Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Computing in
Civil Engineering 2017
Information Modeling and Data Analytics
Selected Papers from the ASCE International
Workshop on Computing in Civil Engineering 2017
Seattle, Washington
June 25–27, 2017
Edited by Ken-Yu Lin, Ph.D.; Nora El-Gohary, Ph.D.;
and Pingbo Tang, Ph.D., P.E.
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
COMPUTING IN CIVIL
ENGINEERING 2017
INFORMATION MODELING AND DATA ANALYTICS
SELECTED PAPERS FROM THE ASCE INTERNATIONAL
WORKSHOP ON COMPUTING IN CIVIL ENGINEERING 2017
June 25–27, 2017
Seattle, Washington
SPONSORED BY
Computing Division
of the American Society of Civil Engineers
EDITED BY
Ken-Yu Lin, Ph.D.
Nora El-Gohary, Ph.D.
Pingbo Tang, Ph.D., P.E.
1801 ALEXANDER BELL DRIVE
RESTON, VIRGINIA 20191–4400
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Published by American Society of Civil Engineers
1801 Alexander Bell Drive
Reston, Virginia, 20191-4382
www.asce.org/publications | ascelibrary.org
Any statements expressed in these materials are those of the individual authors and do not
necessarily represent the views of ASCE, which takes no responsibility for any statement
made herein. No reference made in this publication to any specific method, product, process,
or service constitutes or implies an endorsement, recommendation, or warranty thereof by
ASCE. The materials are for general information only and do not represent a standard of
ASCE, nor are they intended as a reference in purchase specifications, contracts, regulations,
statutes, or any other legal document. ASCE makes no representation or warranty of any
kind, whether express or implied, concerning the accuracy, completeness, suitability, or
utility of any information, apparatus, product, or process discussed in this publication, and
assumes no liability therefor. The information contained in these materials should not be used
without first securing competent advice with respect to its suitability for any general or
specific application. Anyone utilizing such information assumes all liability arising from such
use, including but not limited to infringement of any patent or patents.
ASCE and American Society of Civil Engineers—Registered in U.S. Patent and Trademark
Office.
Photocopies and permissions. Permission to photocopy or reproduce material from ASCE
publications can be requested by sending an e-mail to
[email protected] or by locating a
title in ASCE's Civil Engineering Database (http://cedb.asce.org) or ASCE Library
(http://ascelibrary.org) and using the “Permissions” link.
Errata: Errata, if any, can be found at https://doi.org/10.1061/9780784480823
Copyright © 2017 by the American Society of Civil Engineers.
All Rights Reserved.
ISBN 978-0-7844-8082-3 (PDF)
Manufactured in the United States of America.
Computing in Civil Engineering 2017
iii
Preface
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Welcome to Seattle, the Emerald City in Washington!
The 2017 ASCE International Workshop on Computing in Civil Engineering
(IWCCE) was held in Seattle from June 25-27, 2017. The workshop was hosted by
the University of Washington with sponsorship from ASCE’s Computing Division.
The workshop is the Computing Division’s major meeting event and is held biannually in the United States, with participation from scholars worldwide.
The workshop has a long history of success and serves as a platform for sharing
research innovation as well as valuable lessons. We introduced several pioneering
changes this year, including the inaugural all-stakeholder meeting for the Computing
Division. We had a strong and engaged Technical Committee which provided
rigorous reviews for the abstracts and full papers, with each submission being
reviewed by at least two members of our Technical Committee.
The 2017 workshop, as a standalone event, received more than 300 abstracts, 184 full
papers, and 32 extended abstracts for the poster and demonstration sessions. The
participation from our growing community has set a record and a total of 162 full
papers were accepted and included in the proceedings. Among these papers, Building
Information Modeling and Civil Information Modeling formed the most popular
technical interests while Energy, Sustainability and Resilience topped the list of
application contexts.
We would like to thank the Department of Construction Management at The
University of Washington for its support of the workshop. We are also grateful for the
guidance from the Computing Division’s Executive Committee and the assistance
from ASCE.
We hope that you enjoyed the technical sessions at the workshop and had a
memorable and meaningful IWCCE experience in Seattle this year.
Ken-Yu Lin, Ph.D.
Chair, Organizing Committee, IWCCE 2017
Nora El-Gohary, Ph.D.
Chair, Technical Committee, IWCCE 2017
Pingbo Tang, Ph.D., P.E.
Vice Chair, Organizing Committee, IWCCE 2017
© ASCE
Computing in Civil Engineering 2017
iv
Acknowledgments
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Special thanks are due to the following individuals at the University of Washington
for their continuous and tireless support throughout the organization of the workshop:
Name
Julie Angeley
Mark Baratta
Brian Vogt
Zhenyu Zhang
Title
IWCCE Local Administrator
IWCCE Local IT Lead
IWCCE Local Web Consultant
IWCCE Secretary
A sincere appreciation goes to the Microsoft Corporation for providing the editors
free access to Microsoft’s Academic Conference Management Service and for
customizing the online platform for the workshop.
The editors would also like to thank the following Technical Committee members for
their assistance and effort with the paper review and selection process:
Name
Abbas Rashidi
Albert Chen
Ali Mostafavi
Amin Hammad
Amir Behzadan
Andre Barbosa
Andre Borrmann
Atefeh Mohammadpour
Auroop R. Ganguly
Baabak Ashuri
Behzad Esmaeili
Bon-Gang Hwang
Brenda McCabe
Burcin Becerik
Carl Haas
Carlos Caldas
Carol Menassa
Changbum Ahn
Chao Wang
Chen Feng
© ASCE
Institution
Georgia Southern University
National Taiwan University
Florida International University
Concordia University
Missouri State University
Oregon State University
The Technical University of Munich
Indiana University-Purdue University Fort Wayne
Northeastern University (United States)
Georgia Institute of Technology
University of Nebraska-Lincoln
National University of Singapore
University of Toronto
University of Southern California
University of Waterloo
University of Texas at Austin
University of Michigan
University of Nebraska-Lincoln
Louisiana State University
Mitsubishi Electric Research Laboratories
Computing in Civil Engineering 2017
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Chien-Cheng Chou
Chimay Anumba
Christian Koch
David Lattanzi
Dong Zhao
Dulcy Abraham
Ebrahim Karan
Eduardo Santos
Esin Ergen
Fadi Castronovo
Farrokh Jazizadeh
Fei Dai
Feng Li
Fernanda Leite
Frank Boukamp
Frederic Bosche
Guangbin Wang
Hanbin Luo
Hubo Cai
Ian Smith
Ioannis Brilakis
Islam El-adaway
Ivan Mutis
Jack Cheng
Jiansong Zhang
Jiayu Chen
Jie gong
Jing Du
Jinyue Zhang
John Messner
John Taylor
Jun Yang
Justin Ker-Wei Yeoh
Koji Makanae
Lu Zhang
Lucio Soibelman
Mani Golparvar-Fard
Mario Berges
Menghan Tsai
Michael Olsen
Ming Lu
© ASCE
v
National Central University (Taiwan)
University of Florida
University of Nottingham
George Mason University
Michigan State University
Purdue University
Millersville University
University of Sao Paulo
Istanbul Technical University
California State University East Bay
Virginia Tech
West Virginia University
Research Institute of Highway (China)
University of Texas at Austin
Royal Melbourne Institute of Technology
Heriot-Watt University
Tongji University
Huazhong University of Science and Technology
Purdue University
Ecole Polytechnique Federale (Switzerland)
Cambridge University
University of Tennessee
Illinois Institute of Technology
Hong Kong University of Science and Technology
Western Michigan University
City University of Hong Kong
Rutgers University
Texas A&M University
Tianjin University
Penn State University
Georgia Tech
Northwestern Polytechnical University (China)
National University of Singapore
Miyagi University
Florida International University
University of Southern California
University of Illinois at Urbana-Champaign
Carnegie Mellon University
National Taiwan University
Oregon State University
University of Alberta
Computing in Civil Engineering 2017
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Mounir El Asmar
Nai-Wen Chi
Nan Li
Nipesh Pradhananga
Nobuyoshi Yabuki
Omar El-Anwar
Oswald Chong
Paul Goodrum
Pin-Chao Liao
Ray Issa
Renate Fruchter
Ren-Jye Dzeng
Reza Akhavian
Rishee Jain
Robert Amor
Rucheng Xiao
Rui Liu
Saiedeh Razavi
Sanghoon Lee
SangHyun Lee
SangUk Han
Semiha Ergan
Seokho Chi
Shang-Hsien Hsieh
Sheryl Staub-French
Steven Ayer
Takashi Michikawa
Tamer El-Diraby
Timo Hartmann
Walid Tizani
Wen Xiong
Xiangyu Wang
Xianzhong Zhao
Xiaowei Luo
Xiaolong Xue
Xuesong Liu
Xuesong Shen
Yelda Turkan
Yimin Zhu
Ying Zhou
Yong Cho
© ASCE
vi
Arizona State University
National Taiwan University
Tsinghua University
Florida International University
Osaka University
Cairo University
Arizona State University
University of Colorado at Boulder
Tsinghua University
University of Florida
Stanford University
National Chiao-Tung University
California State University East Bay
Stanford University
University of Auckland
Tongji University
University of Florida
McMaster University
University of Hong Kong
University of Michigan
University of Alberta
New York University
Seoul National University
National Taiwan University
University of British Columbia
Arizona State University
RIKEN
University of Toronto
Technical University of Berlin
University of Nottingham
Southeast University
Curtin University
Tongji University
City University of Hong Kong
Harbin Institute of Technology
Carnegie Mellon University
University of New South Wales
Oregon State University
Louisiana State University
Huazhong University of Science and Technology
Georgia Institute of Technology
Computing in Civil Engineering 2017
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Youngjib Ham
Yunfeng Chen
Zhenhua Zhu
Zheng Yang
Zhiliang Ma
vii
Florida International University
Georgia Southern University
Concordia University
Stanford University
Tsinghua University
Finally, the editors would also like to thank the following Poster and Demonstration
Organization Committee members for their help with the related review process:
Name
Cheng Zhang
Hamid Abdirad
Jiawei Chen
Kadir Amasyali
Kaijian Liu
Lufan Wang
Luming Shang
Vamsi Sai Kalasapudi
Xuan Lv
Zhenyu Zhang
© ASCE
Institution
Arizona State University
University of Washington
Arizona State University
University of Illinois at Urbana-Champaign
University of Illinois at Urbana-Champaign
University of Illinois at Urbana-Champaign
University of Washington
Arizona State University
University of Illinois at Urbana-Champaign
University of Washington
Computing in Civil Engineering 2017
viii
Contents
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Asset and Facility Management
Optimizing Railroad Bridge Networks Management Using Mixed Integer
Linear Programming and Genetic Algorithm ......................................................... 1
Amirhosein Jafari, Guillermo Pérez, Fernando Moreu, and Vanessa Valentin
The Utilization of an Asset Safety Identification Tool (ASIT) to
Support Safety during Facilities Management ...................................................... 10
Eric M. Wetzel, Jason Lucas, and Walid Y. Thabet
Risk-Aware Multi-Objective Optimization of Capital Structure for
Private Financing in Infrastructure Projects ......................................................... 18
Shuai Li and Hubo Cai
Integration of BIM and Utility Sensor Data for Facilities Management ............. 26
Kamal Suprabhas and Hazar Nicholas Dib
Unsupervised Recognition of Volumetric Structural Components from
Building Point Clouds............................................................................................... 34
Jingdao Chen, Yihai Fang, and Yong K. Cho
Defining a Taxonomy for Virtual 3D City Model Use Cases with a
Focus on Facility Asset Management—A Virtual Campus Case Study .............. 43
Zhouqian Jiang, John I. Messner, and Craig R. Dubler
BIM and CIM
BIM to Facilities Management: Presenting a Proven Workflow for
Information Exchange .............................................................................................. 51
Alireza Borhani, Hyun Woo Lee, Carrie Sturts Dossick, Laura Osburn,
and Marc Kinsman
Concepts for Formal Modeling and Management of Building Design
Options ....................................................................................................................... 59
Hannah Mattern and Markus König
Integrating BIM and Optimization Techniques for Enhanced Tower
Crane Planning ......................................................................................................... 67
Yuanshen Ji, Bharathwaj Sankaran, Jiyong Choi, and Fernanda Leite
© ASCE
Computing in Civil Engineering 2017
Applied BIM: AMT and MTSP Integrated Approach for the
Interior Patrol Routing Problem ............................................................................. 75
Chun-Hao Chen and Albert Y. Chen
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
A Semi-Automatic Approach to Detect Structural Components from CAD
Drawings for Constructing As-Is BIM Objects ..................................................... 84
Qiuchen Lu and Sanghoon Lee
Method of Bridge Structural Analysis Based on Bridge Information
Modeling .................................................................................................................... 92
Rucheng Xiao, Yu Lian, Bin Sun, Xinwei Zhao, Zhao Liu, and Pingbo Tang
Post-Earthquake Fire Simulations of Buildings Considering the
Seismic Damage of Sprinkler Systems .................................................................. 101
Zhen Xu, Zongcai Zhang, and Xinzheng Lu
A Framework for Rule-Based Validation of IFC Space Boundaries for
Building Energy Analysis ....................................................................................... 110
Huaquan Ying and Sanghoon Lee
3D Model-Based Quantity Take-Off for Construction Estimates ...................... 118
Pengxiang Alex Han, Ming-Fung Francis Siu, Simaan AbouRizk, Di Hu,
and Ulrich Hermann
Model-Based Benchmarking for Healthcare Projects:
System Requirements and Demonstration ........................................................... 125
Jiyong Choi, Fernanda Leite, and Daniel P. de Oliveira
Linking BIM and GIS Models in Infrastructure by
Example of IFC and CityGML .............................................................................. 133
S. Vilgertshofer, J. Amann, B. Willenborg, A. Borrmann, and T. H. Kolbe
Automated Wood Construction Cost Estimation ................................................ 141
Temitope Akanbi and Jiansong Zhang
Investigating Building Sustainability by Applying Sensitivity
Analysis of Impact Factors during Design Stage ................................................. 149
C. Zhang and L. Ong
Refinement of the Visual Code Checking Language for an Automated
Checking of Building Information Models Regarding Applicable
Regulations .............................................................................................................. 157
Cornelius Preidel and André Borrmann
© ASCE
ix
Computing in Civil Engineering 2017
Semantic-Rich 3D CAD Models for Built Environments from
Point Clouds: An End-to-End Procedure ............................................................. 166
Yeritza Perez-Perez, Mani Golparvar-Fard, and Khaled El-Rayes
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Investigation of Leveraging BIM Standards to Facilitate Sustainability
Evaluations from Early Stages of Design ............................................................. 175
Issa J. Ramaji, Pelin Gultekin-Bicer, Raphael W. Crowley, and J. David Lambert
Improving RSSI-Based Indoor Localization Performance by
Integrating BIM ...................................................................................................... 184
Hainan Chen, Xiaowei Luo, and Jian Guo
Framework of Dynamic Daily 4D BIM for Tracking Construction
Progress through a Web Environment ................................................................. 193
Jaehyun Park and Hubo Cai
Benefits of Real-Time Data Driven BIM for FM Departments in
Operations Control and Maintenance .................................................................. 202
Omid Davtalab
Exploring the Body of Knowledge for Building Information Modeling
Implementation Using the Delphi Method ........................................................... 211
Wei Wu, Glenda Mayo, Raja R. Issa, Tammy McCuen, and Deke Smith
A Comprehensive Identification and Categorisation of Drivers,
Factors, and Determinants for BIM Adoption: A Systematic Literature
Review ...................................................................................................................... 220
Ahmed L. Ahmed, John P. Kawalek, and Mohamad Kassem
A Case Study in Data Visualization for Linked Building Information
Model and Building Management System Data .................................................. 228
Jennifer I. Lather, Robert Amor, and John I. Messner
An Automated Reconstruction Approach of Mechanical Systems in
Building Information Modeling (BIM) Using 2D Drawings ............................... 236
Chi Yon Cho and Xuesong Liu
BIM-Integrated System for Automated Value Analysis of Buildings................ 245
Lu Zhang and Nora M. El-Gohary
Keyword-Driven Model View Generation for Civil Infrastructure
Projects .................................................................................................................... 254
Tuyen Le and H. David Jeong
© ASCE
x
Computing in Civil Engineering 2017
xi
BIM-Based Construction Noise Hazard Prediction and Visualization for
Occupational Safety and Health Awareness Improvement ................................ 262
Weile Wei, Chao Wang, and Yongcheol Lee
Data Analytics
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Sentiment Analysis for the Construction Industry: A Case Study of
Weibo in China ....................................................................................................... 270
L. Y. N. Tang, Y. M. Zhang, F. Dai, Y. J. Yoon, and Y. Q. Song
Using Data Driven Methodologies to Identify Patterns in
BAS Data to Support Facility Operations ............................................................ 282
Gokmen Dedemen, Marjan Vakilinezhad, and Semiha Ergan
Towards Automated Inference of Occupant Behavioral Dynamics Using
Plug-Load Energy Data .......................................................................................... 290
Andrew J. Sonta, Perry E. Simmons, and Rishee K. Jain
Deep Active Learning for Civil Infrastructure Defect Detection and
Classification ........................................................................................................... 298
Chen Feng, Ming-Yu Liu, Chieh-Chi Kao, and Teng-Yok Lee
Automated Recognition and Localization of Parking Signs Using
Street-Level Imagery .............................................................................................. 307
Qazaleh Mirsharif, Théophile Dalens, Mehdi Sqalli, and Vahid Balali
Similarity-Based Dependency Parsing for Extracting Dependency
Relations from Bridge Inspection Reports ........................................................... 316
Kaijian Liu and Nora El-Gohary
Data-Driven Residential Building Energy Consumption Prediction for
Supporting Multiscale Sustainability Assessment ............................................... 324
Lufan Wang and Nora M. El-Gohary
Stakeholder Opinion Classification for Supporting Large-Scale
Transportation Project Decision Making ............................................................. 333
Xuan Lv and Nora M. El-Gohary
Predicting Leaks in Natural Gas Distribution Networks Using
Generalized Linear Models .................................................................................... 342
Yasamin H. Tari, Burcu Akinci, Mario Bergés, and Matteo Pozzi
Building Energy Use Modes and Thermal Comfort ............................................ 350
Kadir Amasyali and Nora El-Gohary
© ASCE
Computing in Civil Engineering 2017
Infrastructure Monitoring, Control, and Analysis
Level-of-Expertise Classification for Identifying Safe and Productive
Masons ..................................................................................................................... 359
Abdullatif Alwasel, Mohammad Nahangi, Carl Haas, and Eihab Abdel-Rahman
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
3D Approach for Representing Uncertainties of Underground Utility Data .... 369
L. L. olde Scholtenhuis, S. Zlatanova, and X. den Duijn
Geometry-Based Optimized Point Cloud Compression Methodology for
Construction and Infrastructure Management ................................................... 377
Jiawei Chen, Cheng Zhang, and Pingbo Tang
3D Thermal and Spatial Modeling of a Subway Tunnel: A Case Study ........... 386
Ghassan Al Lafi, Zhenhua Zhu, Thikra Dawood, and Tarek Zayed
Representation Requirements for Laser Scan Based Probabilistic
Condition Assessment of Bridges .......................................................................... 395
Varun Kasireddy and Burcu Akinci
Tracking Structural Deformations via Automated Sample-Based Point
Cloud Analysis ........................................................................................................ 403
Bahman Jafari, Ali Khaloo, and David Lattanzi
An Automatic Robust Point Cloud Registration on Construction Sites ............ 411
Pileun Kim and Yong K. Cho
An Autonomous Video Analysis Method for Crack Detection on Metallic
Surfaces Based on Texture Recognition and Bayesian Data Fusion .................. 420
Fu-Chen Chen, Mohammad R. Jahanshahi, Rih-Teng Wu, and Chris Joffe
A Preliminary Study on Disaster Waste Detection and Volume
Estimation Based on 3D Spatial Information ...................................................... 428
Hyung Taeck Yoo, Hyunwoo Lee, Seokho Chi, Bon-Gang Hwang, and Jinwoo Kim
Smart City and Transportation Systems
EMS Response Actions in Mass Casualty Incidents ............................................ 436
Chang-Chi Chou and Albert Y. Chen
Video-Based Indoor Human Detection for Decision-Making of the
Installation Locations for Automated External Defibrillators ........................... 444
Ching-Chun Chen and Albert Y. Chen
Partition Problem for Optimizing the Deployment of Incident Response ........ 450
Zhi-You Dai, Jhihfu Kang, Ning Li, Likai Yang, and Yu-Ting Hsu
© ASCE
xii
Computing in Civil Engineering 2017
Building Smart Transportation Hubs with Internet of Things to
Improve Services to People with Disabilities........................................................ 458
Jie Gong, Cecilia Feeley, Hao Tang, Greg Olmschenk, Vishnu Nair, Zixiang Zhou,
Yi Yu, Ken Yamamoto, and Zhigang Zhu
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
A Data Integration Framework for Urban Systems Analysis Based on
Geo-Relationship Learning .................................................................................... 467
Zheng Yang, Karan Gupta, Archana Gupta, and Rishee K. Jain
© ASCE
xiii
Computing in Civil Engineering 2017
1
Optimizing Railroad Bridge Networks Management Using Mixed Integer Linear Programming and
Genetic Algorithm
Amirhosein Jafari1; Guillermo Pérez2; Fernando Moreu3; and Vanessa Valentin4
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
1
Dept. of Civil Engineering, Univ. of New Mexico, 210 University Blvd NE, Albuquerque,
87106. E-mail:
[email protected]
2
Dept. of Civil Engineering, Univ. of New Mexico, 210 University Blvd NE, Albuquerque,
87106. E-mail:
[email protected]
3
Dept. of Civil Engineering, Univ. of New Mexico, 210 University Blvd NE, Albuquerque,
87106. E-mail:
[email protected]
4
Dept. of Civil Engineering, Univ. of New Mexico, 210 University Blvd NE, Albuquerque,
87106. E-mail:
[email protected]
NM
NM
NM
NM
Abstract
Railroad management entities in the U.S. are developing new tools to improve the management
of railroad bridge networks, in order to comply with new federal regulations on bridge safety and
to increase their profitability. Decisions about maintenance, repair, and replacement (MRR)
actions are currently prioritized by rating the bridges based on structural inspections and
predictions about the estimated costs of operations. This study proposes a framework for the
management of railroad bridge networks that: (1) utilizes a consequence-based management
approach that considers relationships between displacements, serviceability levels and bridge
MRR decisions; and (2) minimizes the expected value of total network costs by determining the
best MRR decisions based on an annual MRR budget. Through this study, two different
optimization methods are explored in two different scenarios: (I) mixed integer linear
programming (MILP) when the impact of bridge location on costs is insignificant resulting in
linear objective function and constraints; and (II) genetic algorithm (GA) when the impact of
bridge location on costs is significant resulting in nonlinear objective function and constraints. A
case study of a network comprised of 100 railroad bridges is used to demonstrate the proposed
framework. The results show that scenario I leads the optimum MRR decision to replace more
bridges. On the other hand, scenario II leads the optimum MRR decision to more repair or
maintain groups of bridges which are closer to each other.
INTRODUCTION
Railroads deteriorate naturally over time mainly due to usage, environmental effects, and aging
(Mishalani and Gong 2009). On the other hand, railroads in U.S. expect to exceed their
capacities over the next 20 years at many locations within their network. Bridges are a critical
component of railroad infrastructure, as researchers use the term ‘bridge network’ for the
transportation network recognizing that the bridge is the most fragile component in the entire
system (Bocchini and Frangopol 2011). Railroad entities in the U.S. are developing new tools to
improve the management of railroad bridge networks, in order to comply with new federal
regulations on bridge safety and to increase their profitability. An important aspect for improving
capacity of network bridges is annual implementation of maintenance, repair and replacement
(MRR) actions. Railroad managers invest in MRR actions every year to allow the operations of
traffic at the required speeds while guaranteeing safe operations. The available financial
resources, however, are insufficient to implement MRR actions for all of U.S. bridge network.
© ASCE
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Computing in Civil Engineering 2017
Therefore, a MRR management system of a bridge network is required for cost-effective
allocation of these limited funds and consequently prioritizing MRR strategies for the bridge
network.
To determine the implementation of MRR actions to railroad bridges in a network, railroads use
bridge inspection reports (American Railway Engineering and Maintenance-of-Way Association
2014). These inspections are regularly made to assess the facility condition and to forecast its
deterioration. Based on inspection outcomes, MRR actions are performed to reduce deterioration
(Mishalani and Gong 2009). One challenge regarding the existing MRR management systems
based on inspections is that decisions made on the basis of infrastructure condition states alone
may cause serious safety consequences (Frangopol and Liu 2007). Using these inspections, the
current strategy is choosing bridges with more urgent work plans first. Another challenge
regarding the existing MRR management systems is that decisions are not necessarily costeffective.
MRR management of bridges network can be formulated as a combinatorial optimization
problem because there always exist objectives to be optimized (Frangopol and Liu 2007). This
optimization problem can be defined either as a single- or multi-objective problem. The most
widely used objective for MRR planning problems is minimizing the present value of life-cycle
costs while satisfying constraints imposed on important bridge performance measures such as
structural reliability (Liu et al. 1997). More recently, multiple and conflicting performance
indicators (e.g., condition, safety, and durability) along with life-cycle cost have also been
simultaneously considered as separate objective functions in the formulation of bridge
maintenance planning problems (Frangopol and Liu 2007; Liu et al. 1997; Liu and Frangopol
2005). In addition to the aforementioned challenges, the number of MRR strategies may be large
posing problems that are computationally intensive. If a railroad network has n bridges with 2
options of (1) performing or (2) not performing MRR over t years, then the total number of
alternative strategies is 2n × t. For long-term decision making, there will be a large number of
alternative strategies. Besides, the process of decision-making for MRR plan is a complex
procedure because the problem contains many inherent interactions among management
objectives, system elements and planning period.
In order to overcome these challenges, a bridges network MRR management system is required
to: (1) assess and forecast the bridge condition considering tools that go beyond regular
inspections, (2) select the most cost-effective MRR strategy for the network, and (3) conduct the
decision-making process in a reasonable time. This study proposes a framework for the
management of railroad bridge networks that: (1) utilizes a consequence-based management
approach that considers relationships between displacements, serviceability levels and bridge
MRR decisions; and (2) minimizes the expected value of total network costs by determining the
best MRR decisions based on an annual MRR budget. Through this study, two different
optimization methods are explored in two different scenarios: (I) mixed integer linear
programming (MILP) when the impact of bridge location on costs is insignificant resulting in
linear objective function and constraints; and (II) genetic algorithm (GA) when the impact of
bridge location on costs is significant resulting in nonlinear objective function and constraints. A
case study of a network comprised of 100 railroad bridges is used to demonstrate the proposed
framework.
© ASCE
2
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Computing in Civil Engineering 2017
OPTIMIZATION TECHNIQUES
Several prior research have used optimization techniques for the management of civil
infrastructure (Liu et al. 1997). Single or multiple performance indicators such as infrastructure
condition, safety and durability along with life-cycle cost are simultaneously considered as
objectives on previous researches (Chen et al. 2015; Hsieh and Liu 2004; Liu and Frangopol
2005; Mishalani and Gong 2009). To tackle the difficult task of finding the best solution for
decision-making in infrastructures network, optimization tools play an important role in
addressing this type of problem.
Since infrastructure investment decision is essentially a 0-1, nonlinear, and constrained problem
(Hsieh and Liu 2004), most of previous research efforts use GA as numerical optimization tools
to solve the problem (Chen et al. 2015; Frangopol and Liu 2007; Liu et al. 1997; Liu and
Frangopol 2005). However, by simplifying the problem to linear equations, a MILP could be
used (Chen et al. 2015; Jacobs 1992). This study uses both MILP and GA for optimization
methods.
MILP. Refers to mathematical programming with continuous and discrete variables and
linearity in the objective function and constraints. The use of MILP is a natural approach
of formulating problems where it is necessary to simultaneously optimize the system
structure (discrete) and parameters (continuous) (Bussieck and Pruessner 2003).
GA. Genetic algorithms, as one type of heuristic algorithms, are stochastic searches and
optimization tools that follow the survival-of-the-fittest principle from the biological
sciences (Goldberg 1989). Since they formally appeared in the 1960s, GAs have been
successfully applied to a wide range of problems because of their ease of implementation
and robust performance (Frangopol and Liu 2007; Jafari and Valentin 2015). GA is a
method for solving both constrained and unconstrained optimization problems.
FRAMEWORK DEVELOPMENT
The developed framework in this study utilizes five main steps: (1) describing the required
components to assess and forecast the bridge condition; (2) inspection to identify the according
MRR actions; (3) calculating network life-cycle costs; (4) formulating the model for
optimization; and (5) using optimization tools find the best MRR strategies. These steps are
described as follow:
Step 1. Model Components
Three components are used to establish the relationship between displacements, serviceability
and bridge MRR actions for each bridge in the railroad network. The first component is the
hazard, assumed as the maximum total transverse displacement of bridges in the inventory,
measured under a loaded train running at the maximum allowable speed for their track class. The
second component are fragility curves elaborated assuming that network bridges have similar
structural properties and that the serviceability of each bridge is independent from the bridge
location within the network. The third component is monitoring that senses the maximum
measurement of transverse displacement for a given bridge under train load, assuming this
maximum displacement represents the bridge condition. These three components are used to
estimate operational costs of bridges based on displacement. For more information about the
integration of aforementioned components, see Moreu et al. (2017).
© ASCE
3
4
Step 2. Inventory
The second step of the framework is defining the bridge inventory, which corresponds here to the
population of bridges owned by the specific railroad for which the MRR policies are being
developed, and their current structural condition. The bridges in the inventory have already been
identified by the railroad to need MRR decisions. This framework assumes that the bridges being
monitored share similar structural properties and operational concerns, so that the measurement
of the hazard of different components of the inventory can be used for relative comparisons
within the inventory.
Step 3. Cost Calculation
The cost to the railroad to maintain the bridge network consists of two components: (i) the MRR
costs for the bridge network, and (ii) the operational costs (OC). OC include the operational
expense (OE), or the expense beyond MRR investments to maintain the bridge network to meet
operational needs, and the lost revenue (LR) to the railroad associated with not doing MRR on
specific bridges. Railroads decrease the speed of trains over bridges of poor condition, assuming
the associated expenses related to traffic delay and decreased network efficiency (Moreu et al.
2017). The total network cost (TNC) is the cost of MRR, plus the OC. Figure 1 shows the
relationship between MRR investments and the TNC associated with maintaining the bridge
network over a specified period of time. Low investments of MRR are associated with high
expenses (OE), i.e. if the bridge network is poorly maintained the cost to operate it will be
higher, whereas large MRR investments increase the bridge condition and reduce the expenses
associated to poor bridge performance (Moreu et al. 2017). The goal is to choose MRR policies
that will minimize the expected value of the TNC.
MRR Costs. The inspection reports recommend different MRR decisions that are based
on structural capacity. The costs of these MRR actions are deterministic and can be
calculated based on the length of the bridge (Moreu et al. 2017).
Operational Costs. The operational costs of bridge per year can be calculated assuming
OEs of unplanned engineering work provided by the railroad as well as LR related to
delay or interruptions to traffic. It is assumed that the operational costs related to the
conditions of the bridge are the only variables in the decision-making, neglecting other
factors including, but not limited to: access to the bridge, financial decisions related to
strategic planning of operations, proximity of related railroad operations to the bridge,
among others.
TNC
Optimum MRR
Cost
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Computing in Civil Engineering 2017
OC
MRR
Annual MRR budget
Figure 1. OC and MRR Cost Relations (Moreu et al. 2017)
Step 4. Model Formulation
Before applying optimization, a decision-making problem is formulated mathematically while
ensuring that its goals and requirements are represented accurately. Binary variables are used as
© ASCE
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Computing in Civil Engineering 2017
decision variables = ( , , … , ) representing the selection of strategies (n is number of
bridges), as shown in Eq. (1):
1
=
(1)
0
The total MRR cost for the bridge network is calculated as shown in Eq. (2):
= ∑
× −
(2)
Where
is the MRR costs of ith bridge and
is the premium to the cost of MRR
based on the proximity between bridges. The assumption for
is as follows:
• $100K: for bridges within 1 mile they have a 100K bonus. This assumes that for
example, if two bridges to be repaired are within a mile of each other, equipment
mobilization expenses will be reduced.
• $50K: for bridges within 5 miles, other partial benefits for sharing some of the equipment
The total OC of the bridges network is calculated as shown in Eq. (3):
= ∑
×
× (1 − ) −
(3)
th
is the operational costs of i bridge and
is premium of OC in bridges
Where
which are close to each other, to avoid penalizing multiple times for slow down the train speed in
a short distance. The assumption for
is as follows:
• 25% of OC for each bridge if there is another bridge with no MRR action within 5 miles
from one side of a bridge
• 50% of OC for each bridge if there are other bridges with no MRR action within 5 miles
from both sides of a bridge
In addition, is defined as operational cost reduction factor for the first year which considers
that maintenance or repair actions may reduce OC due to improved safety. The value of for
replacement is 100%, which assumes that replacing a bridge will eliminate all of its OC in the
first year. The value of
in this study is 100% for replacement, 1% for repair, and ≈0% for
maintenance.
Assuming that the MRR decision should be made based on annual MRR budget, the problem can
be formulized as minimizing OC of the network from Eq. (3), subject to having MRR costs from
Eq. (2) less than or equal to annual MRR budget. The problem can also be applied to multiple
year scenarios (short term). An interest rate of 6% is applied following Moreu et al. (Moreu et
al. 2017).
Step 5. Optimization Methods
Two different scenarios are explored in this study:
I. When the impact of bridge location on costs is insignificant (
=
≈ 0) which results in linear objective and constraints and requires MILP to
find the solution;
II. When the impact of bridge location on costs is significant which results in nonlinear
objective and constraints and requires GA to find the solution.
MILP is able to find the best solution quickly, however it requires an assumption of eliminating
impact of location of bridges network on the costs. To consider the realistic impact of the
location of bridges, more time is required to find an optimum solution using GA.
© ASCE
5
Downloaded from ascelibrary.org by RMIT UNIVERSITY LIBRARY on 01/04/19. Copyright ASCE. For personal use only; all rights reserved.
Computing in Civil Engineering 2017
DATA COLLECTION
A case study of a network comprised of 100 railroad bridges is used to demonstrate the
framework. The data consists of recent timber railroad bridge replacements in the Midwest, but
their specific properties are modified to maintain confidentiality of the company providing this
information. All bridges are in the main route (high traffic), but different traffic levels are shown
for different bridges. The inspection reports recommend different MRR decisions that are
included in the data set. MRR decisions are based on structural capacity and MRR costs are
calculated based on the length of the bridge. MRR costs for each bridge are independent of the
service condition of the bridge. Among these 100 bridges, 30 bridges are assigned for
replacement, 50 bridges are assigned for repair, and 20 bridges are assigned for maintenance.
Without budgetary limit, the railroad would implement all recommended MRR actions to the
entire population of timber bridges to increase the capital, the capacity, the efficiency and the
safety of their railroad network, since replacing timber railroad bridges eliminate risks and
increase capacities. However, the MRR cost of all 100 bridges is $190.3M which is neither
affordable nor urgently needed.
RESULTS AND DISSCUSSION
MRR policies can be modified for a fixed annual MRR budget and minimizing operational costs.
For three consecutive years, the different MRR and operational costs values are calculated and
added together for a network of 100 bridges of similar properties as those described in previous
section. It is assumed that the bridges that are replaced will not have any operational costs on the
current year and future years. Therefore, the bridges which are replaced in year n will be
removed from the decision space from year n+1 and next years. For bridges that are repaired, the
operational cost (calculated by in Eq. 3) is decreased in the current year but will have
operational costs after that if no additional repair or maintenance action is implemented. The
scenarios described in Step 5 of the methodology section are described below.
First Scenario – MILP (I): To find the best amount of annual MRR budget for the case
study (optimum point in Figure 1), the simulation is run for different annual MRR budgets
of $1M to $15M with intervals of $1M. The results are shown in Figure 2a. In this case, the
optimum amount of annual MRR budget for three consequence years is between $3M to
$4M. In addition, the model is run for a $10M annual MRR budget for a 3 –year
consequence management of the bridge network. The optimum MRR decisions for three
years are shown in Figure 3a. In this case, a total of 6, 1, and 2 bridges are selected for
replacement for the first, second, and third year, respectively. Therefore, 9 out of 30 bridges
will be replaced by annual MRR budget of $10M based on the optimum decision.
Second Scenario – GA (II): Similarly, to find out the best amount of annual MRR budget,
the simulation is run for different annual MRR budgets of $1M to $15M with the intervals
of $1M. The results are shown in Figure 2b. In this case, the optimum amount of annual
MRR budget for three consequence years is between $3M to $5M. In addition, the model is
run for a $10M annual MRR budget for a 3-year consequence management of the bridge
network. The optimum MRR decision for three years are shown in Figure 3b. As it is
shown, a total of 2, 1, and 0 bridges are selected for replacement for first, second, and third
year, respectively. Therefore, only 3 out of 30 bridges will be replaced by annual MRR
budget of $10M based on the optimum decision.
© ASCE
6