Automated Site Data Acquisition for Effective

Project Control

Ali Montaser

A Thesis

In the Department of

Building, Civil, and Environmental Engineering

Presented in Partial Fulfillment of the Requirements

For the Degree of

Doctor of Philosophy (Building Engineering) at

Concordia University

Montreal, Quebec, Canada

December 2013

© Ali Montaser, 2013

 CONCORDIA UNIVERSITY

SCHOOL OF GRADUATE STUDIES

This is to certify that the thesis prepared

By: Ali Montaser

Entitled: Automated Site Data Acquisition for Effective Pr oject Contr ol

and submitted in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY (Building Engineering)

complies with the regulations of the University and meets the accepted standards with respect to
originality and quality.

Signed by the final examining committee:

Chair
Dr. C.Y. Su

External Examiner
Dr. C. Haas

External to Program
Dr. A. Hammad

Examiner
Dr. T. Zayed

Examiner
Dr. Z. Zhu

Thesis Supervisor
Dr. O. Moselhi

Approved by
Dr. M. Elektorowicz, Graduate Program Director

December 5, 2013
Dr. C. Trueman, Interim Dean
Faculty of Engineering & Computer Science
 ABSTRACT

Automated Site Data Acquisition for Effective Project Control

Ali Montaser, Ph.D.

Concordia University, 2013

Tracking and control of construction projects depend primarily on the accuracy,

frequency and time required to collect actual onsite data of construction operations that

characterize the work progress. Earned Value Analysis (EVA) is being used for reporting work

progress and for forecasting project status at completion and at any future time horizon. Critical

to its reliable application is accurate and timely data for quantifying the budgeted cost of work

performed. Automated site data acquisition has received considerable attention in recent years

to circumvent the limitations of manual data collection. The limitations of existing models lie

in their inability to measure the progress of different types of construction activities from

inception to completion in near real-time.

The objective of this research is to study and analyze the characteristics of automated

data acquisition technologies in construction. This thesis is carried out with a focus on

automating the process of data collection and project control. An automated model is developed

that integrates project visualization-information aspects, automated site data acquisition and

earned value analysis. The developed model consists of two main frameworks; one is for data

acquisition and the other for data analysis and processing. Data acquisition is carried out using

the integrated automated data acquisition technologies embedded in the tablet PC used in this

research as well as others, referred to here as independent, automated data acquisition

technologies. The developed model makes full use of 4D BIM to provide visualization and

pertinent information of activities in progress. The developed model embraces the human

iii
factors to augment the visualization-information aspects, localization technologies and

development of progress templates. The developed model demonstrated the significance of

data fusion of a wide range of automated site data acquisition technologies and

visualization-information technologies.

A prime focus is placed on extensive field studies and experimental work. Field studies

on construction jobsite are conducted utilizing a wide range of technologies such as 3D

Scanner, RFID and GPS. Extensive experimental work is conducted to develop deployment

protocols for utilization of Ultra High Frequency (UHF) passive RFID in localization and

material tracking. The field and lab work resulted in a number of observations, findings, and

lessons learned for deployment of passive RFID in construction. The results presented in this

study demonstrate the potential use of short range RFIDs in location estimation and material

tracking in a cost-effective manner for indoor construction operations. The developed method

for location identification and material tracking using RFID technology can be used to obtain

information required for scalable, near-real-time decision-making and timely tracking of

project status.

The developed methods and algorithms are implemented in prototype software. It

consists of two computational frameworks; BIM+ and Control+. BIM+ is Two-Tier software,

which utilizes an object-oriented BIM model. So, it can be used as an advanced tool for data

acquisition through the user’s tablet PC. Control+ is Three-Tier Web-Based software for

processing the captured data from the tablet PC and independent automated data acquisition

technologies. The developed model, methods, algorithms and software constitute a step

ahead of current progress reporting applications and expand upon automated site data

acquisition technologies and visualization-information technologies use in construction.

iv
To my adorable mother, beloved family; Mona, Ahmed, Abdallah and Mohamed

v
 ACKNOWLEDGMENT

All praise and thanks are due to Allah for giving me the patience and perseverance to

successfully accomplish my Ph.D. study. This thesis is dedicated to my family for their endless

support and encouragement throughout my life. Their unconditional love has been source of

inspiration. I could never achieve this work without their endless encouragement and help.

I would like to express my sincere gratitude and thanks to many people who have

helped me during my graduate studies at Concordia University. I am greatly thankful to my

supervisor Dr. Osama Moselhi for his for his invaluable guidance and patience with me

throughout my studies. It was a great honor to work under his supervision. I would like also to

thank Dr. Tarek Zayed and Dr. Zhenhua Zhu for their help during my work.

I would like gratefully to acknowledge the support of Dr. Saiedeh N. Razavi, Dr. Adel

Alshibani and Dr. Samir El-Omari for sharing their knowledge with me. Finally, I would like

to thank all my colleagues in the Automation and Construction Laboratory. And special thanks

to Sasan Golanraghi, Hani Alzraiee, Ibrahim Bakry, Ahmed Atef, Ahmed Eweda, Tushith

Islam, Magdy Omar, Ramy Elfaizy, Fady Halim and Mohammed Gomma for their

encouragement and assistance. Your friendship is irreplaceable. I feel so lucky to have you in my

life. Many thanks to my friend Ashraf El Hakeem for his sincere support and help during the

software programming part in this research.

vi
 TABLE OF CONTENTS

LIST OF FIGURES ....................................................................................................................................... x

LIST OF TABLES ...................................................................................................................................... xiv

1 Chapter 1: INTRODUCTION .................................................................................................................. 1

1.1 General ..............................................................................................................................................1
1.2 Research Motivation and Problem Statement...................................................................................2
1.3 Objectives ..........................................................................................................................................5
1.4 Research Philosophy ..........................................................................................................................6
1.5 Research Methodology ......................................................................................................................7
1.6 Thesis Organization .........................................................................................................................10

2 Chapter 2: LITERATURE REVIEW......................................................................................................... 12

2.1 General ............................................................................................................................................12
2.2 Project Tracking and Control ...........................................................................................................13
2.2.1 Integrated Tracking and Control .................................................................................................15
2.2.2 Measuring Project Performance .................................................................................................16
2.2.3 Traditional Forecasting Methods ................................................................................................17
2.2.4 Simulation as Forecasting Tool ...................................................................................................19
2.3 Automated Data Acquisition Technologies......................................................................................21
2.3.1 Bar Code ......................................................................................................................................23
2.3.2 Radio Frequency Identification (RFID) ........................................................................................24
2.3.3 Global positioning System (GPS) .................................................................................................28
2.3.4 Multimedia ..................................................................................................................................29
2.3.5 Tablet Computers ........................................................................................................................30
2.3.6 3D Imaging and Photogrammetry ...............................................................................................31
2.3.7 Other Types of Sensors ...............................................................................................................34
2.4 Localization ......................................................................................................................................35
2.4.1 Signal Measurements ..................................................................................................................37
2.4.2 Localization Algorithms ...............................................................................................................38
2.4.3 Indoor Localization Technologies ................................................................................................40
2.4.4 RFID Indoor Localization .............................................................................................................42
2.5 Data Visualization ............................................................................................................................43
2.5.1 Geographic Information System (GIS) .........................................................................................44
2.5.2 Building Information Modeling (BIM) .........................................................................................45
2.6 Summary and Identified Limitations ................................................................................................48

3 Chapter 3: PROPOSED MODEL ........................................................................................................... 50

3.1 General ............................................................................................................................................50
3.2 Proposed Model ...............................................................................................................................54
3.3 Data Acquisition Framework ...........................................................................................................58

vii
 3.3.1 Tablet PC Data Acquisition ..........................................................................................................58
3.3.1.1 Building 4D BIM ...............................................................................................................62
3.3.1.2 User Location Identification ............................................................................................66
3.3.1.3 Zone Configuration ..........................................................................................................76
3.3.1.4 Real Time Visualization ....................................................................................................77
3.3.1.5 Tablet PC Database ..........................................................................................................81
3.4 Data Analysis and Reporting Framework ........................................................................................87
3.4.1 Performance Measurement Module ..........................................................................................89
3.4.1.1 Material Localization and Tracking ..................................................................................89
3.4.1.2 Productivity and Quantity Measurement ........................................................................96
3.4.2 Earned Value Analysis Module ..................................................................................................119
3.4.3 Central Database .......................................................................................................................126
3.4.4 Reporting Module .....................................................................................................................130
3.5 Summary ........................................................................................................................................130

4 Chapter 4: FIELD STUDIES ................................................................................................................ 131

4.1 General ..........................................................................................................................................131
4.2 Case Study......................................................................................................................................132
4.3 Preparations for Field Work ...........................................................................................................134
4.4 Independent Automated Data Acquisition Technologies ..............................................................137
4.4.1 3D Scanner ................................................................................................................................137
4.4.2 Fixed GPS ...................................................................................................................................141
4.4.3 CCTV Camera .............................................................................................................................148
4.5 Tablet PC Automated Data Acquisition Technologies....................................................................149
4.5.1 Camera ......................................................................................................................................151
4.5.2 Barcode .....................................................................................................................................153
4.5.3 Mobile GPS ................................................................................................................................153
4.5.4 Mobile RFID ...............................................................................................................................158
4.6 Summary ........................................................................................................................................162

5 Chapter 5: EXPERIMENTAL STUDIES OF PASSIVE RFID INDOOR FOR LOCALIZATION ........................ 164
5.1 General ..........................................................................................................................................164
5.2 Zone Identification Using K-NN Algorithm .....................................................................................165
5.2.1 Field Study .................................................................................................................................167
5.2.1.1 Tag Deployment Topologies ..........................................................................................167
5.2.1.2 Closeness of Passive Reference Tags .............................................................................168
5.2.1.3 Stationary vs. Moving Reader Agent .............................................................................170
5.2.1.4 Zone Configuration ........................................................................................................171
5.2.1.5 Speed of Moving Reader ...............................................................................................173
5.2.2 Laboratory Experiments ............................................................................................................174
5.2.2.1 Closeness of Passive Reference Tags .............................................................................174
5.2.2.2 Studies on RSSI ..............................................................................................................175
5.3 Location Identification Using Two-Step Algorithm ........................................................................179
5.4 Deployment Protocols Using Proximity Algorithm.........................................................................188
5.5 Summary ........................................................................................................................................193

viii
6 Chapter 6: COMPUTER IMPLEMENTATION ...................................................................................... 196
6.1 General ..........................................................................................................................................196
6.2 Developed System..........................................................................................................................196
6.2.1 System Platform ........................................................................................................................198
6.2.2 BIM+ ..........................................................................................................................................203
6.2.3 Control+ ....................................................................................................................................233
6.3 Summary ........................................................................................................................................256

7 Chapter 7: CONCLUSIONS AND FUTURE WORK ................................................................................ 259

7.1 Conclusions ....................................................................................................................................259
7.2 Research Contribution ...................................................................................................................260
7.3 Limitations .....................................................................................................................................263
7.4 Future Work ...................................................................................................................................264

References ........................................................................................................................................... 266

8 Appendix A: Tracking Scraper-Pusher Fleet for Cut-Fill Earthmoving Operations ............................. 276

9 Appendix B: Central Database ......................................................................................................... 278

10 Appendix C: Data Collection ............................................................................................................ 290

11 Appendix D: Productivity of Earthmoving Operations in Planning ................................................... 297

12 Appendix E: Computer Implementation........................................................................................... 302

ix
 LIST OF FIGURES

Figure 1-1: Research methodology ........................................................................................... 9

Figure 2-1: Chapter 2 overview ............................................................................................... 13
Figure 2-2: Earned value method (Li, 2004) ........................................................................... 16
Figure 2-3: self-learning model (Li et al., 2006) ..................................................................... 19
Figure 2-4: 1D and 2D bar code (McCullouch and Lueprasert, 1994).................................... 23
Figure 2-5: RFID system components..................................................................................... 26
Figure 2-6: Portal structure with 4 fixed RFID readers (Song et al, 2006) ............................. 28
Figure 2-7: 3D scan and 3D model object recognition (Bosche et al, 2009) .......................... 33
Figure 2-8: Switch limit main components (ABB, 2013) ....................................................... 35
Figure 2-9: (a) Trilateration (b) Angulation localization algorithm ........................................ 38
Figure 3-1: Chapter 3 overview ............................................................................................... 50
Figure 3-2: Research vision ..................................................................................................... 52
Figure 3-3: Developed Model ................................................................................................. 55
Figure 3-4: Data acquisition framework overview .................................................................. 59
Figure 3-5: Data acquisition framework architecture .............................................................. 60
Figure 3-6: Indoor data acquisition (Montaser and Moselhi, 2012 c) ..................................... 63
Figure 3-7: Linking the 3D BIM model to project schedule ................................................... 64
Figure 3-8: The process of 4D BIM model visualization ........................................................ 65
Figure 3-9: GPS user location identification ........................................................................... 67
Figure 3-10: RFID reader receives signals .............................................................................. 69
Figure 3-11: The process of deploying RFID reference tags .................................................. 70
Figure 3-12: User location identification using trilateration algorithm ................................... 71
Figure 3-13: User location identification using trilateration algorithm with imperfect path loss
model ....................................................................................................................................... 72
Figure 3-14: Trilateration algorithm (Montaser and Moselhi, 2013 a) ................................... 73
Figure 3-15: Proximity method concept .................................................................................. 75
Figure 3-16: Proximity algorithm (Montaser and Moselhi, 2013 a) ....................................... 75
Figure 3-17: Zone Configuration with time ............................................................................ 76
Figure 3-18: Area plan and floor plan ..................................................................................... 77
Figure 3-19: Real time visualization ingredients..................................................................... 78
Figure 3-20: The process of real time visualization (part 1) ................................................... 79
Figure 3-21: The process of real time visualization (part 2) ................................................... 80
Figure 3-22: Tablet PC database entity relationship (ER) diagram......................................... 83
Figure 3-23: The process of capturing data using the tablet PC.............................................. 86
Figure 3-24: Data analysis and reporting framework overview .............................................. 87
Figure 3-25: Data analysis and reporting framework architecture .......................................... 88
Figure 3-26: Diagrammatic sketch for material location identification .................................. 90
Figure 3-27: Material location identification using Proximity algorithm ............................... 91
Figure 3-28: Material location identification using Trilateration algorithm ........................... 92
Figure 3-29: Material localization and tracking sub-module .................................................. 94
Figure 3-30: Estimating “Units Activities” percentage complete ........................................... 95
Figure 3-31: Flowchart for the process of tracking earthmoving operation ............................ 98
Figure 3-32: Flowchart for the process of the captured data processing ................................. 99
Figure 3-33: GPS/GIS based tracking (Montaser et al., 2012).............................................. 103
Figure 3-34: GPS captured data processing part 1 ................................................................ 104
Figure 3-35: GPS captured data processing part 2 ................................................................ 105

x
Figure 3-36: RFID gate based tracking (Montaser and Moselhi, 2012 b .............................. 108
Figure 3-37: RFID captured data processing (Montaser and Moselhi, 2012 b) .................... 110
Figure 3-38: Cycle time in GPS/GIS and RFID gate tracking method ................................. 111
Figure 3-39: RFID reader captured data with excavator ....................................................... 112
Figure 3-40: RFID reader captured data with loader............................................................. 113
Figure 3-41: The change in control sensor captured data...................................................... 114
Figure 3-42: RFID and control sensor data integration (Montaser and Moselh, 2013 b) ..... 114
Figure 3-43: RFID/control sensor captured data processing ................................................. 116
Figure 3-44: RFID/Control Sensor method and GPS/GIS or RFID Gate method ................ 117
Figure 3-45: The process of stochastically forecasting productivity ..................................... 124
Figure 3-46: Specific purpose earthmoving operation DES model....................................... 125
Figure 3-47: Entity Relationship (ER) diagram for the central database .............................. 127
Figure 4-1: Chapter 4 overview ............................................................................................. 131
Figure 4-2: Genomics office building ................................................................................... 132
Figure 4-3: Sample of project progress reports ..................................................................... 133
Figure 4-4: Genomics building 3D BIM model (Montaser and Moselhi, 2012 a) ................ 134
Figure 4-5: 4D BIM NavisWorks© model ............................................................................ 136
Figure 4-6: RIEGL LPM 100 VHS LADAR ........................................................................ 138
Figure 4-7: (a) GPS unit (b) Genomics building (Montaser et al., 2011).............................. 141
Figure 4-8: GPS based daily productivity (Montaser et al., 2012) ........................................ 143
Figure 4-9: Cycle time components PDFs............................................................................. 145
Figure 4-10: ExtendSim simulation model............................................................................ 145
Figure 4-11: Simulation results vs. upper and lower limit planed estimates......................... 147
Figure 4-12: Sample of CCTV camera captured data ........................................................... 149
Figure 4-13: Sample of CCTV camera captured data ........................................................... 150
Figure 4-14: Sample of Indoor camera captured data ........................................................... 151
Figure 4-15: 4D BIM image vs. actual image for outdoor zone............................................ 152
Figure 4-16: 4D BIM image vs. actual image for indoor zone.............................................. 152
Figure 4-17: Items with barcode inside Genomics building jobsite ...................................... 153
Figure 4-18: GPS hardware and screen shot ......................................................................... 155
Figure 4-19: Post-processed GPS data using GPS PathFinder software ............................... 156
Figure 4-20: User route during outdoor filed experiment ..................................................... 158
Figure 4-21: RFID hardware ................................................................................................. 159
Figure 4-22: RFID reader screen shots .................................................................................. 160
Figure 4-23: RFID data acquisition from reference tags and material tags........................... 160
Figure 4-24: Different test beds setup and pictures ............................................................... 161
Figure 5-1: Chapter 5 overview ............................................................................................. 165
Figure 5-2: Two deployment topologies .............................................................................. 168
Figure 5-3: On-site scenarios for tag closeness ..................................................................... 169
Figure 5-4: Different scenarios for a stationary reader agent ................................................ 170
Figure 5-5: Scenarios for different zone configurations........................................................ 172
Figure 5-6: The trajectory of the moving target (reader)....................................................... 173
Figure 5-7: Concordia university construction automation lab ............................................. 174
Figure 5-8: Average RSSI vs. distance for four tags ............................................................. 177
Figure 5-9: Regression analysis............................................................................................. 178
Figure 5-10: Two-step algorithm (Montaser and Moselhi, 2013a) ....................................... 179
Figure 5-11: Test beds 1, 2 & 3 setup and site pictures Case Study I ................................... 181
Figure 5-12: Test beds 6 setup Case Study II ........................................................................ 182
Figure 5-13: Localization error for both algorithms.............................................................. 183
xi
Figure 5-14: Comparison between trilateration and proximity algorithm ............................. 184
Figure 5-15: Material tracking using trilateration and proximity algorithm ......................... 184
Figure 5-16: Box plot for each test beds results (Montaser and Moselhi, 2012 d) ............... 185
Figure 5-17: Summary of statistical analysis results for user localization ............................ 186
Figure 5-18: Summary of statistical analysis results for material localization...................... 187
Figure 5-19: CDF for estimated error (Montaser and Moselhi, 2013 a) ............................... 188
Figure 5-20: Test bed setup (Montaser et al., 2013).............................................................. 189
Figure 5-21: Errors observed under different durations in all test beds ................................ 191
Figure 5-22: Errors observed in the 9 different predefined locations.................................... 192
Figure 5-23: Errors observed in all test beds with different durations .................................. 192
Figure 5-24: Errors observed in all test beds due to metal proximity ................................... 193
Figure 5-25: Errors observed in all test beds due to the number of deployed tags................ 193
Figure 6-1: Chapter 6 overview ............................................................................................. 196
Figure 6-2: Developed system architecture ........................................................................... 200
Figure 6-3: ER diagram for the designed Microsoft Access central database....................... 202
Figure 6-4: BIM+ use-case diagram ...................................................................................... 205
Figure 6-5: Linking Microsoft Project file to BIM+ (sequence diagram) ............................. 206
Figure 6-6: Linking Microsoft Project file to BIM+ ............................................................. 208
Figure 6-7: Mapping 3D Revit objects to project activities (sequence diagram) .................. 209
Figure 6-8: Mapping 3D Revit objects to project activities (Part 1) ..................................... 210
Figure 6-9: Mapping 3D Revit objects to project activities (part 2) ..................................... 211
Figure 6-10: The added attribute “Group” to Revit object .................................................... 212
Figure 6-11: Selected objects to create Group....................................................................... 213
Figure 6-12: Assigning MS Project activity to Group name ................................................. 213
Figure 6-13: Show and hide activities options ...................................................................... 214
Figure 6-14: Show specific activity ....................................................................................... 216
Figure 6-15: Show and hide activities manually ................................................................... 217
Figure 6-16: Show and hide activities by date (4D BIM) ..................................................... 219
Figure 6-17: Pictures from the generated 4D Revit visualization ......................................... 220
Figure 6-18: Area plan vs. floor plan for the basement floor at Genomics building............. 221
Figure 6-19: Area plan for outdoor zones at Genomics building .......................................... 222
Figure 6-20: C# .Net source code for the process of linking dividing activity to plans ........ 223
Figure 6-21: Linking dividing activity to floor plan and area plan ....................................... 224
Figure 6-22: Getting user zone automatically options........................................................... 226
Figure 6-23: User indoor Location visualization and data acquisition.................................. 227
Figure 6-24: User outdoor location ....................................................................................... 229
Figure 6-25: User outdoor Location visualization and data acquisition................................ 230
Figure 6-26: Selecting user location manually (Part 1) ......................................................... 231
Figure 6-27: Selecting user location manually (Part 2) ......................................................... 232
Figure 6-28: Control+ login page .......................................................................................... 235
Figure 6-29: Control+ project menu – Weather Conditions .................................................. 236
Figure 6-30: Control+ resource menu ................................................................................... 237
Figure 6-31: Truck+ system architecture .............................................................................. 239
Figure 6-32: Truck+ sample of class diagram ....................................................................... 240
Figure 6-33: Truck+ web page layout ................................................................................... 241
Figure 6-34: Truck+ use-case diagram .................................................................................. 242
Figure 6-35: User enters project number ............................................................................... 243
Figure 6-36: GPS points layer ............................................................................................... 244
Figure 6-37: Earthmoving operation spatial boundaries sketches ......................................... 246
xii
Figure 6-38: Generated cycle times and its respective components...................................... 248
Figure 6-39: Generated tabular report for hauling trucks cycles ........................................... 248
Figure 6-40: Generated tabular report for earned value performance indices ....................... 249
Figure 6-41: Generated graphical report for Cost Performance Index (CPI) on daily basis . 249
Figure 6-42: Stochastic forecasting overview ....................................................................... 251
Figure 6-43: Simulating different fleet configuration ........................................................... 252
Figure 6-44: Control+ BIM/Walkthrough menu ................................................................... 254
Figure 6-45: Control+ CCTV/Time-Lapse menu .................................................................. 255
Figure 6-46: Control+ Central database menu ...................................................................... 257
Figure 7-1: Research contribution summary ......................................................................... 263
Figure A-1: The entity relationship diagram (ER) for the developed database..................... 277
Figure C-1: Vertical and horizontal attributes ....................................................................... 292
Figure C-2: Outdoor scanning positions................................................................................ 293
Figure C-3: First floor scanning positions ............................................................................. 294
Figure C-4: Different resolution 3D scanner images............................................................. 295
Figure C-5: Outdoor scanning and curtain walls glazing created noise ................................ 296
Figure D-1: Google Earth generated road profile.................................................................. 299
Figure E-1: Add to visible button (Sequence diagram) ......................................................... 302
Figure E-2: Add to hidden button (Sequence diagram)......................................................... 303
Figure E-3: Add all to hidden button (Sequence diagram).................................................... 304
Figure E-4: Add all to visible button (Sequence diagram) .................................................... 304

xiii
 LIST OF TABLES

Table 2-1: Piling progress measurement template .................................................................. 17

Table 2-2: CYCLONE modeling elements (Halpin 1973) ...................................................... 20
Table 2-3: Comparison between RFID and Barcode............................................................... 25
Table 2-4: Characteristics of radio frequencies ....................................................................... 27
Table 2-5: Indoor positioning technologies (Li and Becerik-Gerber, 2011) ........................... 41
Table 3-1: GPS data after processing (Montaser et al., 2012) ............................................... 106
Table 3-2: RFID data after processing (Montaser and Moselhi, 2012 b) .............................. 109
Table 3-3: Cycle time calculations (Montaser and Moselh, 2013 b)..................................... 115
Table 3-4: Comparative study ............................................................................................... 118
Table 3-5: Example of “Units Activities” objects percentage complete estimation ............. 120
Table 4-1: Sample of GPS unit captured data (Montaser et al., 2011) .................................. 142
Table 4-2: GPS productivity and the process duration (Montaser et al., 2012) .................... 147
Table 4-3: Sample of GPS corrected data ............................................................................. 157
Table 4-4: Sample of RFID captured data ............................................................................. 162
Table 5-1: The process of applying K-NN algorithm ........................................................... 166
Table 5-2: Error rate for the two deployment topologies (Razavi et al., 2012)..................... 168
Table 5-3: Error rates of tag deployment distance (closeness) on site .................................. 169
Table 5-4: Error for stationary reader locations vs. a moving reader .................................... 171
Table 5-5: Error rates for different zone configurations........................................................ 172
Table 5-6: Error rate for the scenarios of a slow moving vs. a fast-moving reader .............. 174
Table 5-7: Error of tag deployment distance (closeness) ...................................................... 175
Table 5-8: Characteristics of test beds (Montaser and Moselhi, 2013a) ............................... 180
Table 5-9: Different characteristics of the experiment (Montaser et al., 2013) .................... 190
Table B-1: Central database attributes .................................................................................. 278
Table C-1: Genomics tablet PC data acquisition schedule .................................................... 290
Table C-2: Genomics 3D scanning schedule......................................................................... 291
Table C-3: LADAR specification .......................................................................................... 292
Table C-4: Outdoor scanning attributes (September 15, 2010) ............................................. 293
Table C-5: First floor indoor scanning attributes (October 15, 2010) ................................... 294
Table C-6: Outdoor scanning with different resolution (November 3, 2010) ....................... 295
Table D-1: Values of variables affecting excavator production rates ................................... 298
Table D-2: Values of the variables affecting the truck production rate ................................ 299
Table D-3: Sample of hauling units’ travel times.................................................................. 300
Table D-4: Values of the durations governing truck production rate .................................... 300
Table D-5: Lower limits and upper limits of productivity .................................................... 301

xiv
 1 Chapter 1: INTRODUCTION

1.1 General

The ever-changing dynamics of construction requires that all parties in a project be

kept appraised of issues that can affect cost, schedule and performance of the work. Effective

project management requires tracking all aspects of construction project (Ghanem and

Abdelrazig, 2006). Tracking and control of construction projects depend primarily on the

accuracy, frequency and time required to collect actual data of construction activities onsite

that characterize the work progress (Moselhi and El-Omari, 2006). Traditionally, time and cost

control functions are done separately using the Work Breakdown Structure (WBS) and the Cost

Breakdown Structure (CBS). Integrating time and cost is essential to meaningful and reliable

progress reporting and effective control of construction projects. Earned Value Analysis (EVA)

developed under the Cost/Schedule Control System Criteria (C/SCSC) by the US Department

of Defense (DOD) is presently used in common practice as valid way to monitor project

performance (Fleming and Koppelman, 1995). The process of actual data acquisition is

essential in generating reliable progress reports using EVA (El-Omari and Moselhi, 2009).

Accurate and timely information on the status of project resources and activities are needed to

allow managers to effectively assess the overall project status and to make the best possible

decisions (Ergen and Akinci, 2007).

Effective tracking system onsite can ensure that sufficient quantities of materials are

available for construction needs, and waste at the end of the project is reduced. Location

identification in construction can be used to identify the location of materials and equipment

and to stay updated on a project’s status. Tracking material location and accessing its related

1
information are challenging tasks. Since, materials are commonly scattered in different areas

and may arrive on a daily basis. Material location can be changed several times before they are

installed in their final locations. Material management has been identified as one of the areas

of construction work with the greatest potential for improvement (Razavi and Haas, 2010).

Most material tracking problems involve: (1) time consuming identification and determination

of location, (2) error-prone data recording and documentation, and (3) delayed information

(Song et al, 2006).

Many project managers have continually relied on traditional methods of reporting

project progress. These methods place considerable load on project teams due to the large

amount of data that needs processing and the interdependency between data pertinent to various

project components. Lack of adequate visual representation often causes construction

managers to struggle with large, unwieldy amounts of data. This may cause costly errors that

result in delays and cost overruns. Such limitations of traditional tools on one side and the

advances in visualization-information technologies on the other have stimulated various

research and development efforts to advance new innovative construction visualization

techniques that depict construction sequence of operations (Waly and Thabet, 2002).

1.2 Research Motivation and Problem Statement

Current approaches for managing and capturing construction jobsite data are usually

labor-intensive and manual. Data collected using labor-intensive methods are usually stored in

a paper-based format, which could be lost or misrepresented. Manual data acquisition methods

may not be reliable or complete, as they are contingent upon motivation, judgment and skills

of site personnel capturing those data. Hence, it is problematic and costly to retrieve and

process those data into useful information. Failure to effectively track and access onsite related

2
information on demand can results in significant delays, rework and additional costs (Navon

and Goldschmidt, 2002). Onsite observations of construction sites have documented that field

supervisory personnel spend 30–50% of their work time recording and analyzing field data

(Cheok et al., 2000a). Most construction projects require approximately 2 to 3 million

individual parts from the beginning to the end. Construction site personnel need to carry a

sizeable amount of project documents such as drawings and specifications with them in order

to accomplish their assigned tasks; otherwise, they have to memorize extensive information

about the structures and facilities (Jang and Skibniewski, 2008).

Emerging technologies have been the driving force behind initiatives to automate the

process of data acquisition for tracking, control and progress reporting in the last two decades

(McCullouch and Gunn, 1993, Akinci et al 2002, El-Omari and Moselhi, 2009 and Razavi and

Haas, 2010). Automated data acquisition technologies have the potential of being effective in

monitoring and tracking of a project progress in real time (Jang and Skibniewski, 2008).

Location identification provides support for decision-making (e.g. timely tracking of the

project status and proactive safety monitoring). Developments in sensing technologies have

enabled the deployment of a variety of technologies for identification, location sensing, and

tracking (Moselhi and Alshibani, 2007).

Global Positioning Systems (GPS) has satisfactorily met the need for outdoor location

sensing (Navon and Goldschmidt, 2002). However, GPS is not reliable when it comes to

indoor areas due to the poor reception of satellite signals. In addition, GPS is still too expensive

for the automated tracking of individual material items (Song et al, 2006). Tablet PC

capabilities have increased significantly in recent years; integrating various automated data

acquisition technologies such as Radio Frequency Identification (RFID), barcode readers, GPS,

wireless communication (Wi-Fi and Bluetooth) and a camera for video clips and digital images.

3
In addition to its traditional advantages, which are durability, mobility, expressivity of display

and ease of use (Brilakis, 2007).

Traditionally, project teams use 2D drawings, Gantt-charts, and sketches to provide the

necessary details of measuring a project’s progress. However, these tools do not provide the

information required to generate 3D visualization of project progress. Building Information

Modeling (BIM) is gaining momentum in construction industry, especially by large-scale

building owners, who required its use on their constructed facilities (McCuen, 2008). BIM

allows a three dimension (3D) simulation of a building and its components. This simulation

goes beyond demonstrating how different building assemblies can be combined in the project.

It can assist in predicting problems, show the construction variables associated with different

building designs, and calculate material quantities (Eastman et al., 2008). The 3D BIM model

is a static model, built for representing a physical building’s geometry and its related

information. However, construction processes are dynamic and as such require a dynamic

representation (Hartmann et al, 2008).

The past two decades have seen advancement in construction automation, especially in

the area of earthmoving operations, structural steel erection, and concret structures (Elzarka

and Bell, 1997, Abeid and Arditi, 2002, Jaselskis and El-Misalami 2003, Yoon et al, 2006,

Golparvar-Fard et al. 2009 a, and Turkan et al., 2011). The limitations of developed models lie

in their inability to measure the progress of different types of construction activities from

inception to completion in near real-time. Existing models are incapable of dealing or even

documenting progress with many activities on a site. Those activities are not directly associated

with the resources location such as inspection, concrete curing, and interior finishing. Most of

these models lack the capability of objectively monitoring nonspatial tasks such as, inspection,

tests, calibration, and commissioning that affect activity progress. Digital Imaging and 3D

4
scanning were used in this area too. However, their use is subject to the motivation and

capabilities of the person gathering and processing captured data. Manipulation to captured

images or point clouds, such as interpretation, merging, stitching and object fitting is required.

These manipulations are performed manually with considerable effort. 3D as-built models

generated by digital imaging and 3D scans do not provide any location identification or

tracking of building components. For informed decisions and objective assessments of the

progress on a construction site, data from a number of sources must be integrated because not

all of the necessary information can be captured using a single technology (Ghanem and

Abdelrazig, 2006, Razavi, 2010, El-Omari and Moselhi, 2011 and Shahi, 2012). An overview

of the existing models indicates room for improvement.

1.3 Objectives

To address the challenges highlighted above, this research aims to study the models

used in tracking and control of construction projects utilizing automated data acquisition

technologies to improve the efficiency of progress reporting. This thesis address identified gaps

in development of efficient automated model for tracking and progress reporting of

construction operations, with visualization capabilities. This is to be achieved through the

following sub-objectives:

1. Study the progress made in the field of automated site data acquisition and indoor

localization.

2. Investigate the use of Ultra High Frequency (UHF) passive RFID for real time indoor

localization and material tracking under various conditions. Analyze changes in the

accuracy of position location as construction progresses and the indoor environment

5
 physically evolves. Scrutinize the possibility of utilizing RFID and/or spatial technologies

(GPS/GIS) for tracking and control of earthmoving operations.

3. Examine the potential use of different data acquisition technologies such as Light Detection

and Ranging (LADAR), Closed Circuit Television (CCTV) camera, RFID (mobile and

fixed), GPS (mobile and fixed), barcode, multimedia, tablet PC and wireless

communication. Evaluate the best way to integrate them for use in tracking and progress

reporting purposes.

4. Develop an automated model and its respective modules and algorithms to maximize the

benefits of 4D BIM visualization and indoor localization in progress reporting.

5. Develop a method to calculate earned value based on automated data acquisition

technologies and progress templates. Design centralized database to organize and manage

collected data.

6. Develop a computational platform (prototype software), which can act as an interface with

the user. This prototype shall make it possible to capture actual data on-site and store it in

centralized database, as well as process the collected data for progress reporting in near

real time.

1.4 Research Philosophy

The research philosophy behind the developments made in this thesis rests on (1)

integration as means for benefiting from the advantages and useful features of available

technologies, (2) automation while allowing for balanced human interaction, (3) lab

experiments and field studies as applicable, (4) understanding of the processes involved with

a focus on improving current practice via development of innovative and practical applications

and methods. Based on that philosophy and in an effort to understand the process of data

acquisition and project control, the following questions was deemed important:

6
 • What are inputs to the process?

• What are outputs of the process?

• What is the function of each input and output?

• How does this function add value to the process?

• What is the sequence of operations inside the process?

• What is the most suitable model that maps the relation between input and

output?

And to automate the process of data acquisition and project control, the following

questions were deemed important:

• What is the purpose of each step in the process?

• Can this step be eliminated?

• Does this step use the most suitable automated data acquisition technology?

• How can this step be simplified?

• Can more than one step be integrated?

The philosophy and related questions listed above were always helpful in guiding the

development made in this thesis.

1.5 Research Methodology

Figure (1-1) illustrates the methodology followed to achieve the objective of this

research. The methodology is summarized in five stages: analysis stage, design stage,

development stage, implementation, verification and validation stage, and conclusion stage.

The analysis stage began with a problem statement and definition of the objectives. Then, it

focuses on performing a comprehensive state of the art review on the following domains:

7
 • Project tracking and control

• Automated data acquisition technologies

• Indoor localization

• Data visualization

From the analysis stage, gaps and limitations in the current literature are identified. To

respond to the limitations and gaps identified in the analysis stage, a research vision was formed

in the design stage. From the research vision, a developed model was shaped to cover the big

picture of the control process without losing the details of its ingredients. The developed model

consists of automated data acquisition framework and data analysis and reporting framework.

Both frameworks were explained in detail in the development stage. While data collection was

being conducted in the implementation stage, the developed model framework and algorithms were

also gradually designed.

8
Figure 1-1: Research methodology

9
 Laboratory experiments and field studies were performed in order to assess the

feasibility of utilizing this model in the real world. The challenges associated with the practical

integration of the developed model within the framework were also addressed at this stage. In

addition, computer prototype software was developed to validate the developed model using

data captured from a construction jobsite. The finding of this research, lessons learned and its

limitations are presented in the conclusion stage.

1.6 Thesis Organization

The thesis consists of seven chapters and five appendices. Chapter 1 presents an

overview of the research that includes introduction, problem statement and motivation. This

chapter also provides a description of the main and sub-objectives of the research. Finally, it

presents a summary of the methodology adopted to achieve the thesis’s objective. The literature

review is presented in Chapter 2.The review covers the fields of project tracking and control,

automated data acquisition technologies, indoor localization and data visualization. The listed

topics are viewed with a focus on how they reflect upon the developed model. Summary of the

limitations and gaps in existing methods are presented at the end of this chapter.

Chapter 3 begins by presenting the research vision; outlining the main components of

the developed model. It also presents the data acquisition framework and its modules. This

framework is designed for capturing data from construction jobsites. It consists of two

modules; independent data acquisition module and a module established through tablet PC data

acquisition, where tablet PC works as integration media for different data gathering

technologies. Chapter 3 then focuses on the analysis and reporting framework, where all the

captured data is stored and processed for progress reporting purposes. This framework consists

10
of four modules; central database, performance measurement, earned value analysis and

reporting.

Chapter 4 focuses on the data collection stage, which is necessary to implement the

developed model. It presents a case study that involves capturing data from a construction

jobsite. Wide ranges of technologies were deployed, in this case, such as LADAR, RFID, GPS,

CCTV camera and tablet PC. The data captured in this chapter is used later to verify and

validate the developed model. Chapter 5 presents in-depth studies for deployment protocols of

Ultra High Frequency (UHF) passive RFID for indoor localization. Experimental work was

conducted on construction jobsite and in a lab environment. K-Nearest Neighbor (K-NN)

algorithm was used as coarse-grained zone detection. In addition, Trilateration and proximity

algorithms were used too for fine-grained location identification. This chapter is validation for

the two-step algorithm developed in chapter 3 for identification user location and material

tracking. The developed software is described in Chapter 6. It covers the designed database,

classes, use-cases, sequence diagram and snapshots from its Graphical User Interface (GUI).

Chapter 7 highlights contributions and limitations of the development s made in the thesis

along with suggested future research work.

11
 2 Chapter 2: LITERATURE REVIEW

2.1 General

Tracking and control of construction projects depends primarily on the nature,

accuracy, frequency and time required to obtain on-site current data of the construction

operations that characterize work progress (Moselhi and El-Omari, 2006). Integrating time and

cost is essential to effectively monitor project management functions. Earned Value (EV)

analysis is conducted by using actual data gathered from construction sites, comparing that

data to the planned values and generating progress reports. Accordingly, the process of actual

data collection is crucial in generating reliable progress reports (Caldas et al., 2006). Various

automated site data acquisition technologies have been utilized to automate the process of data

collection required to perform tracking, control and progress reporting in construction projects.

These technologies include barcode technology (Tserng and Dzeng, 2005), RFID (Song et al

2006), pen-based computers (Elzarka and Bell, 1997), LADAR (Bosche and Haas, 2008),

Photogrammetry (Styliadis 2007), and multimedia technology (Abeid and Arditi, 2002).

Researchers have also developed a web-based interface to facilitate the information flow

between project participants during the control process (Moselhi et al. 2004).

Location identification provides support for the decision-making process and is crucial

for the timely tracking of project status. Recent advances in automated data acquisition

technologies have enabled the use of sets of sensors for identification, location sensing, and

tracking the movements of objects. Location identification in construction projects is a recently

evolving area in which the location of resources can be used to acquire precise information

about project status (Ergen et al., 2007b, Jang and Skibniewsk, 2007, Moselhi and Alshibani,

2008 and Razavi and Haas, 2010). Lack of adequate visual representation causes construction

12
managers to struggle to manipulate tremendous amounts of data, which may lead to costly

errors and/or project failure. Utilizing visual representations to explain the complex structure

of data and eliminate confusing factors has definite advantages (Eddy, 2007). This Chapter

provides a literature review on the implementation of EV analysis as a tool to track and control

construction projects. It also presents an overview on automated data acquisition technologies

and their implementation in the construction industry. A state of the art review for indoor

location sensing and RFID localization in construction is included. A section for data

visualization in the construction industry was illustrated focusing on GIS and BIM, as shown

in Figure (2-1).

Figure 2-1: Chapter 2 overview

2.2 Project Tracking and Control

Controlling construction projects is a challenging task without a reliable and effective

control system. Keeping a project on its planned duration and budget is the main objective of

any control system. Tracking and control include three main aspects: (1) collecting data from

the construction site; (2) monitoring progress; and (3) comparison of actual performance to the

planned baseline and determination of any variations in performance. The project manager can

13
then make decisions based on the results of the analysis. Performance evaluation methods are

used to calculate the variance of time and cost from its control baseline and to assess the value

of deviation (Moselhi, 1993). The baseline is subject to change, and the control system is

required to be flexible to accommodate any variations in the planning method.

Over the years, a number of methods have been developed to facilitate overall project

control. Traditionally, time is tracked and controlled through schedules that are linked to

measures of physical progress, while cost is controlled using the percentage complete S-curve

method (Li, 2004). The earned value method integrates both time and cost. It has proven to be

the most reliable tool for the tracking and control of construction projects. It requires periodic

monitoring of actual expenditures and of the scope of accomplishments, and allows for the

calculation of cost and schedule variances, along with the performance indices. Performing EV

on a regular basis during the course of a project enables early detection of variations from the

baseline and this give the management team the chance to organize an appropriate response

(Jung and Kang, 2007).

Materials control systems are responsible for ensuring the availability of resources

needed for a project; also, they are utilized to handle the waste during a project. Due to their

immense importance, Nasir (2009) identified the shortcomings of materials control systems as

the most common and substantial factor compromising construction productivity, with an

estimated productivity reduction of nearly 40%. In summary, effective materials tracking and

control can help increase productivity, avoid delays, reduce the time needed for materials

management, and reduce the cost of materials due to decrease of waste. Tracking the exact

location of materials had been considered too costly, but recent improvements in automated

data acquisition technologies have made it technically and economically feasible (Razavi,

2010).

14
2.2.1 Integrated Tracking and Control

EVM works by producing three curves depicting the time versus cost relation through

the project’s life. These curves are used to display and evaluate the project performance. These

curves typically have S shape, and they namely are: Budget Cost for Work Scheduled (BCWS),

Actual Cost for Work Performed (ACWP), and Budget Cost for Work Performed (BCWP), as

shown in Figure (2-2). The curve representing the base line planned project is the BCWS curve.

The actual expenditure to date is represented through the ACWP curve. The third and final

curve is the BCWP; it represents the budgeted cost of the work that is performed to date, which

is the actual value earned for the project. The BCWP is the actual addition to the traditional

control system, which only included the BCWS and ACWP (Li, 2004).

The actual progress is measured by BCWP, which is associated with the percentage of

work completed. BCWP might differ from the BCWS because the actual quantity installed can

be different from the budgeted quantity and/or the actual start date can differ from the

scheduled start and/or the unit cost per day can be varied from those budgeted. According to

this method, cost and schedule variances are routinely computed during the construction stage

for performance evaluation. The cost variance (CV) is the difference between the BCWP and

the ACWP, Equation (2.1). Favorable cost conditions are observed when the Cost Performance

Index (CPI) is greater than one as shown in Equation (2.2), which indicates that the BCWP

(Earned) is more than the ACWP (Spent), and vice versa. The schedule variance (SV) reflects

the difference between the BCWP and the BCWS, Equation (2.3). Favorable schedule

situations occur when the Schedule Performance Index (SPI) is greater than one as shown in

Equation (2.4), indicating that what is being accomplished (BCWP) is more than what is

scheduled (BCWS) (Li, 2004).

CV = BCWP – ACWP (2.1)

15
 CPI = (2.2)

SV = BCWP – BCWS (2.3)

(SPI) = (2.4)

Figure 2-2: Earned value method (Li, 2004)

2.2.2 Measuring Project Performance

Handa and Barcia (1986) pointed out that several methods are available to assess actual

progress. The units completed method is used for activities that involve repeated production of

easily measured work (e.g. earth moving operation). Incremental milestone is a method used

for activities that are sequential with clearly defined milestones. The Start-Finish method is

used for activities that do not have interim milestones and/or are difficult to quantify in terms

16
of time and cost. Supervisor’s opinion is used for activities where detailed analysis is not

necessary. The Cost ratio method is used for long-term activities with allocated bulk cost/time.

Finally, the weighted units method is used for long-term activities that involve multiple

subtasks with different units of measurement (CII, 1987).

Progress measurement template, developed by Moselhi (1993), can be used effectively

to evaluate performance and to estimate the Percentage Complete (PC) for an activity. For

activities with relatively long durations, interim control points could be introduced. The

template divides the activity into control points based on its time and cost characteristics, as

shown in Table (2-1). However, the user can overwrite these templates by introducing

specifically designed templates that best suit the project at hand. This method has been

employed in this research.

Table 2-1: Piling progress measurement template

Work Content Cumulative % Earned

Rig in position 40

Drive and inspect 80

Trim and finish 95

Hand over 100

2.2.3 Traditional Forecasting Methods

Forecasting project time, cost and productivity is a key function of construction project

tracking and control and it is necessary for the evaluation of project status. Forecasting

accurately is a difficult task. This difficulty is due to the impact of various factors on project

such as weather conditions, change orders, varying project risks, etc. Different techniques have

been utilized to build forecasting models, all working under different assumptions and having

different limitations, as shown from Equations from (2.5) till (2.9) (Hassanien, 2002, Moselhi

17
et al, 2004, and Li et al, 2006). Most common forecasting techniques are based on linear trend

analysis. Fleming and Koppelman (1994) assumed that the final cost is affected by both the

cost efficiency factor and the schedule efficiency factor.

EAC = (2.5)

D= (2.6)

Where

BAC = Budget at Completion,

D = Estimated duration at completion,

Db = Planned duration

EAC = (2.7)

D= (2.8)

– ( )
EAC = ACWP (to-date) + (2.9)

Hassanein and Moselhi (2003) proposed a model to forecast linear infrastructure

projects time and cost at completion. Their model enables removing specific time periods

during which exceptional conditions are known to have prevailed and are not likely to be

repeated beyond the reporting date. Li et al. (2006) proposed a “Self-Learning Adjustment”

method. In this method, the difference between predicted results and actual occurred values are

18
used to improve the accuracy of the forecasted results, as shown in Figure (2-3). Alshibani

(2008) proposed calculating the BCWP as the ratio of the installed to the budgeted quantities

at a given point of time. His total project cost and duration can be forecasted using one of two

assumptions. The first assumption is that the current performance at the report date will prevail

until the end of the project, and the second one is that the remaining portion of the project will

be carried out as originally planned.

Figure 2-3: self-learning model (Li et al., 2006)

2.2.4 Simulation as Forecasting Tool

Shannon (1998) defined simulations as “the process of designing a model of a real

system and conducting experiments on this model for the purpose of understanding the

behavior of the system and/or evaluating various strategies for the operation of the system”.

Simulation of construction operations allows construction planners and estimators to predict

productivity and to evaluate construction operations before starting site work (Alzraiee et al.,

2012). Literature reveals considerable work on computer simulation for modeling repetitive

19
cyclic operations (Zayed, and Halpin, 2001, Marzouk and Moselhi 2004, and AbouRizk, 2010).

Discrete Event Simulation (DES) method is the most famous type of simulation. It serves as a

powerful tool in analyzing and breaking down complex problems. Consequently, it grabbed

the biggest share of researchers’ attention, focusing particularly on construction operations.

DES models a system as a network of processes that are sometimes preceded by

queues, where state changes occur at discrete points of time. In DES model, each unit flows

through the system and reserves resources to complete a job. When, the job is accomplished,

the unit releases those resources. Queues temporarily hold units in case resources are not

readily available. These actions are called events (Hajjar and AbouRizk, 1999). Tremendous

efforts have been exerted by many researchers in the construction field to model and simulate

various construction operations. Halpin (1973) developed a powerful modeling element called

CYCLONE, which simplified the simulation and modeling process for users with limited

simulation background (Table 2-2). CYCLONE was later used as a base for other simulation

systems. In 1996, Martinez produced a more evolved simulation tool (STROBOSCOPE). This

tool is capable of handling uncertainty not only in time, but also for different resource

quantities. Once again, in 1999, Martinez and Ioannou developed the system (EZSTROBE).

EZSTROBE’s main objective was to reduce the complexity associated with their previous

model (STROBOSCOPE).

Table 2-2: CYCLONE modeling elements (Halpin 1973)

Element
COMBI NORMAL QUEUE FUNCTION COUNTER ARROW
Name

Modeling
Element
Symbol

20
 Traditionally, simulation works through utilizing historical data in the form of

statistical distributions to represent different processes. The highly dynamic environment of

construction projects adds a sense of uniqueness to each project, which questions the validity

of using historical data of other projects to represent a project at hand. Hence, there is a need

for a method that puts together the capabilities of traditional modeling of construction

operations and real time field data acquisition. Many researchers explored real time simulation,

such as Guo et al. (2008) and Song et al. (2008). So far, most researchers focused their work

on earthmoving construction operations (Montaser et al., 2011, Montaser et al., 2012, Song

and Eldin, 2012, Akhavian and Behzadan, 2012 and Vahdatikhaki et al., 2013).

2.3 Automated Data Acquisition Technologies

The accurate and timely collection of information about the status of projects’

resources and activities is necessary to support managers in the effective evaluation of project

status and in making timely decisions during the execution of on-site work. Effective tracking

and control systems require the collection of accurate data from a construction site in a timely

manner through a cost efficient methodology. An ineffective tracking and control system can

result in costly delays, the need for rework and other additional direct costs (Jang and

Skibniewski, 2008). Tracking and control process is a challenging and demanding task in view

of the dynamic nature of on-site operations including material delivery and utilization. This

process requires the collection of a very large amount of data from construction sites in order

to report on a project’s overall status so that corrective actions can be taken if needed (Shahi et

al., 2013). Although different techniques have been used to automate data collection, the cost

of such techniques is still quite high. Current approaches for capturing, storing and managing

construction jobsite data are mostly manual and labor-intensive methods. Data collected using

manual methods is not reliable or complete, as it is based on the motivation, judgment and

21
skills of the site personnel. Furthermore, data collected via those methods is typically

transferred and stored in a paper-based format. The data will be difficult to search and retrieve,

which makes data processing into useful information expensive and unreliable. In other words,

potentially useful data may not be accessible to the project’s parties when needed, and/or data

becomes obsolete (Moselhi and El-Omari, 2006).

Clearly, cost-effective automated data collection is needed. It can increase

productivity, reduce cost and increase profit to contractors. The efficiency of large-scale

construction operations’ tracking and control systems is primarily due to the nature, accuracy,

frequency and time required for collecting the onsite data (Moselhi and El-Omari, 2006).

Current technological advancements in data acquisition systems have made the consistent

management of construction projects feasible by utilizing more accurate and reliable data.

Recently, tablet PC capabilities have been enhanced significantly, in addition to their

traditional advantages of durability mobility, display expressivity and ease of use (Elzarka and

Bell, 1997). Tablets already integrate various automated data acquisition technologies such as

barcode readers, RFID readers, wireless communication (Wi-Fi and Bluetooth), GPS and

cameras for video clips and digital images. Recent technological advancements in data

acquisition systems have made the consistent management of construction projects and their

information more feasible. Bar code and Radio Frequency Identification (RFID) are used to

automatically identify, track and facilitate generating status documents. GPS, on the other

hand, has proven to be an appropriate technology to locate assets’ positions for outdoor

operations, and with accuracy within centimeters (Navon and Goldschmidt, 2002). Using

cameras for data collection allows users to capture a broader perspective of construction

activities and operations. Camera-captured images can reduce the time required for inspection

22
by allowing this task to be done remotely (Abeid and Arditi, 2002 and Brilakis, 2007). The

captured data can be transferred to a project management team via an internet connection.

2.3.1 Bar Code

Bar code is an automatic identification solution that streamlines identification and data

acquisition. A bar code system consists of three components: tags, a reader and a printer. Bar

code tags are classified as either one-dimensional or two-dimensional, as shown in Figure (2-

4). A one-dimensional bar code is typically used as an identifier code on an object, while two-

dimensional bar code can contain a whole file of information about an object. The second

component is the laser scanner (reader), which is capable of reading data coded on a bar code

label. Different types of scanners are available: hand held scanners, pen scanners, wall or table

mounted scanners, fixed-position scanners and Personal Digital Assistant (PDA) scanners with

a built-in barcode reader. The last component is the printer to print bar code tags (Moselhi and

Shehab-Eldeen, 2003). Cheng and Chen (2002) integrated bar code and GIS for monitoring

construction progress. The scheduled components of the erection progress could be tracked

through systematic monitoring of the construction process. Tserng and Dzeng (2005) have

integrated bar code technology and PDA as a way to facilitate the process of data collection

and information sharing between project participants.

Figure 2-4: 1D and 2D bar code (McCullouch and Lueprasert, 1994)

23
2.3.2 Radio Frequency Identification (RFID)

RFID works in a way that is similar to bar code technology. Whereas in RFID, data

can be stored in tags and retrieved with readers. But these readers can also communicate with

the tags using radio frequency waves. Table (2-3) is a comparison between RFID and bar code

technologies. With its lower cost and increased capabilities, RFID continued to gain industry

acceptance. The basic premise behind RFID systems is marking items with tags. These tags

contain transponders that emit messages readable by RFID readers, see Figure (2-5). RFID tags

store an identification number or code and other object-related information. RFID tags contain

writable memory, which store information to be readable by various RFID readers. Tags can

be distinguished as passive or active, according to their power source (Jaselskis and El-

Misalami 2003).

Passive tags need to be activated by the electromagnetic energy the reader emits and

depend on that power to operate. Therefore, they have shorter read ranges and smaller data

storage capacities. Active tags rely on internal batteries for their power supply, which

significantly enhances their read ranges and enables additional onboard memory and local

sensing. However, using a local power source limits their lifetime to 5–10 years and increases

the cost. RFID tags have various specifications, such as their power source, frequency, read

range, read rate, data storage capacity, size, operational life, and cost (Akinci et al, 2002).

24
 Table 2-3: Comparison between RFID and Barcode

RFID Barcode

Fast (Could reach 100 object per

Read rate Slow ( 1 object per scan)
scan)
High (Could exceed 10m for
Read Range Small (centimeters)
UHF passive Tags)

Labor-intensive scanning No Yes

Storage capacity Large Small

Survive in harsh conditions Yes No

Long-term data storage Yes No

Require line of sight No Yes

Ease of positioning for sensing Easy Difficult

Cost of a tag Cheap Cheapest

Ease of obtaining information Easy Easy

Knowledge of items' exact

Easy Difficult
position

Write Possible Impossible

Yes (Active Tags)

Batteries (Power Source) No
No (Passive Tags)

Many according to frequency

Varity of Types (LF, HF, UHF and Microwave 1D and 2D only
Frequency)

Data security High None

Ability to be integrated with

Yes (GPS, infrared and
other automated data Yes
ultrasound)
acquisition technologies

25
 RF Waves
RFID Tag Mobile RFID Reader

Figure 2-5: RFID system components

A reader retrieves information about the ID number from a database and acts upon it

accordingly (Ergen et al, 2007a). An RFID reader can simultaneously communicate with

different RFID tags. Thanks to this capability, RFID readers can capture the contents of an

entire shipment as it is loaded into a warehouse (Jaselskis et al., 1995 and Ergen et al, 2007b).

RFID frequency is key element, one that determines the characteristics of the signals traveling

between a reader and tags. The available frequencies include low frequency (LF), high

frequency (HF), ultra-high frequency (UHF) and microwave. Presently, UHF is the most

widely used, because it offers simple and affordable solutions. The characteristics of each

frequency are summarized in Table (2-4) (Tzeng et al., 2008, and Li and Becerik-Gerber,

2011). Jaselskis et al. (1995) presented an RFID system designed to track material delivery

vehicles, material-handling equipment and tracking concrete delivery vehicles. They suggested

also that RFID card could be issued to construction workers on-site to track their work. Akinci

et al. (2002) proposed the use of RFID technology in tracking precast concrete pieces and

managing information associated with them throughout a supply chain.

26
 Table 2-4: Characteristics of radio frequencies

Ultra High Ultra High

Low High
Frequency Frequency (UHF Microwave
Frequency Frequency
(UHF Active and Semi Frequency
(LF) (HF)
Passive) Passive)
Frequency
125-135 kHz 13.56 MHz 865-956 MHz 433-864 MHz 2.45-5.8 GHz
Range

< 0.5 m < 1.0 m <5 m 100 m (Semi > 100 m

Read Range
(Passive) (Passive) (Passive) Passive and Active) (Active)

Metal/fluid
Very low Low High High Very High
impact

Data transfer
Low Medium High High High
rate
Read Only / Read Only /
Read Only / Read
Type Read Only Read Only Read and Read and
and write
write write

Memory Small Small Small/Large Small/Large Small/Large

RFID gates are used at specific locations to identify the arrival dates of materials onsite.

The major task of a recognition point is to report the identification of items to the system for

further processing (Goller and Brandner, 2011). Utilizing RFID system at laydown yards and

portal gates along a project’s supply chain paves the way for time savings in material

identification, increasing accuracy and shortening time for establishing information on material

availability for craftwork planning and reducing searching time for misplaced material (Song

et al, 2006). Additionally, RFID gate can be used to register personal protective equipment

(PPE) items and generate a warning for missing items (Helmus, 2007). Song et al (2006)

developed a gate system using fixed RFID readers to identify pipe spools delivery and obtain

other information about these spools, such as purchase order number and arrival time at the

storage yard, as shown in Figure (2-6). Lee et al (2008) developed an RFID gate sensor system,

27
which uses passive RFID tags and wireless network technology for an intelligent logistics

management system at construction sites.

Figure 2-6: Portal structure with 4 fixed RFID readers (Song et al, 2006)

2.3.3 Global positioning System (GPS)

GPS is at present fully practical Global Navigation Satellite System (GNSS). GPS

technology has attracted researchers looking for effective ways to track the outdoor location of

construction labor and equipment on jobsites (Navon and Goldschmidt, 2002 and Navon et al,

2004). The construction industry has already embraced the use of GPS in control of site

surveying and earthmoving operations. GPS can be used to track the location of workers and

equipment over a wide range of geographic and geometric scales. The use of GPS for

earthmoving operations has reportedly led to an increase of 22 % in productivity and 13 % in

cost savings for short haul distance projects. However, for long haul distance, the advantages

were less, a productivity increase of 5.57% and cost savings of 4.79% (Han et al 2006). Caldas

et al. (2006) developed a system integrating GPS and GIS, in which GPS was used to record

the position of pipe spools on a construction project. Moselhi and Alshibani (2007 & 2008)

presented a new model for the tracking and control of earth moving operations. Their model

28
uses spatial technologies, including GPS and GIS, to facilitate automated data acquisition.

Their proposed system is designed for highway construction projects.

Song et al (2006) examined the feasibility of applying RFID technology to automate

the tracking of materials on construction projects. Using RFID combined with GPS, it was

possible to determine the two dimensional location of materials. The objective was not only to

track the movement of the object with the use of RFID but to find the exact location of that

object. GPS would first identify the precise location of the material and then there would be no

need to install many sensors to track the material's movement. The supervisor walks around

the site, and the position of the reader at any time is known. Despite GPS have proven to be a

significant addition to existing sensing and tracking tools, GPS usage is limited to outdoor

projects. This is attributed to the substantial degradation in GPS signal in indoor environment.

2.3.4 Multimedia

Information can be collected in the form of digital photographs, video clips, and/or

sound and stored on a computer. During the reporting period, multimedia information is

extremely useful for its visualization capabilities (Liu 1995). Cameras are used in construction

projects to determine, assess, document, and correct work tasks. Cameras can provide project

managers with a broader perception of construction activities and operations. It allows

managers to use them for real time visualization wherever internet connectivity is available,

especially from a distance and at a standardized viewpoint. Pictures and videos are valuable

information tools to highlight problem areas (Abeid and Arditi, 2002). Camera users can log

into a web user interface and see if building sections or components have been completed or if

rework is needed, allowing for the quick identification of issues or problems while construction

is still underway. Images can reduce the time needed for inspection by allowing this task to be

done remotely (Brilakis, 2007). Time-lapse recording allows long operations to be viewed in a

29
short time. Researchers applying multimedia technology in the construction industry have

focused on support areas such as as-built information, claim management, and inspection

tested. Liu et al (1995) developed a system called M-LOG to store and retrieve as-built

information. Construction logs report document project information regarding weather, work

accomplishment, on-site resources including labor, equipment, and materials.

Abeid et al (2003) introduced an automated real time monitoring system (PHOTO-

NET II) for construction projects. It accepts digital images taken from different cameras, stores

them and links them to a database. The pictures taken from cameras are placed on a website

from which a computer can capture and store them in a database. Site visits can thus be both

reduced and optimized. Another significant documentation advantage stems from the

standardization of site pictures. Using cameras for tracking workforce, materials, and

equipment across a site can show a number of key imbalances in a project. Time-wasting, task

completion times, and inefficiencies can be recognized and adjusted for better optimization of

project resources. These forms of utilization are in addition to the more common usage of

reducing theft and vandalism of site equipment and material (Bohn and Teizer, 2009).

2.3.5 Tablet Computers

Tablet computers (Tablet PC) are tools that support automated data collection. Tablet

computers are becoming less costly and more powerful and suit the construction industry very.

The burden of moving around in a construction site with many drawings and papers can be

eliminated with the application of tablet PC. The user can access drawings, specifications, and

bills of quantities on the spot without losing any time searching for a document. The portability

of tablet PC suits construction fields. Data entry is performed by writing on the screen with a

pen, or by selecting the desired option from pop-up lists. Thus, there is no need to fill out forms

or to remember information and then enter it later into a computer. Data entry is done at the

30
place where it is collected (McCullouch and Gunn 1993). In a survey, Alemany (1999) showed

that supervisors who used computers at work saved time on paperwork and spent more time

on supervision.

This could have multiple positive impacts on productivity and quality. Providing

construction workers with tablet PCs that can help them locate tools, equipment, and materials,

send requests for information and access relevant schedule information. Eventually tablet PCs

allow workers to spend more time on direct work and less idle time waiting for answers or

looking for tools and materials (Elzarka and Bell, 1997). McCullouch and Gunn (1993) worked

on a system to automate the process of data acquisition from construction sites using tablet PC.

The system was designed to obtain information pertinent to material consumption and labor

working hours. Thabet (1997) developed an automated model for data acquisition; it integrated

a CAD model where project drawings are stored, with a pen-based computer. This data

acquisition model included object-oriented database designed to relate electronic forms, used

with pen-based computer, with the drawing of each activity.

2.3.6 3D Imaging and Photogrammetry

3D imaging refers to systems that are used to measure or to capture existing conditions.

In the past, this type of measurement was conducted with single point measurement systems

such as total stations, or camera-based systems that relied on photogrammetric techniques.

Currently, these instruments are generally categorized as laser scanners and optical range

cameras. 3D images are used for a range of applications from the creation of accurate as-built

of process plans undergoing modification to the rapid surveying of highways and mines. Laser

Detection and Ranging (LADAR) is a 3D laser scanner that is mainly used for spatial

measurement. Other applications include surveying, earthmoving operations, monitoring the

31
progress of concrete casting, highway alignment, paving operations and construction quality

control (Lytle, 2011).

LADAR scanning yields the collection of 3D points or “point clouds”. Point clouds are

given in x, y, z, coordinates. LADAR technology works by sending illumination pulses to an

object where the distance between that object and the LADAR is called the “Range”. Every

pulse makes a point in a point cloud that represents a range image. Bosche and Haas (2008)

and Bosche et al (2008 & 2009) developed a novel method of inferring the presence of model

objects in range images. Their approach focused on the comparison of a 3D image of a

construction scene with a simulated scan of a 4D building model using similar scan parameters.

The approach was successfully demonstrated in a steel construction project, as shown in Figure

(2-7). Turkan et al., (2011) and Turkan (2012) presented a system integrating 3D object

recognition technology and schedule information into a combined 4D object-based

construction progress tracking system. During the construction of a reinforced concrete

structure, they performed and extensive field study to investigate the performance of the

system.

32
 Figure 2-7: 3D scan and 3D model object recognition (Bosche et al, 2009)

Photogrammetry is the extraction of the geometrical properties of an object from

photographic images (Styliadis 2007). The value of photo images is that they can obtain

33
information about texture and color, which is an advantage to photogrammetry over laser

scanning (Zhu et al., 2010). Golparvar-Fard et al. (2009 a & b) developed a methodology for

construction progress monitoring that leverages the large number of photographs that are

already taken on construction sites for production documentation. By analyzing imagery taken

daily, a time-based visualization can be generated which compares the 4D as-built data with

the 4D as-planned data within a common user interface. Photogrammetry and 3D scanning was

integrated to track changes for work accomplished. Integrating 3D imaging and

photogrammetry mitigates the limitations associated with each of them individually, such as

the number of scans required and the time needed for each scan to produce satisfactory results

during the 3D modeling process (Moselhi and El-Omari, 2008 and El-Omari, 2008).

2.3.7 Other Types of Sensors

There is a variety of automated site data acquisition technologies that could be used in

the process of tracking and control. In this section, two types are highlighted due their

significance in the context of this research. They are control sensor and switch limit. Control

sensors are a type of sensors that detects the occurrence of events and reports those events. It

could be connected to an equipment control system via microcontroller. Accordingly, when

the equipment operator makes a certain action, the control sensor designated events are

triggered, and it is registered by the microcontroller memory. The other type of sensor is the

switch limit, which is shown in Figure (2-8). Mechanical switch limits are switches that are

mechanically activated. Therefore, they have some sort of arm, lever, knob and plunger, which

is automatically activated by making contact with another object. As the object makes contact

with the actuator of the switch, it eventually moves the actuator to its “limit” where the contacts

change state. Example of switch limit is the one connected to the fridge door, when the fridge

34
door is opened, it activates and switches on the light and later shuts off the light when the door

is closed (ABB, 2013).

Figure 2-8: Switch limit main components (ABB, 2013)

2.4 Localization

The aim of object localization is to determine the location information of the object of

interest through spatial relationship observations between the object and known references. To

collect and estimate spatial object information, localization system needs to be implemented,

which can then periodically or continuously observe the object (Tzeng et al., 2008). Global

Positioning System (GPS) has certainly met the need for outdoor location sensing and can now

provide positions with centimeters accuracy under non-stationary situations (Navon and

Goldschmidt, 2002). However, GPS has some limitations. First, GPS is not a valid solution for

indoor or urban outdoor areas, due to the poor reception of satellite signals. In addition, GPS

is still expensive to be used extensively for automatic tracking of individual items. Tagging

hundreds of materials with GPS for tracking would be excessively expensive along with other

costly means for identification and monitoring (Song et al, 2006).

35
 Localization could be categorized into two types; these are coarse-grained localization

and fine-grained localization. Coarse-grained localization is also commonly referred to as

range-free or connectivity-based localization. In contrast with fine-grained localization, coarse-

grained localization does not utilize any measurement technique, such as the techniques

described later on. Coarse-grained localization uses sensors, called anchors, having known

locations. The known locations of these anchor sensors are used to estimate the locations of

other sensors. This estimation is based on connectivity information, through knowing which of

the anchor sensors is within the communication range of this sensor. The connectivity method

does not attempt to measure the distance from an object to the reference points. However, it

determines whether the object is near one or more anchor sensors in known locations. This

category of localization has a minimum requirement for information and consequently does

not yield high accuracy.

The K-Nearest Neighbor (K-NN) algorithm is recognized as one of the most widely

used methods for solving this type of problems. K-NN is used for classifying objects locations

based on the closest anchor sensors. In case the sensors communication connectivity exists

between the reader and any of the anchors sensors, the reader reads the tag. Such a read event

contributes one neighborhood to the problem of the zone identification. Combining this

neighborhood restricts the possible zone reader. Therefore, the K nearest neighbors is the last

K reads that occurred prior to the most recent read (Duda and Hart 1973). On the other hand,

fine-grained localization utilizes manipulation on signal measurements received from the

localization sensor and hence it is more accurate. The following section describes in detail the

different signal measurements used and the algorithms used for fine-grained localization.

36
2.4.1 Signal Measurements

Signal measurements used for localization are Received Signal Strength (RSS), Angle

of Arrival (AOA) and Time of Arrival (TOA). TOA is a localization method that can measure

signal traveling time between the signal source and signal receiver. The system has to predefine

the velocity of the signal. If there are several signal receivers, which measure the same signal

at various points, the localization system can use a variation of TOA and/or the Time

Difference of Arrival (TDOA). TDOA measurement takes into consideration different signal

travel times from the single source to different receiver locations. GPS, Ultrasound signals, and

Ultra-Wide Band (UWB) are localization technologies that use time measurement (Hightower

and Borriello, 2001). AOA is a signal measurement method that considers the direction of

signal propagation. For precise measurements of signal arrival angles, the receiver contains

different signal-receiving arrays. However, if the signal can be reflected by the surroundings,

it becomes less accurate than TOA. Thus, AOA is more suitable in outdoor environments and

open spaces than indoor environments or congested areas (Patwari et al, 2005).

RSSI is one of the most common measurement signal types. Signal strength has a close

relationship to the distance between the sender and receiver. A certain signal, such as Radio

Frequency (RF), ultrasound or vibration decreases as it travels through media, and the

localization system can then determine spatial information using the degree of signal

attenuation. In order to estimate the distance between fixed and mobile devices, RSSI

measurements are made so that the distance can be estimated using a propagation (path loss)

model. However, RSSI implies a certain degree of error due to the multipath effect. Signal

strength measurement localization systems have two main advantages, cost effectiveness and

straightforward implementation (Choi, 2011). RSSI signal measurements technique was used

in this research.

37
2.4.2 Localization Algorithms

Signal measurement (TOA, AOA, or RSSI) localization techniques estimate the

location of a signal source. Three main algorithms have been developed to locate a target:

Triangulation, Proximity and Scene Analysis. Triangulation is an algorithm for determining

the location of an object, based on geometric properties and mathematical formulation. There

are two different techniques of triangulation: Trilateration and Angulation, as depicted in

Figure (2-9).

(a) (b)
Figure 2-9: (a) Trilateration (b) Angulation localization algorithm

Trilateration uses distances and angulation uses angles to determine the position of an

object in a coordinate system. Trilateration determines the position of an object by measuring

its distance from different reference positions. Trilateration systems measure distances with

different signal measurements methods: TOA, TDOA, or RSSI. A localization system needs

at least three distances from different receivers, the system can then draw three circles; each

circle’s radius is equal to a measured distance, and the center for each circle is the known signal

receiver’s position. The intersection of the three circles represents the signal source’s location,

as shown in Figure (2-9 a). Angulation is a signal direction (or angle) based-localization

38
method to determine the position of an object. In a 2-dimensional Cartesian coordinate system,

two angles and the distance between the reference points are used, as shown in Figure (2-9 b)

(Gonçalo and Helena 2009). Trilateration algorithm was applied in this research.

Proximity algorithm requires the measurement of the nearness of a number of

neighboring points, which are close to the target and whose locations are fixed and known.

When a tag or a reader is attached to a target, it continuously communicates with the readers

or tags deployed in the area, and the received signal strength or time of arrival is observed and

used to determine the nearness of these reference positions. The measured nearness, along with

the corresponding known locations, is used to estimate the location of a target. Thus, the

proximity guarantees the most straightforward and easiest implementation for object

localization. RFID and Infrared based localization systems have adopted the proximity

approach (Li and Becerik-Gerber, 2011). Proximity algorithm was also used in this research.

Scene analysis technique utilizes a pre-observed data set of the monitored scene to

estimate the location of a signal source. Fingerprinting is the most well-known scene analysis

based-location sensing technique. The system compares current received signal measurements

and prior observed measurements from known locations. Then, the system estimates the source

location as the one with the most matched fingerprints. Pre-application mapping is required,

during which the location sensing infrastructure, the readers or tags, are deployed in the sensing

area, and the signal strengths of the reference positions are measured and recorded to establish

a signal strength fingerprint map (Woo et al, 2011). This approach requires additional

information data storage to maintain the pre-observation data. The signal strength is the most

common measurement with the scene analysis technique (Fu and Retscher, 2009). It is very

difficult to apply this algorithm during construction due to the changing dynamics of

construction operations.

39
2.4.3 Indoor Localization Technologies

Wide ranges of technologies are used for indoor location sensing. Ultrasound and

infrared were used as localization technologies. An ultrasound signal is emitted from an object,

and the localization system can determine the source location using calculations of the signal

flight time to multiple receivers (Priyantha et al, 2000). Inertial navigation systems (INS) allow

a user carrying a package of motion and rotation sensors and a microprocessor to sense its own

location independently of external infrastructure. Unfortunately, INS systems accumulate

errors and require periodic recalibration (Fox et al, 2003). Various GPS-based solutions have

been introduced to expand the use of GPS to indoor environments. High-sensitivity GPS, for

example, uses highly signal-sensitive receivers developed for weak-signal conditions. As well,

Assisted GPS (A-GPS) sends assisting information, such as satellite orbit data, to the receiver

to speed its satellite acquisition time and improve performance (Diggelen, 2002).

Ultra Wide Band (UWB)-based solutions use short pulses for communication between

tags and receivers, and ensure high accuracy by requiring the deployment of a large amount of

infrastructure and actual communication for positioning at long range (Aryan, 2011). Wireless

Local Area Network (WLAN)-based solutions can be easily set up and require few base

stations. They use the RF signal strength and the magnitude of the electric field as an indicator

of the distance between a source and a receiver. This distance information is then used to obtain

the user’s location by triangulation (Woo et al, 2011). RFID-based solutions require the

installation of RFID readers and reference tags around the building, and rely on the signals

traveling to estimate the distance or proximity of the tags. These reference tags have a well-

known location, and they transmit the location of the target objects to the readers (Choi, 2011).

Hybrid positioning systems could provide more advantages to improve positioning accuracy

as well as cost benefits such as combining two systems instead of one for outdoor activities

40
and another one for indoor activities. Jiang et al (2010) proposed a system that provides outdoor

positioning using GPS and UWB to provide indoor positioning.

Li and Becerik-Gerber (2011) conducted a comparative study for different indoor

location sensing technologies, taking into consideration the following factors: accuracy,

affordability, line of sight, wireless communication, context independence, onboard data

storage, power supply, and wide application in the building industry (Table 2-5). They

concluded that RFID technology is considered the most suitable indoor location sensing

technology.

Table 2-5: Indoor positioning technologies (Li and Becerik-Gerber, 2011)

Wide
On- Built-
Afford No line of app. in
Wireless Contex board in
Tech. Accuracy . sight the
Comm. t Inde. data power
($/m2) required building
storage supply
industry

GPS 1-2 cm 380 X X X

INS 1.1-4.15 m 20 X X

Infrared 30-50 cm 17 X X X X X

UWB 6-50 cm 140 X X X X

4.53-6.89
WLAN 3 X X X
m
1.55-3.11
RFID 25 X X
m

RFID technology has onboard data storage capacity, which enables the tags to act as

distributed databases and decentralizes the tagged objects’ information, making it available

wherever the objects exist. It also supports wireless communication, which allows the

identification and data transfer to be done in a non-contact fashion. Choi (2011) made the same

41
conclusion, since passive UHF RFID-based localization overcomes the limitation of

conventional indoor localization systems, such as their high cost of installation and

maintenance, because of its relatively inexpensive price, the absence of a power source, and

the lightweight of passive RFID tags. These conclusions justify the applicability of UHF RFID

technology in the construction industry, which is a dynamic work environment.

2.4.4 RFID Indoor Localization

RSSI is the main signal measurement used with RFID indoor localization. TOA and

AOA signal measurements are not utilized with RFID indoor localization due to the multipath

effect that disturbs the received signal strength. Trilateration, proximity and scene analysis

algorithms could be used with RFID technology for indoor localization. However, scene

analysis is not suitable for building under construction. Those algorithms are performed based

on the attenuation methods; using path loss models relate the received power to the distance.

Indoor RF wave’s propagation is characterized by reflections, diffractions and dispersion in

the internal structures. Transmitted signals follow different paths to the receiver, which results

in fluctuations in the received signal. This effect is known as multipath propagation, and is

mainly affected by the type of materials used in the construction of the building and by the

surrounding objects (Gonçalo and Helena 2009).

There are two approaches for utilizing RFID in indoor location sensing systems. The

first approach is the “mobile tag fixed reader” and is used by tagging moving targets and

placing RFID readers at fixed locations (Borriello et al, 2004). In this approach, when a reader

is reading one tag, the system verifies that the mobile target carrying the tag is inside the read

range of the readers. This approach is feasible for applications, in which mobile targets move

around well-partitioned and separated areas through checkpoints. Readers are usually placed

at checkpoints to detect the movement of the targets. However, when the accuracy

42
requirements are higher, or the number of checkpoints increases, this “mobile tag fixed reader”

approach becomes impractical. In contrast to this approach, a “super-distributed tag

infrastructure” approach is used (Chon et al, 2004). In this approach, a large number of tags

are distributed in the area, and each mobile agent carries an RFID reader. Context-aware

information can automate the delivery of spatial information to on-site personnel, with which

targets, including building components, equipment, tools and people, can be easily located and

target-specific information can be accessed (Li and Becerik-Gerber, 2011). In this research

“super-distributed tag infrastructure” is used.

2.5 Data Visualization

In construction industry, the lack of adequate visual representation means that project

team members have to struggle with large amounts of data, which can lead to costly errors or

even project failure. Construction progress visualization is needed to explain the complicated

data and eradicate confusing factors (Eddy, 2007). Construction Industry Institute (CII) studies

showed that the cost of rework could range from zero to 25% of the installed cost of

construction. However, the use of visualization representations such as 3D models by

management team can reduce that amount of rework by an average of 65%. The efficient use

of space on a construction site is a site-specific, difficult task and is often left to the intuition

of the construction manager. Poor spatial planning has been deemed one of the major sources

of productivity loss (Thabet and Beliveau 1997).

On large construction projects, project management teams often suffer from not being

able to mentally visualize a project complexity. Thus, they are frequently not able to make the

best decisions to mitigate risks. Visual representation is one way to overcome this problem.

Visual models represent the spatial aspects of schedules and communicate schedules more

43
effectively than traditional methods such as bar charts (Eastman et al., 2008). By themselves,

3D models do not have the ability to show the exact status of a project at a specific period, and

so are of little help in progress control. In order to create a construction schedule from 2D

drawings, planners have to visualize the sequence of construction in their minds. This is an

extremely difficult task since workspace logistics and resource utilization are highly dynamic

by nature. Most site organizations plan their works based on the conceptual site layout and

drawings, which are rarely updated during the project duration. Thus, site managers have not

fully benefited from recent advancements in visualization-information technologies (Chau et

al., 2005).

3D models have some basic information that results from shapes, sizes and locations.

This is the geometric database. Each element has geometric attributes. In addition to the

geometric attributes, which describe the physical model, each element in a model can have any

number of non-geometric attributes associated with it. One non-geometric attribute might be

an activity ID from a time schedule (Aslani et al., 2009). Golparvar-Fard and Pena-Mora (2007)

proposed several semi-automated vision-based approaches to improve and promote the

communication of progress information and decision-making.

2.5.1 Geographic Information System (GIS)

GIS is essentially a combination of computerized mapping and database management

systems (Hassanien, 2002). It is used to represent and analyze the geographic features present

on the earth’s surface. GIS database is a collection of information about objects and their

relationships to each other. It includes a wide range of information such as geographic,

environmental, political and social (Bansal and Pal, 2009). GIS allows maps to be drawn from

its database and data is referenced from the maps. When a database is updated, the associated

map can be dynamically updated. The primary purpose of a GIS is to display, query and analyze

44
spatial data. GIS is used in many applications as a decision-making and problem-solving tool.

Considerable work has been done using GIS for various purposes such as monitoring project

schedules (Min and Jiann, 2002), automated data acquisition (Hassanien and Moselhi, 2002),

and the remote controlling of earthmoving equipment (Marco and Fioerenzo, 2003).

Hassanien and Moselhi (2002) used GIS to develop a model for automating data

acquisition and analysis for planning and scheduling highway construction projects. Min and

Jiann (2002) have developed an automated schedule monitoring system for precast building

construction. The system integrates a barcode, GIS, and a database management system to

assist engineers in controlling and monitoring the erection process on a real time basis. The

erection process was selected because it is the most critical activity in precast building

construction. Macro and Fioerenzo (2003) developed a remote control system for a fleet of

earthmoving and construction equipment. The system uses GIS, CAD and GPS for controlling

and monitoring operations that involve heavy equipment.

2.5.2 Building Information Modeling (BIM)

The National Building Information Model Standard (NBIMS) defines BIM as “A

digital representation of physical and functional characteristics of a facility. A BIM is a shared

knowledge resource for information about a facility forming a reliable basis for decisions

during its lifecycle; defined as existing from earliest conception to demolition. A basic premise

of BIM is collaboration by different stakeholders at different phases of the life cycle of a facility

to insert, extract, update or modify information in the BIM to support and reflect the roles of

that stakeholder” (NIBS, 2007). Building information modeling (BIM) is gaining momentum

in the architecture/engineering/construction (AEC) industry, especially since owners such as

the US Army Corps of Engineers (USACE) implemented initiatives requiring future projects

involving the design and construction of facilities include BIM. In 2005, the United States

45
General Services Administration (GSA) announced that starting 2006, AEC firms would be

required to provide a building information model with their designs (McCuen, 2008).

The goal of adopting a BIM is to provide a comprehensive view of the building by

including (drawings, specification, details, etc…) in a single-source model (Krygiel et al.,

2008). BIM databases contain physical and functional characteristics of a structure since a BIM

model is composed of smart objects rather than lines, arcs, and text. All of these characteristics

are mainly due to BIM’s capability to realize the building through all of the stages in the form

of a database (Fu et al., 2006). BIM users can obtain information about any single element, or

for all elements in a project, to inform decisions about the project. Examples of data included

with an element are material quantities, costs of elements, time considerations related to the

element, building performance, operations and maintenance, and several other items essential

in the lifecycle of facility (McCuen, 2009). Motamedi and Hammad (2009 a & b) and

Motamedi et al. (2011) integrated BIM and RFID for lifecycle management of facilities

components. They added the data taken from BIM database to RFID tags attached to the

components. The idea of making components data readily available on the tags provides easy

access for facility mangers through real-time connection to the central database or a portable

device.

With its potential to assemble the whole project virtually before any actual construction

begins, BIM adds a level of accuracy to both quantity and quality issues. Building materials

can be demonstrated in real time scenarios rather than requiring manual analysis (Krygiel et

al., 2008). By drawing building elements only once for a project in a plan view, the projections

of all elevations and sections are generated automatically. One of the direct benefits is the

reduction in drawing time; therefore, designers can mainly focus on other design issues

(Krygiel et al., 2008). 3D BIM model is a static model, built to represent a building. This could

46
be seen as a shortcoming for the construction process since it is a dynamic process and merits

a dynamic presentation.

4D BIM is a visual representation that combines an object oriented 3D BIM model

with time. 4D BIM is information visualization that is easier to understand than traditional

methods. Traditionally, project managers use 2D drawings, bar charts, and sketches to clarify

the construction design, but these visualizing methods do not integrate the temporal or spatial

dimensions. 4D BIM models are a form of visual representation of a project that also takes into

consideration the temporal aspect of how project teams plan to actually build a project,

according to construction schedules (Hartmann et al., 2008). 4D BIM could be used

strategically by on-site management for progress visualization and presentation, locating

equipment such as material hoists, checking access/openings for equipment, storage

visualization and the utilization and estimation of quantities.

Moreover, 4D BIM can assist site personnel at brainstorming sessions and discussions

about access, storage and sequencing of works (Chau et al., 2005). Better visualization

facilitates team collaboration in removing logical errors in construction operations. Owners of

the constructed facilities may have little experience in construction projects, and are often

unable to truly participate in the construction plan development process unless a simple method

of visualization and communication is made available to them (Kang et al., 2007). 4D BIM

visualization seems to be an effective way of enhancing the many different types of human

perception and it can help anticipate potential construction conflicts during the operational

stages (Chan et al., 2004, and Staub-French et al., 2008).

47
2.6 Summary and Identified Limitations

The literature was reviewed in areas pertinent to automated data acquisition

technologies, indoor localization and data visualization. The literature review was conducted

with prime focus on the impact of those areas on the development of efficient project control

process. The following gaps and limitations are identified accordingly:

• Geometry restriction assumptions for loading and dumping areas and lacks the full

utilization of GPS data to identify the travel and return durations such as the model of

Alshibani (2008).

• Inability to differentiating between individual hauling trucks and absence of complete

representation of earthmoving cycle such travel, dumping and return durations as in the

model of Rezazadeh Azar and McCabe (2011).

• Expensive technology implementation such the On Board Instrumentation System (OBIS)

of CAT Product Link (2013) and its black box format, which prevents users from

accessing its respective algorithms and modifying it as they see fit.

• Inability of predicting fleet performance deterministically or stochastically using OBIS.

Limited work on the use has been done utilizing data sensed from equipment itself such

as the utilization of control sensor.

• Inability of measuring progress for different types of construction activities such as the

models of Ghanem and Abdelrazig (2006), Chin et al. (2005) and Shi (2009). These

models were limited to the delivery and erection of steel structures and did not provide

any location identification for the tracked material.

• The generated 3D models built from images or point clouds are unable to identify objects

that have been delivered to the site or the objects that have been installed but away from

their designed location by a distance greater than the tolerance specified such as the

48
 models of Golparvar-Fard et al. (2009b), Cheok et al. (2000b), Bosché and Haas (2008)

and Turkan (2012). The data acquisition of these models has some subjectivity as it

depends on the motivation and capabilities of the site personnel capturing data.

• All the previous models lack the capability of data fusion among more than one

technology and lack the full utilization of BIM technology.

• Inability to provide real time localization due to the mathematical and computational

complexity such as the model of Razavi (2010). This model was unable to track indoor

activities or material delivery.

• El-Omari (2008) developed model did not present the integration methods or algorithms

that were used to track labor, equipment and material. In addition, the integration was

limited to calculating the quantity of excavation completed.

• Shahi (2012) developed model was limited to just data extraction from the 3D BIM model

elements, with no visualization capabilities. The model was limited to indoor material

tracking utilizing UWB and did not provide material any tracking for material delivery or

outdoor material location.

• All previously mentioned models are incapable of tracking progress of finishing and non-

spatial activities such as inspection, concrete curing, and interior finishing.

• The limitations of existing models lie in their inability to measure the progress of different

types of construction activities from inception to completion in near real-time.

• Most forecasting techniques utilized in those models for repetitive construction operations

use deterministic approaches or stochastic approaches that are based on historical data of

similar past projects. Utilizing such data impacts the accuracy of simulation results for it

does not account for the unique characteristics of the construction operation being

considered.

49
 3 Chapter 3: PROPOSED MODEL

3.1 General

The aim of this chapter is to provide a comprehensive description of the developed

model. This chapter presents the methods and algorithms of the developed model as well as

the steps of the process used to integrate automated site data acquisition technologies,

visualization-information technologies and earned value analysis. The model was developed

to support the tracking, control and progress reporting of construction jobsites. Figure (3-1)

depicts this chapter’s main sections. The developed model is built on integration and

automation of methods and algorithms to support the process of project control. Figure (3-2)

outlines the concept of and vision behind its development. The developed model combines

project scope, time, resources and location for project’s controls purpose. The model is

developed to embrace the human factor and expert knowledge in the use of specially designed

progress templates.

Figure 3-1: Chapter 3 overview

50
 The process of project control starts during the planning stage. Where project scope is

defined clearly and then the project is broken down to complete the Work Breakdown Structure

(WBS). WBS is divided also until it reaches the project’s activity level. The integration

between project activities and time yields project schedule, which is activities time phased

plan. Creating project schedule is the first step in the area of project control. A further step is

taken by integrating scope (Activity), time and earned value analysis to generate project

baseline (S-Curve), as shown in Figure (3-2). Project activities consist of objects that are

spatially scattered in the project. Therefore, every object has a specified location that is

identified spatially in the project’s 2D plans and 3D models or project map for outdoor

environment. These 3D models and maps were used alongside with the project baseline (S-

Curve) for project control. However, this method of control lakes an important ingredient,

which is the information part. In addition, this method had the limitation of being visually

static. A huge leap happened by the introduction of Building Information Modeling (BIM) and

Geographic Information System (GIS). Both technologies enriched visualization models by

integrating information from GIS geospatial database along with BIM database.

Researchers and industry professionals were eager for more project control. Therefore,

they focused more on project assigned resources. This step was established to identify the

progress made by resources on activity or task level. The integration of time and activities’

resources produces the concept of productivity. Many manual methods were developed to

identify resources productivity such as work sampling, craftsman questionnaire and delay

survey. A further step was taken by integrating activities’ resources and its changing location

for tracking purposes. This integration was initially done through manual methods. For

example, tracking if material had been delivered or not and tracking material’s locations on

site. Another huge leap happened in the last two decades by introducing automated site data

51
acquisition technologies to construction industry. This helped a lot in integrating resource’s

changing location and time for tracking, productivity analysis and consequently controlling

project's time and budget.

Figure 3-2: Research vision

Figure 3-3: Research vision

Literature reveals that most of the researches and industry professionals’ efforts dealt

with the different aspects of project control as separate and isolated problems. For example,

earthmoving activities were dealt with in literature from two perspectives. Some efforts

focused on tracking the quantity of excavated soil only, which is the output of the earthmoving

operation. Others focused only on tracking the hauling trucks, which is the input of

earthmoving operation. Similarly, in case of material, the efforts were focused on tracking

52
delivery or outdoor location or indoor location but without giving much attention to how that

material was handled until commissioning. Therefore, the area of project controls still needs

more attention in controlling both input and output of the process. Accordingly, literature, for

most of its parts, lakes integrated models that combine projects’ different aspects for the

purpose of control.

The developed model in this research stems from Figure (3-2), which forms the

research vision. The developed model combines project scope, time, resources and location for

project’s controls purpose. This is done through the fusion of earned value analysis,

visualization aspects and automated site data acquisition technologies. The developed model

was not developed currently to replace the human factor. However, it was developed to

embrace the human factor combined with construction knowledge through using specially

designed progress templates.A number of characteristics have been identified and deemed

necessary for the model to be effective for project controls, they are listed as follows:

• The model should work with minimum interaction between the user and its different

modules. In case of any user’s interaction, it should reduce the user subjectivity to the

minimum limit.

• The model should be structured in modules, each of which should then be broken down

into separate sub-modules according to their functionality. The model’s automated

modules should be mainly working independently. This modularity will guarantee

manageable modifications on specific modules.

• The model should be capable of extracting and exchanging data from/with visualization-

information technologies such as BIM and GIS.

53
• The model should be able to handle different types of project data. In addition, it should be

able to represent projects with various hierarchical breakdown structures.

• The model should capture the onsite data in an electronic format in real time and be capable

of handling large amounts of data on daily basis. The processing and analyzing of the

captured data should be in near real time.

• The model should be cost effective and be capable of integrating cost and schedule

functions. In addition, it should compute the cost and schedule variances at each activity

level and project level.

• The model should automate the process of performance measurement and be accurate in

forecasting project cost and time. The model should automate and facilitate progress

reporting for project team members. The model should be multi-user that allows different

users to access the project captured data and progress reports.

• The model should be able to support the reusing of attributes and to support the sharing of

attribute values. The Internet-based model already allows overall data reusing and sharing

among team members.

• The model should keep all the project-performed data especially repetitive work as

historical data for future reference. In addition, the data, which have a repetitive nature, are

utilized to develop standard templates for domain-specific works.

• The design of the model has to be flexible enough to allow the extensibility of it in the

future to meet various requirements.

3.2 Proposed Model

The developed model consists of two main frameworks; each encompassing methods,

procedures and algorithms as illustrated in Figure (3-3). The first framework is the data

acquisition framework. It is responsible for capturing onsite data. The second framework is the

54
data analysis and reporting framework, which is utilizes for storing, processing and analyzing

the captured data for control and reporting purposes.

Figure 3-4: Developed Model

The data acquisition framework captures onsite data through two main components.

The first component is done through an onsite monitoring person, equipped with a tablet PC.

The tablet PC has the 4D BIM of the project to provide visualized presentation that represents

the status of construction work. As well, the tablet PC includes data acquisition technologies

to enable the onsite personnel of capturing data in real time. The tablet PC has barcode reader,

RFID reader, GPS, Wi-Fi, Bluetooth, a camera for images and videos and a speaker

microphone for voice recording.

55
 The second component of the data acquisition framework includes the “Independent

Automated Data Acquisition Technologies”. It consists of diverse technologies such as GPS

units, RFID readers fixed on project gates, LADAR and Closed Circuit Television (CCTV)

cameras. This component is flexible enough to host additional data acquisition technologies

besides the ones that already exist. Both components of the data acquisition framework collect

the onsite progress data and send it to the data analysis and reporting framework, which is on

a web server containing the previously captured data from the automated data acquisition

technologies and all project information. The captured data is used to represent the actual

progress, which is then compared to the as-planned baseline progress, using earned value

analysis (EVA) to measure the project’s performance. The data analysis and reporting has four

main modules. These modules are central database, performance measurement module, earned

value analysis module and reporting module. Through that framework, the data is stored and

analyzed. Then, the updated data is generated to be sent back to the data acquisition framework

on the tablet PC for the next update cycle. Reports at project, activity, and resource levels are

produced and sent via internet to project team members.

The project activities tracked trough this model are clustered into three main categories:

• “Units Activities”: This category includes electromechanical equipment, curtain walls,

precast concrete panels, etc. These activities consist of units tracked through tracking its

attached low-cost UHF passive RFID tags. Materials belonging to these activities are

tagged using RFID labels or RFID encapsulated rugged tags, which are attached via screws

or epoxy adhesive according to the specifications of each material. The process of material

tagging is conducted at the fabricator's shop with ID similar to the one in 3D BIM model.

For packed material such as wood doors and frames, gypsum board panels, etc., it is tagged

using printed label RFID tags. Two tags are used, one for the pack as a whole to include

56
 information about the packed quantity at delivery and another printed label tag attached to

each item in the pack. On the other hand, materials such as bricks and bulk ones are not

tracked using such method. However, other techniques are used to track this category.

Tracking the tagged material “Units Activities” is achieved by identifying its location

through its RFID tags. The tracking process starts from the point of on-site delivery of the

material until arriving to the installation zone. Upon arrival to its related zone, progress

template is then used until the commissioning of the activity.

• “Bulk Activities”: This category includes earthmoving operations, reinforced concrete

activities etc. Tracking these activities requires tracking the recourses utilized in these

activities as well as tracking the output of the activity. For instance, in earthmoving

operations, three different tracking methods were developed utilizing different

technologies. The first tracking method utilizes GPS mounted on hauling trucks and

Geographical Information System (GIS). The second tracking method deploys RFID gates

in dumping and project areas. Finally, RFID and control sensor are integrated to track also

earthmoving operations. The three methods achieve the part of tracking the input to the

activity. Tracking the output is done through erecting weighting sensors to hauling trucks

to quantify the amount of soil inside the truck. Else, LADAR and cameras are used to scan

the site on daily basis to quantify the total amount of excavated soil.

• “Finishing and Non-Spatial Activities”: This category includes painting, floor tiling,

calibration, testing and commissioning etc. Activities in this category are tracked by

utilizing the integration of 4D BIM model visualization, the tablet PC’s user real-time

localization and supported by progress templates on the tablet PC.

57
3.3 Data Acquisition Framework

As stated earlier the Data Acquisition Framework consists of two main separate sources

for capturing the required data: 1) the independent automated data acquisition and 2) the tablet

PC data acquisition, as depicted in Figure (3-4). The first collects data using GPS, RFID gates,

CCTV cameras and LADAR. Then, the data is transferred automatically to the web server for

storage and processing. This process is totally automated with no human intervention. While

the second source, the tablet PC, collects data with human intervention. The site personnel is

equipped with the tablet PC and ready to periodically go through the indoor and outdoor

project zones, according to the required reporting frequency, to collect the progress data. Then,

the captured data is transferred to the web server, as shown in Figure (3-5). The following

section focuses on the process of tablet PC data acquisition.

3.3.1 Tablet PC Data Acquisition

The first step is building a 3D BIM model that includes all project components and

information. The indoor and outdoor construction areas are divided into exclusive zones, stored

in the 3D BIM model. Then, the 3D BIM is integrated with the project schedule to generate the

4D BIM model (Montaser and Moselhi, 2012 c). The user walks both inside and outside onsite

collecting the progress data utilizing the automated data technologies capabilities of the tablet

PC. For indoor activities, the tablet PC is utilized to identify the location of the site personnel

collecting data.

58
Figure 3-5: Data acquisition framework overview

59
Figure 3-6: Data acquisition framework architecture

60
 This location identification is done utilizing the RFID reader and the low-cost

previously attached reference passive RFID tags, as indicated in Figures (3-6). Having

identified the user’s current location, the following steps are followed:

• Based on the location identified, the zone is retrieved from the 4D BIM model and isolated

in 3D to visually display to the user on the tablet PC the current zone. In each zone, the

user then utilizes the tablet PC capabilities to capture data in various formats such as

scanned RFID data and barcode data, as well as audio records. The user can take snapshots

with the camera and write comments and notes about these zone related activities.

• The user also fills out progress templates related to the current zone. These templates

contain lists of the ongoing activities and unchecked finished activities that guide the user

to where the input is needed. Moreover, these templates could have the as-planned pictures

extracted from the 4D BIM, visually representing the zone and highlighting the ongoing

activities. The user utilizes the templates for data input for those activities where applying

automated site data acquisition technologies would be costly and technically impossible.

For example, if the barcode is not accessible or the illumination is too low to take an image.

• The captured data from each zone is linked to its respective spatial zone in the tablet PC

database. This data is also used for documentation purposes and then synchronized with

the web server for online data storage. The 3D BIM thus becomes richer in its as-built data

and the user can access and search this well organized and documented data easily. This

data gained two new attributes, which is the date/time it was captured in and the zone it

was captured from. Then, project manager can access and analyze this data for different

purposes such as progress, safety, timing and taking corrective actions if needed.

• After finishing the first zone, the model highlights for the user the next zone to go and

proceed for data collection. The user could choose the next zone manually or follow the

61
 model recommendations. The former steps are repeated until all zones have been covered

and confirmed by the model. Detailed description of the above steps is described in the

following five subsections.

3.3.1.1 Building 4D BIM

4D BIM integrates the building project 3D model and its construction schedule. A

realistic project baseline schedule should be developed, including project activities and their

early start (ES) and early finish (EF) dates. The 3D BIM model, first, imports the planned data

directly from scheduling software such as Microsoft Project, in database format. Figure (3-7)

depicts the flowchart of the procedure used to link the 3D BIM model to project schedule. It

then maps the 3D BIM model objects to project schedule baseline activities. As such, different

groups of these 3D BIM objects are assigned to its respective activities. Zone progress

templates are created for the user based on mapping each activity in progress objects to its

related zone. Figure (3-8) shows the flow chart for manipulating 4D BIM for visualization

purposes and the template creation process (Montaser and Moselhi, 2012 a). For each group, a

series of decisions are made to classify whether it belongs to a finished activity, an activity in

progress or an activity to be executed in the future. Before making these decisions, the user

enters the current date. The system compares the current date to the ES and EF dates of each

group, and controls each group’s visibility accordingly. Future activities, those that have not

yet started, are hidden from the developed 4D BIM.

62
Figure 3-7: Indoor data acquisition (Montaser and Moselhi, 2012 c)

63
 Figure 3-8: Linking the 3D BIM model to project schedule

For finished activities, the 4D BIM checks if each group has been inspected and

checked as being finished, or if according to the as-planned schedule it is finished but not yet

checked or inspected. The 4D BIM makes the finished and checked activities visible in their

final forms, and displays unchecked finished activities in a red highlighted form. The activities

currently in progress are displayed in yellow. The activities displayed in red and yellow are

divided spatially among identified zones. Therefore, the tablet PC provides the user with the

zones where progress data should be captured. In addition, a progress template is generated for

each active zone to indicate the status of the red and yellow activities objects as well as the

predefined states for each object in a zone, for example objects erected, inspected or

commissioned (Montaser and Moselhi, 2012 c).

64
Figure 3-9: The process of 4D BIM model visualization

65
3.3.1.2 User Location Identification

The identification of site personnel’s location is necessary to allow integrating this

location with the 4D BIM model to obtain real time location-visualization for the user. For

material localization, identification of site personnel's location is vital for project control and

progress reporting. A two-step algorithm was developed to identify both user and material

locations. The first step in the algorithm is to identify the user current location. Then, the

second step in the algorithm is to identify the material location making use of user locations.

For outdoor zones, tablet PC’s GPS is used to identify user current location. However, it is not

feasible to use GPS for indoor zones. Therefore, Ultra High Frequency (UHF) passive RFID

was explored in this research to locate user and material's for indoor zones. In addition, RFID

is integrated with GPS for outdoor material tracking. Tablet PC’s RFID reader is utilized for

that purpose. User location includes two cases, the case of an outdoor zone and the case of an

indoor zone. The case of outdoor zone location identification is straightforward.

The user starts operating the GPS to receive satellite signals to identify the current

location, as depicted in the flowchart Figure (3-9). GPS signals can be affected by several

sources of error during transmission. The errors sources include (1) satellite and receiver clock

biases (2) ionospheric and tropspheric refraction (3) satellite ephemeris errors (4) multipathing

(5) instrument miscentering (6) antenna height measurement (7) satellite geometry (8) receiver

noise. GPS errors and correction procedure are described in detail in (Blewitt, 1997).

Therefore, a real time data correction is performed. The procedure used in this research is

Differential GPS (DGPS) to obtain more accurate results for outdoor localization. Chapter 4

will provide more information about the DGPS procedure followed and the methods used for

errors correction.

66
 Figure 3-10: GPS user location identification

GPS data at a point is in the form of latitude ( ), longitude ( ) and height (h), which is

mainly a form of astronomical observations (Satellite reference coordinate system). Hence, this

data needs to be converted into the Universal Transverse Mercator (UTM) coordinate system,

which is a geodetic coordinate system (X, Y, Z). Equations (3.1) till (3.4) make the

transformation between the two coordinate systems. In these equations, North latitudes are

considered positive and South latitudes are negative. Similarly, East longitudes are considered

positive and West longitudes are negative.

67
 X = (RN + h) * cos * cos (3.1)

Y = (RN + h) * cos * sin (3.2)

Z = [RN * (1-e2) + h] sin (3.3)

RN = (3.4)

In these equations, the World Geodetic System 1984 (WGS84) is used. WGS84 is the

standard for use in cartography, geodesy, and navigation in Canada. It comprises a standard

coordinate frame for the Earth (the datum or reference ellipsoid). RN is the radius in the prime

vertical of the ellipsoid. In the WGS84, e is the eccentricity of the reference ellipsoid and equals

to 0.08181919084, in Equation (3.4). In Equation (3.4), a is the semi major axis of the ellipsoid

and equals to 6,378,137 m. Further transformation is performed to transform the captured data

into the 3D BIM model local coordinates system. This is done through calculating the vector

3D transformation matrix, refer to the procedure described in Leick (1995). At the end, the user

current location Lj (Xj, Yj) is obtained, as depicted in the flowchart Figure (3-10).

The second case for user location identification is the case of an indoor user. This

section is focuses on the use of RFID for localization for indoor zones. In the current two-step

algorithm, RFID technology is utilized for capturing spatial data of indoor operations. A

number of low cost passive RFID tags are distributed on the jobsite. Each passive RFID tag is

used as a reference point with known location (Landmark) within a predefined zone. In this

research, reinforced concrete columns, shear walls and wall edges were used as landmarks

(Montaser and Moselhi, 2012 d). The indoor construction area is divided into exclusive zones.

The known locations of reference tags are used to estimate the location of the user, based on

the Received Signal Strength Indicator (RSSI) transmitted from reference tags. The user at a

68
given location operates the RFID reader at a time to and captures the signals received from the

reference and material tags, as shown in Figure (3-10). Where, the squares at the corners

represent the reference tags. This process is repeated at a set of time intervals; referred to here

as ∆t. In the field experiments, ∆t ranged from 15 to 60 seconds. The reason for such estimation

of ∆t for the RFID is explained later in Chapter 5.

Figure 3-11: RFID reader receives signals

The location of reference tags is identified with subscript (i), and the location of the

user (site personnel who carries the mobile RFID reader) is identified with subscript (j).

Subscript (k) is used for tracked material temporal location onsite and subscript (f) is used to

represent the final location of a tracked material, obtained directly from the project’s 3D BIM

model. Figure (3-11) illustrates the process of deploying RFID reference tags, which starts by

assigning RFID reference tags to each zone’s landmarks. The coordinates (xi,yi) of each

reference tag (i) are then stored in the tablet PC database. All reference tags are deployed at

the same height. This step is performed once per floor and is used as the input for location

identification.

69
 " # $ % &

Figure 3-12: The process of deploying RFID reference tags

Tag ID is used to distinguish RFID reference tags from material tags. Tablet PC

database is used to filter these tags based on their respective IDs, and to ensure the use of

signals from related reference tags only. If a tag ID is for tracking material, then it is stored in

the database to be processed later along with its related data. However, if the tag ID is not a

reference tag or a material tag, then it is ignored along with its associated information.

Trilateration and Proximity algorithms were used to identify both user and material locations.

Both algorithms are considered as RSSI based localization algorithms. The following section

explains in detail how both were utilized to achieve the first step of the two-step algorithm to

identify the user current location for indoor zones (Montaser and Moselhi, 2013 a).

In the “Trilateration Algorithm”, user’s location Lj is identified location with

coordinates (Xj,Yj) using signals captured from reference tags, as shown in Figure (3-12). As

a result, the coordinates of the unknown user is obtained as in Equations (3.5) and (3.6).

70
 #$% &% '% ((& &) )*#$ & ' ((&) &% )*#$) &) ') ((&% & )
= − (3.5)
" '% (& &) )*' (&) &% )*') (&% & )

($% &% '% )(' ') )*($ & ' )(') '% )*($) &) ') )('% ' )
+= − (3.6)
" &% (' ') )*& (') '% )*&) ('% ' )

Figure 3-13: User location identification using trilateration algorithm

However, with imperfect information and inaccuracy in the path loss model (i.e. R2

1) the circles do not intersect at a single point. In this situation, the intersection of the three

circles is an area, as shown in Figure (3-14). The centroid of that area is the user location as

calculated in Equations (3.7) and (3.8) (Sarda, 2011). The algorithm applies trilateration

method after filtering reference tag data, as in the flowchart shown in Figure (3-14). The

trilateration method requires a path loss model to convert RSSI to distance (d). The developed

path loss model is explained in Chapter 4.

71
 Figure 3-14: User location identification using trilateration algorithm with imperfect path
loss model

2 2 2 2
,-.1 −.2 1+- 22 − 21 1+#+22 −+21 (3∗2∗#+23 −+22 (−,-.2 −.3 1+- 23 − 22 1+#+23 −+22 (3∗2∗#+22 −+21 (
X1= (2∗ 2 −2∗ 3 )∗#2∗+2 −2∗+1 (− (2∗ 12 −2∗ 2 )∗(2∗+3 −2∗+2 )
(3.7)

($% $ )*#' '% (*(& &% )*'∗("∗'% "∗' )

Y1= ("∗&
(3.8)
"∗&% )

The trilateration method determines the position of the user by measuring the distance

from several reference tags. All the readings collected from each reference tag are averaged

and converted into equivalent distance (di) using the path loss model. When the localization

algorithm identifies at least a set of three distances from three different reference tags, the

algorithm generates circles; their centers are the known positions of the reference tags (xi,yi).

The intersection of these circles determines the expected signal source’s location. If the

intersection is not at a point, but rather an area, the center of gravity (C.G.) of that area is used

instead. The coordinates of the user could be calculated (Xj,Yj) accordingly. It is worth noting

that, in many cases, there is more than one intersection area.

72
 Figure 3-15: Trilateration algorithm (Montaser and Moselhi, 2013 a)

In such cases, the following heuristic rule is applied: “If more than one area of

intersection exists, then the C.G. of the area formed by the circles having the least summation

73
of their radii is used”. This rule acts as a useful filter, removing noisy out-of zone-signals. The

user moves to the next location in the same zone or to another zone on the jobsite and the same

procedure is repeated. The generated location Lj (Xj,Yj) is stored in the tablet PC database

with its corresponding time for further utilization in the second step of the algorithm for

material localization. Figure (3-13) depicts the user standing in a zone surrounded by three

reference tags with coordinates (x1, y1), (x2, y2) and (x3, y3) and corresponding average signal

strength (RSSI) 1, (RSSI) 2 and (RSSI) 3. This provides sufficient data to determine the user

location (X1,Y1).

In the “Proximity Algorithm” the same data collection steps that were followed in the

trilateration method are repeated in this algorithm. However, the proximity algorithm uses

RSSI as a weighting method to express how near the reader is to the reference tags. RSSI is a

measurement of the power present in a received radio signal. Therefore, the higher the RSSI

number (or the less negative in some devices), the stronger the signal is, meaning that the user

is closer to that tag. For example, if there are two reference tags and the user received RSSI

equals 30 from the first tag and 40 from the second tag, as shown in Figure (3-15 a). Then the

user location is on the line connecting the two reference tags and more near to the RSSI 40 tag,

as shown in Figure (3-15 b). However, if there are a third tag and the user receiving 35 RSSI

from it; see Figure (3-15 c). Then the user location is in the middle between the three tags and

the location is calculated according to the weighted average of RSSI received from each

reference tag, as shown in Figure (3-15 d). The readings collected for each reference tag were

averaged and converted into a related weight (Wi), which represents how much closer the

reader is to that tag, as shown in Figure (3-16). The coordinates of the user (Xj,Yj) are

calculated using Equation (3.9).

74
 89 ∗ 89
:;% ? ∗
Xj = :;%
89
& => = 89
(3.9)
:;% :;%

a b

c d
Figure 3-16: Proximity method concept

Figure 3-17: Proximity algorithm (Montaser and Moselhi, 2013 a)

75
3.3.1.3 Zone Configuration

Project zones are main ingredient to the developed model for the part of localization

and 4D BIM integration. However, project zones are not a static boundary. On the contrary, it

evolves with time during the project construction. During construction of a given floor in a

project, there is an important time (t) that changes the approach for zone configuration. This

time marks the commencement of floor dividing activity (i.e. partitions) such as gypsum board

or brickwork interior walls activity. Before the beginning of this activity, the floor is an open

big space. However, when this activity starts, the floor is divided into architectural spaces

(rooms and corridors). For each floor, two plans are created, one before the beginning of floor

dividing activity and is called “Area Plan”. The other plan is called “Floor plan”, which is used

after the floor dividing activity starts, as depicted in Figure (3-17).

Figure 3-18: Zone Configuration with time

Floor plan resembles the architectural zoning of the floor after construction of the

dividing activity. However, in the area plan, the user defines the zones without having to abide

by the spaces that will be formed after the floor dividing activity commences. Before the

beginning of the dividing activity, most of the work in the floor is electromechanical work.

76
Figure (3-18) shows the difference between both plans, where the area pan has four zones and

the floor plan has eight zones. The concept of area plan is used also for outdoor areas and roofs,

where there is no architectural zoning for this area. The concept of area plan creates fewer

zones and saves time and effort during inspection and data collection. However, the area and

floor plan model only the 2D requirement for identifying the zone. For indoor zones, the height

from the slab soffit until next floor slab soffit is considered the third dimension of the zone.

For outdoor zones, the third dimension equals the height from the foundation level until the

roof level.

Area Plan Floor Plan

Figure 3-19: Area plan and floor plan

3.3.1.4 Real Time Visualization

Real time visualization allows the user to visualize the current zone in the developed

4D BIM model to capture progress data. The ingredients for real time visualization are zone

configuration, 4D BIM model, activity (objects) and the user identified current location, as

shown in Figure (3-19). Therefore, the steps for this type of visualization start by specifying

the date for date collection, as depicted in Figures (3-20) and (3-21).

77
 Figure 3-20: Real time visualization ingredients

4D BIM model visualization changes according to the specified date. Afterwards, the

user should specify whether he is capturing data in an indoor or outdoor environment. RIFD or

GPS is then utilized based on the user decision to specify the user location, zone and floor if

he is indoor. In case the user is outdoor, area plan is utilized for real time visualization. If the

user is indoor environment, the utilization of floor plan or area plan is decided based on the

specified date entered earlier. Then, the user’s current zone is isolated in the 4D BIM model.

The isolation process is conducted through applying bounding volume to the current zone. The

shape of that volume is decided in 3D based on the zone configuration described earlier. The

application of zone bounding volume supports the process of identifying objects from the 4D

BIM model that resides inside that volume.

78
Figure 3-21: The process of real time visualization (part 1)

79
Figure 3-22: The process of real time visualization (part 2)

80
 Afterwards, a list of progressing objects appears for the user based filtering only the

objects that belongs to activities in progress inside the current zone bounding volume. Then,

the user is asked to specify those objects’ status. The user can select object from that list based

on specific criteria. For example, the user can select circular concrete column with certain

diameter or select all circular concrete columns with certain diameter or all concrete columns

in that zone. This facilitates the data collection process for the user. Since, the selected object

or objects belongs to the same activity, then progress templates provide for the user an easy

way to specify objects’ progress status.

After using progress templates for tracking objects under progress in this room, the

user can use the tablet PC automated data acquisition technologies such as using barcode

reader, taking images or videos for the zone, recording voice or taking hand written notes.

Afterwards, the user can choose to move to the next location in the same zone to capture

additional data. Knowing that, the user should at least capture data from each zone from three

different locations not at the same line, as will be explained later in section (3.4.1.1) for material

localization. When the user finishes capturing data from all locations in a zone then the

developed model checks if data has been captured from all project zones or not yet. If not all

zones are inspected, the model guides the user to the next step. This loop continues until the

user finishes all project zones and captures their respective data through automated data

acquisition technologies and progress templates.

3.3.1.5 Tablet PC Database

This database is considered the core of the tablet PC data acquisition, and is designed

to organize and store all the data captured from the construction site. The database consists of

sixteen entities (or tables) such as project, zone, reference tags, tablet PC, GPS files, RFID

reader file and user location entities. Figure (3-22) shows the entity relationship diagram for

81
the developed database. Entities are basic objects with an independent physical or conceptual

existence. Relationship types involve one-to-one (1:1), one-to-many (1:M), and many-to many

(M: N) relationships. Key attributes are used to distinguish entities. Each entity has a unique

identifier called a primary key, where a key can be a single attribute or a combination of several

attributes (a composite key). In addition, an auto number data type could be employed as a

primary key for the entities.

The ID’s of the projects along with their characteristics (duration, location,

budget…etc.) are considered as the attributes for the project entity. The project entity is

considered as the origin of the database and all other entities stem from it. Project entity

attributes are user input. The project consists of many zones that are classified as indoor or

outdoor. The zone ID and the floor number define the Zone entity primary key. These zones

are extracted from the 3D BIM model and its floor and area plans. Since, the project has many

zones. The relationship between project entity and zone entity is one to many. In each zone,

RFID reference tags are attached on each zone’s landmarks. Therefore, the connecting

relationship between the zone entity and reference tag entity is one to many as well. The RFID

reference tag entity has three main attributes. The user enters those attributes and they are tag

ID, xi and yi, which represent the coordinates of the tags with respect to the BIM model local

coordinate system. It should be noted that as more than one tablet PC can be used for a project,

the relationship between project entity and tablet PC entity is one to many. All tablet PCs

incorporate different ID numbers along with their automated data acquisition technologies such

as GPS and RFID reader. The user enters tablet PC attributes.

82
Figure 3-23: Tablet PC database entity relationship (ER) diagram

83
 Project entity is connected to activity entity through one to many relationship. Activity

entity represents the project scope of work on time phase. Consequently, it is actually the

project schedule and units of measurement and quantity of work. The scheduling tool provides

the attributes for this entity such as planed dates and description. The activity’s quantity is

extracted directly from the 3D BIM model. Each activity is divided spatially in the 3D BIM

model into objects. Therefore, the relationship between the object entity and activity entity is

many to one. This entity and its related attributes are extracted from the 3D BIM model. The

extracted data includes geometrical properties of the object such as length, surface area and

volume. Therefore, it is necessary to select one of these attributes that exactly matches the units

of measurement of the activity. For example, if the activity quantity is measured in (m3) then

the quantity of the object extracted from the 3D BIM should be in (m3) too. This procedure

facilitates the aggregation of objects quantities to match exactly the activity’s total quantity. It

is also worth noting that the activity is divided into 3D BIM objects, which is spatially scattered

in the 3D BIM zones. Therefore, this relationship among activities, objects and zones is crucial

specially when mapping activities in progress to 4D BIM model zones.

When the tablet PC is used in outdoors, the tablet PC’s GPS generates GPS file entity

identifying the outdoor user location. The relationship between the tablet PC and GPS

generated file entity is one to many. Since, the user utilizes the tablet PC in many locations and

every location has its own GPS file. The entity data is filled by the GPS captured data. Real

time correction and transformation algorithms are applied to the GPS captured data. Then, the

identified outdoor user location (Xj,Yj) is deposited in the user location entity. On the other

hand, when the tablet PC is used indoors, the tablet PC’s RFID reader generates RFID reader

file. RFID reader file entity is considered the main repository for the captured RFID data. It

consists of six attributes and they are serial number, tag ID, data file, date, time and RSSI. The

84
serial number is automatically generated to represent how many records the RFID reader

captured. The user current location (Xj,Yj) is identified through Integrating both the data from

RFID reader file entity along with reference RFID tags entity and applying Trilateration or

Proximity algorithm in real time. The relationship between the tablet PC and RFID generated

file entity is one to many because, the user utilizes the tablet PC in many locations and every

location has its own RFID file. The identified indoor user location is stored in the user location

entity. This RFID file entity is used later for material indoor localization and tracking. The user

location entity is the result of processing the RFID reader file or GPS file. It also specifies the

time consumed by the user in this location through “time from” and “time to” attributes. User

location entity is connected by many to one relationship to zone entity.

After identifying the user’s location (indoor or outdoor), the user utilizes the tablet to

visualize his current zone location via the 4D BIM model. Then, progress templates interact

with the user, requesting the status of the zone objects that was mapped earlier to the

progressing activities. The user then utilizes the tablet PC’s capabilities, as indicated in Figure

(3-23) for documentation of progress. Progress template entity represents the steps for

construction activity objects. Therefore, it is connected to activity entity with many to one.

Progress templates represent the construction knowledge of the. Since in many cases, it proves

to be more efficient and straightforward in jobsite implementation than automated sensors.

85
 Figure 3-24: The process of capturing data using the tablet PC

The data captured are presented in the database using six entities. They are sound,

video, image, hand written note, barcode and 3D scan. The first five entities belong to the tablet

PC acquisition and the 3D scan entity belongs to the independent data acquisition as described

earlier. The designed database allows for 2 types of 3D scanners; one that requires activation

from a tablet PC and the other does not. In the first type, the 3D scan entity has to be linked to

the user location. In case the 3D scanner works independently, the 3D scan entity has to be

linked to the project entity. Since, the user in the same location can capture many data such as

images then the relationship between the automated data acquisition technologies entities and

the user location entity is many to one. The data collected, in different formats, have two new

main attributes, time and location. This procedure facilitates data analysis and retrieval for

progress reporting as well as for the remote monitoring of the project status.

86
3.4 Data Analysis and Reporting Framework

The data analysis and reporting framework consists of four modules: 1) Performance

measurement module, 2) Central database module, 3) Earned value analysis module and 4)

Reporting module as outlined in Figure (3-24). The process of the developed data analysis is

outlined in Figure (3-25). Each of its components is described below. The central database

plays a vital role in data exchange. The data exchange takes place between performance

measurement module and earned value analysis module. The performance measurement

module is responsible for analyzing the project performance through “material localization and

tracking” sub-module and “productivity and quantity measurement” sub-module. The

reporting and analysis framework is integrated in a web-based platform. The web-based

platform interacts with the reporting module inside the web server to generate reports for

project team members. The progress templates and the project schedule are updated through

the central database utilizing the processed information.

Figure 3-25: Data analysis and reporting framework overview

87
Figure 3-26: Data analysis and reporting framework architecture

88
3.4.1 Performance Measurement Module

Performance measurement module consists of two sub-modules; “Material localization

and tracking” and “Productivity and quantity measurements”. The material localization and

tracking sub-module is used to track the “Units Activities”, starting from the time their delivery

onsite. After the unit material objects have been located on the final installation zone, the

second sub-module, productivity and quantity measurement, takes over. It extracts activities’

different percentage completion from progress templates and extracts activities’ quantities

from 3D BIM model to facilitate percentage complete calculation. The productivity and

quantity measurement sub-module is used alone to process the data related to “Bulk

Activities”, “Finishing and Non-Spatial Activities”. This sub-module calculates the quantities,

productivity and percentage complete for those activities. The two sub-modules are described

below.

3.4.1.1 Material Localization and Tracking

After delivery of material on site, it is distributed for use at different locations. For both

indoor and outdoor zones while the user is in location Lj, the tablet PC’s RFID reader receives

signals from material tags. Identifying material's location is the second step in the developed

two-step algorithm. The user's location is calculated as described earlier using RFID for indoor

zones or using GPS for outdoor zones. However, the algorithm cannot instantaneously

calculate the material location (Lk), since it needs signals from different user locations. The

trilateration algorithm determines the position of the material by measuring its distance from

previously identified user locations. All material readings collected from each user location is

averaged and converted into equivalent distance (dj), using the developed path loss model.

When the localization algorithm identifies a set of at least three distances from different user

locations, the algorithm generates circles and the centers are the known position of the user

89
(Xj,Yj). Similarly, by applying the same procedure of the first step of the user localization

algorithm for indoor zones but replacing reference tags (i) by user location (j) and applying the

same heuristic rule, the material location is calculated. Figure (3-26) shows the schematic

diagram for applying these procedures for indoor or outdoor zones.

Figure 3-27: Diagrammatic sketch for material location identification

90
 As shown in Figure (3-26), the user moves to another location, for example L2 (X2,Y2)

and the RFID reader again receives signals from material (k). Also, when moving to a third

location L3 (X3,Y3), the reader keeps on receiving a signal form the same material (k). Upon

receiving a signal from material (k) from at least three different locations, the location of

material (k) is calculated as (Xk,Yk). Figure (3-27) shows the flowchart for the procedure of

material location identification using the proximity algorithm and using Equation (3.10) to

calculate the material coordinates. Figure (3-28) shows the flowchart for the proposed method

of material location identification using the trilateration algorithm (Montaser and Moselhi,

2013 a).

89
B;% A∗ A 8C
B;% EA∗ A
Xk = 8C
& =D = 8C
(3.10)
B;% A B;% A

Figure 3-28: Material location identification using Proximity algorithm

91
Figure 3-29: Material location identification using Trilateration algorithm

92
 This process is repeated daily during the user’s walkthrough for data collection. At that

point, material location could be identified and tracked on daily bases. The gate RFID reader

detects the material delivery. While, the user walkthrough detects the material temporal

location, Lk, using GPS and RFID for outdoor zones and RFID only for indoor zones as shown

in Figure (3-30). The material location is compared to material final location (Lf) on daily

bases to know whether the material is placed and/or installed in its final location or still in

handling processes. Upon the arrival of the material to its final zone, the productivity and

quantity measurement module takes over. Figure (3-30) shows the procedure for estimating

“Units Activities” percentage complete based on material localization and tracking sub-module

93
Figure 3-30: Material localization and tracking sub-module

94
Figure 3-31: Estimating “Units Activities” percentage complete

95
3.4.1.2 Productivity and Quantity Measurement

This sub-module is automated to track the status and productivity “Bulk Activities”

and “Finishing and Non-Spatial Activities”. It is also used in tracking and monitoring of “Units

Activities” when the material reaches its final installation zones. For example, in case of

earthmoving operations this sub-module is applied to track and monitor the activity resources

such as hauling trucks. It estimates the productivity and the related quantities of the activities

in daily basis to generate the percentage complete for that activity. Regarding other types of

activities such as painting, gypsum board partitions etc., the integration of 4D BIM model, user

location identification and progress templates provide the user with the required forms and

queries in binary format (Done or Not) to be filled in different zones and their respective

locations. For example, when the user location is identified and referenced to the 4D BIM

model, the model interacts with the user requiring information for the object status. The

interaction of the user in this case is simply stating whether the item on the measurement

template of the activity being tracked is done or not. Depending on the inputs, the central

database interacts with the BIM model to extract the quantities and aggregate it to reach the

percentage completed at the activity level.

As an example of “Bulk Activities”, earthmoving operations were considered. For

earthmoving operations, three different tracking methods were developed using different

independent automated data acquisition technologies. The first tracking method utilizes GPS

mounted on hauling trucks and Geographical Information System (GIS) for visual

representation and data analysis. The second tracking method uses RFID gates in dumping and

project areas and attaching low cost passive RFID tags to hauling trucks. The RFID gates

identify and record the time each truck enters or exits one of these gates. The third method uses

RFID and control sensor (or Switch Limit) to track the hauling trucks. Low cost passive RFID

96
tags are attached to hauling trucks and fixed RFID readers are attached to loaders or excavators.

The read range of the used RFID tag is centimeters, to be activated only when a loader with an

attached RFID reader is loading a truck. On the other hand, control sensor is connected to the

truck control system and operated by the motion of its movable bed. The function of control

sensors is to record the signal time when the truck operator gives an order to the truck control

system to raise or lower truck bed.

The hauling trucks are tracked based on the interaction between the central database

and the automated site data acquisition technologies. This process starts by assigning

automated site data acquisition technology, as shown in Figure (3-31). The assigned tracking

technologies start reporting the hauling truck captured data. Then, the data is transferred to the

web server. The data is then processed automatically, without any human intervention, to

calculate total cycle time for each truck and its respective loading, travel, dumping and return

durations. Truck characteristics and excavated soil properties are retrieved from the central

database to calculate the actual productivity of the hauling truck. Actual productivity is a

function of hauling truck number of trips made at certain period, which requires calculating

truck different cycle time durations. Consequently, calculating cycle time accurately is key

element for estimating onsite actual productivity. The three developed methods are capable of

capturing large volume of data on daily basis, processing and analyzing the captured data,

calculating onsite productivity, measuring actual performance based on earned value concept

and comparing the actual versus planned to derive variances if any, and deterministically

forecast time and cost for the operation.

97
 Figure 3-32: Flowchart for the process of tracking earthmoving operation

Figure (3-32) shows a detailed flowchart of the procedure for processing and analyzing

the data captured (Montaser et al., 2012). Central database provides the list of project working

hauling trucks and retrieves all respective data. For each truck, five main events are identified

and their corresponding registered times T1, T2, T3, T4 and T5 respectively. By identifying

those five events, the cycle time component is calculated as following:

• Loading Time (L) = Registered time of Event 2 - Registered time of Event 1

• Travel Time (TR) = Registered time of Event 3 - Registered time of Event 2

• Dumping Time (D) = Registered time of Event 4 - Registered time of Event 3

• Returning Time (R) = Registered time of Event 5 - Registered time of Event 4

98
Figure 3-33: Flowchart for the process of the captured data processing

99
 Trucks loading time provides a clear picture of the number of trucks inside the loading

area and whether it matches the number of loaders (excavators). It assists in identifying whether

there are trucks queue or not and how long is the queue. Similarly, dumping time could help in

detecting dumping area congestion. Such information supports project managers in detecting

the causes behind any unacceptable performance. Accordingly, the total cycle time (C) is

calculated by the summation of loading time, travel time, dumping time and return time. For

each hauling truck, the capacity and project soil properties are retrieved from the central

database to calculate the truck productivity and to estimate the quantity of hauled excavation.

It is assumed that the truck is loaded with its full capacity; according to truck manufacturer

data and soil type. Those steps are repeated for each cycle and each truck and the total cycle

time and its components are identified and appended to the central database. The first step is

to identify the desired project. Then, the central database provides the list of project working

hauling trucks and retrieves all respective data. Since, the number of hauling trucks, the number

of trips per truck and the capacity of each truck are known; the daily-excavated quantity could

be calculated. In conclusion, the total quantities are calculated using Equation (3.11).

Qt = F M (Number of Truck Calculated Cycles)t X Truck Capacity (3.11)

Where:

Qt = is the quantity of excavated soil at day t

n = Number of hauling trucks utilized in the project

GPS/GIS Based Tracking method integrates GPS and GIS to track hauling trucks and

consequently the earthmoving operation onsite. GPS units are mounted on hauling trucks to

identify its locations during the earthmoving operation cycle. The data captured using GPS is

100
automatically transferred to onsite server and subsequently to the central database. GIS works

as web-based graphical interface, where the user interacts and defines the earthmoving

operation spatial boundaries and query reports. The GPS starts reporting the hauling truck

location and respective data. GPS data are in the form of truck positions over short time

intervals. GPS captured data is truck identification ID, date, time, longitude and latitude. Then,

the data is transferred to the main server, where the central database resides (Montaser et al.,

2011 and 2012).

Truck cycle time consists of four main components, which are loading, traveling,

dumping and return. The loading time is the time that the truck is inside the loading area. Travel

time is the time spent between exiting loading area until entering dumping area. Dumping time

is the time that the truck is inside the dumping area. Return time is the time required for the

truck to travel back from dumping area to loading area to start new cycle. In order to allow the

user to add layers to the GIS map, a drawing tool should be developed for each cycle

component (i.e. loading, travel, dumping and return). Using the drawing tools, the user draws

the polygon around the project area. The drawing tool should be flexible enough to enable the

user to draw regular, irregular and circular shapes. Therefore, if the GPS data point resides

inside this polygon, it means that the truck is in the loading process. Similar procedure is done

for the dumping area. For the travel and return road, the user draws a poly line. This ploy line

represents the optimum travel and return roads, which should be calculated during planning

stage of earthmoving operation. Optimum means, it is the roads that achieves the minimum

travel and return durations and consequently the maximum productivity. If the drawn ploy line

inside streets, it automatically retrieves the street dimensions (centerline and width) from the

GIS map database. On the other hand, if the ploy line is on unpaved roads, an offset with

predetermined value is applied to the poly line to form a polygon. Figure (3-33) shows a

101
diagrammatic sketch of the data capturing scheme and classification of the components of the

earthmoving operation. The user defines the polygons and poly lines one time at the beginning

of the project and its corresponding data is saved automatically in the central database.

Each GPS data point is mapped on the GIS map, which is used to display

geographically, as a separate layer, the moving truck. The truck data point is classified as the

truck crosses from one polygon to another and the time that polygon was crossed is identified.

Then, it is classified as loading, travel, dumping, or return. Figures (3-34) and (3-35) depict the

process of classifying the points and identifying the cycles that the truck has done through a

series of decisions. The five main events in each cycle are identified to be used afterwards for

cycle time calculations. Table (3-1) shows the GPS data after processing.

It should be noted that loading (dumping) time include the exact direct loading

(dumping) time in addition to the truck maneuvering time in loading (dumping) area. Also, If

the truck is outside the optimum travel and return roads, another rule is applied to calculate the

travel and return time. The rule is based on the truck direction. If the direction is from the

loading area to dumping area, it is classified as travel and vice versa for return. The time spent

outside the optimum travel and return roads is calculated and considered as cycle time

inefficiencies of the operation and it is reported to the project manager. This procedure provides

for the user a capability to define if-then rules in the central database, which could help in

situations such as driver break time or to fuel the truck.

102
Figure 3-34: GPS/GIS based tracking (Montaser et al., 2012)

103
Figure 3-35: GPS captured data processing part 1

104
Figure 3-36: GPS captured data processing part 2

105
 Table 3-1: GPS data after processing (Montaser et al., 2012)

Date Time Latitude Longitude Classification Event Time (min) Cycle # Cycle Time

08/04/2010 8:12:06 45.45667 -73.64062 Loading 1

08/04/2010 8:15:05 45.4566 -73.64045 Loading

08/04/2010 8:15:11 45.4543 -73.63966 Travel 2 0:03:05

08/04/2010 8:17:11 . . .

08/04/2010 8:21:11 . . .

08/04/2010 8:26:11 45.45729 -73.62556 Travel

08/04/2010 8:26:20 45.45602 -73.62363 Dump 3 0:11:09

1 0:25:05
08/04/2010 8:26:11 . . .

08/04/2010 8:27:58 . . .

08/04/2010 8:28:06 45.45634 -73.62273 Dump

08/04/2010 8:29:11 45.45738 -73.62531 Return 4 0:02:51

08/04/2010 8:30:11 . . .

08/04/2010 8:37:46 . . .

08/04/2010 8:37:11 45.45532 -73.64124 Return 5 0:08:00

RFID Gate Based Tracking method is developed to calculate the total cycle taken by

the truck and the total number of cycles needed to accomplish the scope of work. This method

has two components; the first is using RFID gates in both dumping and loading zones and the

second is by attaching low cost passive RFID tags to hauling trucks. The main RFID hardware

components used are RFID fixed reader and RFID encapsulated tags or RFID label tags. For

passive RFID label tags, it is not rugged. However, it can be used inside the truck itself. It can

be mounted directly on the windshield glass at the front of trucks. Rugged encapsulated tag is

mounted on the driver’s door of trucks. An illustration of the method is shown Figure (3-36).

The developed method uses fixed RFID readers for the gate system at the loading and dumping

106
areas. This allows collecting RFID singles from entering and departing trucks into and from

the two areas. Once an empty truck enters the loading area, the fixed RFID reader (RFID

Reader 1) receives signals from the RFID tag attached to the truck. This process is repeated in

a similar sequence, but using Reader 2 at the dumping area. In between, i.e. over the hauling

and return roads, no signal is received. RFID reader generates a file with five fields: a) RFID

tag ID, which was read, b) the number of times this tag was read, c) RSSI) d) date, e) time. By

fusing the data from both RFID Readers 1 and 2 for individual truck, the activities durations

within a complete cycle can be clearly defined (i.e., loading time, travel time, dumping time

and return time). The data collected from both RFID Readers 1 and 2 is transferred to the

computer onsite and subsequently transferred to the server in the head office of the contractor

(Montaser and Moselhi, 2012 b).

Figure (3-37) illustrates the method to identify the main five events upon performing

the data fusion of the two readers. Event 1 represents the commencement of the loading

process. The approximation here has to do with limiting signal range of the passive tag used in

the developed method. When a truck leaves the loading area after finishing loading the

excavated material, RFID reader 1 registers Event 2. Event 3 is registered from RFID Reader

2 when the hauling truck enters the dumping area, which also confirms the end of the hauling

activity. Upon dumping the excavated material, the hauling truck exits dumping area and RFID

Reader 2 registers the commencement of Event 4. A cycle is completed upon return of the truck

to the loading area (i.e. commencement of Event 5). Table (3-2) shows an example of the cycle

time calculations computed by the developed method.

107
Figure 3-37: RFID gate based tracking (Montaser and Moselhi, 2012 b

108
 Table 3-2: RFID data after processing (Montaser and Moselhi, 2012 b)

RFID
Tag Time Cycle Cycle
Date Time Reade Event Type
ID (min) # Time
r

24.11.2010 8:12:06 BF 1 1

24.11.2010 . BF 1

24.11.2010 . BF 1

24.11.2010 . BF 1

24.11.2010 . BF 1

24.11.2010 8:15:11 BF 1 2 0:03:05 Loading Time

24.11.2010 8:40:27 BF 2 3 0:25:16 Travel Time

24.11.2010 . BF 2

24.11.2010 . BF 2

24.11.2010 . BF 2

24.11.2010 . BF 2

24.11.2010 8:42:38 BF 2 4 0:02:11 Dumping Time

24.11.2010 9:03:43 BF 1 5 0:21:05 Return Time 1 0:51:37

109
Figure 3-38: RFID captured data processing (Montaser and Moselhi, 2012 b)

110
 Both GPS/GIS and RFID gate methods assumes that the loading and dumping times

are calculated once the truck enters or exists the loading or dumping areas. This assumption

adds to exact direct loading or dumping time additional time for truck maneuvering in the

loading or dumping area as shown in Figure (3-38).

Figure 3-39: Cycle time in GPS/GIS and RFID gate tracking method

RFID/Control Sensor Based Tracking method, low cost passive High Frequency (HF)

RFID tags are attached to hauling trucks and fixed RFID readers are attached to excavators

(loaders). RFID is utilized to detect truck proximity to excavator (loader) for loading, which

improves tracking method accuracy for loading time. The read range of the used RFID tag is

less 1m, to be activated only when the excavator (loader) with the attached RFID reader is

loading the truck. Then, the RFID reader starts capturing the RFID signals from the truck tag.

On the other hand, control sensor is connected to the truck control system and operated by the

motion of its movable bed. The function of control sensors is to record the signal time when

the truck operator gives order to the truck control system to raise or lower truck bed during the

dumping process. In addition, switch limit could be utilized instead of the control sensor to

detect the movement of truck bed in dumping process to calculate the exact dumping time. The

control sensor sends its data via cable to microcontroller. Truck microcontroller has a wireless

communication module that could send and receive data to/from RFID readers. The data

111
captured by RFID from hauling truck’s passive tag and control sensor is transferred wirelessly.

Fusing the data captured from RFID reader and control sensor is used to identify loading,

travel, dumping and return time that constitute the hauling truck cycle time and consequently

the productivity of the fleet (Montaser and Moselhi, 2013 b).

Figure (3-39) depicts RFID hardware implementation method, where the passive tag is

attached to the hauling truck bed, which is the nearest location to the excavator during loading.

RFID reader is fixed to the excavator front with the antenna attached to the excavator bucket.

This setup in addition to tag read range allows the RFID reader to receive tag signals when the

excavator is loading the truck. The signal is represented by a zigzag line to indicate the loading

process. The excavator is stationary in its location; loading its bucket and dumping the soil in

the truck bed. This process is repeated until the hauling truck is loaded with its full capacity.

In this state, the response of excavator’s RFID reader is continuous zigzag line during the

loading process. The time from the beginning of the zigzag line till its end represents the

loading time in this cycle.

Figure 3-40: RFID reader captured data with excavator

112
 The same RFID setup is used in case of using a loader to load the hauling truck.

However, the loader sequence of work is different as shown in Figure (3-40). As, it moves to

the soil stock pile forward and backward per each bucket load. For each loader bucket load,

the RFID reader response is continuous zigzag line then straight line due to the loader

movement away from the truck to the soil stock pile till it enters the reading range once again.

Accordingly, the time from the beginning of the first zigzag line until the end of the last zigzag

line represents the loading time in this cycle. In this process, the RFID reader generates a file

with five fields: a) RFID tag ID, which was read and represents the truck ID, b) the number of

times this tag was read, c) received signal strength, d) date, e) time. Figure (3-41) shows a

graphical representation of control sensor’s captured data when the hauling truck bed is raised

for dumping excavated soil and lowering it afterwards. The change in control sensor status

happens due to the hauling truck bed motion. The control sensor captured data are date, time,

truck ID and status (On or Off).

Figure 3-41: RFID reader captured data with loader

113
 Figure 3-42: The change in control sensor captured data

Figure (3-42) illustrates the main five events that describe the entire earthmoving

process, upon performing the data fusion of the two sensors. These five events represent a

complete cycle in the operation being modeled. Event 1 represents the commencement of the

loading. As long as the hauling truck is within the excavator’s (loader’s) read range, the reader

keeps receiving signals from the truck. Event 2 registered at the end of the zigzag line in case

of an excavator, or at the end of last zigzag line in case of a loader. Event 3 is registered from

control sensor when the truck operator gives order for dumping which is indicated by raising

the truck bed. Upon dumping the excavated material, the operator gives order to lower the truck

bed then the control sensor registers the commencement of Event 4. A cycle is completed upon

return of the truck to the loading area (i.e. commencement of Event 5).

Figure 3-43: RFID and control sensor data integration (Montaser and Moselh, 2013 b)

114
 Figure (3-43) shows the process of raw data processing. Then the total cycle time can

be calculated by the summation of loading time, travel time, dumping time and return time.

These steps are repeated for each truck in the project as shown in Table (3-3). For each truck,

the number of cycle times, total cycle time duration and its components durations are identified

then appended to the central database. Unlike the previous two methods for tracking the hauling

truck, this method detects the exact loading and dumping time without adding any other

duration, like queuing and maneuvering to get into position. The three developed methods

allow project teams to check jobsite conditions remotely and study the efficiency of the planned

operations. It provides them also with tools for detecting potential problems in loading areas,

dumping areas and travel hauling and return roads. The RFID/Control Sensor method could be

integrated with the GPS/GIS or RFID Gate method to provide more information regarding the

loading, travel, dumping and return areas and the integrated method accuracy will be enhanced

considerably as shown in Figure (3-44).

Table 3-3: Cycle time calculations (Montaser and Moselh, 2013 b)

Control Cycle
Truck RFID Time
Date Time Sensor Event Type Time
ID Reader (min)
Status (min)
24.11.2010 8:12:06 230 230 1
24.11.2010 . 230 230
24.11.2010 . 230 230
24.11.2010 . 230 230
24.11.2010 . 230 230
24.11.2010 . 230 230
Loading
24.11.2010 8:15:11 230 230 2 0:03:05
Time
Travel
24.11.2010 8:40:27 230 On 3 0:25:16
Time
Dump
24.11.2010 8:42:38 230 Off 4 0:02:11
Time
Return
24.11.2010 9:03:43 230 230 5 0:21:05 0:51:37
Time
24.11.2010 9:03:47 230 230 1

115
Figure 3-44: RFID/control sensor captured data processing

116
 Figure 3-45: RFID/Control Sensor method and GPS/GIS or RFID Gate method

Table (3-4) depicts the comparison among the three different tracking methods. GPS
utilizes satellite wireless communication and requires direct line of sight. Hence, RFID
technology is considered more useful than GPS in tracking hauling units in dense downtown
areas populated with high-rise buildings, which obstruct satellite signals. Knowing that, the
number of trucks, excavators and site entrances (exits) are considered variables for the initial
cost. It is assumed that the costs of maintenance and software implementation are equal for the
three methods. The following notations were used to conduct initial cost comparison among
the three methods:

Number of hauling trucks = X

Number of excavators (loaders) = Y

Number of site exits or entrances = E

RFID Reader Cost = C

RFID Tag Cost = R

GPS unit Cost = G

Control Sensor Cost = S

117
 Table 3-4: Comparative study

RFID/Control
Criteria/Method GPS/GIS RFID Gate
Sensor

Detect cycle time Yes Yes Yes

Detect # of cycles Yes Yes Yes

Calculate productivity Yes Yes Yes

Detect exact loading time No No Yes

Detect exact dumping time No No Yes

Can detect labor and material entering or

No Yes No
exiting the project
Track effectively the truck along travel and
Yes No No
return roads
Effective for tracking building earthmoving
Yes Yes Yes
operations
Effective for tracking highway earthmoving
No No Yes
operations
Can detect truck weight
No No No

Initial Cost
GX RX+CE (S+R)X+CY

Regarding the assumption of hauling truck is loaded with it is full capacity; a weighting

sensor could be attached under the hauling truck to estimate the weight of soil in the truck. As

well, a digital camera could be used at the project gates to take images for the truck and estimate

the volume of excavated material in the truck. Similarly, CCTV camera images around the site

could be used to estimate the volume of excavated material in the truck. The developed three

methods are used to estimate the input to the earthmoving operation. However, the output of

the operation could be measured using 3D scanner and photogrammetry, though scanning the

site on daily basis as proposed by El-Omari (2008).

118
 Other machine-based sensors could also be integrated with these methods to identify

idle time and broken equipment time, which will provide an alerting system for project team.

Near real time control of on-site earthmoving operations, facilitates early detection of

discrepancies between actual and planned performances and support project managers in taking

timely corrective measures. In case of utilizing scraper in highway construction, Montaser and

Moselhi (2013 c) developed a method to track Scraper-Pusher fleet for Cut-Fill highway

earthmoving operations, refer to Appendix A.

3.4.2 Earned Value Analysis Module

Measuring project’s performance in a timely and effective manner is an important task.

It allows managers to diagnose and identify areas that need more awareness and to take

corrective actions if needed. This section presents the utilization of earned value analysis in the

data analysis and reporting framework. This module works on object level and aggregates

different values until it reaches the activity level. Similarly, the same procedure is followed to

aggregate different values at the activity level to reach the project level. In case of “Units

Activities” combination of material localization and tracking sub-module and productivity and

quantity measurement sub-module are utilized. Table (3-5) shows an example for estimating

percentage complete for a fan coil unit activity, which is considered an example of “Units

Activities”.

This procedure estimates percentage complete on object level, knowing that each

activity consists of many objects. So, an aggregation process is established based on objects

until it reaches the activity level using Equation (3.12). This aggregation is done using two

different methods. For activities that are measured by number of units, as the case is in

electromechanical installation, the aggregation is based on the object’s relative cost in activity

119
under study. The second method of aggregation is for activities that are measured in quantity,

as in m3, m2 or linear m, Equation (3.13) is used to aggregate from object level to activity level.

Table 3-5: Example of “Units Activities” objects percentage complete estimation

Cumulative %
Task Work Done Sub-Module
Earned

Delivered to site 60%

Material Localization
Inside Building 65% and Tracking

Inside Final Zone 70%

HVAC (Fan Coil

Fixed to the Ceiling 75%
Units)

Ducting and Fittings 80%

Productivity and
Quantity Measurement
Testing 90%

Handing Over 100%

% Complete for Activity = 8 M (Wi X Cumulative % Earnedi) (3.12)

% Complete for Activity = 8 M (Qi X Cumulative % Earnedi) (3.13)

Where:

Wi = is the object relative cost

Qi = is the quantity of object

i = the object ID

n = Number of activity objects

120
 In case of “Finishing and Non-Spatial Activities”, progress template is utilized to

estimate percentage complete from productivity and quantity measurement module. In case of

“Bulk Activities” such as earthmoving operations, the quantities calculated from tracking

hauling units is used to calculate percentage complete as mentioned earlier in productivity and

quantity measurement module. The cumulative excavated quantities until day t equals to Qt

plus the cumulative excavated quantities at (t-1). This quantity is divided by the total

excavation quantity to identify the actual percentage complete. Percentage complete is used

afterwards in earned value analysis for progress reporting purpose and to estimate the actual

cost and duration. BCWP is calculated according to the percentage complete calculated on

daily basis.

According to earned value analysis, three main indicators are used for performance

analysis; namely as: Budget Cost for Work Scheduled (BCWS), Actual Cost for Work

Performed (ACWP), and Budget Cost for Work Performed (BCWP). BCWS for each activity

is known from the planning stage and saved in the central database. Therefore, BCWP is

calculated according to Equation (3.14). ACWP is entered by the user from actual on site

expenditures related to each activity. Upon the completion of estimating the BCWS, BCWP

and ACWP for each activity, this module progresses with calculating Cost Variance (CV),

Schedule Variance (SV), Cost Performance Index (CPI) and Schedule Performance Index

(SPI), as indicated in Equations from (2.1) until (2.4). These indexes are used to identify the

possible cause(s) of unacceptable performance. Then, the module indicates whether the activity

status for time (behind schedule or on schedule or ahead of schedule) and cost (over budget or

on budget or under budget). The module deterministically forecasts the cost Estimated at

Completion (EAC) at any future set date, as indicated in Equations from (2.5) until (2.9) using

two assumptions: (1) the established performance at the report date continues until completion

121
(EAC1) or (2) the remaining work is performed as originally planned and the previous

performance was exception (EAC2). These assumptions are providing practical scenarios for

the project manager to select from.

BCWP (Activity) = % Complete X BCWS (3.14)

Until now, all calculation were conducted on activity level, to take it up to the project

level and to do earned value analysis Equations (3.15, 3.16 and 3.17) are used to calculate

project BCWS, BCWP and ACWP. Similarly, this module progresses with calculating project

performance indexes (CV, SV, CPI, SPI) but for the project level as indicated in Equations

from (2.1) till (2.4). Additionally, this module estimates the time and cost status for project at

the reporting date and forecasts it at any future set date as indicated in Equations from (2.5)

until (2.9).

Project BCWS = 8 M (BCWS) (3.15)

Project BCWP = 8 M (BCWP) (3.16)

Project ACWP = 8 M (ACWP) (3.17)

Where:

i = Activity ID

n = Number of activities

Most forecasting methods for repetitive cyclic construction operations use

deterministic approaches or stochastic approaches but based on historical data of similar past

projects. These methods did not consider uncertainty. Utilizing such data is an obstacle for

122
accurately simulating the status of these operations on current projects. For example in case of

earthmoving operations, the cycle time differs as the travel time of trucks can be affected by

several dynamic factors, such as weather conditions, operating conditions in the excavation

area and traffic on travel roads. Accordingly, Discrete Event Simulation (DES) is utilized in

this research to forecast activity future performance. This method is applied only to repetitive

cyclic activities such as earthmoving operation. Therefore, the developed method demonstrates

the stochastic value of hauling trucks captured data and uses the actual collected data of the

elapsed periods. Since, this data naturally inherits in itself variables of stochastic nature such

as weather. DES is used to forecast fleet productivity and consequently the forecasted activity’s

duration and cost. It is also used to experiment with different fleet configuration and its impact

on forecasted time and cost (Montaser et al., 2011 and 2012). The captured data for loading,

hauling, dumping and returning are used to generate representative Probability Distribution

Functions (PDF). Those PDFs are used as an input for the DES model to evaluate the current

operation and to highlight the performance bottlenecks and to experiment the fleet

configuration if needed.

In addition, if the activity during the elapsed period experienced delay, corrective

actions could be planned based on generating different fleet configurations using DES. This

method is automated but requires human intervention in selecting different scenarios such as

the number of excavators, hauling units and their capacity. It is built based on the interaction

between the system and the central database where the past period cycles time and its

component has been stored. Figure (3-45) shows the method general flowchart, starting by

generating the PDF for cycle time components from the captured data retrieved. PDF should

be generated for each cycle time component (i.e. loading time, travel time and dumping time)

that best fits the process including its associated parameters (e.g. mean value and standard

123
deviation). The fitted functions are then used as an input for the simulation model to generate

and evaluate different scenarios (Montaser et al., 2011 and 2012).

Figure 3-46: The process of stochastically forecasting productivity

In this research StroboScope discrete event simulation engine was utilized, where it is

installed on the web server. Specially designed DES model for earthmoving operations is

deployed as depicted in Figure (3-46). This DES model is a typical model and used for data

manipulation and processing to calculate the stochastic forecasted productivity and to

experiment different fleet configurations in case of any corrective action. The DES model

retrieves the earthmoving operation data such as total quantity of earthwork, number of hauling

trucks and excavators, hauling truck capacity and PDF for different cycle time components

from the central database. The DES reports to the user the forecasted fleet productivity and its

124
associated cost and time. The user interacts with the DES model to change the fleet

configuration only. Then, different scenarios are generated for the user, which could help in

time or cost optimization and making informed decisions. The simulation model equations are

from (3.18) till (3.25)

Figure 3-47: Specific purpose earthmoving operation DES model

Excavator Idle Ratio = The time weighted average of the content of Excavator Queue /

Number of Excavators (3.18)

Excavator Utilization = 1 – Excavator Idle Ratio (3.19)

Truck Idle Ratio = The time-weighted average of the content of Truck Queue / Number

of Trucks (3.20)

Truck Utilization = 1 – Truck Idle Ratio (3.21)

Time of operation in hours = Simulation Time / 60 (3.22)

125
 Production rate (m3/hr) = The current content of Dumped Soil Queue / Time of

operation in hours (3.23)

Total cost of operation = [Truck cost ($/hr) * Number of Trucks + Excavator cost ($/hr) *

Number of Excavators] * Time of operation (hr) (3.24)

Unit cost ($/m3) = Total cost of operation / The current content of Dumped Soil Queue

(3.25)

3.4.3 Central Database

Construction operations are information intensive. The optimum monitoring of such

operations requires storing and retrieving a large amount of data on daily basis. Therefore, a

central database has been designed to support the management functions of the developed

model and to minimize the user interaction. The performance measurement and the earned

value analysis modules interact with the central database. It is specially designed to organize,

store and manage captured data from construction sites, providing queries to retrieve the

information needed for reporting the project status. Relational database management systems

are better suited for this type of model (Figure 3-47), since it contains structured data,

relationships and schema design.

The central database developed here is considered the core of the data analysis and

reporting framework. The database consists of forty one (41) entities (or tables). Tablet PC

database explained in section (3.3.1.5) is considered part of the central database but on the

tablet PC. However, the central database structure is on a web server to facilitate on-line data

storage, retrieval and analysis. Resources entity involves the main three resources in any

construction project, which are equipment, labor and material.

126
Figure 3-48: Entity Relationship (ER) diagram for the central database

127
 The three entities are connected to resource entity with child to parent relationship as

each one of them inherits attributes from the parent entity. The resource entity in addition to

its three child entities are user defined. The main scope of the equipment entity in this research

is to resemble the fleet configurations such as excavator and truck fleet. Both entities have the

main attributes for the fleet configuration such as capacity. Also, truck and excavator entities

are user input. Truck GPS entity is connected to the truck entity with one to one relationship,

and in case of tracking all equipment this relationship could be connected to the equipment

entity instead of only truck entity. Truck GPS is defined by the user. Truck GPS file entity is

considered the output of the mounted GPS, and it is a separate file for each day’s data. So, this

entity is connected by many to one relationship to truck GPS entity. Road entity is important

for describing the truck optimum travel and return route.

Road entity is connected to project entity and dumping area entity with many to one

relationship and it is user defined. This entity is also important when comparing its points with

the output from the truck GPS file. Dump area entity represents the area that serves the project

for dumping excavated soil. It is connected to project entity with many to many relationship.

Since, many projects can use the same dump area or one dump area could serve many projects

and this entity is user defined.

Soil entity represents project soil characteristics as reported in the project geotechnical

report. Soil entity is connected to the project entity with one to many relationship based on an

assumption that each project has one soil type only. Gate RFID reader entity is connected to

the resource entity with many to many relationship as resource can be read by many readers.

For example, the same truck can be read at least by two readers, one in the project area and the

other in the dump area. This entity is also connected to the project and dump area entities with

many to one relationship as one project or dump area could have more than one gate and

128
consequently many RFID gates. RFID gate entity is utilized for registering the material

delivery time and date. Additionally, it is used to record labor entering and exiting from the

construction jobsite to calculate each labor time sheet. In case of using control sensor to track

the exact dumping time, then control sensor entity is connected to the truck entity with one to

one relationship. The control sensor output is stored in the control sensor file entity and it is

connected to control sensor entity with many to one relationship.

To track the truck proximity to the excavator, an RFID reader is mounted on the

excavator. So, the relationship between the excavator RFID reader entity and excavator entity

is one to one. The results of the excavator RFID reader is deposited in the excavator RFID

reader file. Truck status entity is considered the output for processing the data from GPS/GIS

or RFID gate or RFID/control sensor tracking methods. This entity has many to one

relationship with the truck, as it reports the truck status on daily basis. CCTV camera and 3D

scanner are parts of the independent automated data acquisition technologies that were utilized.

Accordingly, each entity was created for each technology and both are connected to project

entity with many to one relationship. The captured data from the CCTV camera is stored in the

in the image entity. The same procedure is followed with the output of the 3D scanner.

Object temporal location entity is the output for processing user location and RFID

localization for indoor zones, or RFID and GPS for outdoor zones. This location is calculated

on daily basis. This entity is connected to user location by many to many relationship. Activity

status entity is connected to activity entity with many to one relationship and this entity is

considered the output for processing the captured on site data from different automated

acquisition technologies supported by the interaction going on between site personnel and

tablet PC. Project status entity is considered the final output for the developed model as it

reports on project level and it is connected to project entity by many to one relationship. For

129
more details about each entity, attributes and how it is utilized in the developed model refer to

Appendix B.

3.4.4 Reporting Module

The purpose of reporting module is to generate textual, tabular and graphical reports to

facilitate project tracking, control and forecasting. These reports are generated to depict

progress at all control levels. Performance reports are generated daily, weekly and monthly

according to the activity and user preferences. A project level management requires an

overview of project progress and milestones. Since the reporting module exists on web server,

all team members could access it via internet. This module provides progress information

including:

- A summary of status of the schedule, budget and expected completion date.

- Report on the current progress since last update to date.

- Report the cumulative progress to date and report on the end forecast.

3.5 Summary

This chapter presents the research vision and the developed model and is frameworks,

modules, sub-modules, methods, procedures and algorithms. The developed model embraced

integration and automation; utilizing a wide range of technologies in adaptive way to facilitate

timely tracking and progress reporting. The developed model integrates project visualization-

information aspects, automated site data acquisition and earned value analysis.

130
 4 Chapter 4: FIELD STUDIES

4.1 General

This chapter describes the field studies conducted to evaluate existing automated site data

acquisition technologies. The fieldwork focused on deployment of these technologies on actual

construction jobsite. These studies led to the identification of challenges associated with utilizing

each technology separately and exploration of the likely advantage of integrating technologies and

using them to estimate project progress. Figure (4-1) presents the chapter overview. The results of

these studies were conducted concurrently with model development and refinement of its

application. The captured data from the field studies helped in the process of model verification

and validation. This chapter has three main sections. It starts by a brief description of the

construction project used in this research. The second section explains the independent automated

data acquisition technologies deployed on the jobsite. The last section describes the tablet PC used

in this research and its integrated automated data acquisition technologies.

Figure 4-1: Chapter 4 overview

131
4.2 Case Study

Field studies were conducted during the construction of the Center for Structural and

Functional Genomics (CSFG) at Concordia University in 2010-2011, as shown in Figure (4-

2). The project is located at the corner of Sherbrook St. West and West Broadway St. in

Montreal. It forms an expansion to the existing Science Complex. Construction of the building

started on April 2010 and completed by June 2011. The building consists of basement floor,

ground floor, first floor, second floor, mechanical floor and roof. The total built up are is 6000

m2. The building is a reinforced concrete structure, except the mechanical and fire escape

stairs, which are of steel structures. The contractual budget was $20 million with contractual

duration of 12 months.

Figure 4-2: Genomics office building

The progress report provided by the prime contractor has five main sections along with

few appendices where few of the most important documents are attached. The daily reports

submitted by the contractor provide the minimum amount of information regarding the project.

It was noticed that no quantities are reported and no man-hours tracked. Although the level of

details in the daily reports is not high, they were used along with the site daily pictures and the

invoiced percentages to update the baseline schedule, as depicted in Figure (4-3). In addition,

132
daily pictures were also used to identify the actual progress and work logic versus the planned

schedule.

Figure 4-3: Sample of project progress reports

133
4.3 Preparations for Field Work

Initial visits to the jobsite of the project were made and the blue prints released for

construction as well as the construction schedule in MS project were reviewed to get familiar

with the various aspects of the project. More than forty (40) jobsite visits were made. Appendix

C provides a summary of the site visits schedule and the type of data captured. The

Architectural and Structural plans provided by Concordia University were in the form of 2D

CAD drawings, and there was no BIM model for the building. A 3D BIM model was created

as a part of the developed model. The BIM model was developed using the project 2D

drawings. Revit Autodesk software was used to develop the 3D BIM model, as shown in

Figure (4-4) (Montaser and Moselhi, 2012 a).

Figure 4-4: Genomics building 3D BIM model (Montaser and Moselhi, 2012 a)

134
 All of the construction data, such as objects’ materials as well as their specifications

were either available in related documents or assigned during site visits to the building. Since

the developed objects have parametric relationship with each other, the sections, elevations,

and perspectives were generated automatically. The model consists of 7450 parametric object.

In early stages of this research, NavisWorks Autodesk was used to generate 4D BIM

model for the project. NavisWorks was used to link schedule with the 3D BIM model

objects. The 3D Revit model objects were exported to Navisworks along with the construction

schedule. Once the Microsoft Project schedule and 3D model were imported into Navisworks

detailed relationships were developed to link each schedule activity to its corresponding objects

of the model. The software automatically simulates the construction sequence as per the

project schedule, as shown in Figure (4-5). The output of the 4D model was in the form of

a simulation video where an animated sequence of construction can be seen. Different

colors were assigned to objects in the developed 4D BIM model to express their respective

status of progress. At any given time, the 4D model can reflect the time schedule in a visual

manner. Objects with user selected colors means that they are already executed, while objects

green highlighted color means that there is a construction activity in progress. At the top left

of the simulation screen a detailed date specifies the name of the activity being executed along

with the exact time and duration. However, the process of utilizing Microsoft Project for

scheduling, Revit for 3D BIM modeling and NavisWorks for 4D BIM modeling was

cumbersome. Since, any changes in the schedule or the 3D BIM model necessitate beginning

the process of exporting both data from scratch again to NavisWorks to generate the 4D BIM

model. In addition, some data were lost when exported from 3D Revit model into NavisWorks

software.

135
Figure 4-5: 4D BIM NavisWorks© model

136
 This challenge dictates making modification in Revit software to accommodate the

option of generating 4D BIM Model internally without using third party software. This

development is described in Chapter 6.

4.4 Independent Automated Data Acquisition Technologies

Independent automated data acquisition technologies are those not embedded or

integrated with the tablet PC. In this research different technologies belongings to this category

were investigated, which are 3D scanner (LADAR), Fixed GPS and CCTV camera. The data

captured from both 3D scanner and CCTV camera were analyzed and their respective

advantages and limitations identified. The data collected using these two technologies, were

not processed in numerical examples. The developed model database was designed to include

all sources of data capturing technologies including LADAR and CCTV. The following sub-

sections explain the deployment of each technology, data captured and their respective

limitations.

4.4.1 3D Scanner

The construction site was scanned using Riegl LPM 100 VHS 3D scanner, as shown

in Figure (4-6). It is a long-range scanner with a range of 200 m (Riegl, 2013). Fourteen (14)

jobsite visits were conducted at different time spans and whether conditions to scan the outdoor

and indoor zones. For every outdoor scanning five different positions were utilized to scan the

building. For indoor scanning two positions were utilized in the first floor and a similar

approach was followed in the second floor. The duration to scan the outdoor zones or indoor

zones used to be in the range of four to five hours. Appendix C provides information regarding

scanning schedule and the studied scanning characteristics. Those characteristics are the

vertical and horizontal range and resolution, duration, 3D scanner location, date the planned

137
and actual number of point clouds. The scanner was utilized to scan construction sites at

difference stages. Many scans had to be performed from different positions to get enough

information. 3D scanner facilitates capturing point clouds for the structure under study. It

provides a geometrical presentation of the as built condition of the project. However, it is still

yet expensive compared to the other automated data acquisition technologies utilized for the

purpose of project control. Its setup is similar to surveying equipment such as total station,

which facilitates its usage in construction sites. The increase of deployment frequency for

tracking progress combined with its heavy weight made its utilization a difficult task. The

scanner used in this research and its attachments weighed about 25Kg. Similar, to other

surveying equipment, the setup of the 3D scanner for each position is time consuming.

Scanning position configuration and its selection is crucial for capturing data. Yet, this process

is done manually and based on user judgment.

Figure 4-6: RIEGL LPM 100 VHS LADAR

In addition, the 3D scanner setup attributes, which are the vertical and horizontal range

and resolution, necessitate a tradeoff between time and quality. The less the resolution (i.e. the

step angle used in the scanning process) the better data quality is acquired but requires more

scanning time and more data to process. The area of planning the process of scanning and

138
selecting scanning positions still needs more efforts to optimize it. The integration of this area

with 4D BIM model would facilitate the process of geometrical optimization of those scanning

positions.

For outdoor scanning, there were some limitations in utilizing such technology.

Weather conditions were one of those limitations. The used scanner minimum operating

temperature is -5o C, which prevented utilizing it outdoor after starting the month of December.

It is known from literature that glass and reflecting surfaces cause a problem during capturing

data through the scanner due to the laser beam not returning to the scanner. However, it was

interesting to find out during the tracking of curtain walls progress that this phenomenon could

make more noise to the captured point clouds. It was noticed that the items that were not

scanned by the scanner and were behind the scanner started to appear in the point clouds inside

the building. These add more noise to captured data and would cause problems in case of using

priori object recognition techniques with the point clouds. This is because noisy data could

coincide with the location of objects that were not constructed yet, which would misleadingly

report its progress.

In addition, filtering this data would be problematic due to the ignorance of its

existence. So, it would be recommended to take pictures for the scanned position all around to

know the surroundings that could appear in the point clouds. Outdoor scanning was also

utilized to track the progress of the building’s concrete structure. However, the data captured

was not enough since the scans capture only the floors slabs’ edge and the structural elements

on the perimeter only and nothing about the intermediate structural elements. This problem

will increase in case of higher and bigger area of buildings. In addition, the scans were

conducted on the finalized concrete structure, which represents a portion of the concrete

activities. Reinforcement bars, formwork, concrete pouring, stripping formwork and concrete

139
curing are difficult to be tracked using the 3D scanner. Scanning formwork and temporary

structures is a cumbersome process. Accessibility to top view was not possible, so scanning

from underneath yielded noisy data arising primarily from the supporting formwork and other

obstructions including formwork and rebar in 3D BIM model is a very difficult process and

could change during construction. This would raise too much questions about the feasibility of

such technology to track cast in situ concrete structures.

For indoor scanning, there were also some limitations for utilizing such technology. As

construction progresses, it becomes very difficult to utilize. For example, at early stages to scan

the whole floor, 4 different scanning positions were required. But after the partitions activities

started the floor was divided into 64 rooms and spaces, which mandated 64 different scanning

positions. Additionally, due to the reflection of 3D scanner laser beams, all insulated HVAC

ducts and glass partitions were noisy or not captured. The scanning was conducted after

working hours, for both indoor and outdoor scanning. This procedure reduced the noise in the

captured point clouds due to the site congestion during working hours due to the movement of

labor and equipment. Nevertheless, there was too much noise due to the stationary equipment

and material, which generated shadows and obstructions to the line of sight of the scanner.

The output of the point clouds were manipulated through Riprofile software from

Reigel. The point clouds were exported as (.Dxf) files to be viewed on CAD software. Many

trials were performed to superimpose the 3D scans on the 3D BIM model. However, it was not

successful because the utilized 3D scanner did not have a GPS. In addition, it was not possible

to integrate GPS because the used scanner was old; it was acquired for earlier research in 2002.

Therefore, the scanned point clouds were not georefrenced. Modeling 3D point is a time

consuming process that lacks the capability of capturing material types, textures, and colors of

scanned objects. In addition, it depends on manual manipulation that depends on user’s visual

140
identification. Additionally, it depends on the accuracy and the degree of detail of the BIM

model, which could have many changes and modifications.

4.4.2 Fixed GPS

In this research, GPS device was mounted on a hauling truck to capture the spatial data

along designated hauling roads for the project. This section presents the verification and

validation to the GPS/GIS based tracking method and stochastically forecasting performance.

Figure (4-7 a) shows the GPS unit and Figure (4-7 b) depicts the excavation process of

Genomics building (Montaser et al., 2011).

(a) (b)

Figure 4-7: (a) GPS unit (b) Genomics building (Montaser et al., 2011)

The GPS unit was attached to the hauling truck during the whole excavation process

of the project. The captured data was reported every 2 minutes and it depicts the exact truck

position in terms of latitude and longitude, as shown in Table (4-1). The fleet used in this

project consists of an excavator serving a number of hauling trucks. The soil to be excavated

is sandy clay with density of 2900 lb/CY and a swell factor equals to 0.81 according to the

geotechnical report.

141
 Table 4-1: Sample of GPS unit captured data (Montaser et al., 2011)

Unit # 213
Speed
Date Time Information Location Heading Latitude Longitude
(Km/h)
7:20:18 Moving 60 SE 45.49302 -73.65086
06/05/2010
7:22:18 Moving 60 SE 45.48339 -73.62866
06/05/2010
7:32:18 Moving 2 W 45.4575 -73.64226
06/05/2010
3446 West
Stopped for Broadway AV,
7:34:18 45.45676 -73.64086
3m Montréal, QC,
06/05/2010 CA
7:38:02 Moving 14 SW 45.45631 -73.6405
06/05/2010
7:38:18 Moving 21 SW 45.45556 -73.64119
06/05/2010
7:42:07 Moving 26 SW 45.45491 -73.6322
06/05/2010
7:42:11 Moving 31 SW 45.45456 -73.63248
06/05/2010
7:42:12 Moving 30 SW 45.45449 -73.63253
06/05/2010
7:42:18 Moving 13 SW 45.45419 -73.63275
06/05/2010
7:44:18 Moving 50 NE 45.45263 -73.63131
06/05/2010
7:46:18 Moving 14 SE 45.45737 -73.62521
06/05/2010
7:50:18 Moving 15 SW 45.45668 -73.6218
06/05/2010
7:52:18 Moving 33 SW 45.45692 -73.62621
06/05/2010
7:54:18 Moving 18 NE 45.45154 -73.63451
06/05/2010
7:58:18 Moving 6 SW 45.45332 -73.63744
06/05/2010
Corner of
Elmhurst AV
Stopped for
8:00:18 and Harley AV, 45.45465 -73.63992
2m
Montréal, QC,
06/05/2010 CA
8:02:18 Moving 14 NE 45.45509 -73.64033
06/05/2010
3448 West
Stopped for Broadway AV,
8:04:18 45.45673 -73.6409
9m Montréal, QC,
06/05/2010 CA
8:13:21 Moving 18 S 45.45509 -73.64147
06/05/2010

142
 The total quantity of excavated soil is 14 000 bCY. In early stages of this research, a

spreadsheet application was developed to facilitate the calculation of cycle times (i.e., the time

for loading, hauling, dumping, and returning). GPS data pertinent to the direction and velocity

are used to confirm the identification of each cycle activity. The developed spreadsheet

application calculates the duration of each of the identified activities, and hence the cycle time.

The collected GPS data consist of 104 complete cycles. In addition, Appendix D provides the

methods used to estimate of upper and lower limit productivity planed for Genomics building

excavation process. An average daily value of the extracted durations is used to compare the

actual GPS productivity to the planned estimated productivity, as shown in Figure (4-8)

(Montaser et al., 2012).

Figure 4-8: GPS based daily productivity (Montaser et al., 2012)

The calculated durations were used to develop Probability Density Functions (PDF)

for loading, travelling, dumping and returning times. “Crystal Ball” software was used to fit

143
the data and construct the four functions (Oracle, 2013). For each cycle component, “Crystal

Ball” identifies the probability function that best fits the process at hands. Its output is in the

form of a graph and its associated statistical parameters describing the distribution (e.g. mean

value and standard deviation). “Crystal Ball” identifies the quality or goodness of each fit using

Anderson-Darling method.

The best-fit distribution produced for the loading time was lognormal distribution, with

a mean value of 4.43 min, and a standard deviation of 2.30 min, as shown in Figure (4-9 a).

The best-fit distribution produced for the travelling time was also lognormal distribution, with

a mean value of 10.03 min, and a standard deviation of 3.43 min, as shown in Figure (4-9 b).

Similarly, the best-fit distribution produced for the dumping time was lognormal distribution,

with a mean value of 5.80 min, and a standard deviation of 1.33 min, as shown in Figure (4-9

c). Finally, the best-fit distribution produced for the returning time was a lognormal

distribution, with a mean value of 9.59 min, and a standard deviation of 2.47 min, as shown in

Figure (4-9 d).

The fitted PDFs were then used as input for a simulation model to forecast

stochastically productivity, time and cost of the process. After that, the PDFs were used to

experiment and evaluate different scenarios to forecast in a timely manner corrective actions

and better appreciate the main characteristics of the simulated operations, including the impact

of the uncertainty associated with the model parameters. In the early stages of this research,

simulation was carried out using the “ExtendSim” software, which contains common

simulation tools and instruments (ExtendSim, 2013). The simulation model is shown in Figure

(4-10). It has four main components: the loading, travelling, dumping and the returning

operations.

144
 Loading Travel

Dumping Return
Figure 4-9: Cycle time components PDFs

Figure 4-10: ExtendSim simulation model

In “ExtendSim” software, each operation block has the capacity of handling one unit

at a time. This is sufficient for modeling the loading operation, where trucks are loaded one by

one. However, for other operations like hauling and returning, it is not possible to model several

145
trucks on the same road simultaneously. So, several operations blocks were needed to model

such operations having a capacity of more than one unit. In this respect, the loading operation

is represented by an operation block preceded by a stack block to present the trucks waiting in

queue before loading. The travelling and returning operations are represented by four different

operation blocks. The dumping operation is modeled using two operation blocks, representing

the capacity of the dumpsite. Uncertainty impact is incorporated in the present method, as the

simulation model is developed based on actual GPS data captured during ongoing onsite

operations, which implicitly account for uncertainty. The number of trucks was increased

incrementally in search for a balance with the production of the excavator and the change in

productivity and total earthmoving duration was monitored. In a similar manner, the

productivity and earthmoving operation duration were also calculated for the upper and lower

limit of the planning stage. The results are calculated as shown in Table (4-2).

The results of estimating productivity for different fleet configurations are shown in

Figure (4-11) for both, the planning upper and lower estimates, and the simulation model

results. The results shows that the simulated mean of the production rate falls between the

upper and lower planned limits and closer to the lower limit, which is taking the same trend

line for the average daily GPS data. The results of the simulation model indicate that maximum

fleet productivity is achieved using 11 trucks, while the planning phase deterministic model

results show that maximum productivity can be achieved using only 8 and 9 trucks for lower

limit and upper limit cases, respectively. It is interesting to note that the unaccounted for

uncertainty impact can significantly influence estimated production rates, particularly in urban

areas.

146
Table 4-2: GPS productivity and the process duration (Montaser et al., 2012)

Production Rate
No of Trucks Duration (hr) Duration (Day)
(bcy/hr)
2.00 53.27 254.98 32

3.00 78.37 173.30 22

4.00 102.97 131.91 17

5.00 125.74 108.02 14

6.00 145.14 93.58 12

7.00 162.14 83.76 11

8.00 173.60 78.24 10

9.00 179.43 74.70 10

10.00 181.81 74.55 10

11.00 182.19 74.51 10

12.00 182.29 74.51 10

13.00 182.29 74.51 10

Figure 4-11: Simulation results vs. upper and lower limit planed estimates

147
 In Chapter 6, StroboScope DES engine was used instead of ExtendSim due to many

reasons. For each truck or excavator, the model should be changed manually to add loading

travel route or loading block, which is a troublesome process. In addition, the user should enter

the PDFs manually and the software itself could not be integrated as a part of automated

module. However, StroboScope engine has the capability to be integrated with other modules.

In addition, the change in the number of trucks or excavators could be changed without

reconfiguring the simulation model manually.

4.4.3 CCTV Camera

The use of CCTV camera with time-laps capability allows capturing long time

operations and displaying them in a relatively short time. This capability assists in productivity

measurements and outdoor material tracking. Genomics building contractor installed CCTV

camera at an early stage in the project, as shown in Figure (4-18). Camera type was Sony CP

380 (Sony, 2013). The CCTV camera captured images were very useful specially in validating

the GPS/GIS based tracking. However, like any other technology, CCTV camera has its own

limitations. One of the major limitations is the weather conditions such as rain, fog and poor

illumination, which obstruct the camera’s line of sight. For example, in Figure (4-12), the

image taken at May 14, 2010 was not clear due to the rain at that day. Yet, CCTV camera is

still one of the best solutions to visually check construction jobsites remotely.

148
 May 1, 2010 May 12, 2010

May 14, 2010 June 1, 2010

Figure 4-12: Sample of CCTV camera captured data

4.5 Tablet PC Automated Data Acquisition Technologies

Tablet PC is considered one of the most important components in the developed model.

This is because it hosts many automated data acquisition and visualization technologies. The

tablet PC used in this research is a rugged Panasonic Toughbook®H2 handheld tablet PC, as

shown in Figure (4-13). It has 10.1" sunlight-viewable LED screen, fast Intel® Core™i5

vPro™ processor and 1.7GHz with Turbo Boost up to 2.7GHz Intel®. In addition, it has 40

GB hard drive, Smart Cache 3MB, 4GB SDRAM (DDR3-1333MHz), twin batteries (6.5 hours

with both batteries), ports such as USB, serial and Ethernet and IP65 sealed all-weather design

(Panasonic, 2013). The data captured from the tablet PC camera was not utilized in the

numerical examples presented in the thesis.

149
 Figure 4-13: The tablet PC utilized during the field studies

The tablet has a docking station to be connected to the tablet in case it is used as laptop.

The tablet is connected through the docking station to the 3D scanner for setting the scanning

attributes and storing the captured data. The tablet has wireless capabilities such as Bluetooth,

3G mobile internet network and Wi-Fi. In addition, barcode and RFID readers are embedded

in the tablet. However, it could only read high frequency RFID tags, which have a maximum

read range of 1m. So, Intermec mobile RFID reader IP30 was acquired to work with ultra-high

frequency tags that were deployed in the developed model (Intermec, 2013). GPS could be

embedded in the tablet PC. However, the tablet used in this research did not have GPS. So,

Trimbel GeoXT mobile GPS was acquired for outdoor localization. The tablet PC, mobile

RFID reader and mobile GPS communicate and exchange data via Bluetooth standard

communication protocol. The following sections explain the utilization of tablet PC

capabilities for data acquisition.

150
4.5.1 Camera

Multimedia information, including digital images and videos are acquired from site to

capture construction progress on site. Images and videos captured with a digital camera on a

daily basis are stored into the developed tablet PC database. Figures (4-14) show indoor images

at different construction time span. Similar to other technologies, the camera also has some

limitations, as its location and other related attributes are subject to site personnel judgment.

August, 30th, 2010 November, 15th, 2010

January, 27th, 2011 April, 14th, 2011

Figure 4-14: Sample of Indoor camera captured data

Then, the experience, camera orientation and overlap between images are considered

influential factors to be considered. Illumination is another factor that constrains camera

151
deployment in certain situation such as electromechanical shafts with poor lighting. The tablet

PC is used to capture the images whether from indoor or outdoor zones. The captured images

get new attributes from the developed model, which are location and date/time. It could be

compared visually to the pictures obtained from the 4D BIM model, as shown in Figures (4-

15) and (4-16). Instead of manual methods of interpretations, object recognition and

photogrammetry could be utilized combined with the 4D BIM for data processing.

Figure 4-15: 4D BIM image vs. actual image for outdoor zone

Figure 4-16: 4D BIM image vs. actual image for indoor zone

152
4.5.2 Barcode

Barcode is used for identifying items. Most of the electromechanical items delivered

to Genomics building were tagged with Barcode, as shown in Figure (4-17). Identifying those

items and retrieving their respective data is a benefit for tracking progress of the jobsite. If the

barcode reader captured an item, its current zone could be identified accurately based on the

location of the user who scanned that barcode.

Figure 4-17: Items with barcode inside Genomics building jobsite

4.5.3 Mobile GPS

In this research, Trimble GeoXT GPS was used to identify site personnel location in

outdoor zones. It is a high performance GPS receiver combined with a rugged handheld

153
computer with a powerful 520 MHz processor, 128 MB RAM, and 1 GB of onboard storage,

as shown in Figure (4-18 a). TerraSync software was installed on the handheld GPS. It collects

and updates geographical data, as shown in Figure (4-18 b). Differential GPS (DGPS)

correction should be applied to the captured GPS data. DGPS requires one or more additional

receivers, called base stations or reference stations (Trimble, 2013), which are located at known

points. Data collected at base stations is used to determine GPS measurement errors and

compute corrections to these errors. Errors are corrected with DGPS either in real time or

during post-processing (Trimble, 2013). Both methods were performed in this research.

For real-time DGPS correction, the base station calculates and broadcasts the error for

each satellite as each measurement is received, which enables corrections while the user is in

the field and collects GPS data. Real time DGPS corrections are generated and broadcasted in

real-time by privately or self-owned GPS base stations, or by a wide range of government

agencies. Real-time DGPS sources include external beacon and radio sources, as well as

Satellite Based Augmentation Systems (SBAS) such as VRS networks in Canada. VRS

networks use multiple base stations to calculate the DGPS corrections that are then delivered

to the user from a Geostationary satellite (SBAS) or from a radio or cellular phone (Can-Net,

2013).

For example, corrections generated by a VRS network are commonly broadcasted over

an internet server. A VRS network uses data from several base stations to provide corrections

that are generally more accurate than corrections from a single base station. Factors that affect

real time DGPS accuracy include how often the corrections are updated, how far the users is

from the base station, and whether the coordinate system used by the correction source matches

the coordinate system used by the GPS receiver (Can-Net, 2013). VRS base station was 50 Km

far from Genomics building jobsite. Therefore, cellular connectivity was added to the Trimble

154
GeoXT GPS handheld via the TDL 3G cellular modem, which was acquired with the GPS, as

shown in Figure (4-18 c). It was connected to the GPS via wireless Bluetooth. TDL 3G provides

continuous network/internet access to real time VRS corrections. In this research, Concordia

University acquired the VRS corrections license service for one-month to experiment with it

through TDL 3G cellular modem (Trimble, 2013).

(a)

(b)

(c)
Figure 4-18: GPS hardware and screen shot

For post-processed DGPS correction, the collected GPS data is transferred to the tablet

PC via wireless Bluetooth. Since, the construction jobsite was in Concordia University Loyola

campus, wireless Wi-Fi connection was available. Measurements from the base station are

downloaded and GPS data processed through GPS Pathfinder Office software version 4.10 that

was earlier installed on the tablet PC, as shown in Figure (4-19). Typically, post-processed

DGPS uses only one base station. Factors that affect the accuracy of post-processed DGPS

include; the type of receiver and antenna used at the base station, the distance between the base

155
station and the user location where the data was collected, the accuracy of the base station

position, and the logging interval at the base station (Trimble, 2013).

Figure 4-19: Post-processed GPS data using GPS PathFinder software

The GPS Pathfinder software provides all the functionality needed to manage and

process data collected using GPS. It provides the tools needed to correct, view, and edit Global

Positioning System (GPS) data collected in the field, and to export it in a format suitable for

GIS, CAD, or database system. Both the wireless Wi-Fi and GPS PathFinder software were

available on the tablet and the GPS PathFinder converted the data into Microsoft Access

database format. Accordingly, this method was integrated with the developed model.

Therefore, the data is first transferred wirelessly from the GPS to the tablet PC and then

processed through GPS PathFinder software. At the end, it is saved on database format (.mdb)

156
to be used later by the developed model to identify the user outdoor location, as shown in Table

(4-3).

Table 4-3: Sample of GPS corrected data

Position ID Longitude Latitude GPS Date GPS Time Avg. Vert. Prec. Avg. Horz. Prec.

1 -73.63999 45.45696 3/17/2011 10:29 AM 0.46281 0.28646

2 -73.63999 45.45695963 3/17/2011 10:29 AM 0.46281 0.28646

3 -73.63999 45.4569598 3/17/2011 10:29 AM 0.46281 0.28646

4 -73.63999 45.45695919 3/17/2011 10:29 AM 0.46281 0.28646

5 -73.63999 45.45695942 3/17/2011 10:29 AM 0.46281 0.28646

6 -73.63999 45.45696021 3/17/2011 10:29 AM 0.46281 0.28646

7 -73.63999 45.45695986 3/17/2011 10:29 AM 0.46281 0.28646

8 -73.63999 45.45696048 3/17/2011 10:29 AM 0.46281 0.28646

9 -73.63999 45.45696043 3/17/2011 10:29 AM 0.46281 0.28646

10 -73.63999 45.45696036 3/17/2011 10:29 AM 0.46281 0.28646

11 -73.63999 45.45696018 3/17/2011 10:29 AM 0.46281 0.28646

12 -73.63999 45.45695952 3/17/2011 10:29 AM 0.46281 0.28646

13 -73.63999 45.45696015 3/17/2011 10:29 AM 0.46281 0.28646

14 -73.63999 45.45695994 3/17/2011 10:29 AM 0.46281 0.28646

15 -73.63999 45.45696013 3/17/2011 10:29 AM 0.46281 0.28646

Then, the developed model converts the location from latitude and longitude to the

BIM model local coordinates system for outdoor real time visualization. The data gathered in

157
one of the field experiments, after the differential correction was performed, were then saved

into a (.kml) file to be visualized in Google Earth, as shown in Figure (4-20). The Navisworks

4D BIM model was exported to the same (.kml) file for visualization.

Figure 4-20: User route during outdoor filed experiment

4.5.4 Mobile RFID

The RFID hardware components used in this research are RFID mobile reader (IP30),

RFID encapsulated tags (IT67), RFID smart label tags and RFID labels tag printer (PM4i), as

shown in Figure (4-21). RFID hardware could collect data in dirty, harsh and hazardous

conditions. For example, the encapsulated RFID tag used, could work in temperatures ranging

from -40o C to 66o C and could be attached using screws, rivets, double-sided adhesive strips

or a variety of other methods. Regarding its memory size, it has a capacity of 512 bit. In

addition, RFID mobile readers could work under similar harsh conditions such as in

temperatures ranging from -15o C to 50o C. They are protected from dirt, dust, oil, other non-

158
corrosive material, and splashing water. Readers’ connectivity could be through Bluetooth,

Ethernet or Wi-Fi and can host applications written in JavaScript. The read range for

encapsulated tags is 5m and 3m for label tags. The encapsulated RFID tag costs approximately

$5 per tag. Passive RFID tags used in these experiments were printed RFID labels, which cost

2 cents each. The tag labels and the printer are shown in Figure (4-21) (Intermec, 2013).

b)RFID encapsulated tag

a)RFID mobile reader

c)RFID label tag d)RFID tag printer

Figure 4-21: RFID hardware

Figure (4-22) depicts screen shots from RFID reader during field operation. RFID

reference tags were fixed at known locations during the construction of the Genomics building.

Additionally, RFID tags were also attached to material on the jobsite for localization and

tracking purposes, as shown in Figure (4-23). Figure (3-24) depicts two different test beds and

its respective setup and pictures. The output of the RFID reader file is in the form of text file

(.txt). Sample of this data is presented in Table (4-4). This file is transferred wirelessly to the

tablet PC and saved as Microsoft Access database formats (.mdb) for processing by the

developed model to identify the user indoor current location. Chapter 5 presents detailed

experimental work conducted utilizing that RFID hardware for localization and material

tracking using different algorithms and deployment protocols.

159
 Figure 4-22: RFID reader screen shots

Figure 4-23: RFID data acquisition from reference tags and material tags

160
Figure 4-24: Different test beds setup and pictures

161
 Table 4-4: Sample of RFID captured data

Serial # Tag ID RSSI Date Time

1 B0 -54.1 1.3.2011 12:08:49

2 AD -56.5 1.3.2011 12:08:49

3 B0 -54.4 1.3.2011 12:08:50

4 AD -57.4 1.3.2011 12:08:50

5 B0 -53.5 1.3.2011 12:08:50

6 AD -56.4 1.3.2011 12:08:50

7 B6 -59.6 1.3.2011 12:08:51

8 AD -58.7 1.3.2011 12:08:51

9 A8 -57.8 1.3.2011 12:08:51

10 B0 -57 1.3.2011 12:08:52

11 AD -57.9 1.3.2011 12:08:52

12 A8 -58.8 1.3.2011 12:08:52

13 B3 -55.4 1.3.2011 12:08:53

14 B3 -55.9 1.3.2011 12:08:53

15 B3 -56.1 1.3.2011 12:08:54

16 B3 -56.1 1.3.2011 12:08:55

17 B7 -58.7 1.3.2011 12:08:55

18 B7 -59.2 1.3.2011 12:08:56

4.6 Summary

This chapter presents the field study conducted on a construction jobsite. This study

encompasses a wide range of automated site data acquisition. The used technologies are Closed

Circuit Television (CCTV) camera, 3D scanner, GPS (mobile and fixed), mobile RFID,

barcode, and tablet PC. This chapter presents the findings of the field study highlighting the

challenges and limitations of utilizing each technology individually on construction jobsites.

162
This study focuses also on the data captured from each technology to be integrated in the

developed model database. The study encompasses excavation and earthmoving operations,

concrete framing, curtain walls, HVAC installations, and internal finishing works. These

studies led to exploration of the likely advantage of integrating different technologies and using

them to estimate project progress.

163
5 Chapter 5: EXPERIMENTAL STUDIES OF PASSIVE RFID

INDOOR FOR LOCALIZATION

5.1 General

This chapter addresses a range of factors that are vital for the deployment of UHF

passive RFID technology in indoor construction projects. The RFID deployment protocols

introduced in this chapter present guidelines for the near-optimal utilization of RFID

technology on construction jobsites. It utilizes both construction site and controlled lab studies

to identify patterns and settings of deployment parameters. In this research, a number of UHF

passive RFID tags are distributed to capture indoor spatial data with the aid of a mobile RFID

reader. Each passive RFID tag is used as a reference point with a known location

(LANDMARK) within a predefined zone. The indoor areas are divided into exclusive zones,

and each zone is spatially covered with a number of passive RFID tags.

This Chapter has three main sections, as shown in Figure (5-1). The first addresses

coarse-grained localization for zone detection using K-Nearest Neighbor (K-NN) algorithm. It

focuses on different deployment protocols for reference tags and provides guidelines for RFID

utilization in indoor zone detection. The second part of this chapter presents a fine-grained

localization through a two-step algorithm to automate the process of location estimation and

material tracking. The two-step algorithm uses Received Signal Strength Indicator (RSSI) for

signal measurements with two different algorithms (Trilateration and Proximity). Finally, the

third part of this chapter discusses in depth factors that influence the accuracy of utilizing

proximity algorithm for indoor localization.

164
 Figure 5-1: Chapter 5 overview

5.2 Zone Identification Using K-NN Algorithm

This study combines the use of UHF passive reference RFID tags and K-Nearest

Neighbor algorithm (K-NN) for indoor course-grained zone detection of the RFID mobile

reader user. The study aims at providing a set of guidelines for the deployment of UHF passive

RFIDs in building construction, with respect to:

• Tag deployment topologies,

• Distance between deployed tags,

• Stationary vs. moving reader user,

• Zone configuration and,

• Impact of moving target speed.

165
 A total number of 10038 data sets for on-site and controlled lab experiments were

acquired and analyzed. The total number of logged data sets was 5269 for construction site

experiments and 4769 for the controlled experiments (Razavi et al., 2012). K-NN is a memory

based classification of tasks. In this research, the time stamp is used as a metric to define the

closest neighbors. Therefore, the K-nearest neighbors is the last k reads that occurred prior to

the most recent read. The K-NN algorithm is then; (1) finds the K closest reference points that

have been read, and according to their log time, and (2) classifies the location of the reader into

one of the predefined zones based on the majority of readings. Table (5-1) depicts the process

of applying K-NN algorithm to the captured data.

Table 5-1: The process of applying K-NN algorithm

Real Time Calculated

Tag ID Date Time Tag Zone Detection
Actual Zone Zone At K = 5

B0 1.3.2011 12:08:49 2 1
AD 1.3.2011 12:08:49 1 1
B0 1.3.2011 12:08:50 2 1
AD 1.3.2011 12:08:50 1 1
B0 1.3.2011 12:08:50 2 1
AD 1.3.2011 12:08:50 1 1 2 1
B6 1.3.2011 12:08:51 2 1 1 0
AD 1.3.2011 12:08:51 1 1 2 1
A8 1.3.2011 12:08:51 2 1 1 0
B0 1.3.2011 12:08:52 2 1 2 1
AD 1.3.2011 12:08:52 1 1 2 1
A8 1.3.2011 12:08:52 2 1 2 1
B3 1.3.2011 12:08:53 1 1 2 1
B3 1.3.2011 12:08:53 1 1 2 1
B3 1.3.2011 12:08:54 1 1 1 0
B3 1.3.2011 12:08:55 1 1 1 0
B7 1.3.2011 12:08:55 4 1 1 0
B7 1.3.2011 12:08:56 4 1 1 0
B7 1.3.2011 12:08:57 4 1 1 0

166
 For correct zone detection, value equal to zero is placed. However for incorrect zone

detection, value equals to one is placed. The error rate is derived from Equation (5.1).

`ab c d e cc e e
Error Rate = (5.1)
f g ab c d h

5.2.1 Field Study

The field study was carried out during the constructing of Genomics building,

described in Chapter 4, for the five chosen deployment protocols, with two to six different

scenarios for each. Six distinctive active work areas on the third floor were defined as zones.

Passive RFID tags, in the form of labels, were installed on the walls in each zone. The zones

had different sizes and shapes; the tags were placed at the same elevation, which is 1.5m. The

studied area covers 132 m2 of the overall floor area. The areas of the zones 1 to 6 are 52m2,

12m2, 12m2, 32m2, 12m2 and 12m2 respectively.

5.2.1.1 Tag Deployment Topologies

A set of experiments were conducted to analyze the effect of the reference tags’

deployment topology on the error of zone detection as shown in Figure (5-2). Particular

attention was placed on utilizing tags on two topologies: corners and mid-walls. The first

topology experiment focused on corner-deployment with 275 data sets and 4 zones having an

approximate total area of 108m2 (Figure 5-2, left hand side). The second Topology experiment

focused on mid-wall deployment, with 846 data sets and 6 zones with an approximate area of

132m2 (Figure 5-2, right hand side). The experimental results showed that the lowest error rate

was achieved when K=40. At that K value, the results indicated that mid-wall deployment

performed better than corner-deployment as shown in Table (5-2). This could be attributed to

the issue of interference at the corners of the structure.

167
 Figure 5-2: Two deployment topologies

Table 5-2: Error rate for the two deployment topologies (Razavi et al., 2012)

Different number of K for K-NN algorithm

Data
Scenarios
Sets #
5 10 15 20 25 30 35 40

Corner deployment
275 39% 31% 27% 28% 31% 33% 35% 35%

Mid-wall deployment 846

39% 36% 33% 33% 31% 32% 29% 28%

5.2.1.2 Closeness of Passive Reference Tags

Identifying the near optimum distance between tags required to cover a zone is the

ultimate objective for studying this parameter. The impact of the distance between adjacent

deployed tags is studied for two scenarios. Scenario 1 represents the deployment of 26 tags

168
using corner deployment topology with an average distance of 1.5-2 times the Read-Range

(RR) and 846 data sets. Scenario 2 adds one tag between any two tags of scenario 1, the added

tags deployed mid-wall. It has overall 40 tags, 589 data sets, and four zones with an area of

108m2, Figure (5-3). The results presented in Table (5-3) indicate that a closeness of 1RR

results is better detection than 1.5 RR.

Figure 5-3: On-site scenarios for tag closeness

Table 5-3: Error rates of tag deployment distance (closeness) on site

Different number of K for K-NN algorithm

Data Set
Scenarios
#
5 10 15 20 25 30 35 40

1.5 RR 846 39% 36% 33% 33% 31% 32% 29% 28%

1RR 589
44% 35% 30% 29% 27% 24% 23% 22%

169
5.2.1.3 Stationary vs. Moving Reader Agent

To assess the impact of the movement of site personnel with the mobile RFID reader

on the results, a number of data logging scenarios were tested on-site. Four scenarios with a

stationary reader in a pre-defined stationary location and a scenario with moving site personnel

were conducted and compared. A total number of 338 data sets were used for the stationary-

reader scenarios and 275 data sets for the moving-reader scenario. The arrows in Figure (5-4)

represent the reader’s stationary locations during data logging. The results indicated that the

data acquired using a stationary reader can be more accurate than that generated using a moving

reader, as shown in table (5-4).

Figure 5-4: Different scenarios for a stationary reader agent

170
 Table 5-4: Error for stationary reader locations vs. a moving reader

Different number of K for K-NN algorithm

Scenarios Data Set #
5 10 15 20 25 30 35 40

Fixed Location1 62 40% 21% 16% 16% 19% 24% 29% 32%

Fixed Location 2 53 17% 15% 15% 17% 17% 17% 17% 17%

Fixed Location 3 39 15% 5% 13% 21% 23% 26% 28% 15%

Fixed Location 4 88 31% 31% 30% 33% 34% 38% 40% 41%

Fixed Location 5 96 47% 35% 25% 24% 15% 11% 16% 19%

Moving reader 275 39% 31% 27% 28% 31% 33% 35% 35%

Fixed-reader location 2 represents the most robust solution, with the lowest error rate

over the entire range of the study. It is interesting to note that the highest error rates were found

with fixed location 4, where the reader is standing in the middle of the smallest zone of

2RR×1.5RR. This perhaps can be attributed to the relatively short distances from the surfaces

and corners to which tags are mounted.

5.2.1.4 Zone Configuration

Zone configurations may vary through the project cycles and can be related to the

sequence of the construction schedule. A single experiment with 1361 data sets was used for

three scenarios with different zone arrangements, as presented in Figure (5-5). In these

hypothetical zone arrangements, zone 2 has three different dimensions: 2RR × 1.5 RR, 2RR ×

3 RR, and 2RR × 4.5 RR. Table (5-5) presents the experimental results for the various zones

configurations in this case study. It should be noted that zone configuration effectiveness is in

direct relationship with the deployed RFIDs’ reading range. It is clear that the third scenario,

where the defined area for zone 2 is larger than the other two, has the lowest error rate. This

171
result can be credited to a lower interference effect from the nearby zones when the zone

dimension is significantly larger than double the reading range.

Figure 5-5: Scenarios for different zone configurations

Table 5-5: Error rates for different zone configurations

Different number of K for K-NN algorithm

Scenarios Data Set #
5 10 15 20 25 30 35 40

2RR × 1.5 RR 1361 29% 29% 28% 24% 25% 26% 27% 27%

2RR × 3 RR 1361 27% 26% 26% 24% 25% 26% 27% 28%

2RR × 4.5 RR 1361 21% 22% 22% 22% 23% 23% 24% 24%

172
5.2.1.5 Speed of Moving Reader

When the site personnel moves at different speeds, there may be different impacts on

the overall accuracy of the acquired site data. The next set of experiments assessed this impact

by considering two scenarios; a fast moving and a slow-moving reader. In both scenarios, the

moving target/reader walks along the trajectory path, shown in Figure (5-6), which has an

overall length of 58m (from point A to point B). The fast-moving reader walked along the path

in approximately 2.5 minutes, and the slow-moving reader passed through the same path in

approximately 6.5 minutes. A total number of 1954 data sets were used in this set of

experiments. The acquired results for these scenarios are presented in Table (5-6). The results

indicate that a slow-moving reader yields a lower error rate. This result also supports the

outcome of a stationary vs. a moving reader agent.

Figure 5-6: The trajectory of the moving target (reader)

173
 Table 5-6: Error rate for the scenarios of a slow moving vs. a fast-moving reader

Different number of K for K-NN algorithm

Scenarios
5 10 15 20 25 30 35 40

Slow Moving (6.5 minutes) 17% 16% 12% 13% 11% 10% 9% 10%

Fast Moving (2.5 minutes) 44% 35% 30% 29% 27% 24% 23% 22%

5.2.2 Laboratory Experiments

Experiments were also conducted in a controlled lab environment to better study the

effects of the closeness of reference tags and of RSSI. These experiments were conducted in

the construction automation lab in the Engineering and Visual Arts building at Concordia

University as shown in Figure (5-7).

Figure 5-7: Concordia university construction automation lab

5.2.2.1 Closeness of Passive Reference Tags

Four scenarios for tag closeness were set up, with distances of 0.5 RR , 1RR , 1.5 RR

, and 2 R, which were tested with 2752, 982, 544, and 491 data sets, respectively. The area

under study has the dimension 4RR×4.5 RR, divided into four equal zones with dimensions of

174
2RR×2.25RR and tags that were mounted on the walls. Table (5-7) summarizes the outcomes

of this experiment.

Table 5-7: Error of tag deployment distance (closeness)

Different number of K for the K-NN algorithm

Data Set
Scenarios
#
5 10 15 20 25 30 35 40

0.5RR 2752 0.4% 0.7% 0.9% 1.2% 1.4% 1.7% 2.0% 2.3%

1RR 982 1.0% 1.9% 2.5% 3.5% 4.2% 5.0% 3.7% 2.4%

1.5RR 544 5.5% 6.3% 7.4% 9.0% 10.5% 12.2% 12.3% 12.9%

2RR 491 2.7% 3.9% 4.9% 6.7% 7.8% 9.2% 10.4% 11.8%

It was found that a tag closeness of 0.5 RR yields the best overall detection rate,

followed by 1RR, 2RR, and 1.5 RR scenarios. It was also observed that in a controlled

environment the overall performance of the detection algorithm is best for a K value of 5, with

higher K values, associated with higher error rates. This pattern remained the same for all K

values in the controlled experiment. In the 1.5RR deployment scenario, three tags were

installed in each zone, two of which were less than the 0.5RR distance from the hypothetical

boundary of these zones. In the 2RR scenario, however, only two tags were used in each zone,

with a relatively greater distance from the zones’ boundary.

5.2.2.2 Studies on RSSI

The Received Signal Strength Indicator (RSSI) is a measurement of the power

present in a radio signal received by an antenna. The RSSI value is the relative received

signal strength in a wireless environment, in arbitrary units. A higher RSSI value (or the

less negative in some devices) indicates a stronger signal. According to the power laws,

175
the RSSI absolute value has an inverse relationship with distance. When radio signal travels

in the space, their strengths diminish with the increase of the distance. According to this

power law, the received signal strength at the receiver can be used for estimating the

distance between the transmitter and the receiver. In the ideal environment, the RSSI at the

receiver can be modeled by the following equation (Zhang et al., 2010), which represents

the path loss model, Equation (5.2):

lm
ij = k ∗ , 3 (5.2)
$∝

Where:

Pr is the received signal strength at the receiver.

Pt is the signal strength at the transmitter.

d is the distance between the transmitter and the receiver.

c is a constant.

∝ is the signal loss coefficient, which depends on the environment .

The accuracy of RSSI-based methods is highly dependent on the accuracy of the path

loss model. The model explained earlier is easy to manipulate, but it generates a significant

error rate because the RSSI is complex in nature. The complexity is caused by various

uncontrolled environmental effects such as multipath fading, shadows, and terrain,

surrounding materials and obstacles. Therefore, other techniques are needed to assure a

suitable level of accuracy. Figure (5-8) presents the results of a set of experiments

conducted in a controlled laboratory environment to acquire a path loss model. The

176
experiments depict the average of RSSI for every 10cm increments in distance between tag

and reader. The experiment has been repeated for four different tags. The model was

developed using 6704 data sets. Each data set consists of a number of signals captured at a

specific distance. The path loss model is regression model, which helps to depict trends in

the existing data. In order to choose the right regression model for the data, three different

regression types of logarithmic, linear and second-degree polynomial were performed and

the results compared and the results presented in Figure (5-9). In this research the linear

regression model was used as path loss model (i.e. relationship between reader RSSI and

the distance from the tag D), as depicted in Equation (5.3).

RSSI = -6.182 D - 32.682 (5.3)

Distance (m)
-28
0 0.5 1 1.5 2 2.5 3 3.5

-33

-38
RSSI

-43

-48

B1 B2
-53 BC C5

-58

Figure 5-8: Average RSSI vs. distance for four tags

177
 Distance (m)
-20
0 0.5 1 1.5 2 2.5 3 3.5

-25
Average Linear (Average) Log. (Average) Poly. (Average)

-30 y = -6.799ln(x) - 41.323

R² = 0.8425

-35
RSSI

-40

y = 1.2411x2 - 10.402x - 30.22

-45 R² = 0.9097

-50

y = -6.182x - 32.682
-55
R² = 0.8839

Figure 5-9: Regression analysis

178
5.3 Location Identification Using Two-Step Algorithm

A two-step algorithm was presented to automate the process of identifying user

location in real time estimation and identifying material location in near real time, refer to

Chapter 3. It uses RSSI for signal measurements and two different localization algorithms

(Trilateration and Proximity) are used. The main components of the developed two-step

algorithm are outlined in Figure (5-10) (Montaser and Moselhi, 2012 d and 2013 a).

Figure 5-10: Two-step algorithm (Montaser and Moselhi, 2013a)

For validating the two-step algorithm and demonstrating the use of its components,

experiments were conducted. A set of experiments were conducted during the construction of

Genomics building at Concordia University. Five test beds were setup at different time spans

and different locations on construction jobsite. Carrying out the tests at different time spans

was required to prove that the proposed algorithm is feasible during the changing environment

179
of construction operations. Other experiments were conducted in the Construction Automation

Lab, at Concordia University. Test bed 6 was setup in a lab environment. Table (5-7) shows

the characteristics of each test bed, including the total number of data sets collected, date,

location, test bed area and number of tags used in the test bed. Figures (5-11) and (5-12) show

the setup and images of all test beds where the rectangles refer to the locations of attached

reference tags. The error is derived from Equation (5.4), which is the distance in meters

between the estimated and actual locations.

Error = o(p − pk)" − (= − =k)" (5.4)

Where

(X, Y) is the actual predefined location

(Xc, Yc) is the calculated location using Trilateration or Proximity algorithm

Table 5-8: Characteristics of test beds (Montaser and Moselhi, 2013a)

Test Bed # Test Bed 1 Test Bed 2 Test Bed 3 Test Bed 4 Test Bed 5 Test Bed 6

Date 01/12/2010 03/12/2010 08/12/2010 01/3/2011 14/04/2011 24/12/2011

Jobsite (3rd Jobsite (2nd Jobsite (3rd Jobsite (2nd Jobsite (3rd Lab
Location
Floor) Floor) Floor) Floor) Floor) Environment
Total Number of
Predefined 18 18 18 15 15 45
Locations

Total No of data set 418 494 451 729 438 5915

Covered Area (m2) 75.24 75.24 75.24 108 120 61

No of Deployed Tags 24 24 24 25 33 25

Average Covered
3.135 3.135 3.135 4.32 3.6363 2.44
Area(m2/#)

180
Figure 5-11: Test beds 1, 2 & 3 setup and site pictures Case Study I

181
 Figure 5-12: Test beds 6 setup Case Study II

Figure (5-13) shows graphical display of the errors in the calculated locations of the

mobile user. The triangles represent the user’s previously marked locations where the user

stands and starts capturing RFID signals. The circles and squares represent, respectively, the

calculated location using trilateration and proximity algorithms respectively. Figure (5-14)

shows graphical comparison between the two algorithms. There are five sensed reference tags.

Four of them related to the user zone and the fifth one is far away from the user zone. Using

trilateration algorithm automatically selects the intersection area of most circles and due to the

fifth reading is away from that intersection it is not be considered in calculations. However,

proximity algorithm considers all readings and do not have this mechanism so the results are

more affected by noise from reference tags that far from the user zone. Figure (5-15) shows

tracking material labeled C8 utilizing user location numbers 14, 15 and 18.

182
Test Bed 3
Test Bed 5

Test Bed 6

Figure 5-13: Localization error for both algorithms

183
 Figure 5-14: Comparison between trilateration and proximity algorithm

Figure 5-15: Material tracking using trilateration and proximity algorithm

184
 Figure (5-16) shows box plot for all test beds. Summary of statistical analysis for both

trilateration and proximity algorithms is displayed in Figure (5-17). It depicts the accuracy of

in identifying locations of the user obtained from all test beds. Similarly, a statistical analysis

was performed for material location identification and the results are summarized in Figure (5-

18) for both algorithms. The developed two-step algorithm yields 100% accuracy for zone

identification of the user and tracked material in all test beds. Cumulative Distribution Function

(CDF) was used to show the localization error and to measure the localization performance at

a given confidence level as shown in Figure (5-19). The CDF also indicated the error

accumulation in material location is due to the dependency on user location. In case of

trilateration algorithm, the error accumulation was increased rapidly while in case of proximity

algorithm, the error accumulation was steady.

Trilateration Proximity

Figure 5-16: Box plot for each test beds results (Montaser and Moselhi, 2012 d)

185
 ! "
# $
% &
'

( )& * " + *&

( )& * " + *&

( )& * " + *&

! "
# $
% & ,
'

( )& * " + *&

( )& * " + *&

( )& * " + *&

Figure 5-17: Summary of statistical analysis results for user localization

186
 ! "

! "
# $
% &
'

( )& * " + *&

( )& * " + *&

( )& * " + *&

! "

! "
# $
% & ,
'

( )& * " + *&

( )& * " + *&

( )& * " + *&

Figure 5-18: Summary of statistical analysis results for material localization

187
 2 3 " ) 4 &* - .&" & .&" &

# #

! 1
- .&" & / &
- .&" & & 0
"# / / &
"# / & 0

Figure 5-19: CDF for estimated error (Montaser and Moselhi, 2013 a)

The results also indicated that trilateration algorithm is relatively more accurate than

proximity algorithm for both user and material localization. However, the trilateration

algorithm suffers from drawbacks such as the dependency on path loss models (location-

environment dependent models), which is not robust enough to represent the characteristics of

radio waves and its interference in a dynamically changing construction environment. In

addition, the computational time required for identifying location using trilateration algorithm

is much more than that of the proximity algorithm due to the mathematical complexity of

trilateration algorithm, which gives advantage to proximity algorithm in real time localization.

5.4 Deployment Protocols Using Proximity Algorithm

This section presents a lab experimental work conducted to facilitate the use of RFID.

As, it addresses the impact of metal media proximity to RFID tags, the reasonable duration for

188
data capturing, number of RFID tags employed and the distance between them. RSSI was used

as the main attribute for signal measurement to process the captured data. Proximity algorithm

was used to analyze the capture RSSI to calculate the RFID reader location. Five hundred and

fourteen (514) experiments were conducted. Figure (5-20) depicts an area of 3m by 3m (Test

Bed) that was setup for the experiments. There are a number of low cost reference tags fixed

on the perimeter of the test bed. The Distance between reference tags is X. Three different

numbers of RFID reference tags are deployed in the experiment 16, 8 and 4 with distance

between tags equals to 0.75m, 1.5m and 3m respectively. The test bed is with distance equal

to Y from metal media. This distance is varied from zero to 3m in increments of 0.75m. Data

is collected using mobile RFID reader from nine different predefined locations from within the

zone at each experiment. The user at the predefined location operates the mobile RFID reader

and captures the signals received from the reference tags (Montaser et al, 2013).

Figure 5-20: Test bed setup (Montaser et al., 2013)

This process is then repeated at a set of time intervals, referred to here as Duration.

Data is collected for different durations of 15, 30, 45 and 60 seconds at every point of the nine

locations. The different scenarios of each set of experiment are summarized in Table (5-9). The

collected data were captured from nine locations for each test bed at four time intervals. Results

189
of the data analysis are studied under four main categories: duration, number of tags, locations

of tags and metal interference.

Table 5-9: Different characteristics of the experiment (Montaser et al., 2013)

Metal
Experiment 15 30 45 60 Total # of RFID
X (m) Y (m) Proximity
# sec Sec Sec Sec Tags Group
Group
1 0.75 3 A 1
2 0.75 2.25 B 1
3 0.75 1.5 C 1
4 0.75 0.75 D 1
5 0.75 0 E 1
6 1.5 3 A 2
7 1.5 2.25 B 2
8 1.5 1.5 C 2
9 1.5 0.75 D 2
10 1.5 0 E 2
11 3 3 A 3
12 3 2.25 B 3
13 3 1.5 C 3
14 3 0.75 D 3
15 3 0 - -

Figure (5-21) shows the errors in all 15 test beds over the 4 data capturing durations

studied. As a general trend observed in all the experiments, there is a peak in error when the

data is captured at zero distance between the test bed and metal media as observed in test beds

5 and 10. In test bed 15, there was zero distance between the test bed and metal media, and

there were only 4 tags present with two tags not functioning due to the proximity to the metal.

The readings of test bed 15 did not result in a meaningful translation of finding the location of

objects.

190
 Mean
1

0.95

0.9

0.85
Error (m)

15
0.8
30
0.75 45
0.7 60

0.65

0.6
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Experiment #

Standard Deviation

0.53

0.48

0.43
Error (m)

0.38 15
30
0.33
45
0.28 60

0.23

0.18
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Experiment #

Figure 5-21: Errors observed under different durations in all test beds

Figure (5-22) shows the best location to capture data is at the center of the test bed

where there is the equal distribution of signals received from the RFID reference tags. The

191
accuracy of localization decreases as the reading locations diverge from the center of the zone.

Results of the experiment revealed that the least error prone sets of readings were captured

while the duration for data collection was minimal (15 seconds). As the time window of data

capturing increased, the error in identifying location increased as well. It is concluded that

shorter duration time leads to less amount reflections of radio waves from the surrounding, as

shown in Figure (5-23).

Mean Standard Deviation

1.2 0.3
1 0.25
0.8 0.2

Error (m)
Error (m)

0.6 0.15
0.4 0.1
0.2 0.05
0 0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
Location # Location #

Figure 5-22: Errors observed in the 9 different predefined locations

Mean Standard Deviation

0.8 0.33
Error (m)

Error (m)

0.78 0.32
0.76 0.31
0.74 0.3
10 15 20 25 30 35 40 45 50 55 60 65 10 15 20 25 30 35 40 45 50 55 60 65
Duration Duration

Figure 5-23: Errors observed in all test beds with different durations

The effect of metal medium was one of the other factors under study. In the test beds

group E that was at zero distance to the metal object, the errors are substantially greater than

the other groups, as depicted in Figure (5-24). Figure (5-25) indicates that the more tags was

192
deployed in the experiment, the higher the error with the main reason being the interference of

many RFID tag waves existing. However, when there are less numbers of tags the risks of not

yielding results was higher especially with the presence of metal media with zero distance to

tags.

Mean Standard Deviation

0.45
0.4
Error (m)

Error (m)
0.84
0.35
0.79
0.3
0.74 0.25
A B C D E A B C D E
Group Group

Figure 5-24: Errors observed in all test beds due to metal proximity

Mean Standard Deviation

0.82 0.34
0.8
0.33
Errot (m)

Errot (m)

0.78
0.76 0.32
0.74
0.31
0.72
0.7 0.3
1 2 3 1 2 3
Group Group

Figure 5-25: Errors observed in all test beds due to the number of deployed tags

5.5 Summary

The experimental results of applying K-NN algorithm for coarse-grained zone

detection indicated that the speed of the moving target/reader has a measurable impact on the

accuracy of the acquired data. A slow-moving reader yielded lower error rates than fast-moving

readers. It was also found that zone configuration effectiveness has a direct relationship to the

193
deployed RFID read-range. This result is credited to the lower interference effects of adjacent

zones. The results indicated that mid-deployment performs better than corner-deployment.

Also, the results obtained from the on-site and controlled experiments show different trends

for the impact of K value on the error rate. The higher the K value, the higher was the error

rate for the controlled experiments, while in the on-site experiment, the higher K values

resulted in lower error rates.

Regarding applying trilateration and proximity algorithms, the results shows mean

error of 1.0 m and 1.9m for user location identification and material tracking using trilateration

algorithm, respectively. The results shows mean error of 1.9 m and 2.6m for user location

identification and material tracking using proximity algorithm, respectively. The developed

two-step algorithm detects user location and material zones with 100% accuracy. The main

limitations of the trilateration algorithm are the need to generate path-loss model for each type

of tag used, the variability associated with deployment of tags, the uncontrolled influence of

noisy signals and potential interference from equipment and/or vehicles located between tags

and between tags and the mobile reader.

Further studies were conducted using proximity algorithm. These studies include the

effects of four factors on localization using RFID tags in 15 different settings. The results prove

that least erroneous settings to capture data proves to be the in the test bed furthest away from

the metal object, where the reader is located at the center of test bed, has the least duration of

data capturing and the least number of surrounding tags. The best duration was found to be the

15 second in the test bed with the least number of tags; as the short amount of time to capture

data did not allow creation of a lot of interference among the emitted signals. Within each test

bed, errors occurred most at points where the received signals were not well distributed in a

360 degree vicinity of the data capturing point. It means that the center point of each test bed

194
resulted in lowest errors and the points located on the extremities led to the highest errors.

Finally, metal objects were found to have major impact on the accuracy of the captured data;

to the level where reliable values for errors could not be calculated in the test beds attached to

metal objects. In summary, the results of the experimental study and related findings are

expected to provide guidelines to the users of RFID technology for localization in indoor

construction. It also demonstrates the potential of utilizing short range RFIDs in location

estimation and material tracking with a cost-effective manner for indoor construction jobsites.

195
 6 Chapter 6: COMPUTER IMPLEMENTATION

6.1 General

This chapter presents the computer implementation of the developed model. It also

highlights the developed system’s frameworks, modules, algorithms and the interaction

between its components. It describes the user interface, which is designed to enable the user to

collect and store data from construction sites and subsequently query reports. Figure (6-1)

depicts chapter overview.

! "

Figure 6-1: Chapter 6 overview

6.2 Developed System

The developed model was implemented in prototype software, which has similar

structure to the developed model, as described in Chapter 3. The tablet PC data acquisition is

implemented and applied through BIM technology. BIM is capable of recognizing building

objects in its fixed asset hierarchy (Family - Type - Object) while being a user-friendly for

196
creating a building’s indoor and outdoor zones. It links those objects to the zone that hosts

them by relating an object’s ID to a zone’s ID. Different software developers such as Autodesk,

Bentley and CATIA have applied the BIM concept. None of these systems’ providers provides

any BIM web application. All of them are standalone applications installed on the user

computer that could be connected to internet.

Each has its pros and cons; however, Autodesk Revit customization capabilities have

been significantly extended over the past few years. Revit Application Programming Interface

(API) allows users to program with any .NET compliant language such as VB .NET and C#

.NET. Revit has thus been selected here to be customized so that the integrated tablet PC data

acquisition system be applied in its API and added as add-In.

The developed Revit add-in is named (BIM+). BIM+ provides the user with 4D BIM

real time visualization for the current zone through utilizing GPS or RFID or even using user

manual selection. It also supports the user in capturing the onsite progress data through utilizing

progress templates or tablet PC embedded data acquisition technologies. The data analysis and

processing framework was implemented as a web-based system that is named (Control+).

Control+ receives the project captured data whether from BIM+ or the independent automated

data acquisition technologies. Then, it stores the data in the central database. Afterwards,

project team members can access Control+ via internet to view generated progress reports.

Web-based systems generally do not suffer from the software incompatibility problems that

are present in other standalone applications. Any computer that has a web browser such as

Internet Explorer or Firefox can access such a system to query and view project progress

reports.

197
6.2.1 System Platform

A typical system application usually consists of three elements: presentation,

application logic, and data. Presentation focuses on interacting with the user. Application logic

performs calculations based on developed algorithms and determines the flow of the

application execution. Data elements manage information that is shared between the users. The

most often used systems are two-tier and three-tier architectures. For example, in web-based

systems, it refers to both architecture as client/server. A client is defined as a requester of

services and a server is defined as the provider of services. Two-tier client/server architecture

groups presentation and application logic components on the client computer. It provides data

sharing through Internet connection (Quatrani, 2002). The structure of three-tier architecture

involves the presentation-tier, the application logic/middle-tier, and the data tier. Adjacent tiers

are connected through the Internet. By contrast, with two-tier architecture, a middle tier was

added between the user interface in the client computer and the database in the web server. The

presentation-tier components manage user interaction and request services. The event handler

embedded in the user interface responds to the user’s request and triggers the middle-tier data

processing components. The middle-tier components perform queuing, executing, and

requesting to the database. ODBC (Open Database Connectivity) connects the data processing

component with the Database to facilitate data entry and retrieval. The major data processing

runs on the middle-tier server (Larman, 1998).

The independence of a middle-tier from presentation and data tiers offers numerous

benefits including:

• Multi-language support: system components can be developed using different

programming languages when needed.

198
• Centralized components: components are centralized for easy development, maintenance,

and deployment.

• Efficient data access: the numbers of database connections are minimized since the

database faces the components only in the middle-tier.

• The middle-tier components can be secured centrally using a common infrastructure.

Access can be granted or denied on a component-by-component basis. The database is

hidden from the client.

• Simplified access to external resources: access to external resources, for example, to other

applications or databases, is simplified.

Taking the above into consideration, two-tier standalone architecture was chosen for

BIM+ system development. However, for Control+ three-tier client-server architecture was

chosen for the system development, as shown in Figure (6-2). BIM+ is two-tier where is the

first tier is established in Revit software. BIM+ tier one represents the user interface and the

application logic. An add-in is developed inside Revit platform to enable the user from

achieving the desired tasks. Tier two in BIM+ is the data tier where the tablet PC database

resides. This tier is responsible for storing data captured using the tablet PC automated data

acquisition and other connected mobile devices such mobile RFID and mobile GPS. The data

captured is extracted from the tablet PC database to be processed through the add-in developed

logic inside Revit to be presented to the user on the Revit interface.

199
 Figure 6-2: Developed system architecture

Control+ is designed to analyze and process all project captured data whether it is

captured using the independent automated data acquisition technologies or through BIM+

system. Tier one in Control+ represents the presentation tier, which is mainly a web-based

interface for the system. The user interacts with the web-based interface through predesigned

quires regarding project progress. The interface results are in a form of predesigned web-based

reports answering the user’s quires. Control+ tier two and tier three reside on the web server.

Tier two is Control+ logic where all developed algorithms and procedures in the data analysis

and processing framework are implemented. The system central database resides in tier three,

where all the captured data for project progress is stored.

200
 Database provides a platform to organize, store and retrieve projects’ planned and

actual performance data in a logical and efficient manner. The Database Management Systems

(DBMS) queries the stored project data from the database using SQL (structured query

language) to generate different management reports for control purposes. Therefore, Microsoft

Access has been chosen as the DBMS for BIM+ tablet PC database and Control+ central

database, as shown in Figure (6-3). Microsoft Visual C# .Net programming language was used

in the system development due to its compliance with Revit APIs, Microsoft Project and

Microsoft Access.

The Unified Modeling Language (UML) is chosen to develop the prototype software

in this research. UML is a language for specifying, visualizing, and constructing of software

systems, where objects are diagrams and documents that comprise the system models. The

Class diagram is the diagram that represents concept, data, and relationships of a system. The

attributes of the classes are the parameters needed to execute the operations. The operations of

the classes execute the tasks (Quatrani 2002). System functionality is the way a system would

behave. It is documented in a use-case diagram. This diagram illustrates the system intended

functions, surroundings, boundaries, internal and external relationships. It includes the

components of actors as well as those of the use-case model, the system boundary, the use-case

communication, and the use-case relationships. Actors are not part of the system. They

represent any user or any external software that interacts with the system. Use-case diagram

describes the dialogues between the system and its actors and represents the functionality

provided by the system (Larmen 1998). .

201
Figure 6-3: ER diagram for the designed Microsoft Access central database

202
 Sequence diagrams of a use-case are developed to represent the implementation

scenarios of the developed prototype system. A sequence diagram is an interaction diagram

that shows how operations are carried out and what and when messages are sent. It is organized

according to implementation time. The time progresses as going down the diagram. The objects

involved in the operation are listed from the left to the right according to their accumulated

time in the message sequence. Sequence diagrams give dynamic views of a system. They focus

on the messages involved in completing a single process and present the mechanism for actions

inside the system (Quatrani 2002). The following sections describe the implementations,

validations, and the uses of the prototype software. It aims at demonstrating the capabilities of

the developed system, according to the description of the developed model in previous

chapters. Typical implementation scenarios, represented by developed classes, use-case and

sequence diagrams, are introduced. In addition, software screen shots and steps are explained.

6.2.2 BIM+

This section presents the developed automated tool BIM+, which was developed using

the “Visual C# .NET” in Revit’s API. Revit was selected to be customized so that the

automated data acquisition using the tablet PC can be applied in its API and integrated as add-

In utility. BIM+ was implemented to the selected case study, described in Chapter 4, as a proof

of concept and for testing the system. The 3D BIM model has all of the parameters and

attributes for building zones and objects. Each parameter and attribute is associated with a

unique ID to avoid conflicts when dealing with data exchange between Revit and tablet PC

database. The generated 3D BIM model feeds the developed BIM+ data, such as the number

of spaces, the area of each space, the object families and family types, in order to identify

objects inside each space, and later facilitate integration of the 3D model with the construction

schedule. Data input by the user(s) and/or data converted or calculated is stored in the tablet
PC database tier. BIM+ is able to extract and store data inside the tablet PC database. Based

on this ID uniqueness feature, all objects’ characteristics could be exported from the 3D BIM

model and imported to the data tier.

In addition to the unique ID that each object has, BIM objects also have two important

characteristics that are family and type. Family represents the main description of objects such

as wall, door, window, etc. While, type represents specific kind of a family object such as

internal wall 1 hour fire rated or wooden door single flush panel. The developed system utilizes

these two aspects of BIM objects to categorize the whole building. Figure (6-4) depicts the use-

case diagram for BIM+. It shows system boundary, functionality and actors. There are three

actors in BIM+, two of them are users and the third is external software. The first actor is the

experienced user, who is familiar with the project 3D BIM model and the construction schedule

of the building being modeled such as project manager or the project scheduler. The

experienced user authentication reveals BIM+ full customization and configuration functions.

The experienced user is responsible for linking the Microsoft Project file to BIM+ and maps

the 3D BIM objects to project activities through utilizing the control visibility options.

Visibility options have three different use-cases that are show by activity, show and hid

activities manually and show activities by date. “Show activities by date” use-case is

considered the 4D BIM model that was customized in Revit. The visibility options given to the

experienced user facilitate the process of modifying or changing activities objects in case of

any design modification or change order. In addition, the experienced user links the floor plans

and area plans with the dividing activity that was extracted from the MS project schedule.

204
 Figure 6-4: BIM+ use-case diagram

The second actor is the regular user. This user is the site personnel responsible for

utilizing BIM+ for data acquisition purposes. Hence, the user utilizes “show activities by date

(4D BIM)”, “get user zone manually” and “get user zone automatically” use-cases. If the user

gets the current zone automatically then two use-cases are generated that are “Use RFID for

Indoor” or “Use GPS for Outdoor”. Whether the user gets the current zone manually or

automatically, “data acquisition” is the next use-case to be used. Then, the data acquisition use-

case branches into two use-cases that are automated data acquisition and progress templates

205
data acquisition. The final actor is Microsoft Project software, which is external software to

the system boundaries. BIM+ should first be customized and configured for the current project

in hand. The experienced user does this step and should link the MS project schedule file to

BIM+. This is done according to the sequence diagram shown in Figure (6-5).

Figure 6-5: Linking Microsoft Project file to BIM+ (sequence diagram)

206
 The user first activates the developed add-in menu in the Revit screen and subsequently

selects from a pull-down menu “Set MS Project File” and completes the process depicted in

Figure (6-6). BIM+ closes the file explorer dialog and puts the selected file path into displayed

writable textbox. The user presses the “Done” button so, BIM+ checks for if the file exists,

which is part of process verification. Then, it saves the new MS project path into the system

configuration file to use it as default MS Project path and closes the change MS Project form.

The second step, performed by the experienced user, is to map 3D BIM objects to its

respective activity, as in the sequence diagram shown in Figure (6-7) and the process operations

highlighted in Figure (6-8). In creating group of elements, the user can link one object or a set

of objects to one activity as diagrammatically shown in Figure (6-9). Prior to that linkage, MS

Project schedule must be activated and verification is performed to make sure that the schedule

has more than one activity. If the MS project file have list of activities, it displays two buttons

“Finish” and “Close”. Then, the system allows the user to select multiple objects from the 3D

Revit model and link them to one of the activities being displayed. After selecting multiple

objects, the user selects the MS Project activity from a pull-down menu “Choose Group Name”

and completes the process depicted in Figure (6-9). Upon completion of the linking process,

the object(s) inherits the attributes of the associated activity such as activity name, early start

and early finish, which are then used for generating the 4D BIM model. This process is repeated

until all 3D Revit objects are mapped to associated MS Project activities.

207
Figure 6-6: Linking Microsoft Project file to BIM+
Figure 6-7: Mapping 3D Revit objects to project activities (sequence diagram)
Figure 6-8: Mapping 3D Revit objects to project activities (Part 1)
Figure 6-9: Mapping 3D Revit objects to project activities (part 2)

211
 Each object in Revit, referred to as element, has a set of attributes such as ID, family,

type. This attributes vary from object to another and they are either generic or specific object.

A new attribute was added to all Revit objects to facilitate the linking process, as shown in

Figure (6-10). It is referred to it as “Group”, which represent one object or a set of objects. The

user selects a set of objects such as supported deep excavation piles, as shown in Figure (6-

11), then; the name of “Group” is made identical to the activity name. The sequence of this

process was designed and implemented to be from outside to inside and from top to bottom.

Upon finishing one group, objects of that group will be hidden to provide access to the other

objects. The mechanics of linking identified objects to the relevant scheduled activity is

performed interactively as shown in Figure (6-12).

Figure 6-10: The added attribute “Group” to Revit object

 Figure 6-11: Selected objects to create Group

Figure 6-12: Assigning MS Project activity to Group name

Controlling visibility of activities is crucial in BIM+. Therefore, three different options

were developed to identify which activities are visible and which are hidden. To activate the

visibility controls, the user activates the add-in menu and selects “Show/Hide Groups”. Then,

BIM+ checks that at least one activity is linked to the 3D Revit objects. Otherwise, the system

pops up an error message instructing the user to map the 3D Revit objects to the MS Project

activities first. If this check is false then the system displays “Group Visibility Control” form.

The form contains three different options, as shown in Figure (6-13). The user has to select one

of the three options.

213
Figure 6-13: Show and hide activities options
 If the user selects “Show/Hide Based on Activity” then the system shows under that

option a drop down list with all activities that were linked to 3D Revit objects. The user selects

one activity that required to be displayed. The system displays the selected activity on the

current view and hides all other activities. In Figure (6-14), the Basement Gypsum Board Walls

activity was shown. If the user selects “Show/Hide Based Manual” then the system displays a

form containing all displayed activities under the visible list and all hidden activities in other

list. The user selects the needed activity to be displayed and press move or selects the needed

activity to be hidden and press move. In addition, the user could select to move all activities

from one list to the other and vice versa. The user presses done button to apply the above

actions to Revit current view. Figure (6-15) sows that all roof and upper roof activities were

hidden from the view as per the selection in the lists. Showing specific activity or activities

become very handy to the experienced user specially, in the cases of design modification or

change orders. Therefore, the user can isolate the activity objects, modify it, and link it again

to the MS Project activity.

The third option is to show/hide activities by date. This option represents the

methodology for applying the 4D BIM concept inside Revit without the need for third party

software. This use-case could be utilized by the experienced user and the regular user. This

use-case considered the first step for the regular user before capturing on site progress data.

When the user selects s the option of “Show/Hide Based on date”, two calendars appear. The

first calendar to specify the start date and the second calendar for the end date. The system

applies the algorithm explained in Chapter 3 for showing and hiding activities on specific date

range.
Figure 6-14: Show specific activity
Figure 6-15: Show and hide activities manually

217
 The system displays the result on the current Revit view, as shown in Figure (6-16).

Figure (6-17) presents images from the 4D Revit BIM model at different locations inside

Genomics building. The final customization and configuration for the project inside BIM+ is

the zone configuration process. For each indoor floor there should be two plans, one is the area

plan and the other is the floor plan. The link between the plans is the floor dividing activity.

Figure (6-18) depicts the area plan and the floor plan for the basement floor. The experienced

user should create the two plans for each floor. In addition, the experienced user should create

area plan for outdoor areas and roof floors, as shown in Figure (6-19). Then, the user manually

starts assigning zones for each plan whether it is floor plan or area plan. This process could not

be automated since it differs according to the user preferences.

After creating all plans and its respective zones, the experienced user starts linking

every floor plan and area plan with its corresponding dividing activity. The user opens the add-

in menu and selects the “Link Room to Area Using Activity”. BIM+ displays form contains

three drop down lists. The first one contains all Rivet floor plans and the second drop down list

has all MS Project activities. The final drop down list consists of all Rivet area plans. The user

selects the floor plan, area plan and the respective dividing activity. BIM+ searches in Revit

configuration file for the selected floor/area/activity and check whether these items have been

linked before or not. If it was linked, then the system send an error message for the user

notifying that item has been selected before. Figure (6-20) depicts the developed C# .Net source

code in Revit API to link the dividing activity with the floor and area plans. The system saves

this linked combination, as shown in Figure (6-21). The process is repeated for each floor. By

the end of this process, the project would be customized on BIM+ for data acquisition.

Afterwards, the rule of the regular user starts, which is the site personnel responsible for

capturing on site progress from indoor and outdoor zones.

Figure 6-16: Show and hide activities by date (4D BIM)
Figure 6-17: Pictures from the generated 4D Revit visualization
Figure 6-18: Area plan vs. floor plan for the basement floor at Genomics building
Figure 6-19: Area plan for outdoor zones at Genomics building

222
 // Link Button Click event handler;

private void btnLink_Click(object sender, EventArgs e)

{
xmlEntity.InsertXmlEntities(new xmlEntity() { ActivityName =
(cboActivities.SelectedItem as MSPActivity).Name, AreaName =
(cboAreas.SelectedItem as CustomView).Name, FloorName = (cboPlans.SelectedItem as
CustomView).Name });
lstXmls = xmlEntity.GetXmlEntity();

MessageBox.Show("Done");
}

// Insert the linked entity to xml file;

public static void InsertXmlEntities(xmlEntity currententity)

{
System.Configuration.Configuration config =
Helper.MSProjectHelper.GetConfig();
string strLinkFilePath = Environment.GetFolderPath(
Environment.SpecialFolder.ProgramFilesX86)+config.AppSettings.Settings["LinkFile
Path"].Value;

XDocument xRoot = XDocument.Load(strLinkFilePath);

XElement eleroot =
xRoot.Descendants("Links").FirstOrDefault();

XElement ele = new XElement("Link");

XAttribute attActivityName = new XAttribute("Activity",
currententity.ActivityName);
XAttribute AreaName = new XAttribute("Area",
currententity.AreaName);
XAttribute FloorName = new XAttribute("Floor",
currententity.FloorName);

ele.Add(attActivityName, AreaName, FloorName);

eleroot.Add(ele);
xRoot.Save(strLinkFilePath);
}

• Xml file Sample:-

<?xml version="1.0" encoding="utf-8"?>

<Links>
<Link Activity="first one" Area="Level 1" Floor="Level 1" />
</Links>

Figure 6-20: C# .Net source code for the process of linking dividing activity to plans
Figure 6-21: Linking dividing activity to floor plan and area plan
 The regular user can then start the data acquisition process by defining the 4D Revit

model date range. The user at that point can visualize a 4D BIM model for the status of the

project at the current date. The user activates the add-in menu and selects “Go to Position

Automatically”. In this case, the user has to specify whether the current zone is indoor or

outdoor, as shown in Figure (6-22). If indoor, then the system starts communicating via

Bluetooth with the RFID reader. The user has to operate the RFID for 15 seconds and close it.

Since, all RFID reader captured data is saved as text file on specified location on the RFID

reader memory, BIM+ accesses this data and stores it in the tablet PC database. According to

the captured data, BIM+ identifies which floor this data belongs to. Also according to the

current date, the system specifies whether the dividing activity for this floor started or not and

then specifying whether to use the area or floor plan for this floor.

BIM+ applies the first step in the two-step algorithm through applying proximity

algorithm to identify the user current location coordinates. After identifying the coordinates

and the used plan, the system identifies the user current zone. BIM+ creates two rivet sections

for the current zone. The created sections are named automatically according to the following

order (Floor name – Zone name – Date – Time). Then, BIM+ orients the Revit 3D view into

the created sections. The system displays the list for all elements exists in selected zone that

belongs to activities in progress only. The user has the option to select single object from the

list. Then, the user can highlight this object in the 3D view or highlight the objects belong to

the same type in the view or all the objects belong to the same family in the view. Since, this

object (objects) belongs to an activity in progress. Then, BIM+ retrieves from the tablet PC

database its respective progress template steps and displays it on the Revit view, as shown in

Figure (6-23).
Figure 6-22: Getting user zone automatically options
Figure 6-23: User indoor Location visualization and data acquisition

227
 The user interactively utilizes the developed progress templates to capture the work

performed at the object level. The user also continues with the automated data acquisition using

the technologies embedded in the tablet PC. The system checks first, if there was an object that

the user did not specify its status using progress template. If this was the case, the system

highlights those objects in the objects list to finish them first. If this was not the case, the system

generates the data acquisition form. The data acquisition form has six buttons that are sound

recording, video recording, image, hand written notes, 3D scan and barcode reading. All of

them, except the 3D scan button, once the user pushed its button it respective technology works

and starts capturing data. When the user pushes the 3D scan button the 3D scanner software

starts working.

Similar procedure is followed if the user is in outdoor zone and selects the outdoor

option to extract the user coordinates from the mobile GPS. However, if the user selects this

option for the first time then some configuration needs to be done, as shown in Figure (6-24).

A pop up window appears requesting the user to enter the coordinates of the building

benchmarks in global coordinates (i.e. Latitude and Longitude) and the equivalent in local

coordinates (i.e. building local coordinates system x, y). This input provides BIM+ with the

transformation parameters required to convert the GPS data from global coordinate system to

the building local coordinate system. Figure (6-25) depicts the user location in outdoor zone.

To add more practicality to BIM+, the user also can select the current zone manually, as shown

in Figure (6-26) and (6-27).

Figure 6-24: User outdoor location
Figure 6-25: User outdoor Location visualization and data acquisition

230
Figure 6-26: Selecting user location manually (Part 1)
Figure 6-27: Selecting user location manually (Part 2)

232
6.2.3 Control+

Control+ is the part of the system that is responsible for data processing and analysis.

It utilizes three-tier architecture. Tier one is the presentation tier, where the user interacts with

web-based graphical interface. Tier two, the application logic, consists of different integrated

modules. Tier two embraces modularity in design, which enhances the system capabilities and

data reuse. Tier three represents the central database of the system. Control+ utilizes Service

Oriented Architecture (SOA) as its software design methodology. It is a design pattern based

on structured collections of discrete software modules, known as services, which collectively

provide the complete functionality of a large software application (Velte et. al., 2009). SOA

allows simultaneous use and easy mutual data exchange between the three tiers and third party

software such as Revit and Microsoft project.

This web-based service is accessible over standard Internet protocols, independent of

platforms and programming languages, and has two main roles. The first is the service

provider, which is the web server. The web server creates a web service and publishes its

interface and access information over the World Wide Web. The web service in that case is

tier one of the system. The second role is the service consumer, which is the software user.

The user accesses the software via internet using desktop computer or mobile device such as

Smartphone, tablet PC or Personal Digital Assistance (PDA). Windows Communication

Foundation (WCF) technology is deployed to implement the SOA web-based software. WCF

is a runtime and a set of APIs in Microsoft .NET framework. WCF is a tool used as intermediate

level between the web application and the web server. It also processes the main functionality

of system.

233
 Figure (6-28) shows the login page for Control+ web page. The user should enter the

user name, password and project number. Based on those entries, the authentication level for

the user is decided (i.e. what reports are available and what are not available). The web page

heading consists of five main menus that are Project, Resource, BIM/Walkthrough,

CCTV/Time-Lapse and Central DB. Under the Project menu five items, which are General

Information, Project Status, Weather Conditions, Updated Schedule and Activity Status. If the

user selects the “Weather Conditions” then the user has to selects the date. The system retrieves

the project location from the central database and access online weather database to retrieve

the weather data, as depicted in Figure (6-29). Control+ generates web-based reports for user

preview based on the data processed and retrieved from the central database and data entered

by the user such as selecting date or specifying certain activity. However, Project menu is part

of this research future work.

Figure (6-30) shows the Resource menu that consists of Material, Equipment and Labor

items. Material item consists of two main reports that are retrieved from the central database

and user date selection. Material Delivery report is generated from the RFID gate data

acquisition system. While Material Location report is generated from RFID and GPS for

outdoor zones and from RFID for indoor zones. Labor item generates labor time through the

data captured from RFID gate data. Similarly, the equipment in site report is generated under

the Equipment item. Resource menu is part of this research future work. However, Truck+

module under Equipment item is fully functioning as will be described below.

234
Figure 6-28: Control+ login page

235
Figure 6-29: Control+ project menu – Weather Conditions

236
Figure 6-30: Control+ resource menu

237
 Truck+ module has been developed to automate the process of estimating, monitoring

and forecasting productivity of hauling trucks. This prototype is based on algorithms of the

GPS/GIS based tracking method and stochastically forecasting performance explained earlier

in Chapter 3. The utilization of SOA design method allows Truck+ to simultaneously use and

easy mutual data exchange between the system modules, central database and other third party

software such as Stroboscope DES engine. The user launch Control+ web-based interface first.

Then, it redirects the user to Truck+ GIS user interface. The GIS web-based user interface is

within the presentation tier, which handles the system communication with the user. Truck+

system architecture is shown in Figure (6-31).

The GIS component of the software consists of ArcGIS bas map, ArcGIS template for

Silverlight and ArcGIS APIs for Silverlight. ArcGIS Base Map is free map service provided

by ESRI (Economic and Social Research Institute) containing the street and satellite layers.

ArcGIS Template for Silverlight is a free template created using Silverlight technology and

contains all basic functionality for any map viewer system such as map panning, zooming and

navigation at any part of the selected GIS map. ArcGIS APIs for Silverlight are many DLLs

(Dynamic-link library), which add many controls that doesn't exists in the default web browser

such as map controls that contain the base map (street and satellite map) and scale bar. It utilizes

the powerful mapping, geocoding, and geoprocessing capabilities provided by ArcGIS server

(ArcGIS, 2013). It is also used to facilitate communication with GPS spatial data and

graphically representing them in the GIS map.

238
 Figure 6-31: Truck+ system architecture

In addition, Microsoft Silverlight is used as a web plug-in to give the web browser

more functionality. This tool is implemented to enable ArcGIS APIs from providing the

mapping functionality of GPS points. Finally, Crystal Reports technology is used to develop

and generate custom reports layout. It facilitates graphically design data extraction from

different data sources such as MS Access central database, ArcGIS base map and user inputs.

Figure (6-32) shows Truck+ sample of class diagram.

239
 Figure 6-32: Truck+ sample of class diagram

For validating Truck+ system and demonstrating the use of its components, case study

was conducted to make use of the construction of Genomics building at Concordia University

that was described earlier in chapter 4. Truck+ Graphical User Interface (GUI) is designed to

be user friendly, straightforward data entry and eliminate redundancy of data. GUI accepts both

graphic and non-graphic data entry. The graphic data entry is done through the GIS provided

drawing tools. The non-graphic data entry is through a set of buttons, dialog windows and radio

buttons. The main screen of Truck+ web page layout consists of one main view in the center

displaying the world map, as depicted in Figure (6-35). The top right corner of the screen

displays magnifying glasses button, overview map button, map legend button and hide/display

map visualization options button. In left side of those four buttons, there are three radio buttons

for map visualizations options, which are street view, topographic view and imagery view. The

lower right corner of the screens shows the scale bar and display-navigation controls. The upper

left corner has two buttons the open project button and the help button.

240
Figure 6-33: Truck+ web page layout

241
 Figure (6-34) depicts the use-case diagram for Truck+. It shows system boundary,

functionality and actors. There are four actors in Truck+, two of them are users and the other

two are external software. The first actor is the experienced user, whom is familiar with the

project such as project manager or the person planned the earthmoving operation. The

experienced user authentication reveals Truck+ full functions. This user is able of drawing the

spatial boundaries of the earthmoving operation (i.e. loading, travel, dumping and return). This

user draws the boundaries one time and it is saved later in the central database. The second

actor is the regular user, which could be any stakeholder with interest in the earthmoving

operation.

Figure 6-34: Truck+ use-case diagram

This user shares Truck+ other functionalities with the experienced user, which are open

project, change project, generate reports and stochastic forecasting use-cases. The other two

actors are external to Truck+ system boundaries, which are external software (i.e. Stroboscope

DES engine and Oracle Crystal Ball software). Both software are used when the user trigger

the stochastic forecasting function in Truck+. When the user press open project button, a dialog

box appears. It requests the user to enter the project number, as shown in Figure (6-35).

242
 Figure 6-35: User enters project number

Upon entering the project number, the GIS interface (Tier1) automatically requests

from the system logic (Tier 2) connection to the central database (Tier 3) for loading the project

respective data. The system searches in the central database for the project number in the

projects entity. If the system found an exact match, then it starts searching for the GPS points

in the truck GPS file entity. In parallel, the system searches also if there are sketch drawn

previously for the spatial boundaries. Then, the system extracts the GPS data and displays it in

the GIS map as a separate layer (blue points) for graphical representation. If the spatial

boundaries were found the system displays it also on separate layers. Truck+ zooms

automatically to the extents of the GPS points, as shown in Figure (6-36). In the middle of the

left side, a form appears to facilitate user interaction for data entry. The user can change the

project by pressing on the change project button. Then, a similar procedure to open project is

followed.

243
Figure 6-36: GPS points layer

244
 Under the change project button, there are four buttons (L, D, T and R), which are the

drawing tools for drawing loading polygon, dumping polygon, travel poly line and return poly

line respectively. Those four buttons are only active for the experienced user and not active for

the regular user. Therefore, the experienced user pushes the (L) button and starts drawing

loading area on its respective layer. The interface enables the used to draw polygon sketch on

the map and waits till the user finishes the polygon. The user can zoom and pan the map during

this process. Afterwards, the interface disables this button and sends the coordinates of the

loading polygon sketch to the central database to be saved, as shown in Figure (6-37). To

change the loading polygon sketch, the user has to select the polygon and then press the reset

button to redraw it. A similar procedure is followed during sketching the dumping area

polygon.

For sketching the travel road, the user presses the (T) button and instead of drawing

polygon, the user drawing dashed poly line. This line is transformed into polygon according to

the road that it was drawn on. If it was street road then the centerline and width is extracted

from the GIS map. However, if it was not on street then a message box appears for the user to

enter the poly line offset from both sides. Then, the created polygon is saved in the central

database. In the same way, the return road poly line is drawn. Each sketched polygon or poly

line layer is added to map legend in lower right corner, as presented in Figure (6-37). These

layers are utilized to classify each GPS point according to the algorithms mentioned in Chapter

3. At this stage, the user should select one of the radio buttons, which to choose whether

reporting for specific data or reporting from project start to specific data. The project start data

is extracted from the central database.

245
Figure 6-37: Earthmoving operation spatial boundaries sketches

246
 According to the user selection, a calendar appears to select dates. Then the user is

required to enter the ACWP for each day, which represents the actual expenditure for

earthmoving operation to perform earned value analysis, as shown in Figure (6-37). The system

checks if the user finished drawing loading, dumping, travel and return sketches. Then, the

system filters the GPS points that belongs to date that the user selected. If the user selected

“Report To specific Date” then system gets all GPS points from project start date to the desired

date. If the user selected “Report for specific Date” then the system gets all GPS points for the

desired date only. The system generates the report for each hauling truck and determines the

cycle’s times and its respective components.

Tabular format reports have been used to present each truck cycle time and it respective

components, as shown in Figure (6-38). Figure (6-39) depicts the number of cycles each truck

did in front of each date, which is utilized in conjunction with the truck capacity to calculate

the excavated material percentage complete. Then, earned value analysis is conducted to

determine the operation performance indices, variances from planned and different

deterministic forecasts scenarios for the operation as shown in Figure (6-40). Graphical format

reports are developed to display information at varying degree of detail such as CPI per day,

as shown in Figure (6-41).

247
Figure 6-38: Generated cycle times and its respective components

Figure 6-39: Generated tabular report for hauling trucks cycles

248
 Figure 6-40: Generated tabular report for earned value performance indices

Figure 6-41: Generated graphical report for Cost Performance Index (CPI) on daily basis

To stochastically forecast performance for cyclic repetitive operations, Stroboscope

DES engine was used. However, DES requires Probability Density Function (PDF) for each

operation. So, Oracle Crystal Ball was utilized to generate best fit PDF of the cyclic operations.

Therefore, Truck+ was integrated with Stroboscope and Oracle Crystal Ball to stochastically

forecast progress of earthmoving operations. In addition, this integration was used also to

experiment the different fleet configuration impact on productivity, time and cost of the

operation. Since, Stroboscope DES engine and Oracle Crystal Ball are external software

utilized by the system. Then, it should be installed on the user computer. All runtime in this

case is done on the user computer not the web server. Also, a typical Stroboscope earthmoving

249
DES model is created to simulate the process. This model is saved in a predefined path on the

user computer too. However, the user can change the model location and browse for the new

location.

The variables in the DES model are the number of hauling trucks, truck capacity,

number of loaders, the quantity of remaining excavation, truck cost per hour and the loader

cost per hour. The system retrieves all this data from the central database. To generate PDF for

each cycle time component, the user specifies the data range (i.e. start date and finish date).

Then, the system extracts the cycle time components from the central database and filters it to

the specified data range. In addition, there is an option to exclude dates from this data range.

This enables removing specific time periods during which exceptional conditions are known

to have prevailed and are not likely to be repeated. Oracle Crystal Ball is utilized afterwards to

generate best fit PDF for each cycle time component based on the captured data specified date

range.

When the user presses simulation reporting, the results is send to an excel file on the

user computer desktop. The results are the operation forecasted loader utilization, truck

utilization, production rate, time, total cost and the unit cost of excavated material. Since, the

forecasting process was done stochastically through PDFs. Then, the results is not crisp values,

it is in the form of average, standard deviation, maximum and minimum values, as shown in

Figure (6-42). In case, the user wants to experiment with the different fleet configuration but

with the same PDFs. Then the user could enter the minimum and maximum number of trucks

and loaders in the form. Similarly, the same procedure is followed but with different scenarios,

as shown in Figure (6-43).

250
Figure 6-42: Stochastic forecasting overview

251
Figure 6-43: Simulating different fleet configuration

252
 Afterwards, the user can go back to Control+ to view the other progress reports. Figure

(6-52) shows the components of the “BIM/Walkthrough” menu. It gives the user the capability

to preview the data captured using BIM+ supported by tablet PC whether from outdoor or

indoor zones. “jQuery” technology was integrated with the design of Control+ to preview the

captured data. “jQuery” is a multi-browser “JavaScript” library designed to simplify the user

preview interaction of HTML (HyperText Markup Language) files. It is free and open source

software that allows the creation of powerful dynamic web pages and web applications.

jQuery's syntax is designed to make it easier to navigate a document, select elements, create

animations and handle events (jQuery, 2013). The user selects first the data type to be

previewed, which are image, video, sound or notes. Then, the user selects the date when the

data acquisition process happened from an interactive calendar. Figure (6-44) depicts the

developed C# .Net source code in BIM/Walkthrough menu.

The user then selects whether the requested data are from indoor or outdoor

environment. According to the previously selected date, Control+ provides for the user a list

of plans for different floors to choose from. Afterwards in the right section of the web page a

preview of area plan or floor plan extracted from Revit software is displayed to the user. The

user selects one zone to preview the captured data. In the lower part of the web page, a preview

window appears to display the type of data requested. Under that window, there are several

navigation controls for the user to utilize. Similarly in the “CCTV/Time-Lapse” menu, the user

selects the data type and the date to preview the captured data from the CCTV camera on site,

as shown in Figure (6-45).

253
Figure 6-44: Control+ BIM/Walkthrough menu

254
Figure 6-45: Control+ CCTV/Time-Lapse menu

255
 The “Central DB” menu is used to facilitate data entry and editing for the central

database. A set of web forms for different tables in the central database have been developed

to support the process of populating the databases. These web forms serve as the interface of

the designed MS Access central database. According to the user authentication level, this forms

provides for the user data inserting, updating, deleting, and viewing capabilities. To input,

update, and/or delete data, the users can trigger the “Central DB” menu residing in Control+

presentation tier. Figure (6-46) depicts a sample of web forms for entering the project

information and resource information. These web forms are designed to provide simple and

user-friendly interfaces. Control+ provides for project team members the capability to preview

and query project captured data and progress reports remotely in near real time.

6.3 Summary

This chapter presents the computer implementation of the developed model. It also

highlights the developed system’s frameworks, modules, algorithms and the interaction

between its components. It covers the designed database, classes, use-cases, sequence diagrams

and snapshots from its Graphical User Interface (GUI). It describes the user interface, which is

designed to enable the user to collect and store data from construction sites and subsequently

query reports. The developed software is coded using “Visual C# .Net” programming language.

It consists of two computational frameworks; BIM+ and Control+. BIM+ is Two-Tier software,

which utilizes object-oriented BIM model. So, it can be used as an advanced tool for data

acquisition through the user’s tablet PC. Control+ is Three-Tier Web-Based software for

processing the captured data from the tablet PC and independent automated data acquisition

technologies.

256
Figure 6-46: Control+ Central database menu

257
 A complete sample (sequence diagrams and source code) of one function in BIM+ is

included in Appendix E. This function is a key function in the development of the 4D BIM

model, where objects in the 3D BIM model are linked to scheduled activities. Similar work

was carried for the rest of the software developed in this chapter.

258
 7 Chapter 7: CONCLUSIONS AND FUTURE WORK

7.1 Summary and Conclusions

This research presents study, concepts and design of automated model for data

acquisition and progress reporting to support time and cost tracking of construction projects.

The developed model integrates project visualization-information aspects, automated site data

acquisition and earned value analysis. Project visualization-information is achieved by

utilization of GIS and 4D BIM that integrates 3D BIM model and project schedule. Automated

site data acquisition technologies are clustered into two categories. The first utilizes tablet PC

that integrates RFID, barcode, GPS, wireless communication and camera technologies. It also

utilizes 4D BIM to provide visualization capabilities to the process of data acquisition. The

second category, referred to as “Independent Automated data Acquisition” integrates other

technologies beyond those housed in the tablet PC. The latter category consists of equipment

mounted GPSs, RFID readers mounted on project gates, 3D scanner and Closed Circuit

Television (CCTV). Data is collected on site and stored in a web-server for later processing to

generate progress reports. At the core of the developed model lies a data analysis and

processing framework that stores, retrieves and processes the captured data to support project

tracking and control. The collected data is used to represent the actual progress, which is then

compared to the as-planned baseline schedule. Earned Value Analysis (EVA) is then used to

measure the project performance.

Field studies were conducted on each technology, which led to identification of

limitations associated with each technology. Therefore, the advantage of integrating different

technologies and the data fusion for the captured data were explored to estimate the project

status for progress reporting purposes. In-depth studies were conducted to determine

259
deployment protocols of Ultra High Frequency (UHF) passive RFID for indoor localization.

Experimental work was conducted on construction jobsite and in a lab environment. K-Nearest

Neighbor (K-NN), Trilateration and proximity algorithms were used for location identification.

Prototype software was developed, using “Visual C#. Net” programming language, to

demonstrate the use of the developed model and validate its methods and algorithms. This

model is expected to enable project team members to track cost and time not only at the activity

and project levels, but also at the resource and object level in addition to generating progress

reports in a timely manner.

7.2 Research Contribution

Figure (7-1) summarizes the research contribution. The contributions of this research

are expected to circumvent the limitations associated with the current practice and existing

tracking and progress reporting models. Specifically, the research conducted explores and

describes the:

1. State of the art review on project control models, automated site data acquisition

technologies, indoor localization and visualization-information technologies were

carried out.

2. Research vision created to augment the essence and purpose of project control. This

vision provided inspiration and guidance in shaping the model developed in this thesis.

3. In-depth experimental work conducted in the lab and field studies performed in search

for efficient integration of wide range of technologies in the developed of automated

model for construction tracking and progress reporting.

4. Utilization of low-cost passive RFID for indoor localization and material tracking. This

exploration constitutes several contributions:

260
 • Developing deployment protocols for user’s indoor zone identification (Coarse–

grained localization) through K-NN algorithm.

• Developing two-step algorithm for indoor location identification (Fine-grained

localization) through utilizing trilateration or proximity algorithms.

• Comparing the performance of trilateration and proximity algorithms to identify

the most suitable one for the developed model.

• Developing deployment protocols for indoor location identification through

proximity algorithm.

• An integrated method was developed for tracking material from site arrival till it

is in the installation zone. RFID gate is used to report material delivery. RFID

and GPS are integrated to track material outdoor, while RFID alone is used to

track material indoor.

5. Development of an automated model that seamlessly facilitates the integration of

visualization-information technologies, automated site data acquisition technologies and

earned value analysis for the purpose of project control and progress reporting. The

developed model is capable of generating reports in different granularity (i.e. project

level, activity level, object level and resource level).

6. Developments of innovative method and its related algorithms to benefit from BIM in

progress reporting. This method constitutes several contributions:

• Developing algorithm for integrating 3D BIM model with project schedule to

generate the 4D BIM.

• Developing new method for zone configuration taking into account the progress

of zone’s floor. This method chooses the zones based on linking floor dividing

activity from project schedule and floor plans.

261
 • Creating a novel method and it respective algorithms to generate user location

real time visualization on tablet PC through integrating the 4D BIM model and

user location whether it is identified using GPS for outdoor zones or RFID for

indoor zones or manually.

• Integrate progress templates and user location real time visualization to make

BIM acts as a data acquisition tool for different types of activities.

7. Design of a new methods and its respective algorithms to track and control earthmoving

operations in near real time using:

• Spatial technologies (GPS/GIS).

• RFID gates

• RFID and control sensor or switch limit

8. Development of automated method to calculate earned value based on automated data

acquisition technologies and progress templates. Development of method to forecast

earthmoving fleet productivity, time and cost stochastically based on Discrete Event

Simulation (DES) and actual data captured from site.

9. Design of tablet PC and central databases to organize and manage captured data in this

research.

10. Development of computational platform (prototype software), which acts as an interface

with the user. This prototype can facilitates the capturing of actual data on-site and store

it in centralized database, as well as process the collected data for progress reporting in

near real time.

11. Utilization of actual construction project to demonstrate the use of developments made

in the course of this research

262
 Figure 7-1: Research contribution summary

7.3 Limitations

The following limitations were taken into account during the creation, validation, and

implementation of the data fusion and data management models developed in this research:

1. The parallel and iterative design, development and implementation of the model, resulted

in some of the model’s objectives were developed after the field implementation. So, it

was not validated such as the utilization of RFID gates and RFID and control sensor to

track earthmoving operations.

2. The model was therefore validated in part through field studies and experimental work

in both field and lab environments. A full case study implementation of the model is

suggested for future work, during which final refinements or design changes to the model

could be introduced.

263
3. The process of capturing data using the mobile RFID or GPS was carried out and

transferred via Bluetooth to the tablet PC manually, which could be done automatically

through hardware communication protocols.

4. In DES stochastic forecasting, it was assumed that all trucks have the same capacity and

same Probability Density Functions (PDF). A new DES simulation engine can be

developed to take these factors into consideration.

5. Control+ developed system was limited to certain processing and reporting

functionalities limit.

7.4 Future Work

The following are some recommendations for future work areas:

1. Development of a reasoning model to map the relationship between the inputs and

outputs of the control process. This model should explain the reasons behind

unacceptable performance and to suggest corrective action accordingly.

2. Expand the utilization of DES stochastic forecasting based on actual data to other

construction operations and explore other methods of forecasting productivity, time and

cost.

3. Extend the use of the developed model to include tracking material through the supply

chain and procurement process.

4. Integrating CCTV outdoor cameras with user’s mobile RFID reader to identify outdoor

user’s location and material tracking. This can be made possible through georeferencing

the CCTV camera images.

264
5. Explore the possibility of extending BIM+ to be a Web-Based solution running through

cloud computing. Extend BIM+ to be fully functional planning and scheduling tool

instead of using scheduling third party software.

6. Complete and extend the functionalities of Control+ to be a complete cloud computing

solution for progress reporting.

265
 References

ABB - Limit Switches - 101 A basic guide to ABB’s limit switch portfolio, (2013) <
http://www05.abb.com/global/scot/scot260.nsf/veritydisplay/8bb7f20332ba74d58525
75d20063da5a/$file/1sxu141184x0201.pdf>.
Abeid, J., Allouche, E., Arditi, D., and Hayman, M. (2003). “PHOTO-NET II: a computer-
based monitoring system applied to project management.” Journal of Automation in
Construction, 12(5), 603-612.
Abeid, J., Arditi, D. (2002). “Time-Lapse Digital Photography Applied to Project
Management.” Journal of Construction Engineering and Management (ASCE), 128(6),
530-535.
AbouRizk, S. (2010). “Role of simulation in construction engineering and management.”
Journal of Construction Engineering and Management (ASCE), 126(10), 1140-1153.
Akhavian, R., & Behzadan, A. H. (2012). “An integrated data collection and analysis
framework for remote monitoring and planning of construction operations.” Advanced
Engineering Informatics, 26(4), 749-761.
Akinci, B., Patton, M., and Ergen, E. (2002). “Utilizing radio frequency identification on
precast concrete components-supplier's perspective.” Proceedings of 19th
International Symposium on Automation and Robotics in Construction (ISARC 2002),
Washington, DC, 381–386.
Alemany, C.H. (1999). “Construction foremen computer use and impacts.” Master’s Thesis,
presented to University of Texas at Austin, Austin, Texas.
Alzraiee, H., Moselhi, O., and Zayed, T. (2012). “A hybrid framework for modeling
construction operations using discrete event simulation and system dynamics.”
Proceedings of Construction Research Congress (CRC 2012), ASCE, West Lafayette,
IN, USA, 1063-1073.
ArcGIS API for Silverlight, (2013). < http://help.arcgis.com/en/webapi/silverlight/index.html>
Aryan, A. (2011). “Evaluation of Ultra-Wideband sensing technology for position location in
indoor construction environments.” Masters’ thesis, presented to University of
Waterloo, Ontario, Canada.
Aslani, P., Griffis, F. H. and Chiarelli, L. (2009). “Building information model the role and
need of the constructors.” Building a Sustainable Future - Proceedings of the
Construction Research Congress (CRC 2009), 467-476.
Bansal, V.K., and Pal, M. (2009). “Construction schedule review in GIS with a navigable 3D
animation of project activities.” International Journal of Project Management, 27(5),
532-542.
Blewitt, G. (1997). “Basics of the GPS technique: observation equations.” in Geodetic
Applications of GPS, pp. 10-54, ed. B. Johnson, Nordic Geodetic Commission, The
Swedish Land Survey.
Bohn, J. S. and Teizer, J. (2009). “Benefits and barriers of monitoring construction activities
using hi-resolution automated cameras.” Proceedings of the Construction Research
Congress (CRC 2009), 21-30.

266
Borriello, G., Brunette. W., Hall, M., Hartung, C., Tangney, C. (2004). “Reminding about
tagged objects using passive RFIDs.” Ubiquitous Computing: 6th International
Conference (Ubi-Comp 2004), 36-53.
Bosche, F., and Haas, C. T. (2008). “Automated retrieval of 3D CAD model objects in
construction range images.” Journal of Automation in Construction, 17(4), 499-512.
Bosche, F., Haas, C. T., & Murray, P. (2008). “Performance of automated project progress
tracking with 3D data fusion.” Annual Conference of the Canadian Society for Civil
Engineering (CSCE 2008), Partnership for Innovation, Quebec City, QC, Canada,
349-358.
Bosche, F., Haas, C. T., and Akinci, B. (2009). “Automated recognition of 3D CAD objects in
site laser scans for project 3D status visualization and performance control.” Journal
of Computing in Civil Engineering (ASCE), 23(6), 311-318.
Brilakis, I. (2007). “Long distance wireless networking for site - office data communications.”
Journal of Information Technology in Construction (ITCON), 12, 154-164.
Caldas, C., Grau, D., and Haas, C. (2006). “Using Global Positioning Systems to improve
materials locating processes on industrial projects." Journal of Construction
Engineering and Management (ASCE), 132 (7), 741-749.
Can-Net, VRS Explained, (2013). <http://www.can-
net.ca/index.php?option=com_content&view=article&id=36&Itemid=230>
Caterpillar, Product Link, (2013). <http://www.cat.com/itpaystoknow>
Chan, K.W., Anson, M., and Zhang, J.P., (2004). “Four-dimensional visualization of
construction scheduling and site utilization.” Journal of Construction Engineering and
Management (ASCE), 130(4), 598–606.
Chau, K.W., Anson, M. and Zhang, J.P. (2005). “4D dynamic construction management and
visualization software 1. Development.” Journal of Automation in Construction, 14(4),
512-524.
Cheng, M. Y. and Chen, J. C. (2002). “Integrating bar code and GIS for monitoring
construction progress.” Journal of Automation in Construction, 11(3), 23–33.
Cheok, G. S., Lipman, R. R., Witzgall, C., Bernal, J. and Stone, W. C. (2000a). “NIST -
construction automation program no. 4: non-intrusive scanning technology for
construction status determination.” Research Report NISTIR 6457, BFRL, NIST,
Gaithersburg, MD, USA, January.
Cheok, G., Lipman, R., Witzgall, C, Bernal, J. and Stone, W. (2000b). “Field Demonstration
of Laser Scanning for Excavation Measurement.” Proceedings of the 17th
International Symposium on Automation and Robotics in Construction (ISARC 2000),
Taipei, Taiwan.
Chin, S., Yoon, S., Kim, Y.S., Ryu, J., Choi, C., Cho, and C.Y. (2005). “Realtime 4D CAD +
RFID project progress management.” Proceedings of Construction Research Congress
(CRC 2005), San Diego, USA.
Choi, J. S. (2011). “Accurate and cost efficient object localization using passive UHF RFID.”
Ph.D. Presented to the Faculty of the Graduate School of the University of Texas at
Arlington, University of Texas at Arlington.

267
Chon, H.D., Jun, S., Jung, H. and An, S.W. (2004). “Using RFID for accurate positioning.”
Journal of Global Positioning Systems, 3(1), 32-39.
Construction Industry Institute (CII), (1987). “Model planning and controlling system for EPC
of industrial projects.” Publication 6-3, The Construction Industry Institute (CII), The
University of Texas at Austin, Austin, Texas.
Diggelen, F.V. (2002). “Indoor GPS theory and implementation.” Proceedings of the IEEE
Position Location and Navigation Symposium, IEEE, Piscataway, NJ, USA, 240–247.
Duda, R. O., and Hart, P. E. (1973). Pattern classification and scene analysis. New York, Wiley.
Eastman, C., Teicholz, P., Sacks, R., and Liston, K. (2008). BIM Handbook: A Guide to
Building Information Modeling for owners, managers, designers, engineers and
contractors. John Wiley & Sons, Inc.
Echevery, D. and Beltran, A. (1997). “Bar Code control of construction field personnel and
construction materials.” Proceedings of the 4th Congress on Computing in Civil Eng.,
ASCE, Philadelphia, PA, USA, 341-347.
Eddy, M.R., and Namhun, L. (2007). “Visualization of Project Control Data: A Research
Agenda.” Proceedings of International Workshop on Computing in Civil Engineering
(ASCE), Pittsburgh, USA, 26-34.
El-Omari, S., (2008). “Automated data acquisition for tracking and control of construction
projects.” Ph.D. Thesis presented to the Department of Building, Civil and
Environmental Engineering, Concordia University, Montreal, Canada.
El-Omari, S., and Moselhi, O. (2008). “Integrating 3D laser scanning and photogrammetry for
progress measurement of construction work.” Journal Automation in Construction,
18(1), 1-9.
El-Omari, S., and Moselhi, O. (2009). “Data acquisition from construction sites for tracking
purposes.” Engineering Construction and Architectural Management Journal,
16(5), 490-503.
El-Omari, S., and Moselhi, O. (2011). “Integrating automated data acquisition technologies for
progress reporting of construction projects.” Journal of Automation in Construction,
20(6), 699-705.
Elzarka, H., and Bell, L. (1997). “Development of pen-based computer field applications.”
Journal of Computing in Civil Engineering (ASCE), 11(2), 140-143.
Ergen, E. and Akinci, B. (2007). “An overview of approaches for utilizing RFID in
construction industry.” Proceedings of the 1st Annual RFID Eurasia Conference, 7-11.
Ergen, E., Akinci, B. and Sacks, R. (2007a). “Life-Cycle data management of engineered-to-
order components using radio frequency identification.” Journal of Advanced
Engineering Informatics, 21(4), 356–366.
Ergen, E., Akinci, B., and Sacks, R. (2007b). “Tracking and locating components in a precast
storage yard utilizing radio frequency identification technology and GPS.” Journal of
Automation in Construction, 16(3), 354-367.
ExtendSim, (2013). < http://www.extendsim.com/sols_simoverview.html>

268
Fleming, Q. and Koppelman, J. (1994). “The essence of evaluation of earned value.” Cost
Engineering Journal (AACE International), 36(11), 21-27.
Fox, V., Hightower, J., Lin, L., Schulz, D. and Borriello, G. (2003). “Bayesian filtering for
location estimation.” IEEE Pervasive Computing journal, 2(3), 24–33.
Fu, C., Aouad, G., Lee, A., Mashall-Ponting, A., and Wu, S. (2006). “IFC model viewer to
support nD model application.” Journal Automation in Construction, 15(2), 178-185.
Fu, Q. and Retscher, G. (2009). “Active RFID trilateration and location fingerprinting based
on RSSI for pedestrian navigation.” Journal of Navigation, 62(2), 323-340.
Ghanem, A. G. and Abdelrazig, Y. A. (2006). “A framework for real-time construction project
progress tracking.” Proceedings of the 10th Biennial International Conference on
Engineering, Construction, and Operations in Challenging Environments, 112-120.
Goller, M. and Brandner, M. (2011). “Experimental evaluation of RFID gate concepts.” IEEE
International Conference on RFID, 26-31.
Golparvar-Fard, M., and Peña-Mora, F. (2007). “Application of visualization techniques for
construction progress monitoring.” Proceedings of the ASCE International Workshop
on Computing in Civil Engineering, Pittsburg, PA, USA, 261(27), 216-223.
Golparvar-Fard, M., Pena-Mora, F., and Savarese, S. (2009a). “D4AR-A 4-dimensional
augmented reality model for automating construction progress monitoring data
collection, processing and communication.” Journal of Information Technology in
Construction (ITCON), 14, 129-153.
Golparvar-Fard, M., Pena-Mora, F., Arboleda, C. A., and Lee, S. (2009b). “Visualization of
construction progress monitoring with 4D simulation model overlaid on time-lapsed
photographs.” Journal of Computing in Civil Engineering (ASCE), 23(6), 391-404.
Gonçalo, G, and Helena, S. (2009) “A novel approach to indoor location systems using
propagation models in WSNs,” International Journal on Advances in Networks and
Services, 2(4), 251-260.
Guo, X., Huang, S.Y., Hsu, W.J. and Low, M.Y.H., (2008). “Yard crane dispatching based on
real time data driven simulation for container terminals.” Proceedings of Winter
Simulation Conference, Miami, FL, USA, 2648–2655.
Hajjar, D., and AbouRizk, S. 1999. “Symphony: an environment for building special purpose
simulation.” Proceedings of 31st Winter Simulation Conference, ASCE, Phoenix, AZ,
USA, 998–1006.
Halpin, D. (1973). “An investigation of the use of simulation networks for modeling
construction operations.” Ph.D. Thesis presented to University of Illinois at Urbana-
Champaign, USA.
Han, S., Lee, S., Hong, T., and Chang, H., (2006). “Simulation analysis of productivity
variation by global positioning system implementation in earthmoving operations.”
Canadian Journal of Civil Engineering, 33(9), 1105–1114.
Hartmann, T., Gao, J., and Fischer, M. (2008). “Areas of application for 3D and 4D models on
construction projects.” Journal of Construction Engineering and Management (ASCE),
134(10), 776-785.

269
Hassanein, A., and Moselhi, O. (2003). “Tracking and control of linear infrastructure projects.”
Proceedings of 5th Construction Specialty Conference of the Canadian Society for Civil
Engineering (CSCE 2003), June 4-7, New Brunswick, Canada.
Hassanien, A., (2002). “Planning and Scheduling Highway Construction using GIS and
Dynamic Programming.” Ph.D. Dissertation, Department of Building, Civil and
Environmental Engineering, Concordia University, Montreal, Quebec.
Hassanien, A., and Moselhi O. (2002). “Automated data acquisition and planning of highway
construction.” Procedings of 19th Symposium on Automation and Robotics in
Construction (ISARC 2002), Gaithersburg, Maryland, USA September 23-25.
Helmus, M., (2007). “Application fields of RFID in health safety and environment
management.” Proceedings of the 1st Annual RFID Eurasia, 95-97.
Hightower, J. and Borriello, G. (2001). “Location systems for ubiquitous computing.”
Computer, 34(8), 57-66.
Intermec Solutions Inc., (2013).”RFID.” <
http://www.intermec.ca/products/rfid/index.aspx>
Jang, W. and Skibniewski, M. J. (2008). “A wireless network system for automated tracking
of construction materials on project sites.” Journal of Civil Engineering and
Management (ASCE), 14(1), 11-19.
Jaselskis, E. and El-Misalami, T. (2003). “Implementing radio frequency identification in the
construction process.” Journal of Construction Engineering and Management (ASCE),
129(6), 680-688.
Jaselskis, E., Anderson, M. R., Jahren, C. T., Rodriguez, Y., and Njos, S. (1995). “Radio-
frequency identification applications in construction industry.” Journal of Construction
Engineering and Management (ASCE), 121(2), 189– 196.
Jiang, L., Hoe, L. N. and Loon, L. L., (2010). “Integrated UWB and GPS location sensing
system in hospital environment.” IEEE International Conference of Sensor Networks.
jQuary, (2013). <http://WWW.jquery.com/>
Jung, Y., and Kang, S., (2007). “Knowledge-based standard progress measurement for
integrated cost and schedule performance control.” Journal of Construction
Engineering and Management (ASCE), 133(1), 10-23.
Kang, J.H., Anderson, S.D., and Clayton, M.J. (2007). “Empirical study on the merit of web
based 4D visualization in collaborative construction planning and scheduling.” Journal
of Construction Engineering and Management (ASCE), 133(6), 447–461.
Krygiel, E., Nies, N., and McDowell, S. (2008). Green BIM: successful sustainable design with
building information modeling. Wiley Publishing, Indianapolis, USA.
Larman, C., (1998). Applying UML and patterns: an introduction to object-oriented analysis
and design. Prentice Hall, Inc., Englewood Cliffs, N.J.
Lee, W., Song, J., Kwon, S., Chin, S., Choi, C., and Kim, Y. (2008). “A gate sensor for
construction logistics.” Proceedings of the 25th International Symposium on
Automation and Robotics in Construction (ISARC 2008), june26-29, 100-105.

270
Leick, A. (1995). GPS satellite surveying. 2nd Edition, John Wiley and Sons, Inc., New York,
USA.
Li, J., (2004). “Web-based integrated project control.” Ph.D. Thesis presented to Building,
Civil and Environmental Engineering Department, Concordia University, Montreal,
Canada.
Li, J., Moselhi, O., and Alkass, S., (2006). “Forecasting project status by using fuzzy logic.”
Journal of Construction Engineering and Management (ASCE), 132(11), 1193-1202.
Li, N., and Becerik-Gerber, B. (2011). “Performance-based evaluation of RFID-based indoor
location sensing solutions for the built environment.” Advanced Engineering
Informatics, 25(3), 535-546.
Liu, L. (1995). “Digital data-collection device for construction site documentation.”
Proceedings of 2nd Congress on Computing in Civil Engineering, Atlanta, USA, 1287-
1293.
Liu, L., Chin, S., Ganeshan, R. and Hichs, D. (1995). “Construction daily log management
system using multimedia technology.” Proceedings of the 2nd Congress on Computing
in Civil Engineering, Atlanta, USA, 1084-1089.
Lytle, A. M. (2011). “A framework for object recognition in construction using building
information modeling and high frame rate 3D imaging.” Ph.D. Thesis presented to the
Faculty of the Virginia Polytechnic Institute and State University.
Marco, G., and Fioerenzo, M., (2003). “Remote control and automatic monitoring of earth-
moving in road construction.” Proceedings of 20th International Symposium on
Automation and Robotics in Construction (ISARC 2003), Netherlands, 21-24 Sept., 74,
119-124.
Martinez, J. C. (1996). “STROBOSCOPE: state and resource based simulation of construction
processes.” Ph.D. Thesis presented to University of Michigan, Ann Arbor.
Martinez, J., and Ioannou, P. (1999). “General-purpose systems for effective construction
simulation.” Journal of Construction Engineering and Management (ASCE), 125(4),
265-276.
Marzouk, M., and Moselhi, O. (2004). “Multi-objective optimization of earthmoving
operations.” Journal of Construction Engineering and Management (ASCE), 130(1),
105–113.
McCuen, T. L., (2008), “Scheduling, estimating, and BIM a profitable combination.” AACE
International Transactions, Annual Meeting of AACE International and the 6th World
Congress of ICEC on Cost Engineering (BIM.01), 1-8.
McCuen, T. L., (2009). “The quantification process and standards for BIM.” AACE
International Transactions, AACE International Annual Meeting (BIM-01), 1-11.
McCullouch, B. and Gunn, P. (1993). “Construction field data acquisition with pen-based
computers.” Journal of Construction Engineering and Management (ASCE), 119(2),
374-384.
McCullouch, B. and Lueprasert, G. K. (1994) “2D bar-code applications in construction.”
Journal of Construction Engineering and Management (ASCE), 120(4), 739-752.

271
Min, Y., and Jiann, C., (2002). “Integrating barcode and GIS monitoring construction
progress.” Journal of Automation in Construction, 11(1), 22-33.
Montaser, A. and Moselhi, O. (2012a). “4D and tablet PC for progress reporting.” Proceedings
of International Symposium on Automation and Robotics in Construction (ISARC
2012), Eindhoven, Netherlands, June 26–29.
Montaser, A. and Moselhi, O. (2012b). “RFID + for tracking earthmoving operations.”
Proceedings of Construction Research Congress (CRC 2012), West Lafayette, Indiana,
United States, 2012, 21-23 May, 1011-1020.
Montaser, A. and Moselhi, O. (2012c). “RFID and BIM for automated progress reporting.”
AACE International Transactions, 56th Annual Meeting, San Antonio, Texas, United
States, 2012, 08-11July.
Montaser, A. and Moselhi, O. (2012d). “RFID indoor location identification for construction
projects.” Proceedings of International Symposium on Automation and Robotics in
Construction (ISARC 2012), Eindhoven, Netherlands, June 26–29.
Montaser, A. and Moselhi, O. (2013a). “RFID indoor location identification for construction
projects.” Journal of Automation in Construction, (In Press).
Montaser, A. and Moselhi, O. (2013b). “Tracking hauling trucks for cut and fill earthmoving
operations” Proceedings of International Symposium on Automation and Robotics in
Construction (ISARC 2013), Montreal, QC, Canada, August 11–15.
Montaser, A. and Moselhi, O. (2013c). “Tracking scraper–pusher fleet operations using
wireless technologies.” Proceedings of Canadian Society for Civil Engineering Annual
Conference (CSCE 2013), Montreal, QC, Canada, May 29 – June 1.
Montaser, A. Azarm, R. and Moselhi, O. (2013). “Experimental study for efficient use of RFID
in construction” Proceedings of International Symposium on Automation and Robotics
in Construction (ISARC 2013), Montreal, QC, Canada, August 11–15.
Montaser, A., Bakry, I., Alshibani, A. and Moselhi, O. (2011). “Estimating productivity of
earthmoving operations using spatial technologies.” Proceedings of Canadian Society
for Civil Engineering Annual Conference (CSCE 2011), Ottawa, ON, Canada, 3, 2318-
2327.
Montaser, A., Bakry, I., Alshibani, A. and Moselhi, O. (2012). “Estimating productivity of
earthmoving operations using spatial technologies” Canadian Journal of Civil
Engineering, 39(9), 1072 - 1082.
Moselhi, O. and Alshibani, A. (2007). “Crew optimization in planning and control of
earthmoving operations using spatial technologies.” Journal of Information
Technology in Construction (ITCON), 12, 121-137.
Moselhi, O. and El-Omari, S., (2006). “Integrating bar coding and RFID to automate data
collection from construction sites.” Proceedings of the Joint International Conference
on Computing and Decision Making in Civil and Building Engineering, June 14-16, ,
Montréal, Canada, 1734-1741.
Moselhi, O. and Shehab-Eldeen, T. (2003). “Use of bar-coding in engineering procurement
and construction.” Report prepared for SNC-Lavalin, Department of Building, Civil
and Environmental Engineering, Concordia University, Montreal, Canada.

272
Moselhi, O., (1993). “Applied earned value for control.” International Symposium of CIB W-
65, Sept., St. Augustine, Trinidad and Tobago.
Moselhi, O., and Alshibani, A. (2008). “Tracking and control of earthmoving operation using
spatial technologies.” Cost Engineering Journal, AACE International, 50(10), 26-33.
Moselhi, O., Li, J., and Alkass, S. (2004). “Web-based integrated project control system.”
Journal of Construction Engineering and Economics, 22, 35-46.
Motamedi, A. and Hammad, A. (2009 a). “Lifecycle management of facilities components
using radio frequency identification and building information model.” Proceedings of
5th International Conference on Construction in the 21th Century, Turkey, 20–22 May.
Motamedi, A. and Hammad, A. (2009 b). “Lifecycle management of facilities components
using radio frequency identification and building information model.” Journal of
Information Technology in Construction (ITCON), 14, 238–262.
Motamedi, A., Saini, R., Hammad, A. and Zhu, B. (2011). “Role-based access to facilities
lifecycle information on RFID tags.” Advanced Engineering Informatics, 25(3), 559–
568.
Nasir, H., (2008). “A model for automated construction materials tracking.” Masters’ thesis
presented to University of Waterloo, Waterloo, Canada.
Navon, R. and Goldschmidt, E. (2002). “Monitoring labor inputs: automated-data collection
model and enabling technologies,” Journal of Automation Construction, 12(2), 185–
199.
Navon, R., Goldschmidt, E., and Shpatnisky, Y. (2004). “A concept proving prototype of
automated earthmoving control.” Journal of Automation in Construction, 13(2), 225–
239.
NIBS. National building information modeling standard, (2013) <
http://www.buildingsmartalliance.org/pdfs/mhc_bim_smartmarket.pdf>
Oglesby, H., Parker, W. and Howell, A. (1989). Productivity Improvement in Construction,
McGraw Hill Companies, New York, NY, USA.
Oracle Inc., (2013) <
http://www.oracle.com/us/products/applications/crystalball/overview/index.html>
Panasonic Inc., Tablets, (2013). <http://www.panasonic.com/business/toughbook/ultra-
mobile-rugged-mca-toughbook-h2.asp?cm_mmc=PSCNA-_-Vanities-_-
toughbook-h1field-_-tb-h2-page>
Patwari, N., Ash, J.N., Kyperountas, S., Hero, A.O., Moses, R.L. and Correal, N.S. (2005).
“Locating the nodes: cooperative localization in wireless sensor networks.” IEEE
Signal Processing Magazine, 22(4), 54- 69.
Peurifoy, R. L., and Schexnayder, C. J. (2006). Construction Planning, Equipment, and
Methods, 7th Ed., McGraw-Hill, New York, NY, USA.
Poku, S., and Arditi, D. (2006). “Construction scheduling and progress control using
geographical information systems.” Journal of Computing in Civil Engineering
(ASCE), 20(5), 351-360.

273
Priyantha, N.B., Chakraborty, A. and Balakrishnan, H., (2000). “The cricket location-support
system.” Proceedings of the 6th Annual International Conference on Mobile Computing
and Networking (MOBICOM 2000), New York, USA, 32–43.
Quatrani, T., (2002). Visual modeling with rational rose 2002 and UML. 3rd Edition, Addison-
Wesley, USA.
Razavi, S. N., Montaser, A. and Moselhi, O. (2012). “RFID deployment protocols for indoor
construction.” Journal of Construction Innovation: Information, Process,
Management, 12(2), 239 - 258.
Razavi, S., N. (2010). “Data fusion for materials location estimation in construction.” Ph.D.
thesis presented to the University of Waterloo, Waterloo, Ontario, Canada.
Razavi, S.N. and Haas, C. (2010). “Multisensor data fusion for on–site materials tracking in
construction.” Journal of Automation in Construction, 19(8), 1037-1046.
Rezazadeh Azar, E. and McCabe, B. (2012). “Automated visual recognition of dump trucks in
construction videos.” Journal of Computing in Civil Engineering (ASCE), 26(6), 769–
781.
Riegl Inc., (2013). <http://www.riegl.com/nc/products/terrestrial-
scanning/produktdetail/product/scanner/2/>
Sarda, A. (2011). “Read counts at multiple attenuation level algorithm as an object localization
technique using passive RFID tags.” Ph.D. Dissertation submitted to Rutgers, The State
University of New Jersey.
Shahi, A. (2012). “Activity-based data fusion for the automated progress tracking of
construction projects.” PhD Thesis, presented to University of Waterloo, Ontario,
Canada.
Shahi, A., West, J. S., and Haas, C. T. (2013). “Onsite 3D marking for construction activity
tracking.” Journal of Automation in Construction, 30, 136-143.
Shannon, R. (1998). “Introduction to the art and science of simulation.” Proceedings of the
30th Conference on Winter Simulation, Los Alamitos, CA, USA, 7-14.
Shi, W., (2009). “Framework for integration of BIM and RFID in steel construction.” Ph.D.
Thesis presented to University of Florida.
Song, J., Haas, C., Caldas, C., Ergen, E., and Akinci, B. (2006). “Automating the task of
tracking the delivery and receipt of fabricated pipe spools in industrial projects.”
Journal of Automation in Construction, 15(2), 166-177.
Song, L., and Eldin, N. N. (2012). “Adaptive real-time tracking and simulation of heavy
construction operations for look-ahead scheduling.” Journal of Automation in
Construction, 27, 32-39.
Song, L., F. Ramos, and K. Arnold. (2008). “A Framework for Real-time Simulation of Heavy
Construction Operations.” Proceedings of Winter Simulation Conference, Miami, FL,
USA, 2387-2395.
Sony Inc., Security Camera, (2013), < http://pro.sony.com/bbsc/ssr/cat-securitycameras/>
Staub-French, S., Russell, A., and Tran, N. (2008). “Linear scheduling and 4D visualization.”
Journal of Computing in Civil Engineering (ASCE), 22(3), 192–205.

274
Styliadis, A. D. (2007). “Digital documentation of historical buildings with 3d modelling
functionality.” Journal of Automation in Construction, 16, 498-510.
Thabet, W. (1997). “An object-oriented cad-based model for construction activity progress
status.” Proceedings of the 3rd Canadian Conference on Computing in Civil and
Building Engineering, Montreal, QC, Canada, 315-321.
Thabet, W. Y., and Beliveau, Y. J., (1997). “SCaRC: space-constrained resource-constrained
scheduling system.” Journal of Computing in Civil Engineering (ASCE), 11(1), 48 –
59.
Tserng, H., Dzeng, R., Lin, Y., and Lin, S. (2005). “Mobile construction supply chain
management using PDA and bar codes.” Journal of Computer Aided Civil and
Infrastructure Engineering, 20, 242-264.
Turkan Y., Bosché F., Haas C.T. and Haas R.G. (2012). “Automated progress tracking using
4D models and 3D sensing technologies.” Journal of Automation in Construction, 22,
414-421.
Turkan, Y., (2012). “Automated construction progress tracking using 3D sensing
technologies.” Ph.D. thesis presented to the University of Waterloo, Ontario, Canada.
Tzeng, C., Chiang, Y., Chiang, C. and Lai, C. (2008). “Combination of Radio Frequency
Identification (RFID) and field verification tests of interior decorating materials.”
Journal of Automation in Construction, 18(1), 16-23.
Vahdatikhaki, F., Hammad, A. and Setayeshgar, S. (2013). “Real-time simulation of
earthmoving projects using automated” Proceedings of International Symposium on
Automation and Robotics in Construction (ISARC 2013), Montreal, QC, Canada,
August 11–15.
Velte, T., Velte, A., and Elsenpeter, R. (2009). Cloud Computing, A Practical Approach.
McGraw Hill Professional.
Waly, A.F., and Thabet, W.Y., (2002). “A virtual construction environment for preconstruction
planning.” Journal of Automation in Construction, 12, 139–154.
Woo, S., Jeong, S., Mok, E., Xia, L., Choi, C., Pyeon, M. and Heo, J. (2011). “Application of
WiFi-based indoor positioning system for labor tracking at construction sites: a case
study in Guangzhou MTR.” Journal of Automation in Construction, 20(1), 3-13.
Yoon, S.W., Chin, S., Kim, Y.S. and Kwon, S.W. (2006). “An application model of RFID
technology on progress measurement and management of construction works.” Proceedings of
the 23rd International Symposium on Automation and Robotics in Construction (ISARC 2006),
779-783.
Zayed, T. and Halpin, D. (2001). “Simulation of concrete batch plant production.” Journal of
Construction Engineering and Management (ASCE), 127(2), 132-141.
Zhang, Y., Yang, L. T., and Chen, J. (2010). RFID and sensor networks: architectures,
protocols, security, and integrations. CRC Press, Boca Raton, FL.
Zhu, Z., German, S., and Brilakis, I. (2010). “Detection of large-scale concrete columns for
automated bridge inspection.” Journal of Automation in Construction, 19(8), 1047–
1055.

275
 8 Appendix A: Tracking Scraper-Pusher Fleet for Cut-Fill

Earthmoving Operations

This appendix presents an automated method for tracking and estimating productivity

of Scraper-Pusher fleet in near real time utilizing wireless technologies for cut-fill earthmoving

operations. RFID and equipment switch limit are the technologies used in the developed

method. Low cost UHF rugged encapsulated passive RFID tags are attached to scrapers and

fixed RFID readers are attached to pushers. The read range of used RFID tag is centimeters, to

be activated only when the pusher with the attached RFID reader is pushing the scraper for

loading operations. Then, the RFID reader starts capturing the RF signals from the scraper tag.

On the other hand, switch limit is connected the scraper bowl to monitor its motion. The

movement of scraper bowl activates the switch limit. The function of switch limit is to record

the time when the scraper bowl opens or closes. The switch limit sends its captured data via

cable to micro controller that is attached to the scraper. Scraper micro controller has a wireless

communication module that sends and receives data to/from RFID readers.

The data captured by RFID from scrapers RFID tag and switch limit are transferred

wirelessly to a specially designed relational database in as web server, as shown in Figure (A-

1). The collected data is analyzed and processed automatically, without human intervention.

Fusing the data captured from RFID reader and switch limit is used to identify loading, travel,

dumping and return times that constitute the scraper cycle time. Consequently, the productivity

of scraper-pusher fleet is estimated for the whole earthmoving operation. The results are

reported to project team members via internet. For further details about the developed

algorithm and system architecture, refer to (Montaser and Moselhi, 2013 c).

276
Figure A-1: The entity relationship diagram (ER) for the developed database

277
 9 Appendix B: Central Database

Table B-1: Central database attributes

Table Attribute Name Data Type Description

ProjectID Number Primary Key (User Input)

Location Text User Input

PlannedDuration Number User Input

PlannedBudget Number User Input

ContractualStartDate Date/Time User Input

ContractualEndDate Date/Time User Input

ActualStartDate Date/Time User Input

ActualEndDate Date/Time User Input

Project
ActualDuration Number User Input

Owner Text User Input

Description Text User Input

ProjectName Text User Input

Type Text User Input

ContractType Text User Input

PolygonID Number Provided by System

SoilID Number Provided by System

Date Date/Time Primary key (Provided by System t)

ProjectID Number Provided by System

Provided by System = # (BCWS
BCWS Number
Activities)
Provided by System = # (BCWP
BCWP Number
Activities)
Project Status Provided by System = # (ACWP
ACWP Number
Activities)
Provided by System = (BCWP –
CV Number
ACWP)
Provided by System = (BCWP –
SV Number
BCWS)
Provided by System = (BCWP /
CPI Number
ACWP)

278
Table Attribute Name Data Type Description
Provided by System = (BCWP /
SPI Number
BCWS)
EAC1 Number Provided by System = (BAC / CPI)
Provided by System = (ACWP +
EAC2 Number
(BAC – BCWP) )
TimeStatus Text Provided by System

BudgetStatus Text Provided by System

Primary key (Provided by System
Temperature Number
from MS Project)
Primary key (Provided by System
Humidity Number
from MS Project)
Primary key (Provided by System
WindSpeed Number
from MS Project)
Primary key (Provided by System
Rain Number
from MS Project)
Primary key (Provided by System
ActivityID Number
from MS Project)
Name Text Provided by System from MS Project

ProjectID Number Provided by System from MS Project

Provided by System from MS Project
Type Text
(Critical or Not Critical)
Provided by System (Last Date of
ActualEF Date/Time Activity Status and should be
confirmed by the user)
UnitCost Number User Input
Provided by System (First Date of
ActualES Date/Time
Activity Status)
PlannedEF Date/Time Provided by System from MS Project

Activity PlannedES Date/Time Provided by System from MS Project

Budget Number Provided by System from MS Project

Duration Number Provided by System from MS Project

Quantity Number Provided by System 3D BIM Model

TotalFloat Number Provided by System from MS Project

FreeFloat Number Provided by System from MS Project

WeatherSensitivity Yes/No User Input

ProgressType Text User Input

UnitsofMeasurement Text User Input

TrackingTechnology Text User Input

279
 Table Attribute Name Data Type Description

Date Date/Time Primary key (t Provided by System)

ActivityID Number Provided by System

Provided by System = (BAC x
BCWS Number
Elapsed Duration / Duration)
Provided by System = (% Complete
BCWP Number
x BCWS)
ACWP Number User Input
Provided by System = (BCWP –
CV Number
ACWP)
Provided by System = (BCWP –
SV Number
BCWS)
Provided by System = (BCWP /
CPI Number
ACWP)
Provided by System = (BCWP /
SPI Number
BCWS)
EAC1 Number Provided by System = (BAC / CPI)
Activity Status
Provided by System = (ACWP +
EAC2 Number
(BAC – BCWP) )
According to Activity progress Type
(in Case of Truck = QuantitytoDate /
PrecentageComplete Number
Total Quantity) or from Progress
Template
Provided by System =
QuantitytoDate Number (TodayQuantity + QuantitytoDate(t-
1))
Provided by System = (ActualES +
ForecastedEF Date/Time
(Duration / SPI))
Provided by System ( in Case of
Truck = # [TruckCapacity xTotal
TodayQuantity Number
Number of Cycles] x Soil Fill
Factor)
TimeStatus Text Provided by System

BudgetStatus Text Provided by System

WorkContentCode Number Primary key (User Input)

WorkContent Text User Input

Progress
Template
Cumulative%Earned Number User Input

ActivityID Number User Input

TabletPCID Number Primary Key (User Input)

ModelNumber Number User Input

Tablet PC
Manufacture Text User Input

ProjectID Number User Input

280
 Table Attribute Name Data Type Description

Serial Number Primary Key (RFID Captured Data)

Data File Attachment Primary Key (RFID Captured Data)

RSSI Number RFID Captured Data

Tablet PC
RFID Reader ReaderDate Date/Time RFID Captured Data
File
ReaderTime Date/Time RFID Captured Data

RFIDTagID Number RFID Captured Data

TabletPCID Number RFID Captured Data

PositionID Number Primary Key (GPS Captured Data)

Data File Attachment Primary Key (GPS Captured Data)

Date Date/Time GPS Captured Data

Time Date/Time GPS Captured Data

Tablet PC GPS
Lattitude Number GPS Captured Data
File
Longitude Number GPS Captured Data

Altitude Number GPS Captured Data

LineID Number GPS Captured Data

TabletPCID Number GPS Captured Data

UserPositionID Number Primary Key (Provided by System)

TabletPCID Number Provided by System

Date Date/Time Provided by System

TimeFrom Date/Time Provided by System

TimeTo Date/Time Provided by System

Latitude Number Provided by System (Outdoor)

User Location
Longitude Number Provided by System (Outdoor)

Xj Number Provided by System (Indoor)

Yj Number Provided by System (Indoor)

Floor# Number Provided by System (Indoor)

Provided by System (Indoor or
ZoneID Number
Outdoor)
Indoor or outdoor (Provided by
Type Text
System)

281
 Table Attribute Name Data Type Description

MaterialPositionID Number Primary Key (Provided by System)

MaterialID Number from 3D BIM Model

Date Date/Time Provided by System

Time Date/Time Provided by System

Latitude Number Provided by System (Outdoor)

Object
Temporal Longitude Number Provided by System (Outdoor)
Location
Xk Number Provided by System (Indoor)

Yk Number Provided by System (Indoor)

Floor# Number Provided by System (Indoor)

Provided by System (Indoor or
ZoneID Number
Outdoor)
Indoor or outdoor (Provided by
Type Text
System)
ZoneID Number Primary Key (from 3D BIM Model)

Floor# Number 3D BIM Model

Zone
ProjectID Number Provided by System
Indoor or outdoor (from 3D BIM
Type Text
Model)
ReferenceRFIDTagID Number Primary Key (Provided by System)

xi Number User Input

Reference
RFID Tag
yi Number User Input

ZoneID Number User Input

SoilID Number Primary Key (Provided by System)

FillFactor Number User Input

SwellSector Number User Input

Soil
Denisty Number User Input

Type Text User Input

Description Text User Input

SegmentID Number Primary Key (Provided by System)

Road ProjectID Number Provided by System

DumpingAreaID Number Provided by System

282
 Table Attribute Name Data Type Description

Type Text Provided by GIS Sub-Module

StartLongitude Number Provided by GIS Sub-Module

StartLatitude Number Provided by GIS Sub-Module

StartAltitude Number Provided by GIS Sub-Module

EndLongitude Number Provided by GIS Sub-Module

EndLatitude Number Provided by GIS Sub-Module

EndAltitude Number Provided by GIS Sub-Module

Distance Number Provided by GIS Sub-Module

Grade Resistance Number Provided by System

Rolling Resistance Number Provided by System

DumpingAreaID Number Primary Key (User Input)

PolygonID Number User Input

Dumping Area Location Text User Input

Owner Text User Input

Capacity Number User Input

ReaderID Number Primary Key (User Input)

ModelNumber Number User Input

Manufacture Text User Input

Gate RFID
Reader
ReaderType Text User Input

ProjectID Number User Input

DumpingAreaID Number User Input

Primary Key (Gate RFID Captured
Serial Number
Data)
Primary Key (Gate RFID Captured
Data File Attachment
Data)
RSSI Number Gate RFID Captured Data
Gate RFID
Reader File ReaderDate Date/Time Gate RFID Captured Data

ReaderTime Date/Time Gate RFID Captured Data

RFIDTagID Number Gate RFID Captured Data

ReaderID Number Gate RFID Captured Data

283
 Table Attribute Name Data Type Description

ResourceID Number Primary Key (User Input)

ActivityID Number User Input

Resource
CostAccount Number User Input

RFIDTagID Number User Input

MaterialID Number Primary Key (from 3D BIM Model)

Manufacturor Text 3D BIM Model

Type Text 3D BIM Model

Material
ResourceID Number User Input

Xf Number 3D BIM Model

Yf Number 3D BIM Model

LaborID Number Primary Key (User Input)

ResourceID Number User Input

FirstName Text User Input

LastName Text User Input

Trade Text User Input

BirthDay Date/Time User Input

Labor
Address Text User Input

HireDate Date/Time User Input

Phone# Number User Input

SIN# Number User Input

Wage Number User Input

OverTimeRate Number User Input

EquipmentID Number Primary Key (User Input)

ResourceID Number User Input

EfficiencyFactor Number User Input

Equipment
Manufacturor Number User Input

LicensePlate Number User Input

PurchaseDate Number User Input

284
 Table Attribute Name Data Type Description

TruckID Number Primary Key (User Input)

ModelNumber Number User Input

EquipmentID Number User Input

Truck EmptyWeight Number User Input

MaxWeight Number User Input

HeapedCapacity Number User Input

GPSID Number User Input

GPSID Number Primary Key (User Input)

ModelNumber Number User Input

Truck GPS
Manufacture Text User Input

TruckID Number User Input

Serial Number Primary Key (GPS Captured Data)

Data File Attachment Primary Key (GPS Captured Data)

Date Date/Time GPS Captured Data

Time Date/Time GPS Captured Data

GPSID Number GPS Captured Data

Truck GPS File Heading Number GPS Captured Data

Speed (KM/h) Number GPS Captured Data

Latitude Number GPS Captured Data

Longitude Number GPS Captured Data

Altitude Number GPS Captured Data

Information Text GPS Captured Data

Date Date/Time Primary Key (Generated by System)

Cycle# Number Primary Key (Generated by System)

TruckID Number User Input

Truck Status
Event1 Date/Time Generated by System

Event2 Date/Time Generated by System

Event3 Date/Time Generated by System

285
 Table Attribute Name Data Type Description

Event4 Date/Time Generated by System

Event5 Date/Time Generated by System

LoadingTime Number Generated by System

TravelTime Number Generated by System

DumpingTime Number Generated by System

ReturnTime Number Generated by System

CycleTime Number Generated by System

Generated by System =
ActualProductivity Number
(TruckCapacity / CycleTime)
ExcavatorID Number Primary Key (User Input)

ModelID Number User Input

EquipmentID Number User Input

Excavator
BucketCapacity Number User Input

MaxDumpingHeight Number User Input

MaxDiggingDepth Number User Input

SoundID Auto Number Primary Key (Generated by System)

Based on User Location (Provided by
Position ID Number
System)
Date Date/Time Provided by System

Sound Record Time Date/Time Provided by System

Based on User Location (Provided by
Zone ID Number
System)
Sound Attachment Data captured by the user

Description Text Entered by the user

VideoID Auto Number Primary Key (Generated by System)

Based on User Location (Provided by
Position ID Number
System)
Date Date/Time Provided by System
Video Record
Time Date/Time Provided by System
Based on User Location (Provided by
Zone ID Number
System)
Video Attachment Data captured by the user

286
 Table Attribute Name Data Type Description

Description Text Entered by the user

ImageID Auto Number Primary Key (Generated by System)

Based on User Location (Provided by
Position ID Number
System)
Date Date/Time Provided by System

Image Time Date/Time Provided by System

Based on User Location (Provided by
Zone ID Number
System)
Image Attachment Data captured by the user

Description Text Entered by the user

HandWrittenNotes ID Auto Number Primary Key (Generated by System)

Based on User Location (Provided by
Position ID Number
System)
Date Date/Time Provided by System
Hand Written
Time Date/Time Provided by System
Notes
Based on User Location (Provided by
Zone ID Number
System)
HandWrittenNotes Attachment Data captured by the user

Description Text Entered by the user

BarCode ID Auto Number Primary Key (Generated by System)

Based on User Location (Provided by
Position ID Number
System)
Date Date/Time Provided by System

Barcode Time Date/Time Provided by System

Based on User Location (Provided by
Zone ID Number
System)
BarCode Text Data captured by the user

Description Text Entered by the user

3DScan ID Auto Number Primary Key (Generated by System)

Based on User Location (Provided by
Position ID Number
System)
3D Scan Date Date/Time Provided by System

Time Date/Time Provided by System

Based on User Location (Provided by
Zone ID Number
System)

287
 Table Attribute Name Data Type Description

3DScan Attachment Data captured by the user

Description Text Entered by the user

ControlSensorID Number Primary Key (User Input)

ModelNumber Number User Input

Control Sensor
Manufacture Text User Input

TruckID Number User Input

Primary Key (Control Sensor
Serial Number
Captured Data)
Primary Key (Control Sensor
Control Sensor Data File Attachment
Captured Data)
File
Date Date/Time Control Sensor Captured Data

Time Date/Time Control Sensor Captured Data

ControlSensorID Number Control Sensor Captured Data

Status Binary (On/Off) Control Sensor Captured Data

ReaderID Number Primary Key (User Input)

ModelNumber Number User Input

Excavator
Manufacture Text User Input
RFID Reader
ReaderType Text User Input

ExcavatorID Number User Input

Primary Key (Excavator RFID
Serial Number
Captured Data)
Primary Key (Excavator RFID
Data File Attachment
Captured Data)
RSSI Number Excavator RFID Captured Data
Excavator
RFID Reader ReaderDate Date/Time Excavator RFID Captured Data
File
ReaderTime Date/Time Excavator RFID Captured Data

RFIDTagID Number Excavator RFID Captured Data

ReaderID Number Excavator RFID Captured Data

CCTVCmeraID Number Primary Key (User Input)

ModelNumber Number User Input

CCTV Camera
Manufacture Text User Input

ProjectID Number User Input

288
 Table Attribute Name Data Type Description

LocationX Number User Input

LocationY Number User Input

LocationZ Number User Input

ImageID Auto Number Primary Key (Generated by System)

Date Date/Time Provided by System

Time Date/Time Provided by System

CCTV Camera
Image
Image Attachment Data captured by the Camera

Description Text Entered by the user

CCTVCamerID Number Provided by System

3DScannerID Number Primary Key (User Input)

ModelNumber Number User Input

3D Scanner
Manufacture Text User Input

ProjectID Number User Input

3DScan ID Auto Number Primary Key (Generated by System)

Based on Scanner Location
Position ID Number
(Provided by System)
Date Date/Time Provided by System

3D Scan File Time Date/Time Provided by System

Based on Scanner Location
Zone ID Number
(Provided by System)
3DScan Attachment Data captured by the scanner

Description Text Entered by the user

ObjectID Number Primary Key (from 3D BIM Model)

Family Text 3D BIM Model

Object Type Text 3D BIM Model

Description Text Entered by the user

Quantity Number 3D BIM Model

289
 10 Appendix C: Data Collection

Table C-1: Genomics tablet PC data acquisition schedule

Technology
ID Date
Images Barcode RFID GPS
1 Sep. 15, 2010
2 Sep. 22, 2010
3 Oct. 06, 2010
4 Oct. 13, 2010
5 Oct. 15, 2010
6 Oct. 20, 2010
7 Oct. 22, 2010
8 Oct. 27, 2010
9 Oct. 29, 2010
10 Nov. 03, 2010
11 Nov. 05, 2010
12 Nov. 10, 2010
13 Nov. 17, 2010
14 Nov. 24, 2010
15 Dec. 01, 2010
16 Dec. 03, 2010
17 Dec. 08, 2010
18 Dec. 14, 2010
19 Dec. 15, 2010
20 Jan. 13, 2010
21 Jan. 20, 2010
22 Jan. 27, 2010
23 Feb. 03, 2010
24 Feb. 10, 2010
25 Mar. 01, 2010
26 Mar. 17, 2010
27 Mar. 23, 2010
28 Apr. 23, 2010
29 Apr. 26, 2010
30 May 11, 2010

290
 Table C-2: Genomics 3D scanning schedule

ID Date Type Remarks

1 Sept., 15, 2010 External 5 different scanning positions

2 Oct., 13, 2010 External 5 different scanning positions

3 Oct., 15, 2010 Internal Second Floor, 2 different scanning positions

4 Oct., 20, 2010 External 4 different scanning positions

5 Oct., 22, 2010 Internal Second Floor, 2 different scanning positions

6 Oct., 27, 2010 External 5 different scanning positions

7 Oct., 29, 2010 Internal First Floor, 2 different scanning positions

8 Nov., 3, 2010 External 1 scanning position, 5 different resolution

9 Nov., 5, 2010 Internal First Floor, 2 different scanning positions

10 Nov., 10, 2010 Internal Second Floor, 2 different scanning positions

11 Nov., 11, 2010 Internal First Floor, 2 different scanning positions

12 Dec, 1, 2010 Internal Second Floor, 2 different scanning positions

13 Dec, 3, 2010 Internal First Floor, 2 different scanning positions

14 Dec, 8, 2010 Internal Second Floor, 1 position, 9 different resolution

291
 Table C-3: LADAR specification

Attribute Specification
Horizontally up to 360 o
Scanning Range
Vertically up to 150 o
Positioning accuracy ± 0.01 gon
Scanning speed 40 gon/sec
Measurement range 2m up to 200m
Measurement accuracy ± 2cm

Horizontal

Vertical
Figure C-1: Vertical and horizontal attributes

292
 Table C-4: Outdoor scanning attributes (September 15, 2010)

Point 1 Point 2 Point 3 Point 4 Point 5

Day Attributes
Vertical
Range & (140o,0.5o) (140o,0.5o) (140o,0.5o) (132o,0.5o) (130o,0.5o)
Resolution
Horizontal
Range & (135o,0.5o) (120o,0.5o) (150o,0.5o) (100o,0.5o) (150o,0.5o)
Resolution
Sep.,
Duration
15, 18 16 20 13 18
(min)
2010
Planned #
of point 76151 67722 84581 53265 78561
Clouds
Actual # of
point 53917 41150 61143 34872 63753
Clouds

Figure C-2: Outdoor scanning positions

293
 Table C-5: First floor indoor scanning attributes (October 15, 2010)

Point 1 Point 2
Day Attributes

Vertical Range & Resolution (145o,0.5 o) (145o,0.5o)

Horizontal Range & Resolution (354o,0.5o) (354o,0.5o)

Oct., 29,
Duration (min) 48 48
2010
Planned # of point Clouds 206901 206901

Actual # of point Clouds 201465 205061

Figure C-3: First floor scanning positions

294
 Table C-6: Outdoor scanning with different resolution (November 3, 2010)

Resolution
Day Attributes o o
5 1 0.5o 0.25o 0.2o
Vertical
140o 140o 140o 140o 140o
Range
Horizontal
130o 130o 130o 130o 130o
Range
Duration
Nov., 2 9 18 34 42
(min)
3,
Planned #
2010
of point 783 18471 73341 292281 456351
Clouds
Actual # of
point 542 13254 53022 211690 295475
Clouds

0.5 o
5o

1o 0.25 o

0.2 o
Figure C-4: Different resolution 3D scanner images

295
Figure C-5: Outdoor scanning and curtain walls glazing created noise

296
 11 Appendix D: Productivity of Earthmoving Operations in

Planning

This appendix provide the methodology utilized to calculate the planned upper and

lower limit productivity of trucks in earthmoving operations, for more details refer to Montaser

et al. (2011 and 2012). The process of estimating the productivity of the excavator(s) and of

the hauling unit(s) is described below. Equation (D.1) is used to estimate the excavator

production rate (Peurifoy and Schexnayder 2006). Since each variable in that equation has a

range of values, see Table (D-1), the application of it leads to a range of production rates for

the excavator. The production rate of a hauling truck is calculated using Equation (D.2) until

(D.7).

3600 6 Q 6 F E
Excavator Production Rate = X X Swell Factor (D.1)
t 60 min/hour

Where

Q is the heaped bucket capacity in loose cubic yard,

F is the bucket fill factor,

T is the excavator cycle time in seconds, and

E is an efficiency factor in minutes per hour

fcaeq e r e ?
Balance number of excavator bucket loads = (D. 2)
aeq e r e ?

Truck load weight = Excavator bucket capacity X Number of bucket loads to the truck X

Density of excavated soil (D. 3)

297
Truck loading time = Number of excavator bucket loads X Excavator cycle time (D.4)

Table D-1: Values of variables affecting excavator production rates

Variable Optimistic Pessimistic Remarks

Bucket Capacity Considered as constant As per contractor selection

Soil Type Considered as constant As per the site investigation report

Fill Factor 110% 100% As per soil type - (Peurifoy and Schexnayder, 2006)

Cycle Time(sec) t min t max Equipment manufacturer

Swell Factor Considered as constant As per the site investigation report

Efficiency
55 40 (Peurifoy and Schexnayder, 2006)
(min\hr)

Travel time = Segment length / Travel speed (D.5)

More details on the calculation of the total resistance and on the use of performance

charts can be found in Peurifoy and Schexnayder (2006). From the appropriate truck

performance chart, the maximum speed associated with the calculated total resistance for each

segment can be found and then the travel time determined using Equation (D.5). The time

required to complete the earthmoving operations can be calculated using Equation (D.8).

Equation (D.7) has many variables, and each variable has a range of values, which leads to a

range of production rates for the truck under consideration. The optimistic and pessimistic

production thresholds are shown in Table (D-2).

fcaeq ?eg f b b
Number of trucks (N) = s e t cu vf b b
(D. 6)

w:9
sdd e e? - 1 ` . d caeq caeq r e ? e? { gg v | e c
xy
Truck Production Rate = fcaeq ?eg b b
(D.7)

Time required = Quantity of excavation / Fleet production rate (D.8)

298
 Table D-2: Values of the variables affecting the truck production rate

Optimistic Pessimistic Remarks

Variable
Truck Capacity Considered as a constant As per manufacturer data sheets

Soil Type Considered as a constant As per site investigation report

Swell Factor Considered as a constant As per site investigation report

RR (lb\ton) 40 65 As per road surface type (for asphalt roads)

Efficiency (min\hr) 55 40 (Peurifoy and Schexnayder 2006)

The project selected for this study was Genomics building, described in details in

Chapter 4. Figure (D-1) depicts the road profile of the haul road as extracted from Google

Earth. Accordingly, a sample of the project data was organized as shown in Table (D-3). The

rolling resistance was considered to be equal to 55 lb/ton for the asphalt roads on the route.

Table (D-4) depicts the range of values affecting the production rate of the hauling units. Using

the data presented above, assuming 8-hour workdays and varying the number of trucks used in

the operation, the upper and lower limits of the production rate of each excavation fleet

considered were calculated presented in Table (D-5).

Figure D-1: Google Earth generated road profile

299
 Table D-3: Sample of hauling units’ travel times

Speed Time
Description Distance(ft) RR (%) GR (%) TR (%)
(mph) (min)
847 2.75 -0.432 2.318 28 0.34375

433 2.75 -0.69088 2.059 30 0.164015

Hauling 945 2.75 -0.179 2.571 27 0.397727

1805 2.75 0.610306 3.360 23 0.891798

815 2.75 0 2.750 26.5 0.349485

850 2.75 0.432 3.182 32.5 0.297203

830 2.75 0 2.750 33.5 0.281547

Returning 1900 2.75 -0.61031 2.140 34 0.635027

2200 2.75 -1.016 1.734 34 0.735294

1755 2.75 -1.508 1.242 34 0.586564

Table D-4: Values of the durations governing truck production rate

Upper Limit Lower Limit

Variable
Load Time (min) 2.75 3.67

Dump Time (min) 1.5 2.5

Travel Time (min) 17.76 21.5

Cycle Time (min) 22.01 27.67

300
 Table D-5: Lower limits and upper limits of productivity

Lower limit Upper limit

No. of Production The Production The
Duration Duration Duration Duration
Trucks Rate Controlling Rate Controlling
(hr) (Day) (hr) (Day)
(bcy/hr) Equipment (bcy/hr) Equipment
2 38.898 349.163 44 67.239 201.993 26
Truck Truck
3 58.348 232.775 30 100.859 134.662 17
Truck Truck
4 77.797 174.581 22 134.479 100.996 13
Truck Truck
5 97.246 139.665 18 168.099 80.797 11
Truck Truck
6 116.696 116.387 15 201.719 67.331 9
Truck Truck
7 136.145 99.760 13 235.339 57.712 8
Truck Truck
8 146.772 92.538 12 268.959 50.498 7
Excavator Truck
9 146.772 92.538 12 269.082 50.475 7
Excavator Excavator
10 146.772 92.538 12 269.082 50.475 7
Excavator Excavator

301
 12 Appendix E: Computer Implementation

This appendix includes sample of sequence diagrams and “Visual C# .Net” source code for

the developed BIM+ software presented in Chapter 6 of this thesis. Similar to BIM+, class

diagrams, sequence diagram were developed for various functions of Control+ and Truck+

prototype software with its respective “Visual C# .Net” source code.

Figure E-1: Add to visible button (Sequence diagram)

302
Figure E-2: Add to hidden button (Sequence diagram)

303
Figure E-3: Add all to hidden button (Sequence diagram)

Figure E-4: Add all to visible button (Sequence diagram)

304
BIM+ Vsual C# .Net Code

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Xml.Linq;

namespace RevitCreateGroupAddins.Entities
{
public class xmlEntity
{
string _ActivityName;

public string ActivityName

{
get { return _ActivityName; }
set { _ActivityName = value; }
}
string _FloorName;

public string FloorName

{
get { return _FloorName; }
set { _FloorName = value; }
}
string _AreaName;

public string AreaName

{
get { return _AreaName; }
set { _AreaName = value; }
}

public static List<xmlEntity> GetXmlEntity()

{
List<xmlEntity> lstXmls = new List<xmlEntity>();

System.Configuration.Configuration config =
Helper.MSProjectHelper.GetConfig();
string strLinkFilePath = Environment.GetFolderPath(
Environment.SpecialFolder.ProgramFilesX86)+
config.AppSettings.Settings["LinkFilePath"].Value;

XDocument xRoot = XDocument.Load(strLinkFilePath);

List<XElement> eles = xRoot.Descendants("Link").ToList();
foreach (var item in eles)
{
XElement xele = new XElement("Link");

lstXmls.Add(new xmlEntity()
{
ActivityName = item.Attribute("Activity").Value,
AreaName = item.Attribute("Area").Value,
FloorName = item.Attribute("Floor").Value
});
}
return lstXmls;

305
 }

public static void DeleteXmlEntities()

{

System.Configuration.Configuration config =
Helper.MSProjectHelper.GetConfig();
string strLinkFilePath = Environment.GetFolderPath(
Environment.SpecialFolder.ProgramFilesX86)+config.AppSettings.Settings["LinkFilePa
th"].Value;

XDocument xRoot = XDocument.Load(strLinkFilePath);

xRoot.Descendants("Link").Remove();
xRoot.Save(strLinkFilePath);
}

public static void DeleteXmlEntity(xmlEntity entity)

{

System.Configuration.Configuration config =
Helper.MSProjectHelper.GetConfig();
string strLinkFilePath = Environment.GetFolderPath(
Environment.SpecialFolder.ProgramFilesX86)+config.AppSettings.Settings["LinkFilePa
th"].Value;

XDocument xRoot = XDocument.Load(strLinkFilePath);

List<XElement> eles = xRoot.Descendants("Link").ToList();

XElement cELe = eles.Where(c => c.Attribute("Activity").Value ==
entity.ActivityName).FirstOrDefault();
if (cELe != null)
{
cELe.Remove();
//eles.Remove(cELe);
xRoot.Save(strLinkFilePath);
}
}

public static void InsertXmlEntities(xmlEntity currententity)

{
System.Configuration.Configuration config =
Helper.MSProjectHelper.GetConfig();
string strLinkFilePath = Environment.GetFolderPath(
Environment.SpecialFolder.ProgramFilesX86)+config.AppSettings.Settings["LinkFilePa
th"].Value;

XDocument xRoot = XDocument.Load(strLinkFilePath);

XElement eleroot = xRoot.Descendants("Links").FirstOrDefault();

XElement ele = new XElement("Link");

XAttribute attActivityName = new XAttribute("Activity",
currententity.ActivityName);
XAttribute AreaName = new XAttribute("Area", currententity.AreaName);
XAttribute FloorName = new XAttribute("Floor",
currententity.FloorName);

306
 ele.Add(attActivityName, AreaName, FloorName);

eleroot.Add(ele);
xRoot.Save(strLinkFilePath);
}

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;

namespace RevitCreateGroupAddins.Entities
{
public class RFIDOutputEntity
{

public int Serial { get; set; }

public string DataFile { get; set; }

public string RSSI { get; set; }

public DateTime ReaderDate { get; set; }

public int RFIDTagID { get; set; }

public int TabletPCID { get; set; }

}
}

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;

namespace RevitCreateGroupAddins.Entities
{
public class CustomView
{
string _Name;

public string Name

{
get { return _Name; }
set { _Name = value; }
}

Autodesk.Revit.DB.View _CurrentView;

public Autodesk.Revit.DB.View CurrentView

{
get { return _CurrentView; }
set { _CurrentView = value; }
}
List<CustomElement> _lstElements = new List<CustomElement>();

public List<CustomElement> LstElements

307
 {
get { return _lstElements; }
set { _lstElements = value; }
}

private double _z;

public double Z
{
get { return _z; }
set { _z = value; }
}

public Autodesk.Revit.DB.ElementId id { get; set; }

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using Autodesk.Revit.UI;
using Autodesk.Revit.DB;
using Autodesk.Revit.DB.Architecture;
using RevitCreateGroupAddins.Helper;

namespace RevitCreateGroupAddins.Addins
{

[Autodesk.Revit.Attributes.Transaction(Autodesk.Revit.Attributes.TransactionMode.M
anual)]

[Autodesk.Revit.Attributes.Regeneration(Autodesk.Revit.Attributes.RegenerationOpti
on.Manual)]
public class WalkThroughAddins : IExternalCommand
{
public Result Execute(
ExternalCommandData commandData,
ref string message,
ElementSet elements)
{
UIApplication uiapp = commandData.Application;
UIDocument uidoc = uiapp.ActiveUIDocument;
Document doc = uidoc.Document;

ViewFamilyType vft
= new FilteredElementCollector(doc)
.OfClass(typeof(ViewFamilyType))
.Cast<ViewFamilyType>()
.FirstOrDefault<ViewFamilyType>(x =>
ViewFamily.Section == x.ViewFamily);

ViewSheetSets vs = doc.ViewSheetSets;
foreach (ViewSheetSet item in vs)
{

308
 }
FilteredElementCollector a
= new FilteredElementCollector(doc)
.OfClass(typeof(SpatialElement));
BoundingBoxXYZ sectionBox = null;
foreach (SpatialElement e in a)
{
Room room = e as Room;

if (null != room)
{
sectionBox = RevitHelper.ListRoomData(room);

}
}
using (Transaction tx = new Transaction(doc))
{
tx.Start("Create Wall Section View");

ViewSection.CreateSection(doc, vft.Id, sectionBox);

tx.Commit();
}
return Result.Succeeded;
}
}

using System;
using System.Collections.Generic;
using System.Linq;
using System.Web;
using System.Data.OleDb;
using System.Data;

namespace RevitCreateGroupAddins.Helpers
{
public class HelperFunctions
{
private static HelperFunctions _HelperInstance;

public static HelperFunctions HelperInstance

{
get
{
if (_HelperInstance == null)
{
_HelperInstance = new HelperFunctions();
}
return _HelperInstance;
}
}

private string DBPath;

private string ConnectionString;
private string ActivityHafar;

309
 public OleDbConnection con;
private string mdbConnectionString;
private string mdbcon;

private HelperFunctions()
{
try
{
System.Configuration.Configuration config =
Helper.MSProjectHelper.GetConfig();

DBPath = config.AppSettings.Settings["DataBasePath"].Value;
ConnectionString =
config.AppSettings.Settings["ConnectionString"].Value;
mdbConnectionString = ConnectionString + "";
mdbcon = mdbConnectionString + DBPath;
}
catch (Exception)
{
throw;
}
}

#region Helper Functions

public void OpenDBConnection()

{
try
{

if (con != null)
{
if (con.State == ConnectionState.Open)
{
return;
}
}
con = new OleDbConnection(mdbcon);
con.Open();
}
catch (Exception)
{

// throw;
}
}

public void CLoseDBConnection()

{
try
{
if (con != null)
{
if (con.State == ConnectionState.Open)
{
con.Close();
con.Dispose();
}

310
 }
}
catch (Exception)
{
throw;
}

private DateTime ConvertStringToDateTime(string input, int monthPosition,

int dayPosition)
{
try
{
string[] arr = input.Split('/');
DateTime dt = new DateTime(Convert.ToInt32(arr[2]),
Convert.ToInt32(arr[monthPosition]), Convert.ToInt32(arr[dayPosition]));
return dt;
}
catch (Exception)
{
return new DateTime();
throw;
}

#endregion
}
}

311