system; (3) vehicle health monitoring and management; (4) safety security of aircraft control system, have been identified for CPS applications to the future air transportation system (Baheti and Gill 2011). Mertins et al. (2012) also proposed a CPS

18 application to aircraft maintenance repair and overhaul, with the aim to simplify and shorten the execution of Maintenance, Repair and Overhaul (MRO)-Processes which impact aircraft availability. To facilitate the application of CPS in the transportation domain, a low-priced intelligent vehicle with wireless sensor networks navigation, was designed (Wan et al. 2011). Furthermore, based on CPS theory, Gong and Li (2013) proposed a fusion framework for urban traffic control system, which aims to avoid traffic congestion and improve traffic efficiency. This study provides a theoretical foundation for CPS implementation in intelligent traffic system.

Safety and security, which are most important to the transportation industry, remain big challenges before CPS can be fully adopted (Smart System 2012). The safety threat means that the use of CPS may distract the attention of drivers and result in a higher risk of accidents. Besides, the safety of transportation systems is also highly impacted by system security when there are system logic problems of unmanned vehicles, system interruptions, and hacker attacks (NITRD 2012). All of these concerns call for a high confident CPS to be used for transportation management (Lintelman et al. 2008).

3.2.1.4 Healthcare Industry

Healthcare increasingly relies on networked medical systems to take care of patients with special circumstances. An example is the use of sensors to track the conditions of patients during operations. However, these techniques should be assembled into a new system configuration, such as CPS, to match specific patient or procedural needs (Baheti and Gill 2011). It is envisioned that the next generation of healthcare systems would be secure and reliable systems with wired and wireless networked medical devices (Arney et al. 2010) and medical information systems (Sha et al. 2009). The application of CPS in healthcare involves the national health information network, electronic patient record initiative, home care, operating room, etc. (Shi et al. 2011). Lee and Sokolsky (2010) discussed current trends and promising research directions in the development and use of high-confidence medical CPS. To cope with the safety issues, Cheng (2008) proposed a method to allow the safe cyber-physical operation of a medical ventilator, a life-critical reactive device, to move breathable air into and out of the lungs of a patient with respiratory difficulties. This study also examined potential research areas so that the safety operation of CPS application in healthcare industry can be improved.

Several challenges for CPS application in healthcare industry remain to be tackled. For example, the complex relationship between a patient’s health condition and the required medical condition calls for an accurate algorithm with several parameters other than the time (Cheng 2008). In viewing the trend of medical CPS, Lee and Sokolsky (2010) identified the challenges raised by CPS, including safety of patients, collaboration and interoperability between several treatment systems to one patient at the same time, a simple patient model covering all needed information and a comprehensive simulation for test and validation, adaptive patient-specific algorithms with special concerns to unique parameters to a certain patient, and open interconnectivity standards for medical CPS. 3.2.2 CPS Applications in the Built Environment The construction industry is in need of continuous improvement in areas such as intelligent safety management, cost and resource management, scheduling, and energy conservation. With development in information technologies, recent research (discussed below) have recognized the potential benefits of CPS to the built environment from various perspectives, including: project delivery process, automatic construction site layout generation, construction progress monitoring, light fixture monitoring and control, and Structural Health Monitoring (SHM). Current CPS applications demonstrate how the built environment could benefit from CPS. (1) Project delivery process Current project delivery processes are inefficient and there is a need of transformation so that greater control can be made through an integrated system. By identifying the limitations of previous efforts on managing the construction process, Anumba et al. (2010) demonstrated the need for CPS to improve the project delivery process, and proposed a CPS approach targeting the integration of virtual models and the physical world through a bi-directional flow of information. This provides reference for future research efforts in this research field. Akanmu (2012) highlighted the mechanism and triggers for the bi-directional flow of information. Through use-case scenarios validation and expert interviews, CPS applicability was demonstrated in enhancing bi-directional coordination between virtual models and the physical construction world. (2) Automatic construction site layout generation

20 In view of the importance of resource layout and activities to the success of construction projects, Akanmu et al. (2014) proposed the use of CPS for construction site layout and developed an automated component level system for optimized and real time site layout. Case studies demonstrated that the developed system provides an effective means for real time construction site space tracking and the automatic generation of construction site layout. (3) Construction progress monitoring Potential benefits of CPS to construction progress monitoring were highlighted by Olatunji and Akanmu (2015) with the development of an adaptive CPS approach for construction progress monitoring and control. Besides, Yang et al. (2015) proposed the use of vision-based method for construction performance monitoring, including the control of construction progress both at the project level (civil infrastructures and elements) and the operation level (construction equipment and workers). (4) Light fixture monitoring and control Building on the previous studies, Akanmu presented an approach to improve light fixture monitoring through CPS integration between virtual models and physical light fixtures. A prototype system was developed and implemented for tracking, monitoring and controlling light fixtures throughout a facility life cycle (Akanmu et al. 2014). Other possible applications were also identified, such as management of urban infrastructure street lights, mechanical systems and other building components such as window blinds, etc. (5) Structural Health Monitoring (SHM) SHM helps to prevent civil structural failure or costs by providing information and assisting in decision making for preventative measurements (Smart System 2012). To improve the evaluation of structural performance, recent research has applied CPS to SHM by integrating the computing elements and structural components. According to Hackmann et al. (2014), SHM represents an important application domain of CPS. To integrate damage detection and energy saving, they proposed a cyber-physical co-design approach to structural health monitoring using wireless sensor networks. This approach can selectively activate nodes in the damaged region so that local damage is detected while allowing the rest of the nodes to remain asleep. Tidwell et al. (2009) pointed out the main challenges of CPS application to SHM, such as the actuator dynamics, complex interactions between physical components and their virtual models, and the computation and communication delays. To improve the hybrid real time testing, Tidwell et al.

21 (2009) provided a highly configurable and reusable middleware framework for real-time hybrid testing. This study improves the high-fidelity real-time testing and promotes CPS applications in this area.

3.3 Overview of Temporary Structures

Temporary structure is a broad term for systems and assemblies used for temporary support or bracing of permanent work during construction, and structures built for temporary use. The former are defined as the elements of civil engineering work which are required to either support or enable the permanent works and found in all areas of construction (Grant and Pallett, 2012). Included are temporary supporting systems (i.e. earthwork sheeting & shoring, temporary bracing, soil backfill for underground walls, formwork systems, scaffolding, and underpinning of foundations). The second category includes temporary or emergency shelters, public art projects, lateral earth retaining structures in construction zones, construction access barriers, temporary grandstands and bleachers, sound system and lighting support structures for parades and public events, and indoor and outdoor theatrical stages (Parfitt, 2009).

Past decades have seen numerous significant collapses related to improper erection and monitoring of temporary structures. In 1973, the improper removal of forms triggered a progressive collapse of the Skyline Plaza (Bailey’s Crossroads, VA), killing 14 construction workers and injuring 34 others. Another example was the collapse of a section of the University of Washington football stadium expansion in 1987 due to premature removal of temporary guy wires. A major scaffold system on a 49-story building in New York’s Times Square scaffold system on 43rd street in Time Square collapsed in 1998 as a result of bracing removal, resulting in the death of one individual, several injuries and hundreds displaced from their residences. 3.3.1 Common Types of Temporary Structures In general, there are six common types of temporary structures, including earthwork shoring/sheeting systems, temporary bracing systems, underpinning of foundations, scaffolding systems, formwork, and temporary performance stages. In this section, each of the six common types of temporary structures are carefully studied for the major causes to their structure failure and case studies.

22 3.3.1.1 Earthwork Shoring/ Sheeting System Sheeting & shoring using systems such as steel soldier piles, sheet piles, and slurry walls, are used to prevent soil movement and cave-ins during the excavation of earth. These systems help minimize the excavation area and protect nearby buildings or structures. Sheeting and shoring system can be categorized into spaced sheeting or close sheeting. The former method involves inserting spaced timber shores, bracing, trench jacks, piles or other material to resist the pressure from surrounding earth. The close sheeting requires continuous solid sheeting along the entire length of excavation (Berry, 2009).

Inappropriate design and installation of earthwork shoring & sheeting systems result in numerous accidents each year making earthworks a substantial risk for workers. With increased concerns, the government has made an effort to reduce the amount of cave-in accidents throughout the past decades. However, recent reports from OSHA show that cave-in accidents have kept occurring at the rate of around 20 accidents each year from 2006 to 2012. Table 2 below shows the trends of the fatal injuries in cave-in accidents. Table 3-2: Fatal Occupational Injuries Due to Cave-in, From 1992 to 2010 (BLS, 2007, 2012)

year 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

amount 42 38 49 33 55 35 46 44 40 36 34 48 41 44 28 30 27 22 28 21 17 of fatal injuries

(a) Causes Three major causes are summarized as follows for the structural accidents of shoring/sheeting system. (1) The lack of shoring/ sheeting system. The lack of a shoring/sheeting system causes the majority of trench collapses. Sometimes, excavations are performed to provide access to pipelines or for other small underground projects. In these instances, the excavations are regarded as “easy” and without potential safety hazards so contractors prefer to not install a shoring/sheeting system. In other cases, construction proceeds ahead of schedule and employees work in the trench before the shorting/sheeting system has been fully installed.

23 (2) Inadequate shoring/ sheeting system. Inadequate shoring/ sheeting system means that a system is improperly installed and fails to meet design expectation. It stands there without protecting workers, and may even make things worse when it collapses with soil towards workers. (3) Material storage. The slope of excavation is unstable due to the instability of soil. The shoring/ sheeting system is designed to support the soil around the trench. Too much external load, such as heavy trucks, material storage, etc., will excessively impact the shoring/ sheeting system. It gets worse when there is no shoring/ sheeting system at all. The slope collapses when it cannot hold the external load from material storage near the trench. (b) Case Studies According to literature review, three case studies of shoring/sheeting system accidents are summarized as follows. (1) Trench collapse due to lack of shoring system Accidents occur frequently when workers are working inside of the trench with no shoring system, when the trench suddenly collapses and buries the workers in it. A trench collapsed in 1996 on the Wang Lee Street in Hong Kong, killing one worker when he was laying a pipe line in the trench. The trench is 1.8 m wide and 2.2 m deep. Investigation revealed that the trench was shored improperly, and there was a presence of water inside of the trench; in 2009, a worker was adjusting a water pipe in the trench with a size of 0.6 m wide and 3.0 m deep. One side of the trench collapsed suddenly, and buried the worker to the level of his chest. The worker passed away the next day. It was investigated and found that there was no shoring system inside the trench, and there was no proper and egress (Cheong, 2013).

(2) Trench collapse before shoring system has been installed in time In 1999, a trench was dug for laying pipelines on the road of Sam Mun Tsai Road, in Hong Kong. And the trench suddenly collapsed when the workers were getting ready to install shoring systems. This accident killed one worker, with one other worker injured (Cheong, 2013)..

(3) Falling objects due to lack of shoring system Another accident occurred in 2012, when the worker was working in an excavation. Part of the

24 excavation collapsed due to the lack of shoring, which caused the falling of a reinforced concrete pile. The pile fell towards the worker, and hit him on the back of head. Soil came together with the pile to get him stuck (Cheong, 2013).

3.3.1.2 Temporary Bracing System Temporary bracing systems are used to keep a structure or other building systems stable before the permanent bracing is installed, or the element becomes self-supporting. It is commonly used in construction of masonry walls, tilt-up precast concrete panels, steel frames, large timber framed walls, and wood trusses. During the construction of wood frames, temporary cross-bracing add lateral stability and help prevent collapse of building structures. During the excavation, there are two main types of temporary bracing systems, namely internal bracing and tie backs. The internal bracing system will hinder movement of equipment and materials and should not be used for deep excavation. One type of internal bracing is rakers, which rest on foundation mat or rock to support the wall. Another one is cross lot bracing, which extends from one side of the excavation to the other side to retain earth wall. As for the tie backs, it is most effective in firm ground. Tie backs provide a clear working space within the excavation, yet it is more expensive than internal bracing systems, and it might extend beyond the property lines of the building site.

A temporary bracing system is important to construction safety, yet it is often neglected. Insufficient bracing is cited as one of the four most common causes of failures in steel structures under construction (Kaminetzky, 1991). As is pointed out by Feld and Carper (1997), perhaps the most dramatic structural failures during construction resulted from a lack of stability. In most of the structural collapses, it is due to the insufficient support of loads that applied at the time of failure (Delatte & Rens, 2002). There is a time during construction before permanent bracing systems have been installed, and the project relied heavily upon the temporary bracing system. The structural load is usually analyzed by conceiving the whole structure as a completed entity, and there is frequently a lack of design or proper implementation of these systems. Often, the specific provisions and requirements of temporary bracing systems are left to the workers on the job site that may not have the qualifications or expertise for proper execution (Feld and Carper, 1996).

25 (a) Causes Three causes are summarized based on past structural failures related to temporary bracing systems. (1) Unexpected natural hazards According to Delatte & Rens (2002), the forces that temporary bracing systems are intended to resist mainly come from the wind. When collapse occurred, contractors often explained the failure as "We just got some unexpected wind gusts". Thus a design that accommodates predictable natural hazards was suggested (Feld and Carper, 1996). However, except for the natural hazards, the so called unexpected circumstance acts only as catalyzer, not the root cause, to a structural failure. In 1986, a concrete wall collapsed in a windstorm in Atlanta, Georgia, killing two construction workers. And it turned out that the concrete block wall was not braced at all. (2) Insufficient or nonexistent bracing system As is pointed out by Feld and Carper (1996), minor structural failures occur every day due to insufficient or nonexistent temporary bracing. Actually, insufficient bracing is one of the four most common causes of failures in steel structures under construction (Kaminetzky, 1991). In Toronto, Canada, a welded steel frame collapsed due to inadequate bracing in 1958. In 1984, a high masonry wall under construction collapsed in downtown Edmonds, Washington, for it was not braced. In 1987, a steel stadium project of the University of Washington football stadium in Seattle collapsed during construction due to inadequate temporary bracing. The insufficient or nonexistent bracing mainly results from human negligence or miscalculation of the load analysis. Contractors often have the attitude that "if we work fast enough, we won't have to brace it, and nothing is likely to happen"(Feld and Carper, 1996). (3) Imbalanced or lateral loading due to construction sequence Construction sequencing is very important to preserve the stability of incomplete structures. During construction, the load imposing on an incomplete structure is unstable, due to installation of components and construction activities. While these lateral loads are expected to be supported by temporary bracing, great changes in load may result in failures. This is the usual cause of many roof structures failures, for the roof structure often collapsed before the permanent bracing system has been placed (Feld and Carper, 1996). Feld and Carper (1996) further talked about the structural failure of seven concrete girders which tumbled over on a highway construction

26 project near Seattle, Washington in 1988. It is found that diaphragms that would have provided stability were not yet in place. Most of the time, it is the contractor who is responsible for determining the bracing and construction sequencing (Delatte and Rens, 2002), which makes it hard to determine the appropriate safety construction sequence. (b) Case Study According to literature review, three structure failures related to temporary bracing systems are presented as follows. (1) Collapse of a steel stadium project in Seattle, Washington, 1987 In 1987, an addition of a football stadium at University of Washington collapsed. An inadequate temporary support system was regarded as the most probable cause of failure by Feld and Carper (1997). According to other investigators, an incomplete system of temporary guying cable was the critical deficiency to the collapse (Manno 2009). (2) Collapse of a radial dome in Louisiana in 1964. The dome was designed as spanning 240 feet, with 36 timber arches. These arches had been placed for support on a tension ring on columns around the perimeter. However, before the installation of the deck, the temporary pipe shore of the compression ring at the top was removed. And one hour later, half of the cables connecting the tension ring and the compression ring failed, resulting in the rotation of the compression ring. Thus the whole roof collapsed, with no one component being preserved. (3) The Chicago City Post Office (November 3, 1993). The new building of the Chicago Post Office partially collapsed in 1993 due to the failure of a temporary connection of temporary erection angle pieces, which were used to secure a beam. This tiny piece of failure triggered the collapse of 70 additional components that had been secured. 2 ironworkers were killed, and 5 others were injured in this accident.

3.3.1.3 Underpinning of Foundations Underpinning of foundation is to install a support to an existing foundation to provide either additional depth or bearing capacity. It is mainly used in the following situations: 1) construction of a new project with a deeper foundation adjacent to an existing building; 2) settlement of an existing structure; 3) change in use of a structure; 4) addition of a basement below an existing structure (Ratay, 1996).

Even the most cautiously installed underpinning will come along with some settlement of the structure, and the difference in settlement from one point to another may cause structural damage (Ratay, 1996). Meanwhile, it is common that underpinning of foundations often causes damage to existing adjacent structures (Peraza, 2007). The consequences may involve injuries and loss of life, extensive property damage, construction delays, and expensive litigation (Peraza, 2008). (b) Causes According to literature review and report on past structural accidents related to underpinning of foundations, five major causes can be summarized. (1) Lack of underpinning The contractor often fails to take into consideration the condition of the foundation of the adjacent building, and conducts construction without underpinning it. Settlement, even collapse of the adjacent building happens frequently in this situation. It is noted that, even the pile driven vibration can damage the foundation of the adjacent building without proper protection. (2) Inadequate underpinning and bracing Due to the improper design or construction method, the underpinning and bracing system may be inadequate. For example, the poor material of underpinning system can be damaged easily by water penetration or collapsed due to the lack of capacity to hold the load overhead. An insufficiently installed underpinning system cannot stay firmly in the ground, thus can frequently collapse and result in settlement or shifting of the building. (3) Over excavation Sometimes, the excavation for underpinning is conducted more than required, and extends toward the adjacent property line. This affects the foundation of the adjoining property. (4) Impact from rubble foundation It would be very difficult for contractors to underpin a rubble foundation. This kind of foundation is composed of large stones, and cannot easily be connected and integrated with underpinning pits. The lack of continuity makes it difficult, and sometimes even unsuitable to underpin a rubble foundation (Peraza, 2006). (5) Impact from soil and groundwater The high level of groundwater makes it necessary to underpin the adjacent foundation, which is often neglected by contractors. It works the same with soils that are susceptible to consolidation

28 or vibration settlement. Meanwhile, the site should be dewatered with the existence of high water table. The dewatering of the site can cause the consolidation of soil, resulting in settlement of buildings (SEAoNY, 2005). (b) Case Study According to literature review, this section presented three case studies about structural accidents related to underpinning of foundations. (1) Severe weather (Peraza, 2006) A high rise complex was to be built near an old four story building, which was supported by rubble foundation. During the underpinning of the foundations, this old building settled excessively, resulting in wide cracks in the settled walls. It was found that the underpinning was well planned and executed, including a well-engineered plan and qualified engineers. However, during the work, a heavy rain greatly impacted on the old rubble foundation, which then resulted in significant settlement. (2) Improper underpinning In the 1980s, three old building collapsed during renovation in Lexington, KY. Although a proper underpinning plan was required, the excessive excavation undermined the footings, causing the building to collapse. No one was hurt, yet the contractor was required to rebuild the building. (3) Human negligence During the examination of the required underpinning location, the consultant missed the underpinning of one basement wall. When the excavator took out the soil, the footing was damaged and induced significant settlement. Although measurements were taken immediately to stabilize the building, the contractor was sued by the owners for millions of dollars in compensation.

3.3.1.4 Scaffolding System Scaffolding is used to provide temporary safe working platforms for the erection, maintenance, construction, repair, access or inspection, etc. of structures or other building systems (Grant and Pallett, 2012). It has been used for 5000 years to provide access areas for building and decorating structures taller than the people who worked on them (Ratay, 1996). The basic components of scaffolding are tubes, couplers and boards.

29 (a) Causes Whitaker, et al. (2003) examined 186 access related cases, and 2,910 incidents recorded in UK from 1997 to 2000, and summarized the most common root causes to the collapse of scaffold. (1) Improper ties to buildings. Scaffolds that are improperly attached to buildings are dangerous. Several scaffolds collapsed when the ties were removed after fitting. Some incidents occurred due to improper fitting or lack of ties. This kind of situation happens when there is a need to remove ties so that access to key areas can be reached. However, this modification of a scaffolding system is done randomly without qualified inspection and analysis. (2) Insufficient bracing within the structure. In the analysis of the 186 access related cases, 62 incidents are related to scaffold collapse (Whitaker, et al. 2003). 35.5% of the scaffold incidents occurred due to insufficient bracing system, which ranked a top cause of scaffolding collapses. (3) Overloading with building material. Some scaffolding collapse due to instability or overloading of materials. OSHA investigated the 16 structural failures related to scaffolding between 1990 and 2008, and revealed that 4 out of 16 incidents occurred due to the overloading of building materials (Ayub 2010). (4) Subsidence of foundations. Foundation provides permanent support of scaffold systems on the place where the system rests. Take the regular scaffolding system for example. The foundation of scaffold may be placed on soils with different capacities. Thus loads from scaffold will cause different settlement of the foundation, which then make the scaffold platform imbalanced, or causes collapsed. (5) Inadequate supervision. Most of the accidents related to scaffolding systems cited the unsafe working system as the general causation. This failure to access or control risk can be caused by deficiencies in the working system, defects of platforms, inadequate supervision, as well as improper work procedures. (b) Case Study In 2002, a suspended scaffolding system was used to restore the façade of John Hancock Center in Chicago, Illinois. Two outriggers were installed on the roof to hold the scaffolding platform, yet one outrigger overturned that afternoon and caused the scaffolding platform to swing back

30 and forth along the facade. The façade was disintegrated, and multiple windows and debris of scaffolding systems fell down, killing 4 people, and injuring 8 others.

3.3.1.5 Formwork Formworks are primarily used for standard poured-in-place concrete construction. They are used wherever the concrete is placed, such as a factory setting for precast sections and building sites. Various materials can be used for formworks, such as wood, steel, plastic, aluminum, etc. Formwork construction is associated with a relatively high frequency of disabling injuries and illness (Hallowell and Gambatese, 2009). With the increasing use of formwork, related safety issues have become serious problems (Shapira, 1999). In high-clearance concrete buildings, formwork collapse is defined as the failure of all or a substantial part of a structure (Kim 2006). Kim (2006) also pointed out that because of the potential collapse of elevated slab formwork during concrete placement, the assessment of the shoring system is essential. As for other formwork related injuries, 5.83% of falls and 21.2% of struck accidents mainly result from the construction of formwork (Huang and Hinze 2003). The preparation of formwork for concrete structures was defined as a dangerous stage of construction (Jannadi & Assaf 1998). (a) Causes Hadipriono and Wang (1986) studied 85 cases related to the formwork system collapse over a 23 years period, and found that almost half of the formwork system failures occurred during the pouring of concrete. The second critical stage is during formwork removal and post concrete curing. According to their study, the causes to formwork systems failures are summarized below (Hadipriono and Wang, 1986). (1) Improper/premature removal of formwork. Untimely removal of formwork is noted as the second most significant event, which is relative to the weak concrete and inadequate removal sequence of formwork. The premature removal of formwork usually comes from the desire to reuse form quickly either because of the pressure of scheduling or budget, while the concrete at that time might haven't attained the expected strength (Feld and Carper 1996). (2) Inadequate design of formwork system. Most of the cases related to design flaw are relative to the inadequate consideration of lateral forces and temporary structure's stability. The lack of a bracing system to deal with lateral forces,

31 like wind load and construction load, fails to prevent the formwork system from collapsing when an excessive load is imposed on it. In practice, the formwork components are reused, and the capacity to withhold a load will be reduced. Yet the designer of the formwork often omits the safety factor and calculates the load using the data of the original capacity. From the procedural perspective, the lack of review of the formwork design is also a big issue. Normally, the design of formwork should have been approved by an engineer before installation. Yet in several cases of formwork system failures, it has been identified that this step has been omitted. (3) Improper shoring of formwork. Several important incidents have occurred due to the improper shoring of formwork. It is found that the improper installation of shores is a significant cause of formwork failure, where impact loads from concrete debris and other effects trigger the collapse of vertical shores during concreting (Hadipriono and Wang 1986). (4) Defective component. Some cases of formwork system failure have been the result of the improper maintenance of formwork components, which then become defective after being reused several times. The capacity of these formwork components has been reduced due to corrosion and damages, yet it is seldom taken into consideration during the erection. (5) Improper connection. The formwork components are usually connected inadequately so that it is easier for workers to dismantle it. However, this lack of proper connection has induced several progressive collapses. Two types of improper connection have been identified. One is the lack of bolts, nails or splicing. Sometimes, there is no connection at all between two components. The other is poor weld quality and faulty wedges. (6) Insufficiently strong foundation. In the studies of the 85 cases, many foundations of formwork system failed to transfer the load to the ground, and some were laid on weak subsoil. These foundations are often constructed from mudsills, concrete pads, and piles, which are susceptible to differential settlement of formwork and overloading of shores, and finally results in collapse. Another problem related to the foundation is insufficient depth of the foundation piles, for it reduces the carrying capacity of the formwork. (7) Lack of inspection of formwork during concreting.

32 Pouring concrete is easily accompanied by formwork collapses, and many failures occurred when the inspector was absent or he just overlooked the problems. The lack of inspection also involves a situation in which the inspector is inexperienced or unqualified. (b) Case Study Three structural accidents are discussed in this section as examples of structures failures related to formwork. (1) Bailey's Crossroads - Skyline Plaza (March 2, 1973) On March 2, 1973, the improper removal of forms supporting the 23rd floor of an apartment building in Skyline Plaza triggered a progressive collapse all the way to the ground floor. 14 construction workers were killed, and 34 others were injured. (2) Harbour Cay Condominium (March 27, 1981) The Harbor Cay Condominium collapsed in Cocoa Beach, Florida, in 1981. One of the main causes was the premature removal of forms. As is stated by a worker on the jobsite, "twenty-two years I’ve been pouring concrete and they’ve never pulled the forms in two days like they did here. They usually set there for a week or 10 days” (Montgomery 1981). 11 workers paid their lives for this failure. (3) Collapse of New York Coliseum (1955) In 1955, in New York Coliseum, an exhibition hall collapsed during construction. It was found that the live load of buggies imposed more load than the formwork could hold. One worker was killed and fifty others were injured in the accident.

3.3.1.6 Temporary Performance Stages Temporary performance stages are defined as a structural assembly that is used for an outdoor performance for less than 90 days of one year (Wainscott 2011). Collapses of temporary performance stages have occurred frequently in recent years. In 2008, two of the stages for the Rocklahoma music festival collapsed, resulting in ten injuries when severe winds struck northeast Oklahoma. In 2009, the main stage of Big Valley Jamboree in Toronto collapsed, killing one and injuring at least seventy people during another wind storm. Additional collapses occurred in 2011, including the well-publicized Indiana State Fair Grandstand, which resulted in multiple fatalities and over fifty injured people in total. More recently, the Downsview Park in Toronto collapsed in 2012, killing one person and injuring three others, while another stage roof

33 collapsed in North Carolina in 2013 during bad weather. These accidents continue to occur with little warning to the general public. (a) Causes Based on past accidents, three major causes can be summarized as the major driving factors to structural failures of temporary performance stages. (1) Poor capability of components. Take the Sugarland stage collapse for example (Wainscott 2011). It has been revealed that four structural failures (jersey barrier, guy line and ratchet strap, fin plate) lead to the collapse of the main stage (2) Insufficient structural connection. The connection is often weak and easily damaged, especially under severe weather. The design or installation of connection is often overlooked, which then results in big problems. (3) The lack of engineering review after the stage is erected. There are few regulations on the responsibilities of engineers during the construction of temporary performance stages. Thus, the engineer is seldom required to inspect the stage after installation. Even if the structure is well designed, there might be a big difference between the actual installation and the requirements of engineers. (b) Case Study Temporary performance stages collapsed frequently every year. In 1990, a singer, Curtis Mayfield was hit by a scaffold of the stage during a show in Brooklyn. In 2008, two of the stages for Rocklahoma music festival collapsed, resulting in ten minor injuries. In 2009, the main stage of Big Valley Jamboree stage in Toronto collapsed, killing one and at least seventy people got hurt. More collapses occurred in 2011, with seven people passing away and around fifty people injured in total. Even after that, the Downsview Park in Toronto collapsed in 2012, killing one person, with three people injured.

3.4 Prevention of Temporary Structural Failures

Government and the industry have recognized the importance of safety of temporary structures. In an attempt to prevent temporary structures’ failures, various regulations and standards, industry practice, and education are provided or required. A detailed examination related to these suggestions and requirements are presented in this section. It can be found that although there are

34 efforts on prevention of temporary structures’ failures, there are still few safety regulations in some area of temporary structures. 3.4.1 Regulations & Standards This section reviews the concurrent regulations and standards of temporary structures, including earthwork shoring/sheeting system, temporary bracing system, underpinning of foundations, scaffolding system, formwork system, and temporary performance stages. (1) Earthwork shoring/ sheeting system Unless excavated entirely in stable rock, OSHA requires that all employees working in an excavation should be protected by supportive system. In OSHA regulations (Standards-29CFR), there are requirements for timber shoring, aluminum hydraulic shoring, pneumatic/ hydraulic shoring, trench jacks and trench shields. It is required that soil type should be examined using specified soil classification methods. In order to assistant in designing a shoring system, the required minimum dimension of shoring members is presented in forms of charts, so that designers can calculate the minimum size of members under specific conditions. Besides, OSHA requires that shoring/ sheeting systems should be designed by a registered professional engineer when it is to be used in an excavation deeper than 20 feet. For the excavation less than 20 feet in depth, a graphic summary of requirements is presented for easy application. (2) Temporary bracing system There are few regulations relative to the safe construction of a temporary bracing system. As for the temporary bracing system for masonry walls, the Code of Federal Regulations (CFR 1926.706(b)) simply specifies it as "adequate bracing", yet no more instructions are provided. The International Building Code (IBC) specifies the temporary bracing system as a secondary member. IBC simply states that inspectors should verify if the temporary bracing is installed as designed if the steel/ wood truss spans no less than 60 feet. (3) Underpinning of foundations Few regulations have mentioned underpinning of foundations. In the international building codes, the underpinning of a foundation is only specified as the required step before removing lateral support of foundations. Only a few local states hold some brief regulations about the underpinning of foundations. For example, New York City requires controlled inspections of the underpinning of foundations. Yet no more specifications are provided as for how to control it during the process.

35 (4) Scaffolding system OSHA requires that scaffolds should be designed by a qualified person, and constructed following this design. Several items are pointed out by OSHA. In terms of capacity, it states that the scaffold should be able to support not only its weight, but a specified maximum load applied or transmitted to it. For the scaffold platform construction, it requires the working platform on a scaffold to be fully planked or decked, except for the one used as walkways or to perform scaffold erection or dismantling. Besides, there are several limitations on the width of space between adjacent units and the space between the platform and the uprights. Other items include criteria for supported and suspension scaffold, scaffold access, use of scaffold, fall protection, and falling object protection (5) Formwork system It is required by OSHA that the formwork should be designed, fabricated, and maintained to be able to support all external loads that may be reasonably placed on the formwork. The design of shoring should be performed by a qualified designer. All shoring and reshoring equipment should be immediately inspected prior to, during and after concrete placement, by a qualified engineer. Once the shoring equipment is found to be weak or damaged, it should be reinforced immediately. Other regulations include the requirements of sills, base plate, shore head, and extension devices. In addition, the formwork and shoring should not be removed until the concrete reaches enough capacity. (6) Temporary performance stages Although structural engineers analyze the structural load for the permanent structures, there are few standards or guidelines on the design of temporary performance stages. In addition, while there are a few local requirements, such as New York and Chicago, no national regulations have about the safe construction and maintenance of temporary performance stages. Severe accidents of temporary performance stages have focused spotlight on the call for regulations of such a structure, and wind-load standards for temporary performance stages are also in high demand.

3.4.2 Recommended Practices To ensure the life safety of workers, the Mason Contractors Association of America (MCAA) published the Standard Practice for Bracing Masonry Walls under Construction. It provides procedures for the design of temporary bracing systems for masonry walls. In addition, a

36 restricted zone is specified, so that workers are forbidden to work in that zone once the wind reaches prescribed speed. This standard practice helps to reduce risk to the life of workers, yet it pays little attention on the prevention of structural failures.

There is a guideline (International Building Code, 2006) for the design, manufacturing and maintenance of temporary outdoor stage roofs, yet this guideline fails to provide regulations relative to the entire structure of temporary performance stages. In addition, this guideline is not promoted and followed nationwide. 3.4.3 Education According to OSHA regulation, employers are responsible to provide safety training and education programs to their employees. Meanwhile, OSHA and local governments provide lots of safety training resources online, including excavation safety training, fall protection, scaffolding, concrete, and masonry. However, there are few training resources related to underpinning of foundations, temporary bracing, and temporary performance stages 3.4.4 Limitations of Conventional Methods Due to the implementation of OSHA regulations, safety training programs, and industry safety practices, the fatality rate of the construction industry has reduced significantly. However, a recent study (Huang & Hinze 2006) on safety records shows that there are still numerous safety problems that need to be addressed. Recent temporary structures failures reveal that there are still some temporary structures, such as indoor and outdoor theatrical stages, which are not covered by any safety regulations (Mckiniley 2011). Furthermore, even with enough safety regulations and training programs, temporary structures failures could not be fully prevented as the workers tend to work under great pressure and make mistakes unconsciously (Fabiano, et al. 2008). Besides, as a passive method of protection, the use of Personal Protective Equipment (PPE) can only try to reduce the degree of injury to workers, instead of avoiding potential hazards.

Other researchers (Kim et al. 2011; Zhang et al. 2015) have explored the benefits of IT-based methods to prevent potential failures of temporary structures. However, most of the research focused on the safety design or plan of temporary structures with limited consideration of the dynamic environment on the construction jobsite. Besides, the proposed methods for real time inspection of temporary structures (Moon et al. 2012) are limited to passive inspection with few

37 interactions between the inspection system and the physical temporary structures. Finally, the applications of CPS in the construction industry are still limited with very few explorations of CPS applications in temporary structures monitoring.

3.5 CPS applicability to temporary structures monitoring

The realization of a complete CPS relies on several supporting technologies to integrate the virtual model with the physical construction (Akanmu et al. 2014). Although there have been no CPS applications for the temporary structures monitoring, some of the supporting technologies of CPS, such as Building Information Modeling (BIM) and Data Acquisition system have been utilized for enhanced monitoring of temporary structures. A brief review of these CPS supporting technologies’ applications to temporary structure helps to identify their benefits and limitations, so as to provide a clear understanding of CPS applicability and potential benefits to temporary structure management. 3.5.1 Use of BIM for temporary structures management For better virtualization of temporary structures, Chi et al. (2012) proposed to develop BIM objects of temporary structures, such as scaffolding system and formworks. These BIM objects will be embedded with design, construction and safety information as references to other parties. Similarly, a safety-rule based BIM for temporary structures (such as a scaffolding system), was developed, with special focus on automatically identifying and eliminating potential fall hazards during the design stage (Zhang et al. 2015). In addition to adding safety regulations to the BIM model of temporary structures, Kim et al. (2011) presented a safety identification system for temporary structures, which identifies and predicts potential safety hazard by simulating construction schedules and checking the location of temporary structures at each step. Besides, Li et al. (2008) proposed to integrate the design and construction of temporary structures through virtual prototyping. In view of the difficulties in identifying appropriate temporary structures to be shared among projects for cost saving, Kim et al. (2014) proposed the use of BIM technology for quick identification of sharing solutions of temporary structures among different projects. All of these efforts have benefited the visualization, design, and safety planning of temporary structures. 3.5.2 Use of Data Acquisition System (DAQ) for temporary structures management A DAQ refers to computer based systems with digital input and output (UEI 2006). With

38 developing technologies, DAQ has been recognized as important to prevent construction failures by providing information, and has been increasingly utilized for temporary structures management (Moon et al. 2012). These efforts include the use of Radio Frequency Identification (RFID), wireless sensor networks, and videos. As early as 2007, Yabuki and Oyama (2007) used RFID to record the usage history of temporary structures, so that project managers can understand how long the temporary structures have been used, in order to decide whether it is safe to keep using them. Ubiquitous sensor network technology can be used to determine the structural performance of temporary structures by analyzing deflection, load, strains, etc. (Moon et al. 2012). It provides a real-time approach to monitor formwork operations as a means of preventing structural failures. Most recently, Jung (2014) proposed the use of video method to detect potential defects of temporary structures. This system will continuously record the images of temporary structures, predict potential structural defects, and send warnings if there are potential hazards. With the development of DAQ, more structural information of temporary structures can be obtained for comprehensive structural analysis.

While the supporting technologies of CPS in temporary structures are still at the early stages of implementation, their benefits and limitations highlight the opportunities for further applications of CPS. Implementation of CPS in preventing failures and promoting safe construction techniques of temporary structures remains promising as discussed below.

3.6 Potential Areas for CPS Applications in Temporary Structures

For the purpose of identifying potential areas for CPS application, temporary structures which have historically been involved in a high record of failures were selected for discussion. These cover two general categories - temporary performance stages and temporary support systems. 3.6.1 Temporary performance stages A temporary performance stage is a structural assembly that is used for an outdoor performance for less than 90 days of one year (Wainscott 2011). Collapses of temporary performance stages have occurred frequently in recent years. In 2008, two of the stages for the Rocklahoma music festival collapsed, resulting in ten injuries when severe winds struck northeast Oklahoma. In 2009, the main stage of Big Valley Jamboree in Toronto collapsed, killing one and injuring at least seventy people during another wind storm. Additional collapses occurred in 2011, including

39 the well-publicized Indiana State Fair Grandstand, resulting in multiple fatalities and over fifty people injured in total. More recently, the Downsview Park in Toronto collapsed in 2012, killing one person and injuring three others, while another stage roof collapsed in North Carolina in 2013 during bad weather (Kleinosky 2012). These accidents are also related to the lack of authoritative standards for temporary structures and performance stages (McKiniley 2011). This makes the need for a proactive monitoring system more urgent (Yuan et al. 2014). 3.6.2 Temporary support systems Temporary support systems serve to help carry or support a structure or provide safety access for workers during the construction process. They are categorized into four types (discussed below): scaffolding systems, earthworks, formwork, and temporary bracing systems.

Scaffolding is used to provide temporary safe working platforms for the erection, maintenance, construction, repair, access or inspection of structures or other building systems (Grant and Pallett 2012). According to the U. S. Bureau of Labor Statistics, approximately eight workers working on scaffolding system are hurt each month at the United States construction jobsite (BLS 2013); sheeting and shoring (using systems such as steel soldier piles, sheet piles, and slurry walls) are used to prevent soil movement and cave-ins during earth excavations. Inappropriate design and installation of earthwork shoring and sheeting systems results in numerous accidents each year, making earthworks a substantial risk for workers; formwork systems are primarily used for standard poured-in-place concrete construction. Formwork construction is associated with a relatively high frequency of disabling injuries and illness (Hallowell and Gambatese 2009). This is now recognized as a serious problem (Shapira 1999); temporary bracing systems are used to keep a structure or building system stable before the permanent bracing is installed or the element becomes self-supporting. Insufficient bracing is cited as one of the four top causes of failures in steel structures under construction (Kaminetzky 1991). The structural load is usually analyzed by conceiving the whole structure as a completed entity, and there is frequently a lack of design or proper implementation of the temporary bracing systems. Often, the specific provisions and requirements of temporary bracing systems are left to the workers on the job site that may not have the qualifications or expertise for proper execution (Feld and Carper 1996).

40 3.7 Potential Benefits and Barriers

Based on the section of “key features of CPS” and “applicability of CPS in temporary structures” discussed above, several potential benefits and barriers to CPS implementation are identified as follows. 3.7.1 Potential benefits CPS offers the potential for improved monitoring of temporary structures through real time coordination between virtual and physical systems. By systematically implementing CPS, a number of potential benefits can be achieved as follows:  Real time inspection: in lieu of inspecting specific influential factors, CPS can monitor the performance of temporary structural components in real time, and provide warnings that can help to ensure structural stability.  Tight coordination between physical component and virtual model: CPS enables bi-directional communication between physical components and their virtual representations. Through the “Physical-to-Cyber Bridge”, the movement of physical components will be detected by sensors and sent to the virtual model, where the difference between designed and actual structure will be highlighted on the model. Through the “Cyber-to-Physical Bridge”, once potential hazards are detected, safety alerts or instructions can be sent from the virtual model to the workers or a mechanism to automatically stabilize temporary structures can be actuated if appropriate.  Remote and multi-party access for management: due to the bi-directional information loop between the physical and virtual components, the structural performance can be remotely monitored using virtual models. CPS enables multiple parties to obtain access to the structural monitoring system from different locations through the interface of virtual models. This function benefits project managers, structural designers, owners and other involved parties for routine monitoring, potential structural problem analysis, instruction to avoid potential dangers, and enhanced collaboration.  Early warning with customized safety level: Implementation of a dynamic CPS environment has the potential to shorten the time interval between the on-set of an initial hazard and potential collapse. In addition to the quick identification of potential problems, the system can prevent hazardous situations by setting a safety factor for the performance of temporary

41 structures. In this way, instead of having alarms when there are already structural failures, an early warning will be sent once there is a trend or high possibility of potential structural failures. The safety factor can be used as recommended or customized if the user wishes to have a higher level of safety.  Knowledge base to predict potential hazards: during the structural monitoring of temporary structures, all the performance data of the temporary structures and instructions at different situations is recorded in database. The relationship between instructions and structural problems can be learned and identified, which enables CPS to automatically provide suggestions for instructions when having similar structural problems. 3.7.2 Potential barriers to CPS implementation The preceding discussion has demonstrated that there are several temporary structures applications that can benefit from the implementation of CPS for monitoring and performance purposes. However, there are also barriers and technical issues to be addressed. Some of these include:  Security: There is growing concern about cyber-attacks on CPS, as computing systems and sensor networks are unable to work effectively under malicious attacks. Furthermore, attacks on CPS used in commercial business or hospital environment might disclose personal information (Cardenas et al. 2008).  Reliability: Random failures in CPS may occur due to system errors, inaccuracy of data, and data interference. Sensed data is susceptible to a reduction in accuracy due to interference from other signals such as Wi-Fi or other electronic devices (Akanmu 2012). However, modern construction job sites involve numerous kinds of electronic equipment which impacts the accuracy of sensed signals. Physical damage to a sensor due to construction operations or impact is also a possibility.  Training of workers: CPS involves the management and installation of new technologies, including hardware such as sensors and actuators on job sites. These technologies need to be implemented, tested and inspected on a continuous basis. It requires that construction personnel are adequately trained in the use of these new technologies.  Financial issues: although the price of sensors has dropped, accurate and quick detection requires a large number of sensors to work simultaneously. In addition, the use of actuators, the connection system between sensors, information platform and actuators, and the training

42 of workers have cost implications that will need to be addressed.

3.8 Summary

This chapter reviewed the key features of CPS, examines current CPS applications in the built environment, common types of temporary structures, and analyzed the applicability, potential benefits and barriers for CPS application to temporary structures. It identified the promising application areas, and discussed how CPS could be applied for safety monitoring of temporary structures.

A number of key conclusions could be drawn from this chapter. Generally, the temporary structures failure remains a big risk to the safety management of construction projects. The inadequate monitoring of temporary structures calls for a more proactive, intelligent monitoring system, such as CPS. CPS provides an effective way of integrating the real world with its virtual representation, and has been implemented in several industry sectors, such as the manufacturing industry, power grids, the transportation industry, and the healthcare industry. More recently, CPS has shown potential benefits to the built environment through its successful application in structural health monitoring and facility management. Therefore, CPS offers an opportunity to address the current problems and safety issues associated with temporary structures, including temporary performance stages and scaffolding. To be specific, a CPS for temporary structures monitoring enables remote control, “on physical component instructions”, and real-time interaction between temporary structures and their virtual representations.

43 Chapter 4 Prototype System Development

4.1 System Development Framework

With the application of CPS, the advanced structures monitoring is enabled through tight integration between physical temporary structures and their virtual models. Possible system architecture of CPS-based temporary structures monitoring is shown in Figure 4-1. It works in the way that first of all, the DAQ system placed on temporary structures on the job site collects and sends structural data (information such as inclination, loading, deflection, etc.) to the database; Second, the structural information is continuously fetched by the virtual models for structural performance analysis; Third, the virtual models are updated based on the most recent structural information. If there is a potential hazard, the components in question will be marked in red color for attention. Meanwhile, a warning will be sent to safety supervisors through portable devices; Once receiving the warning message, supervisors and project managers review the structural performance of the scaffolding system in the virtual model and send instructions to workers; Finally, workers take action to correct the dangerous situation based on the instructions. In addition to this solution, various system architectures of a CPS-based temporary structures monitoring may work well with seven core framework elements (discussed below) to be considered.

Figure 4-1: Possible System Architecture of CPS-based Temporary Structures Monitoring

44 Framework Element 1: DAQ System: a DAQ system refers to computer based systems with digital input and output (UEI 2006). With developing technologies, DAQ systems have been recognized as important to prevent construction failures by providing information, and have been increasingly utilized for safety management of temporary structures (Moon et al. 2012). By transferring physical conditions into digital signals, a DAQ system provides an effective way in detecting structural information of temporary structures. Examples of DAQ system include RFID, wireless sensor networks, 3D laser scanners, and photography (Akanmu et al. 2014). The selection of DAQ system is based on the type of information to be collected, information accuracy requirement, and the installation of DAQ equipment.

Framework Element 2: Physical-to-Cyber Bridge: the Physical-to-Cyber Bridge is the process of sending information of physical components to the virtual model. By linking each individual physical component and its corresponding virtual model, the virtual model becomes alive with the real time structural information, which also enables the real time structural performance analysis and automatic visualization of structural deficiency at the virtual model.

Framework Element 3: Virtual Modeling Platform: the virtual modeling platform provides the user interface of CPS-based temporary structures monitoring, as well as the virtual model displaying exactly the performance of physical structures. There are various virtual modeling platforms that could be utilized for CPS-based temporary structure monitoring. Examples of such virtual modeling platforms include Revit Autodesk, Navisworks Autodesk, and Tekla BIM software. All of the three virtual modeling platforms mentioned above support the development and management of 3D virtual model with digital information.

Framework Element 4: Cyber-to-Physical Bridge: the Cyber-to-Physical Bridge is the process of adjusting or monitoring the physical structures through its virtual model. Based on the real time performance analysis of temporary structures, the CPS-based temporary structures monitoring system can monitor or take appropriate corrective control of the physical structures through the virtual modeling platform. This can be achieved through the controlled activities of actuators attached on the physical components or the guided performance of supervisors and construction workers on the job site.

Framework Element 5: Actuators or Portable Devices: an actuator is the controlling instrument installed at the physical structure which behaves according to the instructions from virtual systems. An example of actuators is the air damper actuator used in ventilation systems. It adjusts its position according to the instruction of ventilations system so that to guarantee the indoor air quality. In addition, the monitoring of temporary structures can also be realized through the performance of supervisors or construction workers. In such a situation, the portable devices play as the communication devices between virtual models and supervisors, which receive warning and instructions from virtual models and aids supervisors in the decision making for safety management of temporary structures.

Framework Element 6: On-cloud Database: on-cloud database serves multiple roles in supporting the CPS Bridge. First of all, it provides the connection between DAQ system and the virtual modeling platforms, where the structural information can be continuously uploaded from the physical components and queried by the virtual models. Second, portable devices can get remote access to the detailed information of the structural performance through the on-cloud database. Third, the historical records of potential structural failures can be stored at the on-cloud database for structural performance overview. Besides, the on-cloud database also enables the remote access to the structural performance of temporary structures by multiple participants (such as owners, structure engineers, project managers, etc.) from different places of the same time.

Framework Element 7: Communication Networks: communication networks serve as the channels for information transmission among the core framework elements discussed above. For example, the DAQ system, on-cloud database, and the virtual modeling platform are connected through the communication networks for real time structural performance analysis. Similarly, the portable devices and virtual modeling platform are seamlessly connected through the communication networks for appropriate warning of potential structure failures. In general, the communication networks include the Local Area Network (LAN), Metropolitan Area Network (MAN), Wide Area Network (WAN), and wireless network (WiFi). To be specific, the LAN is widely used at small physical areas, such as offices; the MAN can be viewed as a bigger version

46 of the LAN, which can be set up by connecting several LANs into a larger internet or a single cable; WAN, usually supported by satellite links, is often used for the internet with large distance; wireless network is the fastest growing communication network used in our daily life, which provide the physical equipment with wireless access to the internet (Studytonight 2014). Due to the various objectives of temporary structures monitoring, several types of communication networks can be used in combination.

4.2 Framework of CPS-based Scaffold Systems Monitoring

Based on the identification of the key system elements of CPS to temporary structures (as discussed in previous section) and the system architecture of CPS application to scaffold system, the framework of CPS-based scaffold systems monitoring, as shown in Figure 2, is summarized as follows:

Identify the Identify Warning Tie Physical Scaffold Key Information Threshold to Virtual Model

Potential Structural Hazards Transfer Information from Prediction in the Virtual Physical Scaffold to Virtual Model Model

Remote Monitoring of Visualization of Structural Scaffold Systems through Deficiency Virtual Model

Figure 4-2: Framework of CPS-based Scaffold Systems Monitoring

4.2.1 Identify the Key Information to Check With increasing development of DAQ technology, a variety of field information, such as temperature and 3D laser modeling, can be achieved with high accuracy. However, not all the information is necessary or helpful to an effective monitoring of scaffold systems. Therefore, the information to be detected for structural performance analysis is critical and should be carefully selected.

The failure mode analysis of scaffold systems summarized by Yuan (2013) concluded that there were four main structural failures of scaffold systems, including base settlement, overloading on

47 scaffold systems, insufficient braces, and wind impact. Therefore, in order to identify and prevent these structural failures, four typical types of structural information are required for structural monitoring, which includes loading information on the scaffold system, displacement of scaffold planks or plank supports, connection between the base and the scaffold post, and inclination of the scaffold posts with lateral load. All of this information works together for comprehensive structural integrity analysis and assists in the prediction of potential hazards. 4.2.2 Establish Warning Threshold In addition to identifying the key information to be checked, one should also indicate their safety requirements for the structural performance by setting a warning threshold as the safety check line. In general, the warning threshold is the predefined value in TSM as the reference point for potential structural failures. By comparing the actual structural behavior information and the warning threshold, one can automatically predict if there is a potential for structural failure or excessive movement.

According to the OSHA standards and industrial best practice, the load placed on the scaffold system should be no more than its rated working capacity; the maximum allowed displacement of scaffold planks is 1/60 of the span; furthermore, disconnection between the bases of the scaffold system and the ground or support surface is not allowed. While there is no specific requirement of maximum allowable inclination of the scaffold system, it can be set by the safety experts or users based on heuristics. In reality, the user can also define higher warning thresholds for more stringent safety requirements. 4.2.3 Tie Physical Scaffold to its Corresponding Virtual Model To integrate the physical scaffold and its virtual model, three main steps was taken appropriately: first, set-up a scaffold system in the physical world; second, develop a virtual model with exactly the same structure and appearance; finally, register each scaffold component with its corresponding digital representative in the virtual model.

Following the three main steps, a single frame scaffold (as explained in the previous section), as shown in Figure 4-3, was set up as a physical scaffold to be monitored; meanwhile, a virtual model of the frame scaffold is developed in Autodesk Revit and imported into Autodesk Navisworks as the virtual platform of the TSM. This is because the Autodesk Navisworks

48 provides open .NET application programming interface, which includes extendable user-defined application which can trigger the system from the outside user interface. Besides, compared to Autodesk Revit, Autodesk Navisworks provides more freedom to developers in managing the properties of virtual models; finally, for recognition of each physical component by the virtual model, the physical components and their corresponding virtual representations are closely linked by registering the name of each physical component with their unique ID in the virtual model.

Figure 4-3: Virtual Model and Physical Set-up of Frame Scaffold

4.2.4 Transfer Information from Physical Scaffold to Virtual Model A DAQ system and an on-cloud database are used for information collection and transmission. To be specific, several sensors (such as load cell for loading information, switch sensor for disconnection information, displacement sensor for displacement of planks, and accelerometer for inclination information), are used for structural information detection. This information is retrieved by NIMAX (National Instrument Measurement & Automation Explorer, which provides access to the information from the sensors), calibrated through Labview (which provides the platform of data calibration), and then exported into the on-cloud database.

The on-cloud database is developed at the Amazon Elastic Compute Cloud (EC2), due to its benefits of virtual computing environment, high security control, and static IP addresses for dynamic cloud computing. The on-cloud database stores both the real time structural information and the historical record of potential structural failures of the scaffold system. To manage and check the status of the on-cloud database for system maintenance, the HeidiSQL platform is used at local computers for database management. 4.2.5 Potential Structural Hazards Prediction in the Virtual Model. A plugin was developed at the Autodesk Navisworks to analyze the real-time structural

49 performance of the scaffold system. With the warning threshold predefined by users, and the continuously received the structural information from the DAQ system, the developed plugin at the virtual model is capable of identifying potential hazards accordingly. For example, if the actual displacement of a scaffold plank exceeds its corresponding warning threshold, the serviceability deficiency of the scaffold plank can be predicted immediately by the plugin at the virtual model. 4.2.6 Visualization of Structural Deficiency A direct reflection of potential problems in virtual model provides an effective way to locate the potential problems. The CPS-based scaffold system monitoring system works in the way that once a potential hazard or performance deficiency is identified, the components in question will be visualized (marked in an easily identifiable color) in the virtual model, with the aim of providing direct and detailed information to assist safety inspectors and project managers in the safety inspection of the scaffold system. Meanwhile, the virtual model will inform the project manager of how the warning and instruction has been sent to the site supervisors, who is not in front of the virtual model at the office but working on the job site, regarding this potential hazard. 4.2.7 Remote Monitoring of Scaffold Systems through Virtual Model The remote monitoring of scaffold systems is achieved through the process of Cyber-to-Physical Bridge. For safety concerns, immediate warnings are expected to be delivered to construction site supervisors/foremen. Although a degree of automatic structural control can be achieved by using actuators (which have been used in the control of civil infrastructures by CPS), they may induce movements that affect the stability of workers working on the scaffold system, which may impair the stability of the system itself.

Therefore, to ensure the safety of construction workers, the monitoring is designed to be completed through interaction with the directed intervention of construction supervisors following notification by the CPS system. In doing this, a mobile APP, has been developed and envisioned to be installed on the smart phone of construction supervisors. Once a potential hazard is identified, the mobile App is activated and it alerts the relevant supervisors through vibration and a warning alarm. Along with it, an overview of the structural performance of the scaffold system is displayed as an instruction to supervisors, so that appropriate corrective

50 actions can be taken to correct the structural deficiencies.

4.3 User Needs Identification

A useful system requires the developer to first of all understand how the developed system is expected to work by end users. Therefore, before comprehensive system design and development, the system requirements of the proposed TSM were analyzed and summarized based on the identification of end user requirements. 4.3.1 End User Requirements Identification The CPS-based TSM system is targeted at assisting construction workers, safety inspectors and project managers with the safety management of temporary structures. The user requirements are identified through the review of literature and accident reports published by OSHA. These user requirements are further adjusted and confirmed through interviews with three professionals in charge of safety management on construction jobsite conducted as a part of this research. Based on the interviews, several key requirements are identified as follows.  Functionality. In general, the CPS-based TSM should be able to identify the most frequently occurring structure failures incorporating a clear user interface and limited errors. For active control of temporary structures monitoring, the end user should be granted some freedom in setting the level of hazards to be tracked. Close communication between the virtual model and the actual physical temporary structures, as well as between humans and CPS-based TSM system are required.  Real time inspection. The dynamic construction environment highlights the importance of continuous and real time inspection of potential hazards. It is recommended by the end users that an effective temporary structure monitoring system should keep an eye on the performance of temporary structures as frequently as possible.  Early warning. It is commonly known that prevention of potential failures is much more important than discovery after actual collapse. In the hope of avoiding potential failures of temporary structures, it is required to have an early warning so that potential problems can be corrected before actual accidents occur.  Immediate instruction to construction workers. Based on the working experience of interviewees contacted as part of this research, it is more important for construction workers,

51 rather than safety inspectors and project managers, to receive immediate instructions and warnings. Due to the urgency of potential hazards, the construction workers are expected to take corrective action or get away from the dangerous working zones immediately. As is indicated by the interviewees, there is commonly an absence of proper management when safety professionals are away from the construction jobsite while potential failures of temporary structures occur. This delayed safety management can lead to serious accidents.  Simple and clear. For safety concerns, the use of CPS-based TSM should be simple to use and the information sent to the end users should be easy to understand. In concern for the education background of some construction workers, a simple system can be promoted more quickly and used correctly. Besides, the simple and clear feature of TSM also assists safety inspectors and project managers in safety management without the need for special or extensive training programs.  Nonintrusive to daily work. Portable monitoring or warning devices should be wearable and nonintrusive to the work on a construction jobsite. While the construction workers are working at the jobsite, it would be impossible for them to carry any large devices, which may be inconvenient or cause rather than of prevent potential dangers to workers.  Remote interaction. To obtain immediate notification and updated information for safety inspectors and project managers, the developed TSM should enable remote interaction between the temporary structures on the construction jobsite and safety inspectors and project managers anywhere, even when they are far away from the construction jobsite. Remote interaction is also important when there are serious structural deficiencies that are out of the control of the project manager and need immediate input from structural engineers. In other words, it would be beneficial if structural engineers could get remote access to the structural performance of the temporary structures for structural analysis and problem diagnosis.  Accuracy and trustworthiness. On the construction jobsite, even a small change or difference in the structural performance may lead to tremendous failures. Therefore, it is important to have an early detection and resolution of the potential problems. For example, the change of inclination of a vertical component should be identified within one degree. 4.3.2 System Requirements In view of the end user requirements above, and the available IT-based methods, the system requirements of a CPS-based TSM are identified through the analysis of end user requirements

52 and the key features of CPS. Details are discussed as follows.  Predefined system logics for inspection of most frequent structure failures. This is intended for simple use by the end users with comprehensive temporary structures monitoring. The key failure patterns of temporary structures should be explored for the design of algorithms for potential failure identification. The predefined system logics also minimize the requirement of professional knowledge at the stage of operation, which make TSM a simple and easy tool for the end users.  User-defined threshold for potential hazards. Along with the suggested threshold predefined according to relative safety regulations and manufacturers’ suggestions, for active management of temporary structures by the end users, the system should provide the end users access to the CPS-based TSM so that the threshold of potential hazards can be self-defined, when more stringent requirements are necessary or desirable.  Real-time and Remote Communication. To enable real time and remote interaction between temporary structures and their virtual models, continuous wireless connections must be set up. Since the temporary structures work in the dynamic construction environment, the inspection of the temporary structure should be in real-time (within a specified time period), so that potential structural defects can be quickly identified and an early warning issued before actual accidents occur.  Portable and wearable devices to receive warnings and instructions. Devices carried by construction workers should be small enough to be carried without affecting their daily work activities. In addition, instead of just a simple alarm without information about what is going on, the wearable devices should be able to provide clear information on potential hazards to notify construction workers where the problems are, and to provide safety inspectors and project managers with detailed information for decision making.  Visualization. Good visualization provides clear understanding of the potential problems. The requirements of visualization include two aspects – visualization of the temporary structures and the visual presentation of response instructions. The virtual model of temporary structures is one of the basic systems of CPS, which provides user interfaces for remote inspection and control of the physical model. The virtual model consists of digital components with unique IDs that can be recognized independently by the computing system. It is important that the virtual model provides basic properties and information on the

53 temporary components, which assists professionals in structural analysis and decision making in preventing potential collapse. Besides, to avoid misunderstanding, the instructions, along with the picture identifying the components in question should be presented clearly to workers.  Data Precision and Resolution. The structural analysis of temporary structures has a high requirement for data accuracy and resolution, because of the potential for failures due to the dynamic construction environment. In practice, there is always system error and signal noise along with the observed information. The high requirement of data accuracy calls for an appropriate data processing method for an accurate and confident level of structural analysis. Besides, the selection of the resolution of the DAQ is based on the level of details of temporary structures performance to be checked. It is recommended to use the one with appropriate level of precision due to concern of noise interference (Moon et al. 2012), so that more detailed structural performance of temporary structures can be captured.

4.4 Choice of Development Environment

The selection of the system development environment involves identification of hardware and software environment for developing the prototype system. Both of these work together as the development platform for the CPS-based TSM system.

4.4.1 Hardware Environment The hardware environment provides the equipment required for system development or simulation. Examples include a set of temporary structures, sensors for structural performance monitoring, and portable devices for communication between TSM and construction workers.  Temporary structures. While there are various types of temporary structures, the scaffolding system has been selected as an example of the CPS-based TSM for two reasons. First of all, scaffolding systems, ranking third among top 10 OSHA violations in the year of 2013, account for a large number of fatalities and injuries in the construction industry (U.S. Department of Labor 2014). Secondly, the principles of structural monitoring for scaffolding systems are similar to those of other types of temporary structures, which indicates that the TSM of scaffolding system can be easily adapted for other temporary structures. Therefore, one of the most commonly used scaffolding systems, a simple set of frame scaffold (5 ft. x 5

54 ft. x 7 ft), as shown in Figure 4-3, was set up in the laboratory for system testing. This scaffold consists of two 5ft x 5ft scaffold frames and two sets of 7ft x 4ft cross braces. Six OSHA certified wooden scaffold planks were placed at the top of the scaffold frame, to provide a working platform to support workers or materials.  Sensors. In order to detect the structural performance of the scaffolding system, four main types of sensors are adopted for data acquisition; these include load cells, switch sensors, an accelerometer, and a displacement sensor. The locations of the various sensors were carefully determined by taking into account the required information for analysis, data accuracy, requirement of sensor installation, and sensor sensitivity. More details are discussed in Chapter 5.  Portable devices. In this study, a smart phone running android operating system is used as the portable device for construction workers, safety inspectors, and project managers. The idea of using a smart phone is to provide the end users information from the CPS-based TSM without requiring additional devices. However, the end users are not restricted to the smart phone for safety inspection of temporary structures. For the user’s convenience, other smart devices (such as tablet devices and smart watches) may also be used, as long as the operating system is compatible with the developed TSM system. 4.4.2 Software Environment The development of CPS-based TSM involves the integration of several systems, including the communication network. In general, the core system is built as a plug-in of Autodesk Navisworks Management, which provides the user interface for end users. This plug-in relies on data collection systems and data communication through a communications network. The choice of each system is analyzed below:  Virtual model and Plug-in system. Emerging BIM software provides a variety of choices for visualization of temporary structures, such as Autodesk Revit and Autodesk Navisworks Management. These software provide an open .NET application programming interface (API), which provides an extendable user-defined application that can trigger system operation from outside the user interface. However, compared to Revit, Navisworks provides more freedom to developers in managing the properties of virtual models, which is highly necessary in the TSM based on the system requirements. Therefore, Autodesk Navisworks

55 Management was selected for the development platform of the virtual model and plug-in system.  Data collection platform. National Instruments hardware were used for data acquisition from the physical mock-up of the temporary structure, its associated system –design platform, which is called Labview, was used as the data collection platform. Lab view provides extensive support for interfacing with a variety of devices, such as sensors and cameras. It also has a large number of functions related to data acquisition, signal conditioning, and a GUI (Graphical User Interface) for quick and stable data acquisition.  Communication network between PC and portable devices. In this context, there are three main methods of communication: point-to-point communication, cloud computing service, and text message by setting up a virtualized sever which sends text message of alarm to cell phones when potential hazards appeal. While it is easy to set up point-to-point communication, it requires the setup of a router and server at the DAQ control system. Besides, this communication method requires a change in IP address every time another computer is used. The use of text messaging avoids the development of application at the portable devices, yet it serves more as a passive warning system with limited information and little interaction from construction workers. Besides, a text message can be easily ignored by construction workers for it only provides one alert/notification, which is not different in sound to other text messages. The use of cloud computing services was identified as the most appropriate communication method for this CPS-based TSM for two main reasons. First of all, the cloud computing service provides a stable and quick communication network between different systems on several devices from different locations without changing settings, such as the IP address. It also provides a larger database, which can be extended and restructured for more complicated applications, such as the storage of historical record of each identified potential hazards during construction.  Database Management System. Heidi SQL, also previously known as MySQL-Front, was utilized to manage the on-cloud SQL from local computers. Heidi serves as a simple and open client for database management. It provides a useful and stable way to edit data, create tables and export data. There is a free version of Heidi SQL, which can be used at no cost to the developers.

56  Development Platform for portable devices. While both IOS and Android platforms are available for mobile app development, the IOS platform only provides a closed source, which means that the developed app can only be used on specific types of devices. On the contrary, the Android platform provides open source which enables mobile apps to be used on any devices. To enable most of the construction workers and safety inspectors to use the CPS-based TSM, the android platform was selected for mobile app development.

4.5 Overview of System Architecture Generally, the proposed CPS-based TSM system consists of physical structures and their virtual models, which are integrated through a CPS bridge supported by a communication network and database. The system architecture is demonstrated in Figure 4-4 and described below:  Physical structures: the physical structures are the ones discussed in section 1: “Introduction”, and include both the temporary support systems (such as scaffolding systems and formwork) and temporarily used structures (such as temporary performance stages).  Virtual model: the virtual models are the virtual representations of the physical structures described above. The virtual models can be developed using 3D/4D virtual prototyping software.  Cyber-to-Physical Bridge: the Cyber-to-Physical Bridge enables the information exchange between the physical structures and the virtual models. The Cyber-Physical Bridge is supported by the communication network and an on-cloud database.  Physical-to-Cyber Bridge: real time information on physical structures is collected using data acquisition systems (DAQ) and transmitted to the cyber system via the communication network.  Portable devices: the portable devices are used as communication tools between the TSM system and construction site personnel who may need to adjust or manipulate the physical structures based on the instructions from the virtual model.

57 database

Communication network

to Virtual Model Physical Structure

CPS Bridge

Portable Devices Virtual Model Remote Control Physical Structure User BIM Data Acquisition Input API Data Updates System

Figure 4-4: System architecture of CPS-based TSM

4.6 Development of TSM The CPS-based TSM system was developed based on the hardware and software environment identified above. The key steps of system developments include the development of the 3D virtual model, plug-in system, on-cloud database, and mobile app. Details of each development step is presented as follows.

4.6.1 3D Model and Plug-in Development. A 3D model of the frame scaffold system was initially developed in Autodesk Revit 2015, as shown in Figure 4-5. This Revit model was then imported into Autodesk Navisworks Management 2015 for further modification. To make the virtual model look more realistic compared with the real scaffolding set and to distinguish the various elements, the color of each component was adjusted, as shown in Figure 4-6.

Figure 4-5: Revit Model of Frame Scaffold

Figure 4-6: 3D Model of Scaffold in Navisworks The plug-in for CPS-based TSM was then developed and launched as an add-in tool shown at the Autodesk Navisworks Management 2015. In doing this, the plug-in program was initially developed and compiled using Microsoft Visio Studio, and saved at the specified document address to be triggered by the software of Autodesk Navisworks Management every time the software is opened. A GUI was developed for user input and interaction. Also, a user input form was created and saved as a text file at the local computer so that the plug-in system can remember the latest user input without the need to re-enter it every time the plug-in is opened.

After entering the user input, the CPS-based TSM starts inspection once the “start” button is clicked. Every 2 seconds, the TSM will collect updated data from the database and identify potential structural failures. It works by post processing the updated information to the required format and compares it to a set of limitations, which is also known as the user-defined threshold for potential structural hazards. Once there is potential for structural failure, the components in question will be highlighted in the virtual model, and a warning notification is sent immediately to the portable devices. 4.6.2 Database Development

59 Generally, an on-cloud database is a database that typically runs on a cloud computing platform via the Internet (Arora & Gupta, 2012). An on-cloud database is necessary to the development of the CPS-based TSM system for three main reasons. First, the cloud database enables remote inspection of the physical structures from the virtual model via the Internet; second, multiple users can gain access to the TSM system via the Internet for real time inspection and data analysis; third, as has been discussed in the section on “Communication Network between PC and Portable Devices”, the communication between the virtual model and the mobile app is supported by the on-cloud database.

4.6.3 Mobile App for Human-Machine Interaction An android mobile app, named TSM, was developed and installed on a smart phone, as shown in Figure 4-7. Every 2 seconds, the mobile app checks if there is any warning or instruction sent from the TSM system. Once a warning or instruction is received, the mobile app is automatically activated with alarms that will keep ringing until “responded to” (tap the notification message) by the end users. Once the notification message of TSM app is clicked, the app is opened with a picture showing the virtual model with the components in question highlighted, so that the end users have a direct understanding of the location of potential problems. Meanwhile, a text message notifying the problem is shown at the bottom of the picture. More detailed information on the structural problems of each component can be obtained by tapping the “detail” button.

Figure 4-7: Mobile App of TSM

60 4.7 Summary

This chapter established a system development framework for the CPS applications in temporary structures monitoring. In particular, it selected scaffolding system as an example of temporary structures to demonstrate how a prototype system could be developed following the framework. In doing this, user requirements and system development environment are identified. Finally, a prototype system, TSM was developed.

It can be concluded that an effective TSM is required to provide real time inspection, remote interaction, early warning, and immediate instruction to construction workers. Besides, the use of TSM should be simple and nonintrusive to the daily work of construction workers; the development of CPS-based TSM involves the setup of hardware, such as the temporary structure, and the choice of an appropriate system development platform (including a virtual modeling system, communication networks, and a database management platform).

61 Chapter 5 Physical Experiments with Temporary Structures

5.1 Experimental Set-up The experimental set-up determines how the experiment hardware is installed or laid out in the lab for prototype system experimental testing. It involves the set-up of a scaffolding system, the layout of sensors, and the set-up of DAQ hardware for experimental tests. 5.1.1 Set-up of Scaffolding System As discussed in Chapter 4, a frame scaffold system (7’×5’×5’) has been selected as an example of temporary structures for experimental test of prototype system (as shown in Figure 5-1). In particular, the frame scaffolding system include two sets of 7 feet × 4 feet diagonal braces, two sets of 5feet × 5 feet scaffold frames, four spring locks for connection of diagonal braces and posts, and six 2inch × 10 inch × 10 feet OSHA certified scaffold planks. In general, the loading capability, OSHA regulated limitation and the shape information of key components can been found in table 3.

Figure 5-1: Set-up of Frame Scaffold System

Table 5-1: Key Components of Frame Scaffold System

Allowable load Allowable Component Lenth(ft) displacement Concentric (lb) Uniform (lb/sq ft) (inch) post 7 2400 - - Scaffold Plank 10 - 50 1.4

5.1.2 Set-up of Sensors The set-up of sensors includes the selection of sensor type and the amount of sensors required for the experimental test. In order to get accurate results, each sensor should be calibrated according to their voltage output and readings. Finally, the layout of sensors is designed for a comprehensive measurement of the structure performance of scaffolding systems during experiment. 5.1.2.1 Sensor Selection

Load cell. In general, the load cell selected can withstand and detect up to 2000 pounds of load placed on it with a resolution of 1 pound (as shown in Figure 5-2). It helps to identify the amount of load supported by or distributed to each scaffold post. The load cell produces a positive signal when the post is pulled up from the load cell, and a negative signal when there is pressure from the post onto the load cell. Five load cells are adopted in total, with each one under the four posts of scaffolding system and one connected to the hydraulic jack for the measurement of lateral load placement.

Figure 5-2: Load Cell

Switch sensor; the switch sensor works with a maximum switching frequency of 2Hz, providing the signal that indicates when there is a significant base settlement of the scaffold system (as

63 shown in Figure 5-3). It works in such a way that the signal is “off” when the body of the switch is attached to the ground surface, and turns to be “on” with a minimum distance of 0.5 inches between the body of the switch sensor and the touching surface. Five switch sensors are adopted in total, with each one at the bottom of the four posts of scaffolding system and one at the connection between diagonal brace and post 1.

Figure 5-3: Switch Sensor

Displacement sensor; the displacement sensor helps to monitor the displacement of the scaffold planks under pressure (as shown in Figure 5-4). It can detect up to 4 inches’ displacement with a resolution of 0.001 inch. Five displacement sensors are adopted in total, with each one at the bottom of the four posts of scaffolding system and one under the middle of plank 3. All of the information is sent to the data acquisition system and stored in the database for structural performance analysis.

Figure 5-4: Displacement Sensor

64 Inclinometer; in general, the inclinometers selected are bi-directional inclinometers, which can detect the inclination in both x and y direction (as shown in Figure 5-5). It can measure the inclination within the range of 10 degrees at the resolution less than 0.001 degree. The inclinometer is used to measure the two dimensional inclination of the scaffolding system. Four inclinometers are used in total with each one on the four posts of scaffolding system.

Figure 5-5: Bi-directional Inclinometer

5.1.2.2 Calibration of Sensors

In order to get the most accurate result and minimize the signal error, each sensor is calibrated. Due to the obvious result of “on” or “off” from switch sensors, the switch sensors are tested quickly in the lab. However, the other sensors that produce analog signals are carefully calibrated by recording its voltage output and reading for linear regression modeling. This information is very useful for the virtual channel developed which will be discussed in the section on set-up on DAQ system. In particular, the calibration results of load cells, displacement sensor, and inclinometers are displayed in Figures 5-6 till 5-8.

Calibration of load cells

Load cell 1 Load cell 2 350 350 300 y = 1.0008x + 0.2342 300 y = 1.0019x - 0.525 250 250 200 200 150 150 100 100 50 50 0 0 0 100 200 300 400 -50 0 100 200 300 400

Load cell 3 Load cell 4 350 400 300 y = 0.9998x + 0.257 y = 0.9993x - 0.8223 300 250 200 200 150 100 100 50 0 0 0 100 200 300 400 0 100 200 300 400 -100

Load cell 5 500

400 y = 2031.3x - 29.32 R² = 1 300

200

100

0 0 0.05 0.1 0.15 0.2 0.25 -100

Figure 5-6: Calibration of Load Cells

66 Calibration of displacement sensor

Displacement sensor 1 Displacement sensor 2 4 4 3 y = 0.8199x - 0.1419 3 y = 0.7933x + 0.0565 R² = 1 R² = 1 2 2 1 1 0 0 0 1 2 3 4 5 0 1 2 3 4

Displacement sensor 3 Displacement sensor 4 4 4 y = 0.8046x - 0.0607 y = 0.7896x + 0.0722 2 R² = 0.9999 2 R² = 0.9997

0 0 0 1 2 3 4 5 0 1 2 3 4

Displacement sensor 5 2.5 2 y = 0.9942x - 0.0283 1.5 1 0.5 0 -0.5 0 0.5 1 1.5 2 2.5

Figure 5-7: Calibration of Displacement Sensors

Calibration of inclinometers

Inclinometer 1_x Inclinometer 1_y 4 3.5 3.5 3 y = -0.0724x + 2.4571 3 2.5 2.5 y = 0.0731x + 2.6517 2 2 1.5 1.5 1 1 0.5 0.5 0 0 -15 -10 -5 0 5 10 15 -15 -10 -5 0 5 10 15

67 Inclinometer 2_x Inclinometer 2_y 4 3.5 3.5 3 y = -0.0749x + 2.519 3 2.5 2.5 y = 0.0743x + 2.617 2 2 1.5 1.5 1 1 0.5 0.5 0 0 -15 -10 -5 0 5 10 15 -15 -10 -5 0 5 10 15

Inclinometer 3_x Inclinometer 3_y 3.5 4 3 y = -0.0723x + 2.4862 2.5 3 2 y = 0.0755x + 2.5463 2 1.5 1 1 0.5 0 0 -15 -10 -5 0 5 10 15 -15 -10 -5 0 5 10 15

Inclinometer 4_x Inclinometer 4_y 3.5 3.5 3 3 y = -0.0761x + 2.4662 2.5 2.5 2 y = 0.074x + 2.496 2 1.5 1.5 1 1 0.5 0.5 0 0 -15 -10 -5 0 5 10 15 -15 -10 -5 0 5 10 15

Figure 5-8: Calibration of Inclinometers

5.1.2.3 Sensor Layout

For accurate and full measurement of the structural performance of scaffolding system, five load cells, five switch sensors, five displacement sensors, and four bi-directional inclinometers are installed on the frame scaffold. The layouts of these sensors are presented as Figure 5-9. In particular, four load cells were placed at the bottom of each of the four posts in order to measure the load information held by the scaffolding post. One load cell was installed at the end of hydraulic jack which is used to place lateral load by the side of post 1. It produces negative

68 measurements when experiencing pressures and produces positive measurements when pulled. Four switch sensors are placed at the bottom of each post with the switch connector touching the steel plates which are used to simulate the ground surface. One switch sensor is installed at the connection between post 3 and the upper side of diagonal brace in an attempt to inspect if the diagonal brace is installed in position. In addition to the use of switch sensors to detect the connection between the bottom of post and the ground surface, four displacement sensors are also installed at the bottom of each post so that each small movement of the four posts can be inspected even before the occur of actual disconnection between scaffolding system and the ground. This information is helpful for the identification of the trend of the structural performance and the prediction of potential disconnection. In addition to the four displacement sensors at the bottom of the scaffolding system, one displacement sensor is place under one scaffolding plank at the middle point. It is supported by a supporter as shown in Figure 5-10. The end of the displacement sensor is connected with a screw inside of the scaffolding plank so that the displacement of the scaffolding plank can be accurately measured. Finally, four bi-directional inclinometers are mounted at each post. The measurement of x and y directions is helpful to the understanding of the general performance of the whole scaffolding structure.

Figure 5-9: Sensor Layout for Experimental Test

Figure 5-10: Supporter of Displacement Sensor under Scaffolding Plank

5.1.3 Set-up of DAQ System DAQ system connects the information exchange between sensors and computers. It makes the information obtained from sensors visible and accessible by the computer systems. The set-up of DAQ system involves the design and installation of DAQ hardware, and the set-up of DAQ virtual channels.

5.1.3.1 Set-up of DAQ hardware

The hardware supporting DAQ system includes power supply for sensors, National Instrument module to collect signal from sensor to the computer. As shown in Figure 5-11, two power suppliers are developed for the inclinometers and other sensors. The National Instrument modules are consisted of two parts, one for the analog signal (the module on the left) and the other for the switch signal (the module on the right). The signal of sensors is initially acquired from the power supplier and collected at the National Instrument models. Finally, all the information is sent to the computer from the National Instrument modules. The power supplier for inclinometers is presented in Figure 5-11 as an example of how the power suppliers are developed to provide power to sensors as well as to acquire signal. For more details, please refer to appendix A.

Figure 5-11: Set-up of DAQ hardware

In particular, the DAQ hardware works in three steps: first of all, sensors are connected to the power supply; second, the power supply get the signal from sensor and send to National Instrument Modules (the switch/ Boolean signal goes to the module on the right side, and the analog signal goes to the module on the left side); finally, all the signal collected by the National

71 Instrument Modules will be sent to the computer for information analysis.

5.1.3.2 Set up of virtual channel for DAQ system

In order to analyze the information collected from sensor through DAQ hardware to the computer, a National Instrument Software is used for data acquisition and analysis. As introduced in Chapter 4, the National Instrument Software serves as the platform to display the information from DAQ system continuously. Due to the fact that we are collecting both switch/ Boolean signal (the signal of “on” or “off” from switch sensors) and analog signal (the signal from load cells, inclinometers, and displacement sensors), each signal is assigned to a virtual channel (as shown in Figure 5-12 and Figure 5-13) and grouped into two tasks, including both switch task (named as scaffold switch) and analog task (named as scaffold analog).

Figure 5-12: Virtual Channel for Switch Sensors at DAQ System

Figure 5-13: Virtual Channel for Analog Signals at DAQ System

As shown in Figure 5-12, the scaffold switch task is consisted of the virtual channels of “switch 1 scaffold base”, “switch 2 scaffold base”, “switch 3 scaffold base”, “switch 4 scaffold base”, and “switch 5 scaffold diagonal brace”; in Figure 5-13, the scaffold analog task is consisted of the virtual channels of “displacement 1 scaffold base”, “displacement 2 scaffold base”, “displacement 3 scaffold base”, “displacement 4 scaffold base”, “displacement 5 scaffold plank”, “load cell 1 scaffold base”, “load cell 2 scaffold base”, “load cell 3 scaffold base”, “load cell 4 scaffold base”, “load cell 5 scaffold lateral load”, “inclination x_ scaffold post 1”, “inclination x_

73 scaffold post 2”, “inclination x_ scaffold post 3”, “inclination x_ scaffold post 4”, “inclination y_ scaffold post 1”, “inclination y_ scaffold post 2”, “inclination y_ scaffold post 3”, and “inclination y_ scaffold post 4”.

In general, each signal from the sensor is assigned to be one virtual channel, and all the virtual channels are then grouped according to the type of signal as different task. By clicking on different task, one can easily get the real time information from the sensors at the platform of National Instrument.

5.2 Experimental Scenarios Design

The widespread use of scaffolding is accompanied with an increasing amount of safety issues, and scaffold work has been defined as one of construction’s highest risk jobs (Hsiao and Stanevich, 1996). Falling from scaffolds accounts for a huge amount of falling issues in the construction industry (Whitaker, et al., 2003). As identified by National Association of Home Builders (2008), 15% of fall fatalities were from scaffolds, ranking third of the fall fatalities in home building during the years 2003 to 2006. In addition, scaffold collapses also serve as a big problem for scaffolding safety management. Analysis of accidents related to scaffolds in the nine years prior to 2012 (as shown Table 5-2) shows that even with improved management, approximately eight workers are hurt every month in scaffolding collapses throughout the US. The scaffolding system is still dangerous and calls for more safety precautions. Table 5-2: Fatal Occupational Injuries Related to Scaffolding, of Year 2003-2011

Characteristic 2003 2004 2005 2006 2007 2008 2009 2010 2011 fall from scaffold, staging, temporary 85 90 82 91 89 68 54 44 - platform floor of scaffolds, staging, or temporary 3 5 8 7 8 3 3 3 - platform scaffolds-staging of structures other than 12 7 7 8 9 8 5 4 5 building scaffolds-staging of building 77 86 78 86 83 68 55 46 64 climbing, descending scaffolds 5 10 7 7 6 3 3 - 4 subtotal 182 198 182 199 195 150 120 97 93

74 5.2.1 Failure Pattern of Scaffolding System In general, a scaffolding system is defined as the temporary safe working platform for the erection, maintenance, construction, repair, access or inspection, etc. of structures or other building systems. (Grant and Pallett 2012). For around 5000 years, the widespread use of scaffolding systems has been accompanied with high record of structural failures and fatalities (Retay 1996; Hsiao and Stanevich 1996). While there were various kinds of structural deficiencies that leads to failures, the key pattern of scaffolding structural failures can be identified according to historical record of related structural failures, which in turn helps to identify the key factors resulting to the most frequent and significant scaffolding structural failures as the key structural performance information to be monitored by TSM. Therefore, this study identified the patterns of scaffolding structural failures through literature review (Whitaker, et al 2003) of several cases related to scaffolding collapses in UK from 1997 to 2000, as well as individual interviews with safety specialties on scaffolding safety management at United States. A total of four typical patterns of scaffolding structural failures are summarized and discussed as follows. (1) Insufficient bracing within the structure. In the analysis of the 62 incidents related to scaffold collapse, 35.5% of the scaffold incidents occurred due to insufficient bracing system, which ranked a top cause of scaffolding collapses (Whitaker, et al 2003). According to the interviews with safety inspectors, the insufficient bracing of scaffolding system also happens when the workers failed to pin the diagonal braces into the correct place, which presents the illusion that the bracing systems are in place by actually just hanging the diagonal braces with no lateral support provided to the scaffolding system. Although it has been required by OSHA standards for adequate bracing systems, there is no detailed instruction on the correct use and inspection of bracing systems. In practice, the use of bracing system is mainly relied upon the experience of construction workers (Feld and Carper 1997). (2) Overloading of scaffolding system. Some scaffolding collapse due to instability or overloading of materials. OSHA investigated the 16 structural failures related to scaffolding between 1990 and 2008, and revealed that 4 out of 16 incidents occurred due to the overloading of building materials (Ayub 2010). As is highlighted by Weesner and Jones (2001), the overloading scaffolding systems may lead to potential fatal

75 consequences. In order to prevent the overloading situation, it is required by OSHA standards that the load place on scaffolding system should be no more than required capacity. Besides, OSHA also highlights the use of safety factor for capacity identification during the design of scaffolding systems, so that to minimize the possibility of overloading of scaffolding systems at the dynamic construction jobsite. (3) Subsidence of foundations. Foundation provides permanent support of scaffold systems on the place where the system rests. Take the regular scaffolding system for example. The foundation of scaffolds may be placed on soils with different capacities. Thus loads from scaffold will cause different settlement of the foundation, which then make the scaffold platform imbalanced, or even collapsed. In addition, such a situation also happens when ground supporting scaffolding system freezes and melts periodically during the winter. Although the use of baseplates and mudsill helps minimize the subsidence of foundations, almost half of the 36 scaffolding system visited by Halperin & MacCan (2004) were improperly supported or with missing baseplates. Besides, with the scaffolding system getting higher, even a small amount of subsidence of the foundation, which is difficult to detect visually, can lead to tremendous structural failures. More of the time, the ground settlement of scaffolding system results in disconnection between the base of scaffolding system and the ground surface. Therefore, the disconnection between scaffolding bases and ground can be used to check if there is potential subsidence of foundations. (4) Plank slipped or broken. Among the 801 incidents relative to scaffolding structures, around 24% of cases resulted from the structural deficiencies (such as plank slipped or broken) of the scaffolding planks (Fattal et al. 1980). For most of the time, scaffolding planks are placed on the top of the scaffolding platform without tying to the scaffold structure, which easily lets the planks slip away, especially when workers are working on the platform and push the planks accidentally. Some other main causations to scaffold incidents include the incomplete planks placed on the platforms. When the plank quality is not satisfied or overloaded, the scaffolding plank may get broken into pieces, which brings hazards to workers working on or near the scaffolding system. For safety inspection of scaffolding system, the feature of planking is recommended as one of the top four factors for rapid check (Halperin & MacCan 2004). Besides, OSHA standards specifies that only OSHA certified planks can be used as scaffolding planks, as well as the requirement of

76 maximum allowable plank displacement. However, the differences between allowable and unacceptable displacements are hard to be distinguished visually. (5) Improper ties to buildings. Scaffolds that are improperly attached to buildings are dangerous. Several scaffolds collapsed when the ties were removed after fitting. Some incidents occurred due to improper fitting or lack of ties. This kind of situation happens when there is a need to remove ties so that access to key areas can be reached. However, this modification of a scaffolding system is done randomly without qualified inspection and analysis (Whitaker, et al 2003). 5.2.2 Experimental Design Based on the identification of key patterns of structures failures of scaffolding systems, five major experiment scenarios are designed to simulate the five major causes to the scaffolding accidents. In each experimental scenario, the layout of sensors and the use of other instruments are designed and demonstrated in details. 5.2.2.1: Overloading of Scaffolding System at Post 3

The objective of this experiment is to simulate the situation when the load placed on top of post 3 is more than allowed. According to the manufacturer’s guide, each post is allowed to work with loading of no more than 2300 pounds. In order to simulate an early warning prior to overload, a maximum value of 1500 pounds was selected. This load was placed on post 3. In order to do that, a forklift with a hydraulic jack is used. By placing the hydraulic jack at the top of post 3 and keep lowering the forklift, the load placed on post 3 increased quickly to the amount of 1500 pounds, as shown in Figure 5-14. In addition, various sensors are mounted on the scaffolding system.

Figure 5-14: Experimental Scenario of Overloading at Post 3

5.2.2.2 Experiment 2: Base Settlement of Scaffolding System at Post 1

In order to simulate the situation when of base settlement happening to a scaffolding system, a hydraulic was placed jack at the bottom of each post, which could be moved up and down using a manual pump. With the other three posts are controlled at the same level unmoved, post 1 was moved down gradually by lowering the hydraulic jack up to 3 inches as a simulated base settlement of scaffolding system at post 1 (as shown in Figure 5-15).

Figure 5-15: Experimental Scenario of Base Settlement at Post 1

5.2.2.3 Experiment 3: Lateral Load of Scaffolding System at Post 1

The lateral load on the scaffolding system at post 1 was created by using a hydraulic jack by the side of post 1 (Figure 5-16). The manual pump connected with the hydraulic jack gives pressure that pushes the hydraulic jack towards post 1 as a simulation of wind load or being hit by moving vehicle. A load cell is also connected at the end of hydraulic jack so that the amount of lateral load placed on post 1 can be measured. A pipe holder is attached as the part to touch and push the post. It is used because the semi-circular shape of pipe holder makes sure that the lateral load placed on the post 1 is stable and in the x direction. The hydraulic jack is mounted on a support on the side, with the aim to hold the jack horizontally and stable under lateral load simulation. Please see Figure 5-16 below for the layout design.

Figure 5-16: Experimental Scenario of Lateral Load at Post 1

5.2.2.4 Experiment 4: Displacement of Scaffolding System at Plank 3

In order to simulate the displacement of scaffolding planks, the forklift is used with its beam placed parallel to plank 3. In the middle of the plank 3, a steel cylinder is placed in between the beam of forklift and the plank 3. It works to guarantee linear and vertical load at the middle of plank 3. As the beam of forklifted moves down, the load placed at the middle of plank 3 increases and forms a displacement of scaffolding plank 3 (as shown in Figure 5-17).

Figure 5-17: Experimental Scenario of Displacement of Scaffold Plank 3

80 5.2.2.5 Experiment 5: Lack of Diagonal Brace at Post 3

As discussed in the previous section, there are potential hazards when the diagonal braces are not in position, such as lack of the brace or not in position. In order to simulate the situation, the diagonal brace connecting the upper side of post 3 is disconnected to see how the scaffold system respond (as shown in Figure 5-18).

Figure 5-18: Experimental Scenario of Lack of Diagonal Brace at Post 3

5.3 Experimental Tests and Results

In order to understand the structural performance of the scaffolding system, five initial experimental tests were conducted with an intention to check if there is any difference between expected and actual structural reaction under different scenarios. The scenarios created for these initial experimental tests include overload at post 3, disconnection of diagonal brace, displacement of scaffolding plank 3, lateral load at post 1, and foundation settlement at post 1. The experimental results are summarized as follows. 5.3.1 Experiment 1: Overloading of Scaffolding System at Post 3 The objective of experiment 1 is to check if the overloading scenario can be simulated following

81 the experimental design as specified above. It also helps to check if accuracy of the loading measurement under the overloading experiment of scaffolding system at post 3. In doing this, a bundle of brick block weighing 200 pounds are placed at the top of post 1 to simulate the presence of a worker or construction materials in real working environment. First of all, the forklift was placed at the middle point of the frame connecting post 2 and post 3 (as shown in the left side of Figure 5-19); by pushing the hydraulic jack touching the metal block, the load was gradually increased till 1500 pounds at the point; after that, the load on the frame was released between post 2 and post 3, and a load of 1500lbs was placed on top of post 3 by adjusting the place of forklift and metal block (shown in the right side of Figure 5-19); the loading information at post 3 was recorded under the overloading circumstance.

Figure 5-19: Initial Test of Overloading at Post 3

The measurement result showed that when the load of 1500 load was placed in between post 2 and post 3, the total load on post 3 was around 750 lbs; when the load of 1500 was placed on top of post 3, the load reading on post 3 was around 1500 lbs. The results indicated that the experimental scenario of overloading can be obtained following the experimental design. Besides, the reading record provides accurate reference to system performance analysis.

82 5.3.2 Experiment 2: Base Settlement of Scaffolding System at Post 1 The objectives of experiment 2 are to understand if the base settlement of scaffolding system can be simulated and measured as by experimental design, which is to be simulated by lowering the hydraulic cylinder at Post 1 by around 3 inches and to be measured by the switch sensors, load cell sensors, and displacement sensors installed at the bottom of each scaffolding post, as well as inclinometers on each post.

Figure 5-20: Initial Test of Base Settlement at Post 1

Similar to experiment 1, a brick bundle weighing around 200 pounds was placed on top of post 1 to simulate the real working environment. First of all, all the posts were raised up to the same level by pushing the hydraulic jack under each post; second, post 1 was gradually lowered by 3 inches by lowering the hydraulic jack under the post 1 (as shown in Figure 5-20); reading of sensors were recorded for analysis.

It was observed that significant inclinations were observed at post 1 and post 4 (the change of inclination of post 1 was 1.9 degree in X direction, 2.4 degree in Y direction; the change of inclination of post 4 was 2 degrees in X direction;); the switch sensors at post 1 disconnected with the steel plate under the simulated base settlement scenario; load decreased by 110 lbs at post 1, but increased by 90 lbs at post 2, increased by 95 lbs at post 3 and increased by 50 lbs at

83 post 4; the change of displacement at post 2 is 0.5 inches. These results indicated that the actual performance of scaffolding system worked as expected under the simulated scenario. Meanwhile, the sensors work well in inspecting the base settlement of post 1.

5.3.3 Experiment 3: Lateral Load of Scaffolding System at Post 1 The objective of the initial test of experiment 3 is to check if the scaffolding system performance as expected under the lateral load created by the hydraulic cylinder at post 1.

Figure 5-21: Initial Test of Lateral Load at Post 1

The process of experiment 3 includes the placement of brick bundle weighing around 200 pounds at post 1; to prevent lateral movement of the whole structure and also to simulate the real situation that the base of scaffolding posts are usually anchored at the ground in the construction jobsite, the hydraulic cylinders under post 1 and post 4 were anchored with the floor; in order to observe significant structure behavior change, one diagonal brace connecting the post 1 and post 3 is also disconnected (as shown in the right side of Figure 5-21); the hydraulic cylinder mounted by the side of post 1 was pumped to place pressure laterally onto post 1 (as shown in the left side of Figure 5-21); meanwhile, structural behaviors and sensor measurements were recorded for analysis.

It was observed that there was significant inclination of post 1 (1.7 degrees in Y direction) under the lateral load of 80 pounds; the switch sensor at the base of post 1 will disconnect with the steel

84 plate. The result indicated that the use of hydraulic cylinder and the anchorage of the base of post 1 works well in simulating the lateral load of scaffolding system at post 1; the structural behavior can be clearly inspected by sensors installed on the scaffolding system.

5.3.4 Experiment 4: Displacement of Scaffolding System at Plank 3 The initial test of experiment 4 is to check if the use of steel cylinder and forklift works as expected in creating displacement at the middle of scaffolding plank; it also helps to check the response of displacement sensors mounted under the middle of the scaffolding plank 3.

As usual, a block bundle weighing around 200 pounds was placed at the top of post 1 in order to keep the same working environment simulated in the other experiments; the forklift was placed with its beam laying parallel to the scaffolding plank; a steel cylinder was placed at the middle of the scaffolding plank (see Figure 5-22); a perpendicular load was placed and gradually increased by lowering the beam of the forklift until the displacement of the scaffolding plank 3 reaches 1.4 inches; observed and recorded the structural performance of scaffolding system.

Figure 5-22: Initial Test of Displacement of Plank 3

The observed result indicated that displacement of scaffolding plank 3 could be simulated well following the experimental design. The reading of the sensors provided an accurate reference of the structural performance which could be used for structure monitoring and analysis.

85 5.3.5 Experiment 5: Lack of Diagonal Brace at Post 3 The initial test of experiment 5 was to check the structure performance of scaffolding system when the diagonal brace at post 3 was not in position. It was also conducted to check if the switch sensor can detect the lack of diagonal brace as expected.

To keep the working environment consistent with other experimental test, the 200 pounds’ brick bundle was placed at the top of post 1. After that, the upper side of the diagonal brace with the post 3 was disconnected; meanwhile, the structure performance and sensor readings were recorded for analysis.

It was observed that once the diagonal brace was disconnected, the switch sensor at the connection between post 3 and the diagonal brace turned off immediately indicating the absence of diagonal brace. The observation confirmed that the lack of diagonal brace could be simulated following the experimental design and it could be detected by the sensors as expected.

5.4 Summary

This chapter demonstrated the design of experimental set-up for prototype system test, including the set-up of temporary structures, DAQ hardware and virtual channel set-up and the selection, calibration and layout of sensors. Based on the failure pattern of scaffolding systems, five major experimental scenarios are designed to simulate the situation with potential structures failures. In order to check if the designed experimental scenario can be simulated as expected and to understand the accuracy of sensors for structural performance analysis, initial tests of the experimental tests were conducted and observed.

It can be concluded from this chapter that the scaffolding system served as an appropriate example of temporary structures for prototype system development; the set-up of DAQ system and sensors should take into consideration the type and amount of signals to be collected; there are five major causes to the scaffolding failures, including the overloading of scaffolding system, base settlement of scaffolding system, lateral load of scaffolding system (including pressure from wind or hit be vehicles), displacement of scaffolding plank, and the lack of diagonal braces; initial test confirmed that the experimental scenarios can be simulated and measured as expected.

86 Chapter 6 Prototype System Operation

6.1 Linking Physical Structures to Prototype System

With the development prototype system and the experimental set-up ready, the next step was to link the two parts through an on-cloud database with real-time information exchange. Generally, an on-cloud database is a database that typically runs on a cloud computing platform via the Internet (Arora & Gupta, 2012). An on-cloud database is necessary to linking the physical structures and prototype systems for three main reasons. First, the cloud database enables remote inspection of the physical structures from the virtual model via the Internet; second, multiple users can gain access to the TSM system via the Internet for real time inspection and data analysis; third, as has been discussed in the section on “Communication Network between PC and Portable Devices”, the communication between the virtual model and the mobile app is supported by the on-cloud database.

In view of the above, an on-cloud database was developed for this study using the Amazon Elastic Compute Cloud (Amazon EC2) service (as shown in Figure 6-1), whose SQL database can be edited and managed through Heidi SQL at local computers (as shown in Figure 6-2). According to the information types processed in the database, three main types of services are adopted in Amazon EC2, including a MySQL database, a Linux File System, and GCM (Google Could Messaging) service. First of all, the on-cloud SQL database serves as the storage of the updated information from physical structures and read by the CPS-based TSM for structural performance analysis. Based on the structural information required for structural performance analysis, a table with the name and structural information of each scaffolding component was established and can be maintained through Heidi SQL, as shown in Figure 6-2. Generally, the information to be collected for the four posts includes the change of two directional inclinations, “loading” and “base connection”. A similar database structure is used for storing information on the scaffolding planks. In particular, the negative value of loading indicates downward force, while the positive value indicates upward force between the post and the load cell. Taking the initial experimental test for example (see the section on “Experimental Test of Base Settlement” for details), when post 3 is lowered 1/8 inch, the corresponding information recorded in database is shown in Figure 6-2. It can also be seen that post 1 has a change of inclination of 1.8° in y

87 direction and 1° in x direction; the load pressure recorded at post 1 is 22 pounds downwards (indicated with the value of “-22” in the database); the base of post 1 is well connected (indicated by the value of “1” for “base connection” in the database). Secondly, a Linux File System provides the storage of pictures from the TSM controlling system once potential hazards have been identified and can be accessed within 2 seconds by the TSM mobile app installed on portable devices. Last, a GCM service is used to push alarm notifications to portable devices. It allows multiple registered portable devices to remain inactive until a potential hazard has been identified. The benefits of the GCM service include the economical use of the battery power of portable devices, easy access to alarm notifications by multiple portable devices at the same time, and real-time potential hazards information.

88 (1) Upload information (2) Fetch information continuously (3.1) Upload picture of virtual model (3.2) Upload information relative to structural defects (4) Fetch data every 100ms (5) Push notification (6.1) Fetch picture of virtual model (6.2) Fetch alarming instructions Portable Devices (6.1) (6.2) (5)

Linux file GCM System (3.1) (4)

MySQL

Amazon EC2

(2) Cyber (1) Physical System (3.2) Structure

Virtual Model Physical Structure

CPS Bridge

Figure 6-1: Structure of On-cloud Database

As seen in Figure 6-1, the structure and work flow of the on-cloud database has been identified. To be specific, first of all (noted as step 1 in Figure 6-1), the real time information of physical structures is updated and stored at the MySQL database; second (noted as step 2), this updated information of physical structures is continuously fetched by the cyber system for safe structure analysis; third, once potential hazards are identified, the picture of the virtual model with

89 components in question highlighted is uploaded to the Linux file (noted as step 3.1) with relative information uploaded to MySQL database simultaneously (noted as step 3.2); meanwhile, the GCM queries the MySQL for potential alarm information every 100 ms (noted as step 4); the frequent query ensures that once there is alarm, the GCM sends alarming notification to portable devices (noted as step 5); Finally, portable devices fetch pictures of virtual models from Linux file (noted as step 6.1) and alarming instructions from MySQL (noted as step 6.2), which are previously uploaded by the Cyber system.

The use of Amazon EC2 service also enables scalable applications of TSM if needed. For example, a more comprehensive analysis of the potential hazards can be obtained through machine learning of the detailed historical record of not only structural performance but also other related information, such as temperature and worker’s health status, to be stored in the on-cloud database.

Figure 6-2: Database for Base Settlement at Initial Stage Viewed from Heidi SQL

6.2 System Work Flow

For better understanding of how the CPS-based TSM works, the system workflow is presented below. Generally, the CPS-based TSM works as summarized in the key steps shown in Figure 6-3 and described below. (1) The structural information of the temporary structures, such as inclination and loading, is collected by the sensors discussed in the section on “sensors” and sent to the DAQ platform; (2) information on temporary structures is collected continuously at Lab-view and entered to the on-cloud database; (3) the CPS-based TSM queries the on-cloud database every 2 seconds for structural performance analysis based on predefined system logics; (4) if no potential hazard is identified, the TSM continuously collects performance information through sensors attached to the temporary structures. However, if potential for structural failure

90 was identified by TSM based on performance analysis, the component in question will be highlighted in the 3D virtual model for further investigation or adjustment. Meanwhile, the warning and detailed information of structural performance deficiencies is sent from the virtual model to portable devices used by construction workers, safety inspectors, and project managers; (5) finally, assisted by the useful information associated with the structural deficiency, the end users are able to take appropriate action to address the problem and thus prevent structural failure. In this way, the CPS-based TSM works continuously following the system workflow discussed in this section, and which supports real time monitoring of the actual performance of temporary structures.

Continuous structure (1) Change of Performance monitoring Sensor

Data Process (2) Lab view

(3) Structural analysis Plugin If

Else (4) Potential Structure Failure

Yes Highlight Warning component

Mobile devices Virtual model (5) Resolve problem Safety Project worker Inspector manager

Figure 6-3: System Workflow

6.3 User-interface of Prototype System

The user-interfaces of TSM prototype system include both the interface at the virtual model of temporary structures and the interface at the mobile portable devices. Among them, the interface at the virtual model of temporary structures, which is also called TSM user interface, serves as the major user-interface for TSM prototype. It enables the user to get access to the TSM system

91 for real-time and remote inspection of the physical temporary structures in the construction job site through their virtual models. More details about the TSM user interface are presented as follows, and further demonstrated at the experimental test of base settlement. The user interface on the mobile portable devices, which is also called TSM mobile user interface, provides warning information about potential structure hazards to the workers from the virtual model. The TSM mobile user interface is demonstrated through an example of experimental test of base settlement. TSM User Interface. As shown in Figure 6-4, the CPS-based TSM interacts with end users through the developed add-in tool in Autodesk Navisworks Management named “CPS Monitor”. The right window displays the virtual model of temporary structures, and provides a clear view to the end users. Once the tool button of “CPS Monitor” is clicked, a window for user input of warning threshold and “Log” pops up. As is shown, the value of the threshold will be the one entered last time, and can be edited by users at any time before starting the structural inspection. Once satisfied with the warning threshold, end users can click the “start” button to run the CPS-based TSM. In the background, the system evaluates the structural performance of temporary structures every 2 seconds with updated information from the physical temporary structures. The response of the TSM to the potential failures is designed as twofold. First of all, for the end users working in front of the computer, the 3D model will highlight the components in question for the user’s attention, and display how the TSM is communicating with the portable devices to make sure the construction workers and other safety inspectors on the construction jobsite have been notified of the potential hazards (as shown in Figure 6-4).

Figure 6-4: User Interface of CPS-based TSM

Experimental Test of Base Settlement. As identified by Whitaker et al (2003), base settlement ranks as one of the key causes of scaffolding accidents. For system operation demonstration, an experiment on the base settlement of the frame scaffold was conducted for two stages: initial stage when there are potential basement settlements at post 3 and second stage when actual basement settlement occurs at post 3.

(a) Initial stage. The initial stage is the situation when there is a trend of base settlement before actual significant settlement occurs. This is simulated by lowering the base of post 3 by up to 1/8 inch. Once the warning threshold for each component has been entered and verified in the threshold window (as shown in the left side of Figure 6-4), the “start” button is clicked to run the TSM system for real time inspection. It is noticed (as shown in Figure 6-5) that the loads recorded at the bases of post 3 and post 4 become positive, indicating that the post 3 and post 4 are moving up against their bases. In two seconds, post 3 and post 4 were highlighted immediately as the deficient components in the virtual model, as shown in the left side of Figure 6-5. Besides, the “Log” (shown in the left side of Figure 6-5) indicates that a warning has been sent to the mobile App “TSM” (as shown in Figure 6-6). A picture which highlights the affected component is displayed on the mobile App of TSM, along with messages (as shown in the left side of Figure 6-5). For more detailed information, supervisors can click on the “details” button to display a check list that specifies how each component is performing (as shown in the right side of Figure 6-5). In this way, supervisors understand immediately

93 that there are potential base settlements at post 3 and post 4. By checking these posts in the physical environment, supervisors can take corrective action or report to the project manager for further investigation.

Figure 6-5: User Interface of TSM at Initial Stage of Base Settlement Test

Figure 6-6: Warning Message on Portable TSM Interface

94 (b) Second Stage. In case the potential base settlement at the initial stage is not prevented by supervisors, the second stage indicates the situation when the base settlement of post 3 actually happens as defined by complete separation of the post from the foundation support. This is simulated by lowering the base of post 3 up to 2’’. At this point, there is a disconnection signal from the switch sensor placed at the base of post 3, along with increasing negative loading information at the base of post 3 and post 4. The imbalanced loading distribution happens due to the severe inclination of the whole structure. The actual danger of post 3 and potential problem of post 4 are updated immediately in the virtual model, with the component with the actual problem (post 3) highlighted in purple color and one with potential hazard (post 4) in red color, as shown in Figure 6-7. Meanwhile, warnings and instructions are also sent to the supervisors.

Figure 6-7: User Interface of TSM at Second Stage of Base Settlement Test

6.4 Integration of Temporary Structure Performance with Prototype System

This section discusses the integration of physical structures and virtual model for a complete CPS prototype, which is known as TSM for advanced monitoring of temporary structures. After that, a series of comprehensive experimental tests of the TSM prototype system was conducted under five experimental scenarios, including overloading of scaffolding post 3, base settlement at post 1, lateral load at post 1, displacement of scaffold plank 3, and lack of diagonal brace at post 3. 6.4.1 Integration of Physical Structures and Their Virtual Models The integration of physical structures and their virtual models is supported by the CPS Bridge, which facilitates the bi-directional physical-to-cyber and cyber-to-physical coordination.

95 (1) Link between physical components and the virtual model. It is important that for accurate prediction of potential hazards and quick location of potential problems that each physical component can be uniquely and accurately recognized in the virtual model. In order to bind each physical component with its virtual representation, the unique component ID in Navisworks is linked with the component ID in the on-cloud database (details about on-cloud database has been discussed in the section on “Linking Physical Structures to Prototype System”), which represents the physical component in the real world. (2) Physical-to-Cyber Bridge. Real time information flow from the physical world to the virtual world is supported through the physical-to-cyber bridge, as shown in Figure 6-8. In particular, sensors mounted on the physical scaffolding system collect structural performance information continuously. This information is then read by the National Instrument data collection devices connected to the sensors to capture meaningful data information. Third, based on the appropriate data, Lab View is used as the data acquisition platform, which conducts data configuration. These data will then be exported calibrated and entered into the on-cloud database to be used in determining potential hazards through analysis. Finally, once potential hazards are identified, the corresponding virtual model element in question will be highlighted using an appropriate color to indicate an ‘alarm’.

DAQ On-cloud database Virtual model Sensor LabView

Figure 6-8: Physical-to-Cyber Bridge

(3) Cyber-to-Physical Bridge. From the virtual model to the physical structure, appropriate monitoring of the physical scaffolding system is enabled through the cyber-to-physical bridge, as shown in Figure 6-9. In particular, the structural performance of temporary structures is continuously updated in the virtual model. The virtual model also highlights the potential components in question for review; and sends warning information to the on-cloud database; the portable devices then receive the warning notification from the database to assist safety inspectors.

Virtual model On-cloud database Portable Device

Figure 6-9: Cyber-to-Physical Bridge

96 6.4.2 Experimental Test of TSM System This section focused on the test of TSM prototype system through five experimental tests, including overloading of scaffolding system, base settlement of scaffolding system, lateral load of scaffolding system, displacement of scaffolding system, and lack of diagonal braces. These five failure scenarios have been designed for experimental set-up and tested in the lab without linking to TSM in the previous sections. Based on the successful test of experimental design of the five failure scenarios of temporary structures, the next step is to check how the physical structures and virtual model is connected and how the close linkage works in identifying potential problems of temporary structures. The full experimental tests of TSM prototype system under the five failure scenarios are critical for TSM system performance analysis and refinement.

6.4.2.1 Experimental Objectives & Overview In general, the experimental test aims to identify how TSM system responses under different hazardous situations. The details of the experimental objectives and overview are discussed as follows. (1) Objectives of Experimental Test While the experimental test is conducted in general for performance evaluation of TSM prototype system, the objectives of the experimental tests include four major parts:  to understand how closely the physical structures and virtual models are linked through TSM;  to check if the TSM system monitors the temporary structures in real-time;  to measure the accuracy of TSM system in identifying potential problem;  to identify potential problems of TSM system to be improved. (2) Overview of Experimental Test The full experimental tests are conducted based on the experimental design of the five major hazardous situations of scaffolding system. This section demonstrated how the full experimental test will be conducted, such as the process, and what type of information should be recorded in order to evaluate TSM performance under the test.  Experiments conducted As indicated in the previous section of experimental design and experimental test, five experimental tests, including overloading of scaffolding system, base settlement of scaffolding

97 system, lateral load of scaffolding system, displacement of scaffolding system, and lack of diagonal braces were conducted for TSM system test.  Process of experimental test The process of the full experimental test differs in each experimental test due to the difference in the design of the scenario and temporary structure response. However, the general process of the full experimental test includes the function test of hardware and software, starting TSM real-time inspection of scaffolding system, applying failure scenario to the scaffolding system gradually, entering the structural information from Labview to database, observing response of TSM in regard to each failure scenarios, and conducting performance analysis in terms of user-interface, real-time inspection and the accuracy in hazards identification. The general process offers a clear guidance of TSM system test during the full experimental test.  Performance to be checked and recorded The performance of TSM to be checked is based upon the objectives of the experimental test. It is noticed that one of the main objectives of the full experimental test is to check how quick and accurate the potential problems can be identified by TSM in advance. Because of this, it is critical to observe the close linkage between physical structures and virtual models, the quickness of TSM system in response to potential problems, as well as the accuracy of TSM system in identifying problems by comparing the actual problem and the problem identified by TSM system.

6.4.2.2 Full Experimental Test The difference between full experimental test and the experimental test demonstrated in Chapter 5 lays in the connection between physical structures and the virtual models. In chapter 5, the scaffolding system is checked under different experimental scenarios without the linkage to the virtual model. The performance of the scaffolding system is checked manually through visionary observation and the reading from sensors and Labview system, so that to understand if the experimental tests are designed appropriately. The full experimental test in this section will create the same five hazardous scenarios with the scaffolding system monitored by TSM prototype system. It is during the full experimental tests that the performance of TSM can be observed and evaluated. (1) Experimental Test 1: Overloading of Scaffolding System at Post 3  Conduct of experimental test 1

98 The overloading of scaffolding system is simulated with the help of the forklift, whose beam is placed on top of post 3. The whole process of overloading is conducted with TSM being activated for potential hazards of overloading problems monitoring. The details of the process are demonstrated as follows in Figure 6-10:

Level the scaffolding system

Set threshold value at TSM

Start TSM inspection system

Drive forklift to the right position of scaffolding system

Gradually apply increasing load on top of Post 3

Check TSM performance during the whole process

Gradually apply increasing load on top of Post 3

Stop when the loading at Post 3 reaches 1500 lbs

Figure 6-10: Process of Experimental Test 1

 Experimental test result As shown in Figure 6-11, the physical scaffolding system and its virtual model, as well as the mobile portable device are used during the experimental test. The beam of the forklift is placed on top of post 3. By lowering the beam of forklift, it applied increasing load to post 3.

Figure 6-11: Layout of Experimental 1

By the time when the load reached 1500 lbs (with the information collected by Labview entered manually to the database), the virtual model of scaffolding system is updated with post 3 highlighted in alarming color immediately (see Figure 6-12). The identification of the overloading problem is accurate (correctly identified that potential problem happens at post 3) and quick (within 2 seconds). Meanwhile, the warning message is received at the portable device, indicating that potential problem is presented at the scaffolding system.

Figure 6-12: Warning of Potential Hazard – Experiment 1

By clicking the warning message, the TSM mobile app was opened to review the warning information (shown in Figure 6-13). In the window of image, the post 3 is highlighted with

100 alarming color, which is the same as the one viewed at the interface of virtual model. In the window of text message, it stated that potential hazards identified at post 3. By clicking the “detail” button, the performance overview, shown in Figure 6-13, marked the potential problem as the overloading at post 3, which was the problem simulated in the experimental test.

Figure 6-13: Warning Information of Experiment 1 at Portable Device

(2) Experimental Test 2: Base Settlement of Scaffolding System at Post 1  Conduct of experimental test In order to simulate the hazardous status of base settlement, each base of scaffolding system is placed on top of a hydraulic cylinder with a maximum height of 3 inches to be adjusted. All the four hydraulic cylinders are raised by 3 inches up initially, with only the cylinder under base 1 being lowered. In this way, a base settlement of scaffolding system at post 1 is created. Continuously collect the information from Labview and enter to the database. The TSM system is turned on during the whole process for system monitoring. Please see Figure 6-14 for details of the whole experimental test process.

101 Raise by 3 inches of the cylinders at the base of the four posts of scaffolding systems

place bund of bricks (with 4 brick blocks) weighing around 200 pound at the corner of Post 1

Set threshold value at TSM

Start TSM inspection

Gradually lower the cylinder under the base of post 1 by 3 inches

Check TSM performance during the whole process

Figure 6-14: Process of Experimental Test 2

 Experimental test result The bases of scaffolding system are raised up initially, as shown in Figure 6-15. Because all of the bases are raised up at the same height, the whole structure is stable and level. After that, the hydraulic cylinder under base 1 was gradually lowered using the manual pump, shown in Figure 6-16. Based on the experimental test in previous chapter, it was identified that the base settlement at base 1 will cause the disconnection between “ground surface” (simulated by the steel plate) and base 2. This full experimental test was conducted to check if the TSM can detect the problem immediately and accurately.

Figure 6-15: Layout of Experimental Test 2

102

Figure 6-16: In progress - Experimental Test 2

As shown in Figure 6-17, by the time that the base settlement of base 1 reached 3 inches (with the information collected by Labview entered manually to the database), the virtual model identified that there was potential problem, highlighted post 2 in alarming color, and sent warning message to the TSM mobile app at portable device.

Figure 6-17: Warning of Potential Hazard – Experiment 2

The TSM mobile app demonstrated that the post 2 has potential problem in both image and text message, as shown in Figure 6-17. The click of “Details” button opened the performance overview, shown in Figure 6-18. It clearly pointed that the potential problem of post 2 is the disconnection at post 2. Therefore, the identification of the potential hazard is the same as the problem intentionally created, which quickly provided valuable information to safety

103 superintendents and workers.

Figure 6-18: Warning Information of Experiment 2 at Portable Device

(3) Experimental Test 3: Lateral Load of Scaffolding System at Post 1  Conduct of experimental test The lateral load is created with the help of a hydraulic cylinder secured by the side of post 1 horizontally. It is attached to a vertical beam which provides stable support for the hydraulic cylinder to apply load to post 1. Due to the smooth floor, the base of post 1 and post 4 were secured at the floor by anchors to prevent lateral sliding and for better evaluation of the lateral load of post 1. The whole experimental process is demonstrated in Figure 6-19.

104 Level the scaffolding system

Place one bund of bricks (with 4 brick blocks) weighing around 200 pound at the corner of Post 1

Secure the base of scaffolding system at post 1 & 4

Set threshold value at TSM

Start TSM inspection

Gradually apply load onto post 1 using the lateral cylinder placed by the side of post 1

Check TSM performance during the whole process

Stop when the inclination of post 1 reaches 2 degrees

Figure 6-19: Process of Experimental Test 3

 Experimental test result Similar to the other experimental tests, the physical scaffolding system, virtual models, and portable devices are presented to check how TSM monitored the structural safety of scaffolding system in response of potential hazards. In order to see the movement of post 1 significantly, the diagonal brace between post 1 and post 3 was disconnected. A hydraulic cylinder is mounted horizontally by the side of post 1 with a holder attached at the top of the cylinder. It is designed as the instrument that touches post 1 while applying horizontal load to post 1 without changing the load direction during the whole process. As seen in Figure 6-20, the base of post 1 and post 4 have been secured using steel plates anchored to the floor. The pump was used to manually adjust the load applied to post 1.

105

Figure 6-20: Layout of Experimental Test 3

After checking everything to be ready, the hydraulic cylinder was pumped towards post 1. As shown in Figure 6-21, post 1 inclines significantly with the increasing amount of lateral load applied on it. The information was entered to the cloud database continuously. When the post 1 inclines around 1.4 degree (the threshold set at TSM), the virtual model noticed the problem and highlighted post 1 in alarming color immediately (shown in Figure 6-21).

Figure 6-21: Warning of Potential Hazard – Experiment 3

At the TSM mobile app, it is easy to notice that the window of image demonstrated that the post 1 has potential problem with alarming color highlighted in the virtual model. The text message

106 confirmed that the potential problem occurred at post 1. Specifically, the performance overview (shown in Figure 6-22) identified that the post 1 inclines more than allowed, indicating that the safety superintendent may need to check the inclination of post 1 as well as taking measurements to prevent severe accidents.

Figure 6-22: Warning Information of Experiment 3 at Portable Device

(4) Experimental Test 4: Displacement of Scaffolding System at Plank 4  Conduct of experimental test The displacement of plank 4 is simulated with the help of forklift and a cylinder placed in between the beam of forklift and the middle of plank 4. The TSM system is used to monitor the performance of the scaffolding system during the whole test. Please refer to Figure 6-23 for details of the experimental test 4.

107 Level the scaffolding system

Place one bund of bricks (with 4 brick blocks) weighing around 200 pound at the corner of Post 1

Place a cylinder on top of the middle of plank 4

Set threshold value at TSM

Start TSM inspection

Use the beam of forklift to increasingly apply load on top of the cylinder

Check TSM performance during the whole process

Stop when the displacement of plank 4 reaches 1.4"

Figure 6-23: Process of Experimental Test 4

 Experimental test result As shown in Figure 6-24, the displacement of plank 4 was simulated by applying load on the cylinder placed at the middle of plank 4 with the help of the beam of forklift. By gradually lowering the beam, the forklift is simulating the working loading placed on top of plank 4. As observed in Figure 6-25, the displacement of plank 4 increased with the increasing load placed on top of it. The information was entered to the cloud database continuously. Once the displacement of plank 4 reached 1.4 inches (threshold value set at TSM, also the maximum allowable displacement required by OSHA), in 2 seconds, the virtual model highlighted plank 4 for potential problem and sent warning message to the TSM mobile app at portable device.

108

Figure 6-24: Layout of Experimental Test 4

Figure 6-25: Warning of Potential Hazard – Experiment 4

At the portable device, the TSM mobile app presented both image and text warning message indicating that the potential hazard occurred at plank 4 (shown in Figure 6-26). In order to see what type of potential problem it is, one needs to click the button of “Details”, which demonstrated that the potential problem of plank 4 occurred due to the plank displacement, as shown in Figure 6-26. In this way, TSM helps in monitoring the displacement of planks and report potential problems once it exceeds the allowable displacement, which is extremely difficult to identify visually.

109

Figure 6-26: Warning Information of Experiment 4 at Portable Device

(5) Experimental Test 5: Lack of Diagonal Brace at Post 3  Conduct of experimental test The situation of a missing diagonal brace often happens when the diagonal braces are not well installed or dismantled by workers during construction work. Both of the two common reasons result in the lack of diagonal braces which presents potential stability dangers to the scaffolding system structures. The simulation of the lack of diagonal brace is conducted by manually disconnecting the diagonal brace and post. It is expected that TSM system will identify the problem immediately when the diagonal brace is disconnected with the post. The whole process of the experimental test is demonstrated in Figure 6-27.

110 Level the scaffolding system

Set threshold value at TSM

Start TSM inspection

Disconnect diagonal brace at post 3

Check TSM performance during the whole process

Figure 6-27: Process of Experimental Test 4

 Experimental test result The test of lack of diagonal brace is straight forward. In order to check how TSM system works in monitoring such a problem, the virtual model of scaffolding system and the physical structures are both presented for observation. As shown in Figure 6-28, the virtual model identified nothing when the scaffolding system stays stable with no missing diagonal braces. After that, the diagonal brace was disconnected at the upper side of post 3 (as pointed by the blue arrow in Figure 6-28). The information was entered to the cloud database continuously. The virtual model changed the diagonal brace 2 in alarming color immediately indicating that there is potential problem at this diagonal brace (see Figure 6-29). A warning message was also received at the same time at the portable device (see Figure 6-30).

Figure 6-28: Layout of Experimental Test 5

111

Figure 6-29: Warning of Potential Hazard – Experiment 5

By checking the warning message at TSM mobile app, it highlighted that diagonal brace 2 is in potential danger with both image and text message indication. Specifically, one can check the type of potential problem of diagonal brace 2 at the page of “performance overview”. As shown in Figure 6-30, TSM identified that the problem of diagonal brace 2 is due to the disconnection with the scaffolding structure.

Figure 6-30: Warning Information of Experiment 5 at Portable Device

112 6.5 Summary

Based on the previous result, this chapter integrated the physical structures and the prototype system. For clear understanding of the TSM system, the system work flow was presented. After that, the user-interfaces of TSM system, including the major TSM user interface and TSM mobile user interface were demonstrated through an example of base settlement experiment. Finally, based on the fully developed TSM prototype, comprehensive experimental tests were conducted under the five major experimental scenarios for prototype performance evaluation and system check.

It can be concluded that the prototype TSM is linked through an on cloud database for real-time information exchange; the development of a CPS prototype involves three key parts: physical structures to be monitored, virtual models (which facilitate human-computer interaction), and a CPS bridge, which enables bidirectional information exchange between the physical structure and its virtual representation; the TSM system provide clear user-interfaces to the end users for structural performance inspection and potential hazards notification; comprehensive experimental tests of the CPS-based TSM prototype system have indicated that the proposed system works well in identifying potential hazards of temporary structures in a real time manner with active interaction with end users.

113 Chapter 7 Prototype System Evaluation

7.1 Evaluation Approach

In general, an evaluation workshop is selected for this evaluation. This section introduces the objective, approach, selection of participants, design of questionnaire, and evaluation process of the evaluation approach adopted for prototype system evaluation and refinement. 7.1.1 Evaluation Objectives and Approach This research explored the potential of CPS for temporary structures monitoring. A prototype system, TSM, is developed as an example of how CPS contributes to the safety monitoring of scaffolding system. However, it is important to evaluation the prototype system in order to get a comprehensive understanding of how the proposed method works in the monitoring of temporary structures in the real world. In general, the objectives of prototype system evaluation are identified as follows:  To check the soundness of project hypothesis and system logics for potential improvement;  to understand the performance of the developed TSM system for safety monitoring of temporary structures;  to identify potential improvements of TSM to be addressed;  to check if TSM provides an effective solution to safety monitoring of temporary structures on the construction job site;  to obtain feedbacks on potential applications of TSM system;  to learn about benefits and limitations of TSM system.

In particular, it is critical to evaluate if the TSM satisfies user requirements, including real-time inspection of temporary structures performance, immediate notification of potential hazards, 3D visualization of problem, remote access for problem analysis, simultaneous access by multiple parties, non-intrusive / minimum impact to the daily work on construction job site, and ease of TSM to be used by workers. In addition to the evaluation of system performance, it is also important to understand the practical concerns in the implementation of TSM on construction job site.

114 In order to achieve the objectives identified above, an evaluation workshop is held for the evaluation of prototype TSM system for three major reasons. First of all, at the evaluation workshop, the prototype system can be presented clearly to the experts in the area of safety management of construction projects; secondly, the evaluation workshop provides an opportunity to get direct interaction with the experts in this area so that it is easier to get more comprehensive evaluation and creative suggestions; finally, the diverse experience of different participants may lead to a wide range of discussions and opinions which are valuable to the prototype system evaluation. 7.1.2 Selection of Participants The safety management of temporary structures involves multiple parties of a construction projects. Therefore, in order to get more comprehensive feedback, the selection of participants is important to the evaluation workshop. In general, a total of 15 experts with 18 years working experience on average were invited for this evaluation workshop. They are consisted of experts from different areas. In particular, three participants are safety superintendents, three participants are safety managers, and nine of the participants are project managers. 7.1.3 Questionnaire Design Questionnaire serves as an important part of the evaluation workshop. It helps collect the opinions from experts and guaranteed the important questions to be answered. It also provides a record of the feedbacks which can be further analyzed in the future. A good questionnaire should be able to ask the right questions clearly and keep the whole questionnaire as simple as possible. Therefore, the questionnaire for this evaluation workshop is designed to be brief with key questions identified. In order to achieve the evaluation objectives, the questionnaire consists of three major parts explained as follows: (1) Research background validation It is critical to validate the research background and hypothesis to make sure that the project is conducted appropriately. Therefore, the first part of the questionnaire is designed for project hypothesis validation in order to check if the background of the research project fits well with the real situations at construction jobsite. Examples include the feedback on the concurrent methods of temporary structures, the importance of improving safety management of temporary structures, the major failure patterns of temporary structures failures, and the design of TSM system logistics (such as the value set for alarming threshold);

115 (2) Prototype Performance Evaluation The second part of questionnaire is about the evaluation of the TSM prototype system performance. Experts were required to evaluate the performance of the key features of TSM to understand how TSM support the structure monitoring through real-time inspection, immediate notification, 3D visualization of problem, remote access to identify problem, access by multiple parties, no-intrusive to work, and easy to use. In addition, the overall effectiveness and usefulness of TSM system were evaluated as well for the safety of temporary structures at construction jobsite; (3) Suggestions and Concerns Based on the previous evaluation, the third part of the questionnaire focused on identifying potential limitations and applications of TSM in the construction job sites. This section is designed as open questions to guarantee the freedom of suggestions and comments. In particular, the experts are invited to provide comments on potential limitations of the TSM implementation in the real construction jobsite. In addition to the application of TSM for safety monitoring of scaffolding system, we are also asking the feedback from experts to see if there is any potential opportunity that the TSM can be modified and extended for safety management of the construction project in other areas.

7.2 Conduct of Evaluations and Outcomes

The evaluation workshop was held in an attempt to collect the feedback and comments of engineering experts on the TSM prototype system. This section summarized the general evaluation process and the evaluation outcomes as follows. 7.2.1 Evaluation Process In order to keep the evaluation workshop productive, the evaluation workshop took around half an hour. In general, the evaluation workshop includes the introduction to the research, demonstration of the TSM prototype system, and questionnaire for prototype system evaluation. It was conducted following five major steps as follows: (1) Research Introduction Due to the fact that most of the engineering experts are not familiar with the research project, it would be necessary to provide them with the overview background information for clear understanding and effective evaluation. Therefore, a brief introduction of the research projects

116 was presented to the participants. The introduction included the research background and objectives, which was aimed to help the participants to get a basic understanding of the project. (2) TSM Introduction Based on the research background, the TSM system was introduced briefly to get the participants familiar with the prototype system to be evaluated. We presented the general system workflow to clarify how the whole system works for safety monitoring. For direct and clear understanding, the user-interface of TSM was demonstrated as well. (3) TSM Demonstration The demonstration of TSM was the most critical part during the evaluation work shop. It demonstrated how the TSM works under different circumstances. In order to provide clear demonstration, a video recording was presented to display how the TSM system responded under five experimental scenarios (including the experiment of overloading, base settlement, lateral loading, displacement of scaffolding plank, and lack of lateral brace). (4) Questionnaire Based on the previous instruction and demonstration of TSM prototype system, a designed questionnaire was provided to the participants for their evaluation of the TSM system. As discussed in the previous section, the questionnaire consists of three major parts, which collects the feedback from engineering experts on research background validation, TSM prototype performance evaluation, as well as suggestions for future improvement. (5) Open Discussion In addition to the structured questionnaire, engineering experts were also invited for open discussions. Three major topics were discussed during the open discussion section. First of all, they commented on the importance of TSM as they observed lots of temporary structural failures events at construction job site and inspection of scaffolding system is time consuming; Among the five experimental scenarios, they highlighted the importance to simulate the situation of wind impact and overloading, as these are the most severe problem to temporary structures; finally, they also discussed potential opportunity to apply TSM in the real project for project safety application in the long run. In general, the evaluation process is presented in Figure 7-1.

117 Start

Step 1 Background Introduction

Step 2 TSM Introduction

Step 3 TSM Demonstration

Step 4 Questionnair

Step 5 Open Discussion

end

Figure 7-1: Process of Evaluation Work Shop

7.2.2 Evaluation Outcomes Based on the survey feedback and open discussions, TSM system is valued as very useful in improving the safety of scaffolding system. Details of the evaluation of TSM system is summarized as follows: (1) In general, how useful is the Temporary Structures Monitoring (TSM) system in improving the safety of scaffolding?

Usefulness of TSM System 12 10 8 6 4 2 0 Extremely Very useful Moderately Slightly Not useful at useful useful useful all

Figure 7-2: Usefulness of TSM System

118 Around 14 out of 15 rated the TSM system as extremely useful or very useful in improving the safety of scaffolding systems (see Figure 7-2). According to the comments from participants, the safety of scaffolding system is in high demand of improvement. However, there is still no similar solution in dealing with the safety issue of scaffolding system as TSM does. They think that the use of TSM increased the opportunity for safety superintendent to monitor high risk situation; real-time inspection for potential hazards is highlighted as one of the importance features that makes TSM system to be very useful. In addition to the concept of TSM to improve the safety of scaffolding system, the smooth user interface also contributed to the usefulness of TSM system.

(2) How likely are you to embrace the use of the TSM system for temporary structures monitoring on the jobsite?

Likeness to Embrace TSM 12 10 8 6 4 2 0 Extremely Very likely Moderately Slightly likely Not likely at likely likely all

Figure 7-3: Likeness to Embrace TSM

The participants will moderately embrace the TSM on the jobsite due to the concern of implementation cost (5 people concerned), durability (2 people concerned), and maintenance of the hardware system (4 people concerned), as seen in Figure 7-3 In particular, the worry about implementation cost includes both the installation of TSM system as well as the fees for well trained workers to use TSM. Therefore, the cost of TSM comes as the biggest problem before TSM can be fully adopted by the industry. In addition to cost, it is also worried that the durability and maintenance of TSM may not be high enough in order to be used repetitively under the rough working environment.

(3) How realistic do you rate the experimental tests (the simulation of the five major failures)?

119 Reality of Five Failure Scenarios 12 10 8 6 4 2 0 Extremely Very realistic Moderately Slightly Not realistic at realistic realistic realistic all

Series1 Series2 Series3 Series4 Series5

Figure 7-4: Reality of Five Failure Scenarios

“Series 1” indicates base settlement of scaffolding system; “Series 2” indicates overloading of scaffolding system; “Series 3” indicates severe displacement of scaffolding planks; “Series 4” indicates lateral load (such as wind or vehicle impact); “Series 5” indicates diagonal brace not in position. Generally, all the experimental scenarios are rated as very realistic (see in Figure 7-4). In particular, the lack of diagonal brace and overloading are rated at the two biggest potential failures of temporary structures on the real construction jobsite.

(4) How do you rate the accuracy of the TSM system in identifying potential hazards during the five experimental tests?

Accuracy of TSM System 10 8 6 4 2 0 Extremely Very accurate Moderately Slightly Not accurate accurate accurate accurate at all

Figure 7-5: Accuracy of TSM System

According to the survey, 14 out of 15 think the TSM provides accurate information in identifying potential hazards of scaffolding systems (see Figure 7-5). It indicates that the TSM is able to provide reliable and accurate warning information for timely prevention of potential failures.

120

(5) How well are the temporary structures linked with their 3D models?

Linkage Performance 10 8 6 4 2 0 Extremely Very well Moderately Slightly well Not well at well well all

Figure 7-6: Performance of Linkage between Physical Structures and Virtual Model

This objective of this survey question is check if the core function of TSM (close link between temporary structures and their 3D models) is obtained with good performance. Based on the feedback, 14 out of 15 people agree that the temporary structures are linked with their 3D model extremely well or very well (see Figure 7-6). It confirms that TSM works well in linking physical structures and 3D model for information exchange.

(6) How much does the linkage between the temporary structures and their 3D models contribute to improving safety management?

Importance of Linkage 8 6 4 2 0 Extremely Very Moderately Slightly Not important important important important important at all

Figure 7-7: Importance of Linkage between Physical Structures and Virtual Model

TSM is proposed as an effective solution to scaffolding safety monitoring through the close linkage between the physical scaffolding system and their virtual models (see Figure 7-7). As is noticed above, the linkage between physical structures and their 3D model is well supported by TSM. The objective of this question aims to identify if the linkage between temporary structures

121 and their 3D models is important in safety monitoring. As shown in figure 66 above, 13 out of 15 people rated the linkage as extremely important for very important. This feedback further confirms that the concept of TSM to monitor scaffolding system through the real time linkage between physical structures and their 3D model is highly valued.

(7) The warning system sends alarms within 2s once a problem is identified. How quick is the alarm for you to take measurements to prevent potential hazards?

Adequacy of Warning Alarm 8 6 4 2 0 Extremely Very adequate Moderately Slightly Not adequate adequate adequate adequate at all

Figure 7-8: Adequacy of Warning Alarm

Due to the speed of temporary structure failures, it is necessary to understand if the warning alarm is issued early enough for the safety super-intendents to take measures before actual hazard occurred. Based on the feedback, 13 out of 15 people think the TSM warning alarm is quick enough so that they are confident that there is still enough time for the safety super-intendents to take action of preventions (as shown in Figure 7-8).

(8) How do you rate the performance of the TSM system in relation to the following features?

122 TSM Performance 12 10 8 6 4 2 0 Extremely Very effective Moderately Slightly Not effective effective effective effective at all

Series1 Series2 Series3 Series4 Series5 Series6 Series7

Figure 7-9: TSM Performance Evaluation

“Series 1” indicates real-time inspection; “Series 2” indicates immediate notification; “Series 3” indicates 3D visualization of problem; “Series 4” indicates remote access to identify problem; “Series 5” indicates access by multiple parties simultaneously; “Series 6” indicates non-intrusive (minimum impact) to work; “Series 7” indicates easy to use. In general, all of the seven features of TSM performance are rated as very effective. In particular, the real-time inspection is rated as the most effective one, followed by the 3D visualization of problem and access by multiple parties (as shown in Figure 7-9).

(9) How do you rate the importance of notifying the following people of potential hazards by the TSM system?

Importance of Warning Notification 15

0 Extremely Very important Moderately Slightly Not important important important important at all

Series1 Series2 Series3

Figure 7-10: Importance of Warning Notification

“Series 1” indicates safety superintendent; “Series 2” indicates construction workers; “Series 3” indicates project manager.

123

The safety warning notification is rated as extremely important (see Figure 7-10). Among the people notified, it is extremely important to notify safety superintendent and construction workers, as project manager may not be around at the construction jobsite when potential hazards identified. Therefore, it is most important to notify workers to take care of themselves, and to notify safety superintendents to take measurements to correct the potential problem. In addition to the safety superintendent, construction workers and project manager, the survey result indicates that it’s also necessary to have other parties to be notified. They include the entire project jobsite; the person to be notified should be flexible based on project requirement; the one who are responsible for the temporary structures and can respond quickly, such as foreman; project engineering; safety engineering; safety office of owners; scaffold supplier or engineer; any third party erector of the scaffold; and safety director.

(10) To what extent, would you trust the information from the TSM system?

Trust Level of TSM 9 8 7 6 5 4 3 2 1 0 Extremely trust Very trust Moderately Slightly trust Not trust at all trust

Figure 7-11: Trust Level of TSM by Engineering Experts

Based on the survey result, TSM is trusted by most of the engineering experts. 8 out of 15 rated the information from TSM system to be very trust. 6 out of 15 trusted the information of TSM moderately due to the concern that the TSM required more tests and verification before it can be fully trusted (see Figure 7-11). They think the trust level of TSM information also depends on how TSM hardware could be installed, the durability of the monitoring component, the maintenance of TSM, as well as the capability of the person who will taking measurements upon receiving the warning information of TSM.

124

(11) How long would you need to learn to use the TSM system?

Ease of Learning TSM 8 7 6 5 4 3 2 1 0 Less than 30 30-60 minutes 60-90 minutes 90-120 minutes Other (please minutes specify)

Figure 7-12: Ease of Learning TSM

As indicated by the feedback, most of the people need around 60 to 90 minutes to learn to use the TSM system (see Figure 7-12). It confirms that the TSM system is easy and quick to learn. Because of this, it is applicable to the real project in terms of training time and cost. The ease of TSM training also helps to make it possible for TSM to be commonly adopted in the construction industry.

7.3 Benefits of TSM Prototype System

The TSM prototype system is highly evaluated by the engineering experts due to the real time monitoring of temporary structures through virtual models. According to the survey, the TSM prototype system benefits the construction industry in the following ways: (1) Reduces temporary structure failures and consequent injuries and deaths 8 out of 15 engineering experts pointed out one of the benefits of TSM as the capacity to reduce the risk of temporary structure failures thereby reducing injuries and deaths. Because of this, it also helps to improve the safety management of the temporary structures on the construction jobsite; (2) Provides real-time monitoring of the temporary structures The real-time monitoring of temporary structures makes sure that the temporary structures are constantly under inspection so that potential problems can be identified quickly enough for correction. (3) Issues early warnings in identifying potential problems.

125 The system’s early notification of potential warning is considered essential in helping workers being aware of potential hazards and assisting safety superintendent in taking actions. (4) Prevents potential costs resulting from a failure. By reducing potential temporary structures failures, the TSM system helps in reducing the amount of temporary structure components which may have been damaged in the temporary structure failures. More important, the cost associated with delays, rework, loss of workdays and worker’s compensation following a failure would also be prevented. Therefore, the use of TSM system highly benefits the construction job site in cost saving following failures. (5) Identifies potential hazards which are unnoticeable by human beings As a project manager pointed, the biggest benefit of the TSM system is identifying those hazards which human beings cannot easily notice or when the stability of the temporary structures is affected by the employees accidentally. (6) Applicability to other areas In addition to the scaffolding system, the engineering experts also expected the valuable use of TSM for safety monitoring of other potential application areas. They include the temporary shoring work, the temporary structures under heavy rains, formworks and shoring system on large concrete pours, building movement, crane outriggers, high-speed wind warnings, electrical hazards, hand-railing and fall protection, permanent structure failures, ladders, theater, shoring system, and temporary bracing system.

7.4 Limitations of TSM Prototype System

While the TSM prototype system benefits the safety monitoring of temporary structures, there are still limitations that need to be taken into consideration. In particular, the following limitations are highlighted by the engineering experts as the ones to be improved: (1) Ease of Installation and Use 3 participants suggested that the installation and use of TSM is still time consuming. It is important to make the whole process to be easier and faster. (2) The Use of Hardware As pointed by 5 of the participants, the hardware, such as the sensors, used for experimental test of TSM for scaffolding system has lots of limitations in practical application. It is suggested to try other type of sensors, such as laser scanners and cameras for photo recognition, to make the

126 TSM system more durable, accurate, and scalable to different projects. (3) Cost Control Although TSM system works effectively in real-time monitoring of temporary structure, the cost of the installation, use, and maintenance of TSM is still a big issue to be tackled before the TSM prototype system can be fully adopted by the real construction projects. (4) Practical Concerns In addition to the limitations mentioned above, the practical concerns of TSM system is highlighted as important for TSM system development. According to the suggestions, the practical concerns of TSM system to be dealt with include the overall magnitude of the use of TSM on a large scale project, the easiness of TSM to be used, the durability of hardware at harsh working environment, the simplification of model recreation, training of the employees, cost evaluation of TSM system application, the size, complexity and type of scaffolding system to be monitored, as well as field test to prove the applicability of TSM in the real world.

7.5 Summary

This chapter described the evaluation of the TSM prototype system. It first of all identified the appropriate evaluation approach for TSM prototype system. Based on the objectives of the evaluation, the evaluation workshop is adopted as the approach; the preparation of evaluation workshop involves the selection of participants for evaluation workshop, the design of questionnaire, and the workshop process plan. In general, around 15 engineering experts in charge of safety management of construction jobsite were invited; the questionnaire is designed with three major parts, including the research background validation, the TSM performance evaluation, and suggestions and comments for improvement. After that, the evaluation workshop was held following the process of research background introduction, TSM system introduction, TSM system user-interface demonstration, questionnaire, and open workshop.

According to the feedback and comments collected from the evaluation workshop, the TSM prototype system is highly rated as a useful system in preventing potential hazards of temporary structures and improving the safety management of construction jobsite. In addition to scaffolding system, other temporary structures are recommended as potential application areas of TSM system for effective monitoring and control. However, in order to have TSM to be adopted

127 in the real projects, several practical concerns, such as cost evaluations and filed test, should be taken care of for further application of TSM on the real construction jobsite.

128 Chapter 8 Discussions and Conclusions

8.1 Summary

Based on the previous chapters, CPS has been deployed for the temporary structures monitoring, with both experimental testing and workshop evaluation. The key findings indicated that CPS provides an effective tool for temporary structures, with potentials to be applied to other areas in the construction industry. The research efforts are summarized, with the related tasks and activities highlighted as follows. 8.1.1 Summary This section summarizes the key objectives identified earlier in the thesis relative to the key findings of the research. 1) Investigate CPS applications in other industry sectors, such as manufacturing, healthcare, transportation, and power grid. According to the literature review, CPS has been increasingly used in several industry sectors for improved management. 2) Assess the applicability of CPS in the monitoring of temporary structures, including a review of the enabling technologies. The literature review confirms that CPS provides an effective approach to addressing the structural integrity problem of temporary structures. 3) Review the safety regulations and requirements for a specific type of temporary structure. It identified the urgent need for an advanced method of temporary structures monitoring. 4) Develop and test a CPS-based monitoring system for the specific type of temporary structure identified earlier. The CPS prototype system was developed as a Navisworks plug-in and provides very close linkage between the physical temporary structures and their virtual representations. 5) Evaluate (and refine) the prototype CPS monitoring system based on feedback from laboratory experiments and industry experts. The evaluation results confirm that CPS works effectively in monitoring temporary structures. 8.1.2 Tasks and Activities In order to validate the CPS approach to temporary structure monitoring, seven major tasks were conducted as milestones towards the research objectives. Each task as well as its methodology and activities are described as follows. Task 1: Investigated CPS Applications in Several Industry Sectors.

129 Methodology – This task was conducted through literature review to identify potential applications of CPS in the built environment. Activities – specific activities under this task include: (a) Summarized key features of CPS, starting with definitions and covering fundamental concepts, application contexts, technological requirements, etc. (b) Reviewed CPS applications in other industry sectors. This reviewed the CPS applications in manufacturing, power grid, transportation, and healthcare. (c) Reviewed CPS applications in the built environment. It analyzed the key areas of the built environment where CPS has been applied, and explore trends of CPS applications in the built environment. Task 2: Analyzed the Applicability of CPS in Temporary Structures. Methodology – a combination of literature review and case studies were conducted in executing this task. Activities- undertook as part of this task include: (a) Reviewed performance failures of common temporary structures in construction based on case studies to fully explore the range of temporary structural failures and the potential role of CPS. (b) Summarized the application of CPS related technologies in temporary structures based on a review of the applications of CPS related technologies in the temporary structures. This resulted in a comprehensive analysis that identifies the benefits and limitations of current technologies. (c) Established the need and potential areas for CPS application in temporary structures. This compared the problems that lead to temporary structural failures with the key features of CPS, with the aim of matching the needs for safe temporary structures with the solutions offered by CPS. (d) Selected a specific type of temporary structure for the development of CPS-based monitoring system based on the analysis of CPS applicability, as well as the urgency and importance of safety monitoring. Task 3: Reviewed Safety Considerations of Selected Temporary Structure. Methodology – this task was undertaken by detailed investigation of available safety regulations, and interviews with experts in the safety design of selected temporary structures. Activities-involved in this task include:

130 (a) Reviewed design and stability of selected temporary structure. This investigated the structural components of this temporary structure, and reviewed the applicable design and construction guidance. (b) Reviewed industry practice and safety regulations (e.g. OSHA). In addition, structural requirements of each component and the interactions among components were analyzed. Task 4: Identified Appropriate CPS Technologies for Temporary Structure Monitoring. Methodology – this reviewed the selection of appropriate technologies according to the objective of this research. A virtual model of the selected temporary structure was developed for further research on its interaction with the physical component. Activities-undertook as part of this task involve: (a) Selected appropriate software for temporary structural modeling based on previous analysis. (b) Developed virtual models of selected temporary structures. (c) Selected appropriate technologies for bidirectional coordination between physical components and their virtual models. Task 5: Developed Prototype CPS-based Monitoring System for Selected Temporary Structure. Methodology – this task involved the design of system architecture and the use of rapid prototyping for system development. Activities-involved in executing this task include: (a) Developed a system architecture and functional specification for a CPS application for monitoring selected temporary structure. (b) Prototype Development. This involved choosing the development environment and the development of the modules for information exchange and control. Task 6: Experimental Test of TSM Prototype Application. Methodology - The developed prototype was tested with laboratory experiments. Activities-undertook in performing this task involved: (a) Set up a module of the selected temporary structure in the laboratory, and designed experimental set-up for evaluating the TSM prototype system. (b) Conducted laboratory test of the developed system. This simulated the safety issues that the selected temporary structure typically experiences in practice, and recorded the performance of the developed CPS monitoring system in detecting and responding to these issues.

131 (c) Reviewed test data and analyze the functionality of the developed CPS system in the selected applications. (d) Refined system based on the outcome of the experiments and identify opportunities for further research. Task 7: TSM Prototype System Evaluation. Methodology - The developed prototype was evaluated at evaluation work shop by engineering experts. Activities-undertook in performing this task involved: (a) Demonstrate TSM user-interface as well as its performance during laboratory tests. (b) Conducted questionnaire. This will collect the feedback on the review on the importance of TSM to temporary structures monitoring, the TSM performance, the easiness to learn TSM, TSM benefits, as well as other practical concerns. (c) Conducted open discussion to have collect the comments of engineering experts on TSM prototype system for system refinement and adjustment for future applications.

8.2 Conclusions

The construction industry remains one of the most dangerous industry sectors and urgently needs improvement. To be specific, the safety hazards relative to the failure of temporary structures has a significant impact on the overall safety performance of the construction industry. Emerging technologies such as CPS offer a promising solution for advanced monitoring of temporary structures by providing real time communication between the virtual system and physical temporary structures. This study explored the applicability of CPS for temporary structures monitoring, identified end user requirements and system requirements, and proposed a CPS-based TSM as a method to prevent potential failure of temporary structures. A frame scaffold set was adopted as an example temporary structure for TSM system demonstration and evaluation. To date, the CPS-based TSM has shown promise in the real time monitoring of temporary structures. Several key conclusions can be drawn as follows.  Safety Management remains a critical issue to the construction industry. In particular, the safety management of temporary structures is in urgent need due to the high rate of accidents and lack of appropriate safety regulations or management solutions.

132  CPS offers an opportunity to address the current problems and safety issues associated with temporary structures. In particular, CPS applications have potential in the design and operation of temporary performance stages and scaffolding systems, and can help to improve safety and avoid structural failures.  An effective TSM is necessary to provide real time inspection, remote interaction, early warning, and immediate instruction to construction workers. The use of TSM should be simple and nonintrusive to the daily work of construction workers.  The development of CPS-based TSM integrates the psychical objects, such as the temporary structure, and their corresponding virtual models through communication networks, and a database management platform.  Experimental evaluation of the CPS-based TSM prototype system has indicated that the proposed system works well in identifying potential hazards of temporary structures in a real time manner with active interaction with end users.  Evaluation results from a workshop of industry experts in safety management demonstrated that the TSM system provide an effective solution to safety management of temporary structures, such as scaffolding system.  In order for widespread use and adoption, the TSM as constructed and tested should be improved to address practical concerns, such as the cost of installation and durability, etc., and get actual field test so that to be implemented in the real construction project.  TSM system offers great benefits to the improvement of safety of temporary structures in various perspectives. This includes real time monitoring and immediate alarming. The CPS-based TSM also provides for remote access to the updated structural performance of temporary structures and can be shared by multiple parties. Other benefits include the visualization of the structural deficiency, and analysis of historical performance of temporary structures.  Other potential applications could be developed based on the TSM system. They include the temporary shoring work, temporary structures under heavy rains, formworks and shoring system on large concrete pours, building movement, crane outriggers, high-speed wind warnings, electrical hazards, hand-railing and fall protection, permanent structure failures, ladders, theater, shoring system, and temporary bracing system.

133 8.3 Contributions

Based on the exploration, development and validation of CSP approach to temporary structures monitoring, this research project makes contributions to knowledge in five areas as identified below: Demonstrate the CPS applicability in safety monitoring of temporary structures; CPS has been highly recognized by several industry sectors due to its potential benefits. Yet its application to the construction industry is still limited. This research reviews the key features of CPS, examines current CPS applications in the built environment, and analyzes the applicability, potential benefits and barriers for CPS application to temporary structures. It identifies the promising application areas, and discusses how CPS could be applied in these contexts and the potential benefits. A prototype system that enables mutual communication between temporary structures and their virtual models; In order to demonstrate how CPS could be applied for temporary structures monitoring, a prototype system named TSM was developed by integrating several CPS enabling technologies, including DAQ system and BIM technology. The developed prototype system supports the bi-directional communication between the temporary structures and their virtual models, which enables the real time monitoring of temporary structures through virtual models as well as other potential benefits. Tested and evaluated the prototype system through lab experiment for structure monitoring of scaffolding system; a series of lab experimental tests and expert’s evaluation are conducted to understand the performance of the TSM prototype system, as well as to evaluate its benefits to the construction industry. Five major structure failures were simulated in the lab in order to test how the TSM responds for prediction of potential hazards. The evaluation of TSM performance by key industry experts also provides an insight on the benefits and limitations of the CPS approach for temporary structure monitoring. A new approach to identify potential hazards of temporary structures and call for human interaction through warning message; this research proposed CPS as a new solution to structural failures of temporary structures. This new approach is carefully examined for its applicability based on literature review and validated through prototype system development and evaluation. In particular, the CPS approach for temporary structures enables the bi-directional

134 communication between physical and virtual models. It serves as an advanced assistance to project managers or safety managers in safety management of temporary structures. Enhanced understanding of the enablers and barriers to the CPS application in temporary structures monitoring; this research explored and validated CPS applicability to temporary structures monitoring through literature review, expert interviews, rapid prototyping, laboratory experiment, and evaluation workshop. Based on the result, it identified how CPS could be further developed for adoption as well as its benefits and limitations. It contributes to a deeper understanding on the enablers of CPS applications to temporary structures monitoring. By evaluation the developed CPS prototype, the research also demonstrated barriers to CPS implementations to the construction industry. It provides reference to future researches in CPS applications to temporary structures monitoring.

8.4 Research Limitations

In spite of the contributions identified above, there are various limitations to be considered for future improvement. A clear understanding of the research limitations also supports the validation of the research project. In general, these limitations are summarized as follows: Limited sample size for the evaluation workshop; due to the limited access to large amounts of industry experts in the construction industry, only a limited number of experts participated in the evaluation workshop. However, all the industry experts were carefully selected in the area closely related to temporary structures monitoring from various parties (such as owners, CM, and general contractor). All of these benefitted the validation and quality of the evaluation result. Only one CPS approach was developed; this study designed and developed a CPS prototype as an example solution to temporary structures monitoring. However, in addition to the TSM prototype, there are various approaches to CPS applications to temporary structures, such as the use of other supporting technologies for mutual information exchange. Instead of enumerating all possible CPS architectures, only one CPS approach was developed as a proof of concept. Manual linkage between Labview and database; due to technical difficulties, the linkage between Labview and the database is not fully automated. At the current stage, the information is collected in Labview and manually entered to the database. However, the data collected in Labview is well calibrated with noise being screened so that to increase the accuracy of the information. The technical difficulties are due to the use of this specific experimental

135 configuration and can be addressed by using alternative systems. Further work is planned to facilitate the automatic linkage between the Labview system and the database system. Only laboratory testing was conducted; based on the research scope, the developed CPS prototype was only tested in the lab. While the laboratory test validated CPS as an effective approach to temporary structures monitoring, many physical limitations were present (see next section) and it would be beneficial to have the CPS prototype system tested in the real world. Therefore, building upon the laboratory test results, future research could focus on field tests for CPS implementations.

8.5 Practical Deployment Considerations

Based on the laboratory test and evaluation workshop, it can be noticed that there are several limitations to the laboratory experiment as well as potential opportunities for TSM implemention in the real world. For better understanding of these parctical deployement considerations, the limitations of laboratory experiment and potential opportunities for field applications of TSM system have been highlighted as follows. 8.5.1 Limitations of Laboratory Experiments The laboratory experiments involved the set-up of a very specific limited scope of temporary structures, the calibration and installation of sensors, the set-up of DAQ system, experimental scenarios design, as well as laboratory experimental test. In general, there are five major limitations related to the laboratory experiment, including the wired sensors, limited locations for measurement, installation of sensors, and the experimental scenario design. Wired sensors; the sensors used for the laboratory experiment are all wired sensors, which means that each of the sensors is connected with several cables, including the power cable, the cable into and out of the DAQ system for input and output signal, as well as the cable connected to the reading meters. With the increasing number of sensors used for laboratory experiment, the large amount of cables makes it difficult for trouble shooting and is error-prone. In addition to the amount of cables, the wired sensors also require accurate measurement of the distance between the sensor and the DAQ system. A short cable connected to the sensor may pull the sensor away from where it was installed or get disconnected during the experiment, resulting to significant errors. This limitation can be well addressed by taking advantage of wireless sensing technologies, such as photography technology, wireless sensors, unmanned aerial vehicles and

136 laser scanners. In addition to the limitations identified above, as pointed by the industry experts, the wired sensors also make it difficult to be durable and scalable in the real project. Limited number of sensors used; due to the difficulty of sensor installation, big size of each sensor, limited space of scaffolding structure for mounting sensors, as well as the limited research budget, only limited number of sensors were used. In the real project, it is difficult for multiple sensors to be mounted at the same place for different structural information, such as the inclination and loading information at the same point. Because of this, the layout of sensors should be carefully designed before installed for complete information collection. The limited sensor makes it difficult to have enough structural information of the temporary structures. Limited locations for measurement; due to the fact that each sensor only measures the structural performance at a particular location, it prevents the structural performance measurement of the whole temporary structure. For example, a switch sensor placed at the bottom of post 1 can only measure the connection information between post 1 and the ground surface, without understanding how the whole structure performances. It would be beneficial if the whole temporary structures could be measured continuously for an advanced structural analysis and potential hazards prediction. Installation of sensors is labor-intensive and time-consuming; for the sake of accuracy of signal acquired from sensors, the installation of each sensor is carefully designed and measured. For example, the inclinometer attached at the scaffold post requires a stable and horizontal mounting surface. An “L” shape metal plate is designed and mounted by the side of each scaffolding post, with the inclinometer secured on top of the metal. The inclinometer is also adjusted to the position with zero degree in both x and y direction before secured. The installation of sensors requires large amount of time and effort, which in turn will slow down the speed of experiment preparation and prevented its adoption to the real project. Only major structural failures are simulated; while there are various structural failure cases, only the five major structural failures are simulated for experimental test. This is due to the nature of laboratory test that each structural failure should be carefully designed and created during the test. Therefore, it’s impossible and unnecessary to enumerate and cover all possible structural failures. By studying the five major structural failure scenarios, the TSM system is tested for the prevention of the majority of hazards. In the future during field testing, the TSM

137 system can be improved to be a complex system identifying most of the potential hazards by learning the historical structure records of temporary structures in the real world.

8.5.2 Potential Opportunities for Field Applications of TSM System In review of the practical concerns pointed by engineering experts as well as the limitations of research project analyzed in the previous section, adjustment is necessary before TSM system can be used in real project applications. There are potential opportunities that provided alternative solutions to TSM to address the practical concerns. They include the use of wireless sensors, the use of camera, laser scanners for quick modeling, Unmanned Aerial Vehicles, and the use of optical fibers. Details are described as follows. Wireless sensors: one of the limitations of the current TSM system is the use of wired sensors, which makes the management of cables time-consuming and error-prone. In order for TSM to be applied in the field, wireless sensor helps effectively in reducing the use of cables as well as increasing stability. Because of these benefits offered by wireless sensors, they have been increasingly adopted for structural health monitoring (Nagayama & Spencer 2007) and personal health monitoring (Milenković 2006). Use of camera: image tracking methods, such as the use of camera for structural performance monitoring, provides an alternative way in structural inspection. Technically speaking, the use of multiple cameras in different directions can be accurate and effective in identifying the change of structural performance. Practically speaking, most of the construction jobsite are surrounded with cameras for schedule recording. Because of this, it is scalable to take advantage of the existing cameras for temporary structures monitoring. Laser scanner for quick modeling: one of the significant differences between lab experiment and field test lies in the complexity of temporary structures to be monitored. While it is easy to build a virtual model for laboratory test, a quick and clean way of modeling is important. As indicated by the engineering expert and literatures, laser scanner measures the distance in millimeter to centimeter accuracy at the speed of thousands to hundreds of thousands of point measurement per second (Tang et al. 2010). Because of this, the use of laser scanner provides accuracy and quick modeling method for virtual models of temporary structures. It also provides an adaptive way updating virtual model layout in response to the adjustment or movement of temporary structures used in the field. Unmanned Aerial Vehicles (UAV): UAV is increasingly accepted in the construction industry

138 due to its advantage of carrying camera or laser scanners to capture and generating point cloud models and automatically detecting, classifying and localizing defects (Skibniewski & Golparvar-Fard, 2016). By using UAV in TSM system, there would be no worries about the wires and cables to the wired sensors, for they can be replaced by cameras carried by UAV. This also avoids the efforts and cost required in sensor installation and maintenance, which have been big concerns for TSM adoption based on the TSM system evaluation feedback. Because of the easy access to everywhere of UAV, it also helps in making sure that the whole temporary structures (including the place hard to be accessed by human beings or sensors) can be inspected by UVA by flying around. In addition to the benefits of quick and easy inspection, the use of UVA also provides a potential solution in assisting safety superintendent in potential hazards locating for further management. For example, one can have UVA flying to the components with potential hazard with correcting message attached. By following the UVA, safety superintendent can quickly identify the component in question as well as the suggested measurements to prevent accident occur. Use of optical fiber: all of the alternative methods mentioned above are still taking advantage of sensor hardware in collecting structural performance information of temporary structures. Because of this, the structural monitoring highly depends on the information collected by the sensors, such as the place the sensor is mounted, the type of information the sensor can be collected, or the place where UVA are collecting information from. The dependence on the location and maintenance of sensors weakens the accuracy of TSM system. Instead of attaching sensors on top of temporary structures, the use of optical fiber offers the solution to make temporary structure sensor itself. It means that the temporary structure is producing performance information by its own component with high accuracy at high speed. It provides more comprehensive and accurate information for temporary structures performance evaluation. Furthermore, once the optical fiber is well mounted at the temporary structures, there is no need for installation, maintenance and dismantle of sensors during the use of temporary structures, which also saves large amount of cost and time in the repetitive work.

8.6 Recommendations for future Research

While the proposed TSM method has provided the solution to closely integrate the virtual model and the physical components of temporary structures for structural monitoring, more applications

139 and extensible benefits can be gained based on it. Possible future trajectories for this research include: Application of CPS-based TSM for Practical Site Deployment. Although the CPS-based TSM has been developed and validated for its applicability to safety monitoring of temporary structures through laboratory experiment, the wired sensors utilized for the TSM in this research project would be time-consuming and error-prone, especially with the increasing number of sensors in the real world. There is the need to adapt the monitoring and alert system to cope with the practical constraints of a busy construction site. For example, the use of wireless sensors or cameras as a means of data collection for TSM. The alert mechanism also needs to be tailored to be audible and visible in a construction site environment. An Internet of Things (IoT) Approach to Linked Temporary Structures Monitoring. Building on the CPS-based TSM system, the interdependencies between temporary structures can be analyzed for safety purpose using an IoT approach. The approach of IoT also enables the communication among individual temporary components, which provides potential benefits such as improved construction coordination, construction site layout monitoring, and structural failures control. Knowledge-rich TSM Identifying Causes to Failures and Recommendations. During the monitoring of temporary structures, all the information on the temporary structure’s performance and system instructions under various conditions are recorded in the database. By taking advantage of binary tree classification model proposed by Liu et al. (2012), the TSM can be more accurate and efficient in providing decision making support to safety super-intendents in predicting causes to problems, and suggestions for instructions when similar structural problems occur in the future. Due to the complex situation with various information to be considered for decision, the knowledge-rich TSM can be further developed based on the interval-valued intuitionistic fuzzy principle component analysis model (Liu et al. 2013), which offers an effective way in addressing the problem resulted from the large number of decision attributes and high correlation between them. Automatic Control of Temporary Structures’ Failures. A comprehensive application of CPS enables the use of actuators which are placed on the temporary structures on the construction jobsite and behave according to the system commands. The use of actuators can provide an automatic control method for immediate safety adjustments without excessive human

140 interactions. Optimized Emergency Exit Guide in 3D Map. Future application of portable devices can be utilized for location identification of each end user. By integrating the virtual model of temporary structures and the locations of each end user on the construction jobsite, the TSM system provides not only the identification of deficient structural components, but also the person that works in danger near the temporary structures in question. Besides, instead of requiring workers to get out of the construction jobsite without clear direction when there’s an emergency, the location-based TSM can provide optimized exit guidance in 3D map by identifying the shortest safe exit route without getting into the area of potential failures.

8.7 Concluding Remarks

This research demonstrated how CPS can be applied to temporary structures monitoring to address safety issues. It described how CPS is applicable to construction industry with various benefits. A framework on CPS application to construction industry was presented as reference to research explorations in this area. In particular, this research developed a TSM prototype system following the presented framework with the aim of improving safety monitoring of temporary structures. By taking advantage of CPS, the TSM prototype system enables real-time and remote inspection of temporary structures for potential hazards identification which would be difficult to be detected visually by safety super-intendent. The experimental test of TSM presented how efficient and accurate the TSM works under different dangerous scenarios, including base settlement, overloading, lack of diagonal braces, displacement of planks, as well as the impact of lateral load (hit by vehicles or wind). Based on the feedback from engineering workshop, the TSM prototype system provides an efficient and useful solution to temporary structures monitoring. The TSM is highly evaluated due to its user-friendly interface, accuracy, quick to identify potential hazards, and easy to learn. According to the engineering experts, the benefits of TSM to temporary structures monitoring include reducing injuries and deaths; providing real-time monitoring of the temporary structures; early warnings in identifying potential problems; preventing potential cost from damaged components; identifying potential hazards which are unnoticeable by human beings; as well as applicable to other areas (such as the temporary shoring work, the temporary structures under heavy rains, formworks and shoring system on large concrete pours, building movement, crane outriggers, high-speed wind warnings,

141 electrical hazards, hand-railing and fall protection, permanent structure failures, ladders, theater, shoring system, and temporary bracing system.)

Based on this research, it is recommended that practical tests, such as field test of TSM, should be explored in order that the TSM system would be commonly adopted in real projects to assist safety super-intendent in temporary structures monitoring as well as the monitoring of other structures as needed. To deploy practical test, the use of camera for image recognition, wireless sensors, laser scanners, UVA, as well as optical fiber provide alternative method in addressing the limitations of TSM in practical applications. In addition to practical exploration, other research could be built based on the key finding of this project. They include the use of IOT technology for linked temporary structures monitoring; knowledge-rich temporary structures monitoring system in decision making support; automatic control of temporary structures by TSM, as well as optimized emergency exit guide in 3D map enabled by TSM. In general, this research demonstrated CPS applicability and benefits to construction industry by presenting a new approach to identify potential hazards of temporary structures through the close linkage between physical structures and their virtual models. It enhanced the knowledge of enablers and barriers of CPS to temporary structures monitoring and offered reference to the explorations of CPS application to construction industry in the future.

142 REFERENCES Akanmu, A., Anumba, C.J., Messner, J.I., Lee, S. (2012). “Component level Cyber- Physical Systems integration: a light fixtures example.” Proc., Computing in Civil Engineering. 437–444. Anumba, C. J., Akanmu, A., Messner, J. (2010). “Towards a cyber-physical systems approach to construction.” Proc., 2010 Construction Research Congress, Banff, Alberta, May 8-10. 528–538. Arney, D., Pajic, M., Goldman, J. M., Lee, I., Mangharam, R., and Sokolsky, O. (2010). Toward Patient Safety in Closed-loop Medical Device Systems. In: the 1st ACM/IEEE International Conference on Cyber-Physical Systems. NY, USA .12-15 April (pp. 139-148). Anderson, A. (2011). Report to the President on Ensuring American Leadership in Advanced Manufacturing. Executive Office of the President. Washington DC. Akanmu, A., Anumba, C., and Messner, J. (2014). Active Monitoring and Control of Light Fixtures during Building Construction and Operation: Cyber-Physical Systems Approach. Journal of Architectural Engineering, 20(2), 04013088. Akanmu, A., Nguyen, D., Rasheed, S. H., & Olatunji, O (2014). An Automated Approach to Construction Site Layout Generation. In: 50th ASC Annual International Conference, Washington, DC. 26-28 March (pp 1-8). Arora, I., & Gupta, A. (2012). “Cloud databases: a paradigm shift in databases.” International J. of Computer Science Issues, 9(4), 77-83. Asher A., Miller, S. (2011). A Practical Guide to Ethnographic Research in Academic Libraries. The ERIAL Project Aliaga, M., and Gunderson, B. (2005). Interactive statistics (3rd ed.). Upper Saddle River, NJ: Pearson Education, Inc Ayub M. (2010). “Structural Collapses During Construction: Lessons Learned 1990-2008.” Structure Magazine. 12-20. Baheti, R., and Gill, H. (2011). Cyber-Physical Systems. The Impact of Control Technology, 12:161-166. Bugday S. (2012). “Grounded theory designs in qualitative analysis.” < http://www.slideshare.net/SehribanBugday/grounded-theory-11487784>

143 Björk, B. C. (1999). “Information technology in construction: domain definition and research issues.” International Journal of Computer Integrated Design And Construction, SETO, London. Volume 1, Issue 1. 1-16. Berry, C., (2009). A guide to OSHA excavations standards. N. C. Occupational Safety and Health Program. Raleigh, NC Benbasat, I., Goldstein, D. K., & Mead, M. (1987). The case research strategy in studies of information systems. MIS quarterly, 369-386. Chi, S., Hampson, K. D., Biggs, H. C. (2012). “Using BIM for smarter and safer scaffolding and formwork construction: a preliminary methodology.” Proc., Modelling and Building Health and Safety, Singapore. (In press) Cardenas, A., A., Amin, S., and Sastry S. (2008). “Secure Control: Towards Survivable Cyber-Physical Systems.” Proc., the 28th International Conference on Distributed Computing Systems Workshops. 1 (a2): (a3). Cheng, A. M. (2008). Cyber-Physical Medical and Medication Systems. Proc., the 28th International Conference on Distributed Computing Systems Workshops. Beijing, China. 17-20 June (pp. 529-532). Creswell, J. W. (2008). Educational research: Planning, conducting, and evaluating quantitative and qualitative research. Upper Saddle River. New Jersey: Pearson. Caplinskas, A. and Vasilecas, O. (2004). “Information systems research methodologies and models.” International Conference on Computer Systems and Technologies. IV 4.1-4.6. Coalition, S.M.L. (2011). Implementing 21st Century Smart Manufacturing, in: Workshop Report, SMLC and US DOE, June. Cheong L. (2013). “Case Studies on Excavation Involving Earthworks.” (Sep 1, 2013) Derler, P., Lee, E.A., Vincentelli, A.S. (2012). Modeling Cyber–Physical Systems. Proc., IEEE. 100 (1): 13–28. Denzin, N. K. and Lincoln, Y. S. (2005). The Sage Handbook of Qualitative Research (3rd ed.). Thousand Oaks, CA: Sage. Denzin, N. K. (1984). The research act. Englewood Cliffs, NJ: Prentice Hall. Delatte N. J., and Rens K. L. (2002). “Forensics and Case Studies in Civil Engineering Education: State of the Art.” J. Perform. Constr. Facil.

144 Efunda, Inc. (2013). “Rapid prototyping: an overview” Feld, J., and Carper, K. (1997). Construction failure, 2nd Ed., Wiley, New York. Fabiano, B., Fabio, C., Reverberi, A.P., Pastorino, R. (2008). "A statistical study on temporary work and occupational accidents." Specific risk factors and risk management strategies. Fellows, R., and Liu A. (1997) Research Methods for Construction, Oxford: Blackwell Science Limited. Fattal, S. G., Mullen, C., Lew, H. S., & Hunt, B. J. (1980). Analysis of scaffolding accident records and related employee casualties. Gong, Y., and Li, S. (2013). "Fusion framework of urban traffic control and route guidance based on Cyber-Physical System theory." J. Highw. Transp. Res. Dev. Engl. Ed. 7, 82–89. Grant, M., and Pallett, P.F. (2012). Temporary Works: Principles of Design and Construction. ICE Publishing. London. Greenwood, G., Gallagher, J., & Matson, E. (2015). Cyber-Physical Systems: The Next Generation of Evolvable Hardware Research and Applications. In: 18th Asia Pacific Symposium on Intelligent and Evolutionary Systems. Singapore. 10-12 November (pp. 1: 285-296). Goldstein, D., Markus, M. L., Rosen, M. and Swanson, E. B. (1986). "Use of Qualitative Methods in MIS Research," Proceedings of the Seventh International Conference on Information Systems, San Diego, CA. 338-339. Hackmann, G., Guo, W., Yan, G., Sun, Z., Lu, C., & Dyke, S. (2014). “Cyber-physical codesign of distributed structural health monitoring with wireless sensor networks.” Parallel and Distributed Systems, IEEE Transactions, 25(1), 63-72. Huang, X., & Hinze, J. (2006). “Owner’s role in construction safety.” Journal of construction engineering and management, 132: 164-173. Hallowell, M.R., and Gambatese, J.A. (2009). Activity-Based Safety Risk Quantification for Concrete Formwork Construction. J. Constr. Eng. Manag. 990–998. Hsiao H. and Stanevich R. L. (1996). "Biomechanical Evaluation of Scaffolding Tasks." International Journal of Industrial Ergonomics 18: 407-415. Huang, X., Hinze, J. (2003). “Analysis of construction worker fall accidents.” J. Constr. Eng. Manag. 129, 262–271.

145 Hadipriono, F. C., and Wang H. K. (1986). “Analysis of Causes of Falsework Failures in Concrete Structures.” J. Constr. Eng. Manag.112: 112-121. Halperin, K. M., & McCann, M. (2004). An evaluation of scaffold safety at construction sites. Journal of Safety Research, 35(2), 141-150. International Building Code (2006). Requirement for Temporary Stages. < https://www.in.gov/dhs/files/20120410121621937(1).pdf > (December 2nd, 2012) Jung, Y. (2014). An Approach to Automated Detection of Failure in Temporary Structures using Image Processing. Journal of Engineering and Architecture, 2(1), 49-61. Jones, T. S., Richey R. C. (2000). Rapid Prototyping Methodology in Action: A Developmental Study. ERT &D. Vol. 48, No. 2．69-80. Jannadi, M. O., & Assaf, S. (1998). “Safety assessment in the built environment of Saudi Arabia.” Safety Science, 29(1), 15−24. Kaihara, T., & Yao, Y. (2012). A new approach on CPS-based scheduling and WIP control in process industries. Simulation Conference (WSC), Proceedings of the 2012 Winter.1-11. Kim, H., Ahn, H., & Kim, W. (2011). 3-D Temporary Facility Visualization Using BIM (Building Information Modeling) Technology. KICEM Journal of Construction Engineering and Project Management, 1, 37-42. Krogh, B.H., Lee, E., Lee, I., Mok, A., Rajkumar, R., Sha, L.R., Vincentelli, A.S., Shin, K., Stankovic, J., Sztipanovits, J. (2008). "Cyber-Physical Systems: executive summary." CPS Steer Group, Wash. DC. Kim, J., Fischer, M., Kunz, J., and Levitt, R. (2014). Sharing of Temporary Structures: Formalization and Planning Application. Automation in Construction, 43, 187-194. Khaitan, S. K., & McCalley, J. D. (2014). Design techniques and applications of cyber physical systems: a survey. Systems Journal, IEEE, 9 (2), 350-365. Karnouskos, S. (2011). Cyber-Physical Systems in the Smart Grid. In 2011 9th IEEE International Conference on Industrial Informatics (INDIN). Lisbon, Portugal .26-29 July (pp. 20-23). Kaminetzky, D. (1991). Design and Construction Failures: Lessons from Forensic Investigations. McGraw-Hill, 204–211. Kleinosky, T. J. (2012). Temporary Structure Failure – Case Studies, from http://failures.wikispaces.com/Temporary+Structure+Failure+-+Case+Studies .

146 Kaplan, B. and Duchon, D. (1988). “Combining Qualitative and Quantitative Methods in Information Systems Research: A CaseStudy.” MIS Quarterly, Vol. 12, No. 4. 571-586 Kim J. L. (2006). "Assessment of Shoring System for Safety of Elevated Slab Formwork Using Nested Factor Designs." Building Integration Solutions. 1-14. Li, H., Huang, T., Kong, C. W., Guo, H. L., Baldwin, A., Chan, N., & Wong, J. (2008). “Integrating design and construction through virtual prototyping.” Automation in Construction, 17(8), 915-922. Lee, E. A. (2007). Computing foundations and practice for cyber-physical systems: A preliminary report. University of California, Berkeley, Tech. Rep. UCB/EECS-2007-72. Liu, W., and Su, J. (2011). A Solution of Dynamic Manufacturing Resource Aggregation in CPS. In: Information Technology and Artificial Intelligence Conference (ITAIC), 2011 6th IEEE Joint International. Chongqing, China. 20-22, August (Vol. 2, pp. 65-71). Lee, I., and Sokolsky, O. (2010). Medical Cyber Physical Systems. In: Design Automation Conference (DAC), 2010 47th ACM/IEEE, Anaheim, CA. 13-18 June (pp. 743-748). Lee, J., Bagheri, B., & Kao, H. A. (2015). A cyber-physical systems architecture for industry 4.0-based manufacturing systems. Manufacturing Letters, 3, 18-23. Liu, B., Shen, Y., Chen, X., Chen, Y., & Wang, X. (2014). A partial binary tree DEA-DA cyclic classification model for decision makers in complex multi-attribute large-group interval-valued intuitionistic fuzzy decision-making problems. Information Fusion, 18, 119-130. Liu, B., Chen, Y., Shen, Y., Sun, H., & Xu, X. (2014). A complex multi-attribute large-group decision making method based on the interval-valued intuitionistic fuzzy principal component analysis model. Soft Computing, 18(11), 2149-2160. Lintelman, S., Sampigethaya, K., Li, M., Poovendran, R., and Robinson, R. (2008). High Assurance Aerospace CPS and Implications for the Automotive Industry. In Proceedings of the National Workshop on High Confidence Automotive Cyber-Physical Systems (CPS). Washington, DC. 18-20 November. Moon, S., Choi, B., Yang, B. (2012). “USN-Based Data Acquisition for Increasing Safety in the Concrete Formwork Operation.” J. Comput. Civ. Eng. 26, 271-281. Mckiniley J. C. (2011). “After Accidents, a Call for Regulation.”

147 ml?_r=0> (August 23rd, 2011) Mertins, K., Knothe, T., and Gocev, P. (2012). Towards CPS Based Aircraft MRO. In Collaborative Networks in the Internet of Services. Springer Berlin Heidelberg. 166-173. Mckiniley J. C. (2011). After Accidents, a Call for Regulation. (October 3rd, 2012) Martin, P. Y. and Turner, B. A. (1986) Grounded Theory and Organizational Research. The Journal of Applied Behavioral Science, vol. 22, no. 2, 141. Montgomery, P. L. (1981). Two Still Missing in Condominium Collapse. (December 3rd, 2011) Milenković, A., Otto, C., & Jovanov, E. (2006). Wireless sensor networks for personal health monitoring: Issues and an implementation. Computer communications, 29(13), 2521-2533. Manno D. (2009). University of Washington Football Stadium Addition Collapse (1987). https://failures.wikispaces.com/University+of+Washington+Football+Stadium+(Manno) (August 20th, 2011) NITRD, 2012. Winning the Future with Science and Technology for 21st Century Smart Systems. Penny Hill Press Inc. Noor, A. K. (2011). Intelligent Adaptive Cyber-Physical Ecosystem for Aerospace Engineering Education, Training, and Accelerated Workforce Development. Journal of Aerospace Engineering, 24(4), 403-408. Newsted, P. R., Huff, S. L. and Munro, M. C. (1998). “Survey Instruments in Information Systems.” MIS Quarterly December. 553-554. National Association of Home Builders. (2012). "Petition for Rulemaking to Amend the Residential Fall Protection Standard: 29 CFR 1926, Subpart M." Nagayama, T., & Spencer Jr, B. F. (2007). Structural health monitoring using smart sensors. Newmark Structural Engineering Laboratory. University of Illinois at Urbana-Champaign. OSHA (2014). “OSHA Scaffold Certificate Course.”

148 (2014). Parfitt, M.K. (2009). "Cranes, structures under construction, and temporary facilities: are we doing enough to ensure they are safe?" J. Arch. Eng. 15, 1–2. Penna, R., Amaral, M., Espindola, D., Botelho, S., Duarte, N., Pereira, C. E., Zuccolotto, M. and Morosini Frazzon, E. Visualization tool for cyber-physical maintenance systems. In: 12th IEEE International Conference on Industrial Informatics (INDIN). Porto Alegre, 27-30 July (pp. 566-571). Peraza D. B. (2006). "Getting to the Bottom of Underpinning." STRUCTURE Magazine. December 2006. 16-20 Peraza D. B. (2007). "Preventing and Dealing with Building Damage Due to Underpinning." Forensic Engineering Symposium. 1-10. Peraza D. B. (2008). "Practices to Reduce Failures: Undermining and Underpinning-Suggestions for Minimizing Damage to Adjacent Buildings During Construction." J.Perform. Constr. Facil. 22: 70-70. Rajkumar, R. R., Lee, I., Sha, L., and Stankovic, J. (2010). Cyber-Physical Systems: the Next Computing Revolution. In: 2010 47th ACM/IEEE, Anaheim, CA. 13-18 June (pp. 731-736). Reeves, S., Kuper, A., and Hodges, B. D. (2008) “Qualitative research methodologies: ethnography.” BMJ 337: a1020: 512-514. Ratay, R.T. (1996). Handbook of Temporary Structures in Construction, 2nd ed., McGraw-Hill. New York. Shi, J., Wan, J., Yan, H., Suo, H. (2011). "A survey of cyber-physical systems, in: Wireless Communications and Signal Processing (WCSP)." Proc., 2011 International Conference. 1–6. Stewart P. (2010). “Four Times Square - Scaffold Collapse (July 21, 1998).” (Oct. 1, 2013). Smart System. (2012). Cyber Physical Systems: situation analysis of current trends, technologies, and challenges. Technical Report. National Institute of Standards and Technology, Columbia, Maryland. Shapira, A. (1999). Contemporary Trends in Formwork Standards-A Case Study. J. Constr. Eng. Manag. 125, 69–75.

149 Sun, Y., McMillin, B., Liu, X. F., and Cape, D. (2007). Verifying Noninterference in a Cyber-Physical System the Advanced Electric Power Grid. Quality Software, 2007. Seventh International Conference on. IEEE. Portland, Oregon, USA. 11-12, October (pp. 363-369). Sha, L., Gopalakrishnan, S., Liu, X., and Wang, Q. (2009). Cyber-Physical Systems: A New Frontier. In Machine Learning in Cyber Trust. Springer, NY. 3-13. Shaughnessy, J., Zechmeister, E. and Jeanne, Z. (2011). Research methods in psychology (9th ed.). New York, NY: McGraw Hill. 161–175. SEAoNY. (2005). "An Underpinning Symposium." (Sep. 1, 2013) Studytonight (2014). “Types of Communication Networks.” < http://www.studytonight.com/computer-networks/types-of-networks > (Jun. 3, 2015) Skibniewski, M., & Golparvar-Fard, M. (2016). Toward a Science of Autonomy for Physical Systems: Construction. arXiv preprint arXiv:1604.03563. Tidwell, T., Gao, X., Huang, H. M., Lu, C., Dyke, S., and Gill, C. (2009). Towards Configurable Real-time Hybrid Structural Testing: a Cyber-Physical System Approach. In: Object/Component/Service-Oriented Real-Time Distributed Computing, IEEE International Symposium. Tokyo, Japan. 17-20, March (pp. 37-44). Thomas, G. (2011). “A typology for the case study in social science following a review of definition, discourse and structure.” Qualitative Inquiry, 17, 6. 511-521. Tang, P., Huber, D., Akinci, B., Lipman, R., & Lytle, A. (2010). Automatic reconstruction of as-built building information models from laser-scanned point clouds: A review of related techniques. Automation in construction, 19(7), 829-843. UEI (United Electronic Industries) (2006). “Data Acquisition Systems Then and Now.” < http://www.ueidaq.com/media/static/other/so/data-acquisition.pdf > (Oct. 20, 2014). U.S. Department of Labor (2014). “Top 10 Most Frequently Cited Standards.” < https://www.osha.gov/Top_Ten_Standards.html > (Jan. 01, 2015). U.S. Bureau Labor Statistics (2003-2012). Occupational Injuries by Selected Characteristics, from http://www.bls.gov/iif/oshwc/cfoi/all_worker.pdf. Wu, N., & Li, X. (2011). “RFID applications in cyber-physical system.” RFID Applications in Cyber-Physical System, 291-302.

150 Wan, K., and Alagar, V. (2012). "Resource modeling for Cyber Physical Systems, in: Systems and Informatics (ICSAI)." Proc., 2012 International Conference. 2541–2546. Wan, J., Suo, H., Yan, H., & Liu, J. (2011). A general test platform for cyber-physical systems: unmanned vehicle with wireless sensor network navigation. Procedia Engineering, 24, 123-127. Willis, J. (2007). Foundations of Qualitative Research: Interpretive and Critical Approaches. Thousand Oaks: Sage Publications. Whitaker, S. M., Graves, R. J., James, M., & McCann, P. (2003). “Safety with access scaffolds: Development of a prototype decision aid based on accident analysis.” Journal of safety research, 34(3), 249-261. Wainscott J.E. (2011). "Outdoor Festivals and Fairs: Temporary Stages." (Oct. 10, 2013) Weesner, L. B., & Jones, H. L. (2001). Experimental and analytical capacity of frame scaffolding. Engineering Structures, 23(6), 592-599. Yabuki, N., & Oyama, T. (2007). “Application of radio frequency identification technology for management of light weight temporary facility members.” In ASCE Workshop on Computing in Civil Engineering, Pittsburgh, 697-704. Yuan, X., Parfitt, K. M, Anumba, C. J. (2014). “The Use of Cyber-Physical Systems in Temporary Structures–An Exploratory Study.” In Computing in Civil and Building Engineering. ASCE, Florida, 1707-1714. Yang, J., Park, M. W., Vela, P. A., and Golparvar-Fard, M. (2015). Construction performance monitoring via still images, time-lapse photos, and video streams: Now, tomorrow, and the future. Advanced Engineering Informatics, 29(2), 211-224. Yin, R.K. (1984). Case Study Research. Sage, Beverly Hills, CA. Yuan, X. (2013). “Overview of Temporary Structural Failures in Construction.” (April. 12, 2015). Zhang, S., Sulankivi, K., Kiviniemi, M., Romo, I., Eastman, C. M., & Teizer, J. (2015). BIM-based fall hazard identification and prevention in construction safety planning. Safety science, 72, 31-45. Zhu, Q., and Basar, T. (2011). Robust and Resilient Control Design for Cyber-Physical Systems

151 with an Application to Power Systems. In: Decision and Control and European Control Conference (CDC-ECC), 2011 50th IEEE Conference. Florida, USA. 12-15 December.

152 APPENDIX A Evaluation of Temporary Structures Monitoring (TSM) System

Thank you for sharing your experience with us! There are 17 questions and one optional question in total.

Personal Information  Company ______ Job Position ______ Years of Working Experience ______

1. In general, how useful is the Temporary Structures Monitoring (TSM) system in improving the safety of scaffolding? Extremely Very Moderately Slightly Not useful useful useful useful useful at all Usefulness of the TSM system      Please provide reasons here: ______

2. How likely are you to embrace the use of the TSM system for temporary structures monitoring on the jobsite? Extremely Very Moderately Slightly Not likely likely likely likely likely at all Adoption of the TSM system      Please provide reasons here: ______

153 3. How realistic do you rate the experimental tests (the simulation of the five major failures)? Extremely Very Moderately Slightly Not realistic realistic realistic realistic realistic at all Base settlement of scaffolding system      Overloading of scaffolding system      Severe displacement of scaffolding      planks Lateral load (such as wind or vehicle      impact) Diagonal brace not in position      Please provide reasons here: ______4. How do you rate the accuracy of the TSM system in identifying potential hazards during the five experimental tests? Extremely Very Moderately Slightly Not accurate accurate accurate accurate accurate at all Accuracy of the TSM system      Please provide reasons here: ______

5. How well are the scaffolding components linked with their 3D models? Extremely Very Moderately Slightly Not well well well well well at all Linkage between the scaffolding      components and 3D models Please provide reasons here: ______

6. How much does the linkage between the scaffolding components and their 3D models contribute to improving safety management?

154 Extremely Very Moderately Slightly Not important important important important important at all Importance of the linkage between the components      and 3D models Please provide reasons here: ______

7. The warning system sends alarms within 2s once a problem is identified. How do you rate the 2 second warning to prevent potential hazards? Extremely Very Moderately Slightly Not adequate adequate adequate adequate adequate at all Adequate for prevention      Please provide reasons here: ______8. How do you rate the performance of the TSM system in relation to the following features? Extremely Very Moderately Slightly Not effective effective effective effective effective at all Real-time inspection      Immediate notification      3D visualization of problem      Remote access to identify problem      Access by multiple parties      simultaneously Non-intrusive (Minimum impact) to      work Easy to use     

9. How do you rate the importance of notifying the following people of potential hazards by the TSM system?

155 Extremely Very Moderately Slightly Not important important important important important at all Safety Superintendent      Construction workers      Project Manager      Please provide reasons here: ______

10. In addition to the people listed above, is there anyone else that should be notified of the potential hazards immediately by the TSM system? ______

11. To what extent, would you trust the information from the TSM system? Extremely Very Moderately Slightly Not trust at trust trust trust trust all Trust level of the TSM      information Please provide reasons here: ______12. How long would you need to learn to use the TSM system? Less than 30-60 60-90 90-120 Other (please 30 minutes minutes minutes minutes specify) Time to learn to use TSM    

13. What are the potential benefits of the TSM system to the construction industry? ______

14. What are other potential application areas of the TSM system? ______

15. What improvements would you consider necessary? ___

156 ______

16. What practical considerations are necessary for implementing the TSM system on a real jobsite? ______

17. If you have additional suggestions, please provide comments below to help us improve the TSM system: ______18. Please complete the following information if you would like to be contacted for further discussion in the future (optional).  Name ______ E-mail ______ Phone ______Thank you again for sharing your experience with us! We highly value your feedback and opinions!

157

CURRICULUM VITAE Xiao Yuan

EDUCATION Doctor of Philosophy, Department of Architectural Engineering The Pennsylvania State University, University Park, PA Master of Science in Management Science and Engineering Tongji University, Shanghai, China Bachelor of Engineering, Construction Cost Chongqing University, Chongqing, China Other Educational Experience, Law School Chongqing University, Chongqing, China

PROFESIONAL CERTIFICATIONS  Occupational Safety and Health Administration (OSHA) 30-hour Certificate  Graduate Student Online Teaching Certificate, provided by The Pennsylvania State University

PUBLICATIONS Articles in refereed journals (published/ under review) 1. Yuan, X., Anumba C. J., Parfitt K. M. (2015). “Review of the Potential for A Cyber-Physical Systems (CPS) Approach to Temporary Structures Monitoring.” International Journal of Architectural Research. 9 (3), 26-44. 2. Yuan, X., Anumba C. J., Parfitt K. M. (2015). “CPS-based Temporary Structures Monitoring by Bridging the Virtual and Physical World.” Automation in Construction. 66: 1-14 3. Xu, D. P., Yuan, X., Wu, D. H., Anumba C.J. (2015). “Model Checking Cyber-Physical Systems (CPS): A Case Study.” Journal of Computing in Civil Engineering. (under review) Peer-Reviewed Conference Proceedings 4. Yuan, X., Parfitt, M. K., & Anumba, C. J. (2014). “The Use of Cyber-Physical Systems in Temporary Structures–An Exploratory Study.” Computing in Civil and Building Engineering, ASCE. Orlando, Florida. pp. 1707-1714. 5. Yuan, X., Parfitt, M. K., & Anumba, C. J. (2015). “Real-time Cyber-Physical Systems (CPS)-based Monitoring of Temporary Structures: a Scaffolding System Example.” Proceedings of 32nd International CIB W78 Conference. Eindhoven, the Netherland. (full paper accepted) 6. Yuan, X., Parfitt, M. K., & Anumba, C. J. (2016). “Rapid Prototyping of Cyber-Physical Systems (CPS) for Temporary Structures Monitoring: A Case Study of Scaffolding System.” Proceedings of 16th International Conference on Computing in Civil and Building Engineering. Osaka, Japan.

158