Prelim Program (v6).xlsx - Transportation Research Forum

Photo credit: Flickr.com Co-hosted by Georgia Tech March 12-14, 2015 Georgia Tech Hotel & Conference Center Atlanta, GA 2015 TRF Program Committee Pat McCarthy TRF Program Vice President Georgia Tech Laurie Garrow Georgia Institute of Technology Baruch Feigenbaum Reason Foundation Art Guzzetti American Public Transportation Association David Ripplinger North Dakota State University Denver Tolliver Upper Great Plains Transportation Institute Sue Hendrickson Upper Great Plains Transportation Institute Journal of the Transportation Research Forum OPPORTUNITY in your inbox OPPORTUNITY TO PUBLISH The Journal of the TRF seeks original manuscripts that are timely and relevant to transportation. Papers may be: • Analytical, presenting either empirical or theoretical research on transportation problems and issues. • Industry issue papers that may be descriptive, addressing current issues in transportation. • Book reviews that provide a critical review of books related to transportation. For more information on the Journal’s editorial policy or to see previous issues, go to http://www.trforum.org/journal/ or contact the editors: Michael W. Babcock (785) 532-4571 James Nolan (306) 966-8412 mwb@ksu.edujames.nolan@usask.ca OPPORTUNITY TO LEARN Learn from leaders and colleagues in Transportation. Three issues are published each year. Recent topics include: • • • • • • • • • • • • • • • Airport choice modeling Infrastructure investment decision making Road crashes Transit efficiency Bridge deterioration Rail rates and revenue Bulk product transportation Road crash factors Centerline rumble strips Tailgating issues Railroad operational performance Aircraft leasing Deregulation and trucking Gasoline pricing Public transit training SECOND FLOOR MEETING ROOM FLOOR SPECIFICATIONS CONFERENCE TWO CONFERENCE THREE CONFERENCE FOUR CONFERENCE FIVE BOARD ROOM BREAK AREA F.E.C. #1 F.E.C. #1 ELEC CONFERENCE CONFERENCE CONFERENCE CONFERSIX SEVEN EIGHT ENCE STAIR NINE M W IDF F.E.C. #1 DN UP DN S CONFERENCE C CONFERENCE A CONFERENCE B OPEN TO BELOW CONFERENCE D CONFERENCE ONE F.E.C. #1 SERVICE ELEV. VESTIBULE BREAK AREA BREAK AREA LAUNDRY CHUTE PREFUNCTION ELEC BUSINESS CENTER UP F.E.C. #1 DN CONFERENCE E SALON I SALON V M GRAND BALLROOM SALON III SALON II GOLD SALON IV SALON VI W F.E.C. #1 WHITE F.E.C. #1 A.V. STOR FRESH AIR INTAKE DN UP BANQUET STORAGE PANTRY IDF FIRST FLOOR MEETING ROOMS 56th Annual Transportation Research Forum MARCH 12-14, 2015 WEDNESDAY, MARCH 11, 2015 4:00 – 6:30 PM Conference One 5:00‐7:00 PM Conference A 7:00 PM Registration Welcome Reception with cash bar and light hors d’oeuvres (paid conference registrants only) Dinner on one's own THURSDAY, MARCH 12, 2015 7:00 AM ‐ 5:00 PM Conference One 7:00 ‐ 8:00 AM Salon IV‐V‐VI Registration Continental Breakfast Welcome/Opening General Session 8:00 ‐ 9:15 AM Salon IV‐V‐VI 9:15 ‐ 9:30 AM The New Era of Technology‐Driven Mobility: Do We Know Where We Are Headed? Allen D. Biehler, Distinguished Service Professor of Transportation Systems and Policy Carnegie Mellon University Break  Presenter 9:30 - 11:30 AM SESSIONS 1.1 - 1.4 Session 1.1: Technological Advances in Vehicle Systems Session Moderator: C. Gregory Bereskin, St. Ambrose University Conference Two How Activities While Traveling Would Affect Travel Behavior in the Autonomous Vehicle Future? Scenario-based Simulations on the Sample of Northern California Commuters  Aliaksandr Malokin, Georgia Institute of Technology Assessing the Implications of Autonomous Trucks for Performance- Based Freight Transportation Planning  Denise Smith, Georgia Institute of Technology Multi-Method Research Approach to Gaining Statewide Acceptance of a Road Usage Charge Program in Oregon  Katherine Schomer, PRR Colleen Gants, PRR Michelle Godfrey, Oregon Department of Transportation John Horvick, DHM Research Ian Nunley, PRR Session 1.2: Uses of Information Technology in Transportation Session Moderator: ManWo Ng, Old Dominion University Conference Four Feasibility of Using Bluetooth Data for Estimating Real Time Delay and Stops at Signalized Intersection Approaches  Golam Sarwar, Agile Assets Inc. Ahmed Abdel‐Rahim, University of Idaho Detection Criteria to Estimate Bluetooth-based Travel Time on Arterial Streets  Srinivas Pulugurtha, The University of North Carolina at Charlotte Md Shah Imran, The University of North Carolina at Charlotte Venkata R. Duddu, The University of North Carolina at Charlotte 56th Annual TRF Program, Atlanta, GA, Page 1 An Effective Data Fusion Algorithm for the Travel Time Reliability Application  Xu Zhang, University of Kentucky Obaidur Rahman Kazi, University of Kentucky Mei Chen, University of Kentucky Optimal Number and Location of Bluetooth Sensors Considering Stochastic Optimal Travel Time Prediction  Hyoshin Park, University of Maryland Session 1.3: Topics in Airline Operations Session Moderator: Carl Scheraga, Fairfield University Conference Eight Airport Congestion Mitigation: A Multimodal Approach  Jared Annexstad, North Dakota State University EunSu Lee, North Dakota State University Airline Consolidation and Hub Abandonment: The Impact on Regional Economies  Michael J. McCormick, George Mason University Estimating Flight-Level Price Elasticities Using Online Data  Laurie Garrow, Georgia Institute of Technology Stacey Mumbower, University of South Carolina Matthew Higgins, Georgia Institute of Technology Airport Efficiency and Energy Use  Svein Bråthen, Molde University College Mozhgan Khazraeianvafadar, Molde University College James Odeck, Molde University College Session 1.4: Measuring the Environmental Impacts of Transportation Activities Session Moderator: Wayne Talley, Old Dominion University Conference Three Forecast of CO2 Emissions from the U.S. Transportation Secor: Estimation from a Double Exponential Smoothing Model  Jaesung Choi, North Dakota State University Economic and Environmental Impacts of Market-Based Measures for the Limitation of Aviation's Full Climate Impact  Janina Scheelhaase, German Aerospace Centre (DLR) Katrin Dahlmann, German Aerospace Centre (DLR) Robert Sausen, German Aerospace Centre (DLR) Hermann Keimel, German Aerospace Centre (DLR) Hendrik Niesse, German Aerospace Centre (DLR) Martin Jung, German Aerospace Centre (DLR) Martin Schaefer, German Aerospace Centre (DLR) Florian Wolters, German Aerospace Centre (DLR) Driving Behavior and Emission Analysis at Yellow Intervals with Advanced Warning Message under Foggy Weather Conditions: A Simulator Test  Johora Munni, Texas Southern University Fengxiang Qiao, Texas Southern University Qing Li, Texas Southern University Lei Yu, Texas Southern University Po‐Hsien Kuo, Texas Southern University 9:30 ‐ 10:30 AM Young Professionals in Transportation, Atlanta Chapter No Host Gathering Conference Seven 11:30 AM - 12:45 PM LUNCH -- Salon IV-V-VI Speaker: Todd Long, Deputy Commissioner for Georgia Department of Transportation 1:00 - 3:00 PM SESSIONS 2.1 - 2.4 Session 2.1: Risk Management in Transportation Session Moderator: David Boate, Gannett Fleming Conference Three Estimating Subjective Beliefs in Naturalistic Tasks with Limited Information  Ming Tsang, Georgia State University 56th Annual TRF Program, Atlanta, GA, Page 2 Material-specific Discount Rate for Risk Analysis in Transportation Engineering Economics  Xiao Qin, University of Wisconsin‐Milwaukee Kai Wang, University of Connecticut Zhiguang Wang, South Dakota State University Aligning Program Level and Project Level Objectives for Risk Management: A Case Study of Process Innovation in a State Department of Transportation  Elie Ji‐Yun Sung, Georgia Institute of Technology Gordon Kingsley, Georgia Institute of Technology Juan Rogers, Georgia Institute of Technology Session 2.2: Infrastructure Investment and Economic Development Session Moderator: Tunhsiang Edward Yu, University of Tennessee Conference Four The Pananma Canal Expansion as a Driver of International Trade  Anthony M Pagano, University of Illinois at Chicago  Onesimo Sanchez, Panama Canal Authority Miguel Verzbolovskis, Panama Canal Eddie A Tapiero, Panama Canal Impact of The Panama Canal Expansion in the U.S. Intramodal and Intermodal Transportation Systems: Optimization Model for U.S. Container Shipment  Ju Dong Park, North Dakota State University Won W. Koo, North Dakota State University An Empirical Study of the Transshipment of the Port of Hong Kong  Dingtong Yang, The Hong Kong Polytechnic University Tsz Leung Yip, The Hong Kong Polytechnic University U.S. Port Readiness for the Panama Canal Expansion  Grace Wang, Texas A&M University at Galveston  Anthony M. Pagano, University of Illinois at Chicago Session 2.3: Logistics Planning and Transportation Supply Chains Session Moderator: EunSu Lee, North Dakota State University Conference Eight Regional Food Logistics: A Stakeholder Process to Inform Multi-System Redesign for Sustainability  Michelle Miller, University of Wisconsin Madison  Marina Denicoff, USDA‐AMS‐Transportation Division Lindsey Day‐Farnsworth, University of Wisconsin Madison Hurricane Evacuation Behavior of the Special Needs Population: Empirical Findings to Support Emergency Logistics Planning  ManWo Ng, Old Dominion University U.S. 95 Freight Multi-Modal Corridor Supply Chain: A Pilot Study  Jeremy Sage, Washington State University Ken Casavant, Washington State University A Field Experiment on Carrier Reservation Prices for Truckload Capacity in the Transportation Spot Market  Christopher Lindsey, Cambridge Systematics Hani Mahmassani, Northwestern University Session 2.4: Mobility Issues in Bus and Rapid Transit Systems Session Moderator: Libby Ogard, Prime Focus LLC Conference Seven A Self-Coordinating Urban Bus Route: Implementation and Field Experience  John J. Bartholdi III, Georgia Institute of Technology Russell J. Clark, Georgia Institute of Technology Donald D. Eisenstein, University of Chicago Loren K. Platzman, independent scholar David W. Williamson, Georgia Institute of Technology The Cost Structure of U.S. Public Transportation:Considering the Impacts of Climate Change  David Ripplinger, North Dakota State University John Bitzan, North Dakota State University 56th Annual TRF Program, Atlanta, GA, Page 3 Accessible Transit Stops and a Link to Passenger Demand: A Tool to Prioritize and Coordinate Accessibility Improvements  Todd Hansen, Texas A&M Transportation Institute Matt Killary, Texas A&M Transportation Institute PANEL Session 1: P3's in Transportation Session Moderator: Baruch Feigenbaum, Asst. Director of Transportation Policy, Reason Foundation Douglas Koelemay, Director, VA Office of P3's Ananth Prasad, former sec., Florida DOT Simon Santiago, Nossaman Daryl Van Meter, Georgia State Innovative Program Engineer Salon IV‐V‐VI 2:45 ‐ 3:00 PM 3:00 - 4:45 PM Break SESSIONS 3.1 - 3.4 Session 3.1: Community Issues in Transportation Session Moderator: David Ripplinger, North Dakota State University Conference Eight Challenges and Enabling Mechanisms for Development of Highway Public-Private-Partnership Projects in the United States  Kia Mostaan, Georgia Institute of Technology Baabak Ashuri, Georgia Institute of Technology Urban-Poverty and Public Transit: Evidence from the Atlanta Metropolitan Region Rahul Pathak, Georgia State University  Christopher Wyczalkowski, Georgia Institute of Technology/Georgia State University Xi Huang, Georgia Institute of Technology/Georgia State University The Boda-Boda as a Tool for Female Empowerment in Kampala, Uganda  Camille Matonis, Harvey Mudd College A Framework to Identify Economic and Community Benefits of the Proposed BRT System in Southeast Michigan  Utpal Dutta, University of Detroit Mercy Ramakrishna R. Tadi, State of California Department of Transportation Sawan Dutta, University of Michigan Session 3.2: Information Technologies for Collecting Transportation Data Session Moderator: Jack Ventura Conference Two Commute Warrior: Android Application for Collecting Longitudinal Travel Survey Data  Alper Akanser, Georgia Institute of Technology Vetri Venthan Elango, Georgia Institute of Technology Ramik Sadana, Georgia Institute of Technology Komal Poddar, Georgia Institute of Technology Yanzhi “Ann” Xu, Georgia Institute of Technology Randall Guensler, Georgia Institute of Technology Developing a Method to Assess National Demand Response Level of Service  Ranjit Godavarthy, North Dakota State University Jeremy Mattson, North Dakota State University Del Peterson, North Dakota State University Patrick Nichols, North Dakota State University Jill Hough, North Dakota State University Comparative Evaluation of Technologies/ Data Sources to Capture Travel Time Along Arterial Streets Rahul C. Pinnamaneni, The University of North Carolina at Charlotte  Srinivas Pulugurtha, The University of North Carolina at Charlotte Venkata R. Duddu, The University of North Carolina at Charlotte RM Zahid Reza, The University of North Carolina at Charlotte Sidewalk Scout: Crowdsourcing Android Application for Collecting Sidewalk Condition Data  Alper Akanser, Georgia Institute of Technology Vetri Venthan Elango, Georgia Institute of Technology Alice Grossman, Georgia Institute of Technology Yanzhi “Ann” Xu, Georgia Institute of Technology Randall Guensler, Georgia Institute of Technology 56th Annual TRF Program, Atlanta, GA, Page 4 Session 3.3: Transportation Efficiency and Productivity Session Moderator: James Odeck, Molde University College Conference Three Productivity and Infrastructure Investment: The Link Between Productivity and Economic Density in Norway  Eivind Tveter, Molde Research Institute Hard Red Spring Wheat Marketing: Effects of Increased Shuttle Train Movements on Railroad Pricing in the Northern Plains  Elvis Ndembe, North Dakota State University Estimating the Technical and Allocative Efficiency of U.S. Class I Railroad: A Data Envelopment Analysis  Elvis Ndembe, North Dakota State University Saleem Shaik, North Dakota State University Revisiting Concentration in Fodd and Agricultural Supply Chains: The Welfare Implications of Market Power in a Complementary Input Sector Metin Cakir, University of Minnesota  James Nolan, University of Saskatchewan Session 3.4: Energy Applications in Transportation Session Moderator: Michael Babcock, Kansas State University Conference Four Piezoelectirc-based Energy Harvesting Technology for Georgia Highway Sustainability  Seonghoon Kim, Georgia Southern University Junan Shen, Georgia Southern University Mohammad Ahad, Georgia Southern University Dukgeun Yun, Korea Institute of Construction Technology Highway Investment, Freight Emissions, and Trade  You Zhou, Washington State University Jeremy Sage, Washington State University Development of Natural Gas Distribution Projects in China Liang_Chieh Cheng, University of Houston  Ran Shi, University of Houston Wanxiang Lu, University of Houston Yixiang Wu, University of Houston Chen Cao, University of Houston Locating Battery Swapping Stations for Smart e-Bus System Sang Hwa Song, Incheon National University  Taesu Cheong, Korea University Seungwon Na, Korea Aerospace University 5:00 - 6:30 PM Conference A Reception & Poster Session Effects of the Rise in Oil Price on Ethanol Fuel Consumption in the U.S. Transportation Sector  Jaesung Choi, North Dakota State University David C. Roberts, North Dakota State University Plug-in Hybrid Electric Vehicles: Optimal All-Electric Driving Range for Minimum Societal Cost  Eleftheria Kontou, University of Florida Yafeng Yin, University of Florida Political Ideology Affects the Willingness to Purchase Electric Vehicles  Yeong Jae Kim, Georgia Tech U.S Freight Transportation Eco-efficiency Performance Measure and Benchmarking State by State  Fesseha Gebremikael, North Dakota State University  Yong Shin Park, North Dakota State University Gokhan Egilmez, North Dakota State University Fuel for Next Generation Refuse Trucks: Electricity or Natural Gas? The Case of Metropolitan Atlanta Area  Dong‐Yeon Lee, Georgia Institute of Technology Valerie M. Thomas, Georgia Institute of Technology Patrick S. McCarthy, Georgia Institute of Technology 56th Annual TRF Program, Atlanta, GA, Page 5 Assessing the Ecosystem Goods and Service of U.S. Modal Freight: Supply Chain Linked to Cradle-to-Gate Ecological Based Life Cycle Model  Yong Shin Park, North Dakota State University Gokhan Egilmez, North Dakota State University Investigating The Changes of Marginal Pavement Damage Cost To Key Variables Using Sensitivity Analysis Tariq Usman Saeed, Purdue University Anwaar Ahmed  Samuel Labi, Purdue University Jackeline Murillo‐Hoyos, Purdue University Kumares C. Sinha, Purdue University A Three-Stage Least Squares Analysis of Post- Rehabilitation Pavement Performance  Md Tawfiq Sarwar, University at Buffalo, The State University of New York Panagiotis Ch. Anastasopoulos, University at Buffalo, The State University of New York Assessment of Seismic Damage of a Bridge Pier Using a Seemingly Unrelated Nonlinear Equations Approach  Thomas Androutselis, University at Buffalo, The State University of New York Panagiotis Anastasopoulos, University at Buffalo, The State University of New York Ionnis Anastasopoulos, University at Dundee Emerging Methods of Signal Timing Optimization  David Hale, Leidos, Inc. Zong Tian, University of Nevada, Reno Public- Public and Public-Private Partnerships in Transportation  Janet Kay Tinoco, Embry‐Riddle Aeronautical University Enhanced Prediction of Transportation Project Costs Using A Risk-Based Approach Tariq Usman Saeed, Purdue University Apichai Issariyanukula, Department of Highways Anwaar Ahmed  Samuel Labi, Purdue University Jackeline Murillo‐Hoyos, Purdue University Kumares C. Sinha, Purdue University How Much Does Above Minimum Design Improve Highway Safety? SeyedAta Nahidi, University at Buffalo, The State University of New York Gary A Jordan, University at Buffalo, The State University of New York  Uger Eker, University at Buffalo, The State University of New York Panagiotis Ch. Anastasopoulos, University at Buffalo, The State University of New York Discovering Causality in Traffic Sensor Readings for Road Accidents Impact Prediction  Liyue Fan, University of Southern California Ugur Demiryurek, University of Southern California Cyrus ShahabiUniversity of Southern California Do the Same Factors Affect Accident Frequences on Highway Segments with Different Traffic Volumes and Traffic Compositions? Ugur Eker, University at Buffalo, The State University of New York  Thomas Androutselis, University at Buffalo, The State University of New York Nima Golshani, University at Buffalo, The State University of New York Panagiotis Ch. Anastasopoulos, University at Buffalo, The State University of New York A Spatio-Temporal Approach for High Resolution Traffic Flow Imputation  Jianjiang Yang, University of Tennessee Lee D. Han, University of Tennessee Shih‐miao Chin, Oak Ridge National Laboratory Ho‐ling Hwang, Oak Ridge National Laboratory A Random Parameters Hazard-based Duration Analysis of Senior Travelers' Activity-based Travel Time and Distance Gary A Jordan, University at Buffalo, The State University of New York  Thomas Androutselis, University at Buffalo, The State University of New York Panagiotis Ch. Anastasopoulos, University at Buffalo, The State University of New York Srinivas Peeta, Purdue University Sekhar Somenahalli, University of South Australia 56th Annual TRF Program, Atlanta, GA, Page 6 Additive Manufacturing: Thoughts and Opinions on its Impact on Future Supply Chains Robert Walton, Embry‐Riddle Aeronautical University  Carl Scheraga, Fairfield University A Real-time Control Bus Dispatching Policy to Minimize Headway Variance  Simon Berrebi, Georgia Tech Kari Watkins, Georgia Tech Jorge Laval, Georgia Tech 6:30 PM Dinner on one's own FRIDAY, MARCH 13, 2015 7:00 AM ‐ 3:30 PM Conference One 7:00 ‐ 8:00 AM Salon IV‐V‐VI 8:00 ‐ 9:15 AM Salon IV‐V‐VI 9:15 ‐ 9:30 AM 9:30 - 11:30 AM Registration Continental Breakfast TRF Business Meeting and Elections Welcome/Opening General Session American Railroads: Decline and Renaissance in the Twentieth Century Robert Gallamore Break SESSIONS 4.1 - 4.4 Session 4.1: Transport Safety Session Moderator: Xiao Qin, University of Wisconsin-Milwaukee Conference Two Modeling Safety Performance Functions for Urban and Suburban Arterial for the State of Alabama  Jaehoon Kim, University of Alabama in Huntsville Michael D. Anderson, University of Alabama in Huntsville Determinants of Crew Injuries in Container Vessel Accidents Yishu Zheng, Virginia Port Authority  Wayne Talley, Old Dominion University Di Jin, Woods Hole Oceanographic Institution ManWo Ng, Old Dominion University Analysis of Fatal Train-Pedestrian Collisions in Metropolitan Chicago 2004-12  Ian Savage, Northwestern University Session 4.2: Transportation Forecasting Session Moderator: John Wells, retired, USDOT Conference Seven Comparison between Trip-Based and Tour-Based Truck Travel Demand Forecasting Models in Birmingham, AL Andrew Sullivan, The University of Alabama at Birmingham William G. Allen, WG Allen David Lee, Georgia Tech  Ehsan Doustmohammadi, The University of Alabama at Birmingham Ozge Cavusoglu, The University of Alabama at Birmingham Virginia Sisiopiku, The University of Alabama at Birmingham Optimal Refueling Policies Models: Applicability in Brazilian Routes Luciano Marques Reduzino, Universidade Federal do Espírito Santo  Marta Monteiro da Costa Cruz, Universidade Federal do Espírito Santo Luiza Rosa Fernandes, Universidade Federal do Espírito Santo Pedestrian Travel Demand Estimation Using Geographically Weighted Regression Model  Xu Zhang, University of Kentucky Mei Chen, University of Kentucky 56th Annual TRF Program, Atlanta, GA, Page 7 Accounting for Natural Gas Vehicles in Regional Auto Markets - Estimates for the State of Texas, U.S.  Chen (Sarah) Xu, University of Houston Liang‐Chieh (Victor) Cheng, University of Houston Session 4.3: Topics in Transportation Supply Chains Session Moderator: Ahren Johnston, Missouri State University The Effect of Information Asymmetry for Freight Network Information Between Shippers and Carriers on the Carrier Selection in a Procurement Exercise  Manasi Katragadda, Iowa State University Yoshinori Suzuki, Iowa State University Conference Four Evaluating the Supply Chains of Cellulosic Transportation Fuel in Tennessee  T. Edward Yu, University of Tennessee Burton C. English, University of Tennessee Lixia He, University of Tennessee James A. Larson, University of Tennessee James Calcagno, University of Tennessee Joshua S. Fu, University of Tennessee Brad Wilson, University of Tennessee Feedstock Import Options Regarding Transportation Systems for Asphalt Production Raj Bridgelall, North Dakota State University  EunSu Lee, North Dakota State University Economic Impacts of Freight Disruptions Due to Flood Related Highway Closures Using an Integrated Traffic Network Equilibrium and Regional Economic Impact Methodology  Amlan Mitra, Purdue University Calumet Rodrigo Mesa‐Arango, Purdue University Xianyuan Zhan, Purdue University Satish Ukkusuri, Purdue University Calument Session 4.4: Systems Approaches to Transportation Session Moderator: Utpal Dutta, University of Detroit Mercy Is Hands-Free Texting a Better Alternative of Text Driving?  Sanaz Motamedi, University of Rhode Island Jay Hone Wang, University of Rhode Island Conference Eight The Impacts of Distracted Drivers on Bicyclists' Fatalities  Sia Macmillan Lyimo, South Carolina State University Analyzing What Works Best for Transportation Education  Sampson Gholston, The University of Alabama in Huntsville MD Sader, University of Southern Mississippi Joan Chadde, Michigan Technological Institute Richard Stewart, University of Wisconsin, Superior Amit Mokashi, University of Wisconsin, Superior HSR as Transit: The Continuing Transportation-driven Evolution of Metropolitan Form  Ryan Westrom, MIT Joseph Sussman, MIT Panel Session 2: Privatizing Transit Session Moderator: Baruch Feigenbaum, Reason Foundation Art Guzzetti, Vice President, APTA Salon IV‐V‐VI Sam Staley, Managing Diretor, Devoe Moore Center, FSU Jim Ritchey, Vice President Public Transportation, Whitman, Requardt and Associates 11:30 AM - 12:45 PM Awards LUNCHEON -- Salon IV-V-VI Introduction and Presenter: Art Guzzetti Speaker and Recipient of TRF's President's Award Keith Parker, CEO MARTA 56th Annual TRF Program, Atlanta, GA, Page 8 1:00 - 2:45 pM SESSIONS 5.1 - 5.5 Session 5.1: Highway User Charges Session Moderator: Joseph Schwieterman, DePaul University Conference Two Impacts of Violations of High-Occupancy-Vehicle Lanes: A Simulation-based Study in Tennessee  Jianjiang Yang, University of Tennessee Lee Han, University of Tennessee Benefit-Cost Analysis of Bottleneck Mitigation Strategies  David Hale, Leidos, Inc. Jiaqi Ma, Leidos, Inc. Michalis Xyntarakis, Cambridge Systematics Characteristics of Car-less Households in California: Evidence from the 2012 California Household Travel Survey  Suman Kumar Mitra, University of California, Irvine Jean‐Daniel Saphores, University of California, Irvine Session 5.2: Transport Capital, Funding, and Economic Growth Session Moderator: Maite Pena-Alcaraz, Massachusetts Institute of Technology Conference Seven Port Planning Benchmarking - Best Practices in Funding Programs  Libby Ogard, Prime Focus LLC Richard Stewart, University of Wisconsin Superior Capital Investment Needs of the Small Railroads  Anne Campbell, North Dakota State University Rodney Traub, North Dakota State University Dynamic Relationships Among Transport Infrastructure Investment, Non-Transport Public Infrastructure, and Economic Growths in the United States: A New Look  Junwook Chi, University of Hawaii, Manoa Jungho Baek, University of Alaska Fairbanks Session 5.3: Operating and Management Performance in the Airline Industry Session Moderator: Alan Bender, Embry-Riddle Aeronautical University Conference Four An Analysis of the Effects of Operating and Financial Leverage on the Major International Carriers: 1990-2013  Carl Scheraga, Fairfield University Richard Gritta, University of Portland Proactive Passenger Management with a Total Airport Management Perspective Axel B. Classen, DLR  Florian Rudolph, DLR Why Revenue Management Is a Good Thing  Emmanuel Carrier, Delta Airlines Session 5.4: Regulatory, Legal, and Competitive Issues in Transportation Session Moderator: Frank Mulvey, Iter Associates Conference Eight Assessing the Objectivity of Transportation Related Regulatory Decisions in Canada  James Nolan, University of Saskatchewan Savannah Gleim, University of Saskatchewan Does Competition Improve Technical Efficiency of Transportation Services? A DEA Application in The Case of Norwegian Car Ferry Sector  James Odeck, Molde University College Edvard Sandvik, The Norwegian Public Roads Administration Bicycle Policy Implications of Uneven Development  Caroline Appleton, Georgia Tech School of Public Policy Eco-Efficiency of U.S. Container Ports Using Data Envelop Analysis (DEA)  N. Muhammad Aslaam Mohamed Abdul Ghani, North Dakota State University Eunsu Lee, North Dakota State University Yong Shin Park, North Dakota State University 56th Annual TRF Program, Atlanta, GA, Page 9 Session 5.5: Global Trends in Transportation and Logistics (NeLT-TRF Industrial Session) Session Moderator: Dr. Yongtae Kim, Hankyong National University Technology Trends in Logistics and Supply Chain Management Dr. Anthony Pagano, University of Illinois, Chicago Salon IV‐V‐VI Design of New Concept Container Terminal Dr. Moohong Kang, KMI Advanced Material-Handling Equipment for Manual-dependent Distribution Dr. Yongjang Kwon, KRRI Development of Cargo Batch Loading and Unloading Systems Dr. Sukk Lee, KRRI Supply Chain Issues in Agricultural Products and Technologies Dr. Sangyoung Moon, Hankyong National University 3:30 - 5:30 PM Tours Tour 1 Delta Museum Tour 2 MARTA, Streetcar/Beltline Tour 3 Transit-Oriented Development (TOD) 5:30 PM TRF Chapter Gathering 7:00 PM TRF Council Meeting 7:00 PM Dinner on one's own SATURDAY, MARCH 14, 2015 7:00:00 AM ‐ 12:00 PM Conference One 7:00 ‐ 8:00 AM Conference Hall Break Area 8:00 - 9:45 AM Registration Continental Breakfast SESSIONS 6.1 - 6.3 Session 6.1: Systems Performance and Safety Session Moderator: C . Gregory Bereskin, St. AmbroseUniversity Conference Six Assessment of Operational Performance of Bus Rapid Transit (BRT) In Lagos State Nigeria  Olayemi Funmilayo Dickson, National Centre for Technology Management Assessing the Vulnerability of Evacuation Plans via Critical Element Detection  Chrysafis Vogiatzis, University of Florida Jose Luis Walteros, University at Buffalo Panos M. Pardalos, University of Florida Benefits of System Safety Management  Alireza Edraki, Gannett Fleming Session 6.2: Highway Infrastructure and Public Sector Decisions Session Moderator: James Nolan, University of Saskatchewan Conference Seven Evaluating the Application of Diverging Diamond Interchange in Athens, Alabama  Tahmina Khan, University of Alabama Huntsville Michael Anderson, University of Alabama Huntsville 56th Annual TRF Program, Atlanta, GA, Page 10 Heavy Vehicle Impact on Rural Two Lane Highway Segments Operating Under Various Levels of Service Conditions  Zijian Zheng, North Dakota State University Pan Lu, North Dakota State University Freight Movement by Heavy Duty Diesel Trucks Emission in Port of Houston Area  Ehson Khademi, Texas Southern University Fengxiang Qiao, Texas Southern University Mehdi Azimi, Texas Southern University Lie Yu, Texas Southern University Xiaobing Wang, Texas Southern University Session 6.3: Railroad Technology and Development Session Moderator: Ian Savage, Northwestern University Conference Eight Shortline Rail in Washington State: Challenges and Oppurtunities on the Road Ahead  Jeremy Sage, Washington State University Ken Casavant, Washington State University J. Bradley Eustice, Washington State University Passenger Energy Technologies for the Near-to-Mid-Term: A Summary Review and Case Study  Raphael Isaac, UC Davis Institute of Transportation Studies Lewis Fulton, UC Davis Institute of Transportation Studies Analysis of Capacity Pricing and Allocation Mechanisms in Shared Railway Systems: Lessons for the Northeast Corridor  Maite Pena‐Alcaraz, MIT Joseph Sussman, MIT Frieght Rail Economic Development  Libby Ogard, Prime Focus LLC 10:00 ‐ 10:15 AM 10:15 - 11:45 AM Break SESSIONS 7.1 - 7.3 Session 7.1: Speed, Time, and Delays in Transportation Session Moderator: Chrysafis Vogiatzis, University of Florida Conference Six A Spatio-temperal Based Algorithm for Short-Term Speed Prediction  Jianjiang Yang, University of Tennessee Lee Han, University of Tennessee Efficient Frontier of Route Choice for Equilibrium under Travel Time Variability with Heterogeneous Travel Preferences  Mahyar Amirgholy, University of Massachusetts Eric J. Gonzales Reroute or Wait it out? Estimating the Time until Spoilage in the Presence of Unexpected Delays  J. Bradley Eustice, Washington State University Jeremy Sage, Washington State University Ken Casavant, Washington State University Session 7.2: Factors Affecting Highway Safety Session Moderator: Srinivas Pulugurtha, The University of North Carolina at Charlotte Conference Seven Weights from Safety Perspective for Interchange Lighting Prioritization Tool  Srinivas Pulugurtha, The University of North Carolina at Charlotte Ravishankar P. Narayanan, The University of North Carolina at Charlotte Correlation Between Crash Rates and International Roughness Index on In-Service U.S. Highways  Ahmed Elghriany, The University of Akron Ping Yi, The University of Akron Peng Liu, The University of Akron Quan Yu, Beijing University of Technology The Use of CHAID Decision Tree Models in Injury Severity Analysis of Large Truck Crashes  Jill Bernard, University of Missouri ‐ St. Louis  Christopher Mondy, University of Missouri ‐ St. Louis 56th Annual TRF Program, Atlanta, GA, Page 11 Session 7.3: Organizational Issues in Transport Session Moderator: Svein Bråthen, Molde University College Conference Eight Norwegian Bus Transport- Does Internal or External Organization of Public Authorities Affect Efficiency  Hilde Johanne Svendsen, Møreforsking Molde AS Arild Hervik, Møreforsking Molde AS James Odeck, Molde University College The Relationship between Leadership Strengths and Proformance of Fleet Managers  Ahren Johnston, Missouri State University John L Kent, University of Arkansas Assessing the Effects of Compressed Work Week Strategy on Transportation System Performance Measures Venkata R. Duddu, The University of North Carolina at Charlotte  Srinivas Pulugurtha, The University of North Carolina at Charlotte 56th Annual TRF Program, Atlanta, GA, Page 12 ABSTRACTS th 56 Annual TRF Program, Atlanta, GA, Page 13 SESSION 1.1: Technological Advances in Vehicle Systems HOW ACTIVITIES WHILE TRAVELING WOULD AFFECT TRAVEL BEHAVIOR IN THE AUTONOMOUS VEHICLE FUTURE? SCENARIO-BASED SIMULATIONS ON THE SAMPLE OF NORTHERN CALIFORNIA COMMUTERS Aliaksandr Malokin, Georgia Institute of Technology, amalokin@gatech.edu One of the great promises of autonomous vehicles (AVs) is ability to free traveler’s attention resources (and associated time) from the driving task to engage in other activities of the interest or priority to the driver. In this future, the time slot previously almost exclusively occupied by travel will dissolve in the time use structure demonstrating overlapping continuity of carried activities. In other words, travel will lose its place as a primary activity of its own allowing location related tasks to dominate the temporal landscape during moving phase. Only this aspect alone would entail seismic shifts in the behavioral considerations of travel modes. One could imagine examples of (involuntary) incorporating commute time into the working hours (i.e. a mobile office), or shifting a night sleep to partly take place during a morning commute (i.e. a mobile bedroom). The modern realization of the multitasking phenomenon, which promises to bring improved productivity and/or satisfaction through overlapping multiple activities “at the same time”, portends changing patterns of time use, and especially (in our context) of travel time use. There is a sizable and growing literature on multitasking in general (e.g. König and Waller, 2010; Circella et al., 2012), and in contexts such as the work environment (e.g. Bluedorn and Martin, 2008; Chesley, 2014) or “media multitasking” (e.g. Wallis, 2010) in particular. Still, the investigation of the activities conducted while traveling, and their impact on traveler’s choices, is a smaller but also expanding area of research (e.g. Mokhtarian and Salomon, 2001; Kenyon and Lyons, 2007; Ettema and Verschuren, 2007; Ohmori and Harata, 2008; Zhang and Timmermans, 2010; Rasouli and Timmermans, 2014). In our previous work, travel-based multitasking (i.e. the engagement in additional activities while traveling) has been shown to affect commuters’ mode choices (Malokin et al., 2015). In this study, we explore the observed effects further by testing more elaborate scenario structures of inclusion autonomous vehicles into commuters’ choice sets. This paper is called to addressed the questions of how AV’s unrestricted ability to be engaged in activities while traveling would influence commuters’ mode choice; and, in particular, how this ability would affect the aggregate travel demand of the studied population. Our model is based on the data collected by surveying more than 2,000 Northern California commuters (see Neufeld and Mokhtarian, 2012, for more details). The person with average characteristics for this sample is a female, around 45 years old, college graduate, and living in a household of 2.7 people owning 2.1 vehicles and earning $75,000 - $99,999 annually. Multiple measures of multitasking are available in the data. For example, respondents were asked to rate each alternative mode based on the “ability to do things I need/want while traveling” (mode-specific perceptions). Also, the survey collected information about which of a number of different activities they performed on a recent commute (mode-specific behavior), and what benefits/ disadvantages they experienced by these activities. Since this information was only available for the chosen mode, it could not be used directly as explanatory variables in our mode choice model. Instead, we (1) modeled the propensity to perform each activity (as functions of individual attributes such as socio-economic characteristics, personality traits, attitudes, and commute mode composition) for the choosers of each mode, and then (2) used those models to predict an individual’s propensity to perform each activity for each mode, whether chosen or not. We will use the resulting predicted propensities as explanatory variables in the mode choice model, either directly or as predictors in the modeling of benefits/disadvantages associated with performing the activities (i.e. applying the similar estimating approach on the other mode-specific variables). The resulting propensities of benefits/ disadvantages modeling then will be used to explore effects of doing activities while traveling on mode choice. Changes in propensities of performing activities and benefits/disadvantages associated with it will define a set of AV scenarios. th 56 Annual TRF Program, Atlanta, GA, Page 14 Among scenarios we plan to test are:  single occupancy (SO) AVs with the operating parameters similar to the existing cars used for the entire commute;  high occupancy (HO) AVs (i.e. dynamic sharing) used for the entire commute; SO AVs used as the first leg to access the main mass transit haul, and HO AVs used as the last leg;  mixture of the these and other scenarios. This is still a work in progress. By early 2015, we anticipate refining the set of scenarios, assessing the outcomes given different technology diffusion rates, conducting sensitivity analyses, evaluating AV public transit market shares, and estimating infrastructure and environmental costs of changing travel behavior due to wider freedom of engaging in activities while traveling associated with the AV adoption. ASSESSING THE IMPLICATIONS OF AUTONOMOUS TRUCKS FOR PERFORMANCE-BASED FREIGHT TRANSPORTATION PLANNING Denise A. Smith, Georgia Institute of Technology, denise.a.smith@gatech.edu Autonomous vehicles are inevitable. The question regarding this technology has shifted from “if” to “when”. Many advocates predict that autonomous trucks, in particular, will be commercially available within the next decade, and perhaps even before autonomous passenger vehicles. Advocates also suggest that these technologies could lead to significant reductions in trucking costs, fuel consumption, and airborne emissions. The actual impacts and benefits, however, are far from clear. While there is an evolving literature on the technical feasibility of safe and reliable automated vehicle technology, there has been very little discussion of how widespread adoption of these technologies will affect freight flows, and hence how they ought to be incorporated into the transportation planning process. This paper focuses on the influence that autonomous trucks, and in particular semi-autonomous, multi-vehicle truck convoys, or platoons, could have on measuring the performance of proposed freight transportation projects, with an emphasis on measuring both economic and environmental performance. Specifically, the paper discusses the implications for highway planning posed by the introduction of such convoys, including how they may impact the two largest components of day-to-day truck operating expenses, namely fuel and labor costs. Performance measures that quantify these impacts are proposed, within the context of an evolving truck technology, including how the potential for improvements in fuel efficiency might lead to reductions in greenhouse gas emissions. MULTI-METHOD RESEARCH APPROACH TO GAINING STATEWIDE ACCEPTANCE OF A ROAD USAGE CHARGE PROGRAM IN OREGON Katherine Schomer, PRR, Kschomer@prrbiz.com Colleen Gants, PRR, cgants@prrbiz.com Michelle Godfrey, ODOT, Michelle.D.GODFREY@odot.state.or.us John Horvick, DHM Research, jhorvick@dhmresearch.com Ian Nunley, PRR, inunley@prrbiz.com In 2013 Oregon Legislators authorized the Road Usage Charge Program utilizing technology to assess vehicles a charge of 1.5 cents per mile, and issue a gas tax refund, for up to 5,000 vehicles to voluntarily participate. To identify public understanding of transportation funding and opinions towards this program, a multi-method research approach was used including focus groups, telephone surveys, stakeholder interviews, online surveys, media content analysis, and statewide public outreach. Results from this approach are being used to inform a statewide communication plan to achieve measurable statewide acceptance for a road usage charge program in Oregon. th 56 Annual TRF Program, Atlanta, GA, Page 15 SESSION 1.2: Uses of Information Technology in Transportation FEASIBILITY OF USING BLUETOOTH DATA FOR ESTIMATING REAL TIME DELAY AND STOPS AT SIGNALIZED INTERSECTION APPROACHES Golam Sarwar, AgileAssets Inc., gsarwar@agileassets.com Ahmed Abdel-Rahim, University of Idaho, ahmed@uidaho.edu The paper investigates the feasibility of using single and dual Bluetooth stations to estimate real time approach delay, travel time and stops at signalized intersection approaches. Three different methods are introduced and examined to estimate the approach delay and link travel time using Bluetooth data. Method 1 estimates approach delay using Bluetooth data from a single station, whereas method 2 and method 3 engage dual Bluetooth stations to estimate approach delay and travel time. Approach delay and travel time values estimated using Bluetooth data are compared to benchmark true delay and travel time values measured from traffic videos recorded in the field during the same time the Bluetooth data are collected. Bluetooth data were also used to identify non-stopped, delayed, and stopped vehicles. Statistical analysis shows that accuracy of average approach delay and link travel time values estimated using Bluetooth data is highly dependent on the number of Bluetooth data points. ANOVA test F-value justifies that the three estimation methods examined in this study are significantly different from each other. This indicates that Bluetooth data can be used to provide travel time estimation that is operationally acceptable. The results of the analyses will allow decision makers to assess the validity of using Bluetooth data for intersection specific measures of effectiveness and to develop guidelines and procedures for Bluetooth based field approach delay, stops and travel time measurements. The study suggests that Bluetooth communication technology provides a cost effective, innovation technology and easy to implement automated field data collection alternative. DETECTION CRITERIA TO ESTIMATE BLUETOOTH-BASED TRAVEL TIME ON ARTERIAL STREETS Srinivas S. Pulugurtha, The University of North Carolina at Charlotte, sspulugurtha@uncc.edu Md Shah Imran, The University of North Carolina at Charlotte, mimran1@uncc.edu Venkata R. Duddu, The University of North Carolina at Charlotte, vduddu@uncc.edu Dynamic travel time information is necessary for both the road users and practitioners. Though there are several methodologies currently in practice for identifying real-time travel time, accuracy of these estimations is not clear. Bluetooth based travel time estimation is becoming increasingly popular as a means of collecting data. However, a Bluetooth detector may detect a same Media Access Control (MAC) address several times. This along with capturing of Bluetooth enabled devices that are not related to travel time computations may result in inaccurate travel time estimates. This paper looks at the detection point of view and identifies a group of combinations of detection criteria for travel time estimation. The criteria are based on device signal strength, time-stamp difference, first to first detection, last to last detection, first to last detection, and last to first detection between two Bluetooth detection zones along selected arterial segments. Results obtained indicate that arterial segments generate more outliers in the data. The travel time computed based on the signal strength of detectors provided more accurate results than other combinations. AN EFFECTIVE DATA FUSION ALGORITHM FOR THE TRAVEL TIME RELIABILITY APPLICATION Xu Zhang, University of Kentucky, xuzhang_uk@uky.edu Obaidur Rahman Kazi, University of Kentucky, obaidur.kazi@uky.edu Mei Chen, University of Kentucky, mei.chen@uky.edu Recently, travel time reliability has been emerging as one of the most concerned aspects of the transportation facility or system. Numerous efforts have been devoted to measuring and monitoring the reliability performance. However, most of current researches are only relying on the traffic information from a single data source. For many th 56 Annual TRF Program, Atlanta, GA, Page 16 cases, data/information collected by individual sources is incomplete and unreliable. With proliferation of travel time collection methods, multiple data sources from different sensors and technologies are usually available now. This provides a valuable opportunity to cross-validate the data and improve the assessment of travel time reliability. In this paper, the data from fixed-location detectors and probe vehicle GPS data on a freeway facility will be used to develop a data fusion algorithm tailored for the reliability application. The benefit of using the fused data in the travel time reliability analysis will also be discussed. OPTIMAL NUMBER AND LOCATION OF BLUETOOTH SENSORS CONSIDERING STOCHASTIC OPTIMAL TIME TRAVEL PREDICTION Hyoshin Park, University of Maryland, hyoshin0724@gmail.com Determining the optimal number and location of sensors is essential to effectively manage traffic on highways. Optimal solutions dealing with dynamic traffic patterns and relocation of sensors have received little attention. In this study, existing fixed sensors are used to estimate travel time prediction errors at candidate locations where we deploy portable sensors. Potential sampling error of each candidate location is also counted in selecting optimal locations. A two-stage stochastic formulation considers uncertainty of traffic conditions based on scenarios generated by principal component analysis and clustering analysis to uncover the underlying spatial correlations and temporal patterns. The first stage decision, determining the optimal number of sensors, is made before the deployment. The second stage, evaluating the expected travel time prediction errors, specifies sensor arrangements in each scenario. A dynamic model has predefined rearrangement stages. At each stage, sensor locations are modified as the pattern of travel time error changes over time, considering sensor acquisition and relocation expenses. The deterministic and stochastic solutions serve as a lower bound and an upper bound for the dynamic solution. Higher relocation expense leads to more sensors being used, while higher sensor costs leads to fewer sensors being used with more frequent relocations. SESSION 1.3: Topics in Airline Operations AIRPORT CONGESTION MITIGATION: A MULTIMODAL APPROACH Jared Annexstad, North Dakota State University, jared.annexstad@ndsu.edu EunSu Lee, Upper Great Plains Transportation Institute, eunsu.lee@ndsu.edu In 1978 the Airline Deregulation Act ushered in a new age of air travel. Airlines were no longer servicing unprofitable routes and flying half full aircraft. They were not burdened by regulatory agencies dictating fares and routes and they had the freedom to operate in a free market system. For some, this meant bankruptcy. For others, adapting new business models, such as the hub-and-spoke system resulted in years of increased market share and profitability. Due to this increased efficiency, fares were dropped and air travel suddenly became an affordable travel option. Due to the nature of the hub-and-spoke model, hub airports are becoming increasingly congested. In fact, out of those 4.8 million flights, almost 23% of those flights were delayed or canceled nationwide. In this research, data from a variety of different sources will be analyzed to determine the feasibility and costeffectiveness of establishing networks comprised of different modes to alleviate airport congestion. The goal of this study is to examine the feasibility of both multimodal and intermodal solutions. From the passenger’s perspective, the journey needs to be simple, cost effective and time efficient. The only way this will work is if the airlines and rail companies work together to provide a seamless transition. By freeing up short routes with high passenger volume, the airlines are able to add more profitable long haul flights. Rail and even bus routes need to be seen as a partner and not a competitor. By interfacing airports with a national rail network and offering code share service, airlines will be able to offload some travelers in favor of more profitable routes. At the very least, inter and multimodal systems can be seen as part of the solution to the overall problem. As was concluded by the FAA’s FACT 2 study, incorporating all available solutions together, the capacity needs of the NAS can be met through 2025. th 56 Annual TRF Program, Atlanta, GA, Page 17 AIRLINE CONSOLIDATION AND HUB ABANDONMENT: THE IMPACT ON REGIONAL ECONOMIES Michael J McCormick, George Mason University, mmccorm9@masonlive.gmu.edu Since the United States' 1978 Airline Deregulation Act, the legacy commercial air transport industry followed two significant and interdependent trends: hub-and-spoke network operations and air carrier consolidation. However, as airlines merge operations through consolidation, they rationalize routes, aircraft fleet and facilities to maximize efficiency and reduce costs. This includes abandoning redundant hub airports. This paper examines the question; What is the regional impact of hub abandonment due to airline consolidation? The working hypothesis is that hub abandonment adversely impacts regional economies. The alternate hypothesis is that hub abandonment removes barriers to entry and allows new entrants to have a contravening positive influence on regional economies. The null hypothesis is that hub abandonment does not negatively or positively impact regional economies. Moreover, regional economy may adjust over time mitigating or absorbing the economic impact of hub abandonment. This paper develops a case study analysis of two abandoned airlines hubs, Pittsburgh International Airport and Lambert-Saint Louis International Airport, contrasted with two viable airline hubs, Minneapolis-Saint Paul International Airport and Charlotte/Douglas International Airport. Due to data availability covering a period of legacy airline consolidation the analysis includes 2000 through 2012 and utilizes three methods: input-output model, fixed-effects panel analysis and econometric comparison. A unique dataset is created from sources including airline passengers, air carrier flights, airport operations, air transport employment and regional gross domestic product. Further, the paper develops recommendations for airport operators and regional planners to understand and predict the economic impact of hub abandonment and develop policies to mitigate or leverage the regional impact. ESTIMATING FLIGHT-LEVEL PRICE ELASTICITIES USING ONLINE DATA Laurie Garrow, Georgia Institute of Technology, laurie.garrow@ce.gatech.edu Stacey Mumbower, University of South Carolina, smpittma@gmail.com Matthew Higgins, Georgia Institute of Technology, matthew.higgins@scheller.gatech.edu We estimate flight-level price elasticities using a database of online prices and seat map displays. In contrast to market-level and route-level elasticities reported in the literature, flight-level elasticities can forecast responses in demand due to day-to-day price fluctuations. Knowing how elasticities vary by flight and booking characteristics and in response to competitors’ pricing actions allows airlines to design better promotions. It also allows policy makers the ability to evaluate the impacts of proposed tax increases or time-of-day congestion pricing policies. Our elasticity results show how airlines can design optimal promotions by considering not only which departure dates should be targeted, but also which days of the week customers should be allowed to purchase. Additionally, we show how elasticities can be used by carriers to strategically match a subset of their competitors’ sale fares. Methodologically, we use an approach that corrects for price endogeneity; failure to do so results in biased estimates and incorrect pricing recommendations. Using an instrumental variable approach to address this problem we find a set of valid instruments that can be used in future studies of air travel demand. We conclude by describing how our approach contributes to the literature, by offering an approach to estimate flight-level demand elasticities that the research community needs as an input to more advanced optimization models that integrate demand forecasting, price optimization, and revenue optimization models. This paper was published in 2014 in Transportation Research Part A. th 56 Annual TRF Program, Atlanta, GA, Page 18 AIRPORT EFFICIENCY AND ENERGY USE Svein Bråthen, Molde University College, svein.brathen@himolde.no Mozhgan Khazraeianvafadar, Molde University College, mozhgan_vfdr@yahoo.com James Odeck, Molde University College, james.odeck@vegvesen.no This study has evaluated the effect of energy use on the efficiency of 28 Norwegian airports by applying the twostage Data Envelopment Analysis for 2009 to 2013. Energy use is considered as one of the inputs in a production function along with other factors like employment, runway length and size of the terminal buildings. Air cargo volume, aircraft movements and the annual number of passengers are considered as output factors. The results indicate that including energy use as input can affect the airports' technical efficiency considerably. Moreover, significant effects of external factors to the airport operations such as airport size, the location of the airport and age of the terminal building are found. The scale efficiency of the airports implies that most of the Norwegian airports are not performing at their optimal scale. This study suggests that since energy use can affect the airport efficiency, it is recommended to be considered as one of the inputs in measuring airport efficiency. The lack of relevant input factors in a DEA, like energy use, may be confused with lack of efficiency. SESSION 1.4: Measuring the Environmental Impacts of Transportation Activities FORECAST OF CO2 EMISSIONS FROM THE U.S. TRANSPORTATION SECTOR: ESTIMATION FROM A DOUBLE EXPONENTIAL SMOOTHING MODEL Jaesung Choi, North Dakota State University, jaesung.choi@ndsu.edu This study examines whether the decrease in U.S. CO2 emissions from the transportation sector since the end of the 2000s will be consistent across all states in the nation for 2012-2021. A double exponential smoothing model is used to forecast CO2 emissions for the transportation sector in the 51 states and the U.S., and its findings are supported by the validity test of pseudo out-of-sample forecasts. We conclude that the decreasing trend in transportation CO2 emissions in the U.S. will continue in most states in the future. ECONOMIC AND ENVIRONMENTAL IMPACTS OF MARKET-BASED MEASURES FOR THE LIMITATION OF AVIATION’S FULL CLIMATE IMPACT Janina Scheelhaase, German Aerospace Centre (DLR), Janina.Scheelhaase@dlr.de Katrin Dahlmann, German Aerospace Centre (DLR), Katrin.Dahlmann@dlr.de Robert Sausen, German Aerospace Centre (DLR), Robert.Sausen@dlr.de Hermann Keimel, German Aerospace Centre (DLR), Hermann.Keimel@dlr.de Hendrik Niesse, German Aerospace Centre (DLR), Hendrik.Niesse@dlr.de Martin Jung, German Aerospace Centre (DLR), M.Jung@dlr.de Martin Schaefer, German Aerospace Centre (DLR), Martin.Schaefer@dlr.de Florian Wolters, German Aerospace Centre (DLR), Florian.Wolters@dlr.de Aviation contributes to climate change by both long-term CO2 and short-term non-CO2 effects, such as ozone and methane changes from NOX emissions or contrail cirrus: In 2005 aircraft induced CO2 contributed 1.6% to the global anthropogenic radiative forcing (Lee et al, 2009), while the sum of the CO2 and the non-CO2 effects amounted to 4.9%. During the next decades international aviation is expected to grow significantly (e. g. Airbus, 2013). Whilst aviation’s CO2 emissions are regulated in a number of countries by market-based measures (e. g. emission charges or emissions trading schemes), this is not the case for most of aviation’s non-CO2 climate impacts. th 56 Annual TRF Program, Atlanta, GA, Page 19 The interdisciplinary research project AviClim (Including Aviation in International Protocols for Climate Protection) explores the feasibility for including aviation’s CO2 and non-CO2 climate impacts in international protocols for climate protection. In addition, the associated economic impacts are studied. Therefore four scenarios for the use of alternative market-based measures are developed and the economic and environmental impacts are modelled. The German Aerospace Centre (DLR) conducted AviClim on behalf of the German Federal Ministry of Research. This paper presents the main results of this research project. Modelling results indicate that a global emissions trading scheme for both CO2 and non-CO2 emissions would be the best solution from an environmental and economic point of view. Under the assumption of a moderate CO2 equivalent price development costs and impacts on competition could be minimized and effects on employment are manageable. At the same time, environmental benefits are remarkable. DRIVING BEHAVIOR AND EMISSION ANALYSIS AT YELLOW INTERVALS WITH ADVANCED WARNING MESSAGE UNDER FOGGY WEATHER CONDITION: A SIMULATOR TEST Johora Munni, Texas Southern University, j.munni1575@student.tsu.edu Fengxiang Qiao, Texas Southern University, qiao_fg@tsu.edu Qing Li, Texas Southern University, liq@tsu.edu Lei Yu, Texas Southern University, yu_lx@tsu.edu Po-Hsien Kuo, Texas Southern University, ckin44@gmail.com Traffic signs and signals are two major traffic control devices that directly manipulate drivers’ decision makings and regulate driving behaviors including speed, braking, stopping, lane-changing. Bad weather, construction infrastructure and machines, and natural obstacles (trees, sun glare) may affect drivers’ visions on traffic signs and/or signals and may result in fatal crashes. It is therefore necessary for drivers to receive right traffic control messages on time so as to avoid possible fatalities and even reduce vehicle emissions. With the advancement of modern technologies, smartphones and tablets are so popular that almost all drivers are carrying en-route. This paper characterizes the impacts of foggy weather conditions on driver behavior at yellow interval and how driver behavior is influenced by the advanced warning message from the Drivers Smart Signal System (DSSS) at high-speed signalized intersection approaches from driving simulator test. The DSSS is an android system based smartphone and tablet application that was developed as a real time traffic signal system. Drivers driving with the advanced warning message from the DSSS can be informed about real time traffic signal status from the smartphone easily even though he/she has difficulty to see the traffic signal. The DSSS is an application that can inform drivers with an advanced yellow warning message just about one second before the yellow onset together with a Smart Yellow Countdown (SYCD). The test was designed in a driving simulator for three scenarios (approximately eight minutes for each scenario): (a) normal weather condition (S-1), (b) foggy weather condition (S-2), and (c) foggy weather condition with provision of DSSS message (S-3). The intersection signal timing was designed as thirty-five seconds green, five seconds yellow, and thirty seconds red for the testing vehicle in simulator. The onset of yellow interval was triggered at four different locations in each scenario (S-1, S-2, and S-3) and the distances from the intersection to the triggering location of yellow onset were 100.55 m, 80.44 m, 120.66 m and 60.33 m, respectively. This means a total of four controlled intersections included in each scenario (12 intersections in total). Colored signal light images were always showing on the smartphone screen, the color of which was synchronized to the displayed signal at the tested intersection in simulator. The SYCD was illustrated on smartphone as well together with the advanced yellow warning message. The posted speed limit of the approach was 72.4 km/hr (45 mph). The fog distance for the test was set as 150 m, which is a little bit farther than the signal visibility and the minimum sight distance (140 m), as is defined in the Manual on Uniform Traffic Control Devices (MUTCD). Twenty participants with similar demographic background were recruited for the tests. Each participant drove four intersections with different triggered locations of yellow onset for each of three scenarios. Drivers were instructed to follow speed limits and traffic signals. The scenario S-1 involved in a normal weather condition, S-2 involved in foggy weather, and S-3 involved in a foggy weather condition with the DSSS assistance. In scenarios S-1 and S-2, drivers followed the signal timings at intersections, while in scenario S-3, drivers should follow the signal timing of the DSSS. The instant driving th 56 Annual TRF Program, Atlanta, GA, Page 20 information (time, traveling speed, acceleration, distance travelled etc.) were recorded from the driving simulator at a frequency of twenty records per second. The parameter “Time to Reach Intersection” (TRI) was calculated by dividing the distance to intersection by the vehicle speed at the moment of yellow onset. In total, there were 240 observations for twenty drivers representing 169 stop and 71 go decisions. The approach speed, the braking time, the average deceleration rate, the probability of stopping and going with respect to TRI and vehicle emissions were analyzed for all three scenarios. The mean approach speed for S-1 (21.05 m/s2) is higher than the approach speed for S-2 (18.61 m/s2, p-value: 6.4357E-33), and the approaching speed of S3 (18.56 m/s2, p-value: 2.3057E-17). The overall 50th percentiles of breaking time for S-1, S-2 and S-3 are 8.25 sec, 7.5 sec and 9.07 sec, respectively. The 15th percentile of deceleration rate is observed as 1.6 m/s2 for normal weather condition, which is higher than the one for foggy weather condition with DSSS (1.1 m/s2), whereas smaller than the one for foggy weather condition (1.9 m/s2). The 50th percentile deceleration rate for foggy weather condition is 2.55 m/s2, which is higher than the accepted normal deceleration rate (2.43 m/s2). The probability of stoppings increases with the increase of TRI for normal weather condition, whereas no trends is observed for foggy weather condition and probability of stoppings is always greater than of goings for any TRI for foggy weather condition with the DSSS. Furthermore, results from the emission analysis demonstrate that the DSSS helps to reduce emission rates at yellow interval when approaching an intersection. While the DSSS message in foggy weather condition cannot fully recover drivers’ behaviors to normal weather condition, it can somehow reduce the risk of safety such as running red lights and sharp decelerations, and may also reduce emissions. SESSION 2.1: Risk Management in Transportation ESTIMATING SUBJECTIVE BELIEFS IN NATURALISTIC TASKS WITH LIMITED INFORMATION Ming Tsang, Georgia State University, mtsang1@gsu.edu In the area of transportation policy, congestion pricing has been used to alleviate traffic congestion in metropolitan areas. A wide range of standard decision theories would attribute reactions to congestion pricing to risk attitudes, risk perceptions and value of time. The focus of this paper is on the risk perception of traffic delays to explain reactions to congestion pricing. In a naturalistic setting using field participants, this experiment studies how accurate participants perceive unknown probabilities, and if perceptions of congestion converge on the true probability with experience when the information feedback is endogenous, such that new information about the risk of congestion can only be obtained if they choose the risky route. The present study varies the objective probability of congestion and the monetary incentives, and asks participants to make route choices using a driving simulator. Subjective beliefs are estimated assuming Subjective Expected Utility or Rank Dependent Utility while controlling for risk attitudes. In this endogenous information environment, the results indicate that participants form more accurate subjective beliefs and experience more belief adjustment in the low probability treatment than in the high probability treatments. MATERIAL-SPECIFIC DISCOUNT RATE FOR RISK ANALYSIS IN TRANSPORTATION ENGINEERING ECONOMICS Xiao Qin, University of Wisconsin-Milwaukee, qinx@uwm.edu Kai Wang, University of Connecticut, kai.wang@uconn.edu Zhiguang Wang, South Dakota State University, Zhiguang.Wang@sdstate.edu Engineering economic analyses (EEAs) such as life-cycle cost analysis (LCCA), benefit-cost analysis (BCA), incremental benefit-cost analysis (IBCA), and present-worth analysis (PWA) apply economic methodologies to engineering problems for decision-making support. When conducting an EEA, it is important to choose sound economic parameters that directly affect the results. It is also critical for policy makers to understand associated uncertainties with an EEA so as to enhance the credibility and accountability of any transportation infrastructure investment decisions. In this study, we have investigated the impact of discount rate on the economic analysis due to the nature of th 56 Annual TRF Program, Atlanta, GA, Page 21 different raw materials, monetary policies, and technological innovations. In particular, a material-specific discount rate has been considered because it may be more appropriate than a general discount rate when the inflation rate varies by material. Accordingly, a stochastic approach (also referred to as a “risk analysis”) may be more reasonable for EEAs than a deterministic approach when such uncertainties exist. We have applied pavement type selection LCCAs as case studies to illustrate the outcomes between the general discount rate and the material-specific discount rate, and to highlight the differences between the deterministic approach and the stochastic approach. Results indicate that if construction materials are significantly different, the material-specific discount rate should be considered for comparing design alternatives at the project level. When input parameters in EEAs are highly uncertain, it is recommended to use a risk analysis that provides a confidence interval for the present value. ALIGNING PROGRAM LEVEL AND PROJECT LEVEL OBJECTIVES FOR RISK MANAGEMENT: A CASE STUDY OF PROCESS INNOVATION IN A STATE DEPARTMENT OF TRANSPORTATION Elie Ji-Yun Sung, Georgia Institute of Technology, sung.ejy@gatech.edu Gordon Kingsley, Georgia Institute of Technology, gordon.kingsley@pubpolicy.gatech.edu Juan Rogers, Georgia Institute of Technology, juan.rogers@pubpolicy.gatech.edu In this paper we explore factors that influence the adoption of risk management practices within state transportation agencies. According to the Project Management Institute, risk is an uncertain event or condition that, if it occurs, has a positive or a negative effect on at least one (project) objective. The research literature on risk management suggests effective strategies require coordinated policies and practices at the agency level, program level, and project level of operations. Previous research on the adoption of risk management practices have typically focused on one of two levels (the project level and, to a lesser extent, program level). In this study we examine the alignment between levels as a factor in successful adoption of risk management practices. Alignment problems are found to be a significant impediment to the adoption of risk management practices by senior and mid-level managers. A case study is developed of a state transportation agency engaged in a process of reviewing and updating risk management practices. This is an embedded case design of preconstruction design engineers and technical specialists working in the context of the project development process (PDP). We compare and contrast the views of two classes of managers responsible for the program level and project level of operations. The case study was developed as part of a larger agency project aimed at formalizing a comprehensive risk management approach for the PDP and developing a register of the highest risk factors at each stage of the process. Semi-structured interviews were conducted with managers responsible for the program level and the project level of operations. Each set of managers was asked to identify and describe the factors influencing the adoption of effective risk management practices within the agency during critical steps in the PDP. We found that decisions related to the use of technological and administrative solutions to deal with risk and improve project delivery are mainly the responsibility of project managers. However, we also found that individual offices within the agency have their own views of risk related to office operations and that these can often be at odds with views of risk management held by program level and project level managers. The analysis of interviews reveals that the consequence of these varying views of risk management lead to alignment gaps between the program level strategic objectives of project delivery and their actual impact after implementation at the project level. Building from this case study we develop a model of the change in risk management strategies and practices within the agency. This is a process innovation model designed to explain the influence of alignment on the interactions between roles or offices at the various levels of operations and the adoption of new technological tools related to the adoption of risk management practices. th 56 Annual TRF Program, Atlanta, GA, Page 22 SESSION 2.2: Infrastructure Investment and Economic Development THE PANAMA CANAL EXPANSION AS A DRIVER OF INTERNATIONAL TRADE Anthony M Pagano, University of Illinois at Chicago, amp@uic.edu Onesimo Sanchez, Panama Canal, Osanchez@pancanal.com Miguel Verzbolovskis, Panama Canal, mverzbolovskis@pancanal.com Eddie A Tapiero, Panama Canal, etapiero@pancanal.com The opening of the Panama Canal in 1914 opened new transoceanic markets and facilitated the development of untapped trade routes, shortening the time and distance between centers of production and consumption and allowing countries to reap the attendant benefits by the introduction of a new way of trading. By the end of the last century, as transportation costs became progressively important in the overall landed price of goods, ship owners began to look at increasing vessel sizes to reduce per-unit transportation costs through economies of scale. As a result of changes in the global fleet that favored the use of larger vessels unable to traverse the Canal’s locks, the Canal expansion program was approved in 2006 by national referendum. Several actors along the global supply chain will benefit once the new Canal opens for business. The container trade will be able to take advantage of greater economies of scale offered by the enlarged waterway, and of Panama´s connectivity: there are 28 liner services moving cargo between high density markets (such as the Asia – U.S. East Coast and South America West Coast – U.S. East Coast routes) and 33 feeder services, serving the Latin American regional trade. Much of the cargo to be transported through the Canal in larger vessels will target major ports in the US East Coast and the Gulf of Mexico, many of which are investing in necessary dredging. This infrastructure upgrade will in in turn aid US dry bulk exporters and boost the shipment of liquefied natural gas (LNG) from the US Gulf to Northeast Asia. Other promising trades are those of Colombian coal and Brazilian iron ore exports to China. In short, the expanded Canal is expected to have a significant impact on world trade patterns. As a result, the expansion project and other efforts to enhance Panama´s transport and trade services conglomerate provide the precise opportunities for new developments in the maritime industry. This paper will examine world trade patterns and estimate how these patterns may change with the opening of the expanded Panama Canal. IMPACT OF THE PANAMA CANAL EXPANSION IN THE U.S. INTRAMODAL AND INTERMODAL TRANSPORTATION SYSTEMS: OPTIMIZATION MODEL FOR U.S. CONTAINER SHIPMENT Ju Dong Park, North Dakota State University, judong.park@ndsu.edu Won W. Koo, North Dakota State University, won.koo@ndsu.edu The main objective of this study is to analyze the impact of the Panama Canal expansion on trade flows of manufactural goods shipped by containers in the United States and evaluate the impacts of alternative ocean and inland transportation systems in shipping containerized cargoes from major exporting countries to major importing regions and countries. More specifically, the study is designed to undertake the following scenarios: (1) Estimate the impacts of transportation costs paid by exporters on the world container trade and the United States’ competitiveness, (2) Analyze the role of the Panama Canal on US trades of container shipment with its major trade partner countries in Asia and Europe, focusing on the impacts of changes in toll rates and delay costs at Panama Canal on US container shipment trade due to the canal expansion project, (3) Evaluate the impacts of competition between intermodal transportation system (rail services) and all-water route (Panama Canal) on US trades of container shipment between Northeast Asia and US East Coast, and (4) Evaluate the impacts of intermodal competition between rail and barge transportation on container shipment flows through US Gulf ports. The authors developed an optimization model for container trade in the United States and present a model framework capturing impacts to intramodal and intermodal transportation systems due to underlying uncertainty in demand, cost structure changes, and potential infrastructure constraints. The framework will be used to analyze the potential impacts of Panama Canal expansion on ocean and inland transportation systems in container trade from th 56 Annual TRF Program, Atlanta, GA, Page 23 major exporting countries to major importing regions and countries on the United States’ competitiveness of container traffic and the global container trade. AN EMPIRICAL STUDY OF THE TRANSSHIPMENT OF THE PORT OF HONG KONG Dingtong Yang, The Hong Kong Polytechnic University, steven.yang@connect.polyu.hk Tsz Leung Yip, The Hong Kong Polytechnic University, t.l.yip@polyu.edu.hk In this paper, we analyze the relationship between direct trade and transshipment throughput. Transshipment is a major component of international shipping. Transshipment is defined as the total inward and outward of containers which are stored in the yard side but not transferred to the land transportation. In the development of global trade, transshipment has been contributing significantly to cost reduction and route optimization. Some of the world famous ports are relying heavily on transshipment for their development. In Singapore port, above 80% of the entire volumes of container throughput are transshipment. Hong Kong is transforming to a transshipment hub in the past ten years. In the past 12 years, Hong Kong's transshipment has increased dramatically. Hong Kong used to be a direct trade dominant port (a gate port with 60% direct trade and 40% transshipment in year 2001). In year 2013, the transshipment volumes share more than 70% of the total throughput. Port throughput has been studied widely. A common approach is to analyze port throughput by using multiple methods and models. These methods could yield a decent prediction power of those forecasts. In literature, the port throughput is a combination of both direct trade and transshipment. Different from previous studies, we attempt to separate the port throughput of Hong Kong into direct trade and transshipment. Analysis confirms that transshipment throughput shows different time series properties from those of direct trade. We further investigate and model the relationships between direct trade and transshipment. This study should be one of first rigorous analyses of transshipment throughput in literature. U.S. PORT READINESS FOR THE PANAMA CANAL EXPANSION Grace Wang, Texas A&M University at Galveston, wangw@tamug.edu Anthony M. Pagano, University of Illinois at Chicago, amp@uic.edu The Panama Canal is currently in the process of a major expansion effort. After the new set of locks is opened at the end of 2015, significantly larger ships can traverse the Canal. The new locks will allow the passage of ships that carry up to 13,200 TEU’s, instead of the maximum of 5,100, which is transported in Panamax vessels. With the expansion, the Canal will allow ship owners to take advantage of economies of scale with the placement of larger ships on a shorter route, reducing the fuel and other operational costs associated with the crossing in new, more efficient post-Panamax vessels. Ship owners will benefit from the connectivity offered by the confluence of 30 liner services moving cargo between markets of highest density (as for example, routes from Asia-to the United States, South America - United States, East Coast West Coast US) and 33 feeder services that serve the regional trade of Latin America and the Caribbean (Pagano, Wang, et al, 2014, IAME). This paper estimates the impact of the expansion on US ports. The expansion may have very dramatic effects on all North American ports, especially those on the East Coast and in the South. As stated in the North American Port Analysis by the Colliers International, shift in GDP growth from Europe and Asia to Latin America, Canada, and Russia implies “North American ports and inland distribution centers will likely experience the most activity in the first port-Panamax decade, as the Gulf coast, Southeast Atlantic coast, and Great Lakes ports that are strategically positioned to benefit” (Colliers, 2013). US waterborne trade by world region from the American Association of Port Authorities also confirms the shift of the global trading pattern and major trading partners with the US. The research covers nine container ports on the East and Southern coasts of the US. The ports examined are the East Coast ports of New York-New Jersey and Baltimore; South-east ports of Virginia, Charleston and Savannah; Florida ports of Miami, Jacksonville, and Everglades; and the Gulf port of Houston. Each of these ports is in a th 56 Annual TRF Program, Atlanta, GA, Page 24 different degree of readiness to accept the larger ships that can navigate the expanded Canal. Some of the ports are in the process of expanding their water-side and land side facilities to accommodate the larger ships. Others may delay expansion and it is not clear when if ever these will be completely ready. In order to 1) measure port readiness to accommodate larger vessels, 2) to assess current capacity across the largest container ports in the North America, and 3) to set up reasonable expectation of growth in terms of volume handled, we identify fast-growing ports and their competitive advantages. The ports must have the water infrastructure in place to handle larger ships and the land infrastructure in place to accommodate the increased volume of container traffic. This paper focuses on identifying existing and future capacity at the ports including infrastructure, land side connections, and land availability. SESSION 2.3: Logistics Planning and Transportation Supply Chains REGIONAL FOOD LOGISTICS: A STAKEHOLDER PROCESS TO INFORM MULTI-SYSTEM REDESIGN FOR SUSTAINABILITY Michelle Miller, UW-CIAS, mmmille6@wisc.edu Marina Denicoff, USDA-AMS-Transportation Division, Marina.Denicoff@ams.usda.gov Lindsey Day-Farnsworth, University of Wisconsin - Madison, ldfarnsworth@wisc.edu Until recently, local and regional food system development efforts have emphasized small scale direct marketing activities, while food freight transportation policy and planning have primarily focused on distribution infrastructure for large-scale commodity products. As the demand for local food continues to increase, innovative and scale-appropriate infrastructure and expertise are needed to respond to the market pull. The key question among the stakeholders continues to be: “How can we better move sustainable food products to meet regional market demand?” This paper highlights the process of collaboration between USDA’s Agricultural Marketing Service (AMS) and the University of Wisconsin-Madison Center for Integrated Agricultural Systems (CIAS) to focus on transportation and distribution issues in order to facilitate effective growth in local and regional food systems. The process brought representatives of regional food supply chains together to discuss their experiences and compare notes. More than one hundred participants heard farmer/shippers, distributors, and wholesale buyers discuss innovations and challenges to moving local food in the Upper Midwest. Researchers, Extension educators, planners and NGOs listened and summarized some of the conversations around communicating value, logistical challenges, and creating more sustainable food supply chains. The process catalyzed a series of discussions and actions within the region. The result of the conference was to bring clarity to several themes identified by the participants as key to developing regional food transportation networks, including defining the meaning of “local” and the resulting market differentiation strategies; fostering relationships to improve logistics; identifying first and last mile issues; and, addressing supply chain scale and infrastructure. Business networking emerged as a central function of the meeting, resulting in the development of new supply chains. Another result is a project to investigate strategies that may optimize food freight operations into the Chicago mega-region. To read the published report, please see: http://www.ams.usda.gov/AMSv1.0/getfile?dDocName=STELPRDC5108089 HURRICANE EVACUATION BEHAVIOR OF THE SPECIAL NEEDS POPULATION: EMPIRICAL FINDINGS TO SUPPORT EMERGENCY LOGISTICS PLANNING ManWo Ng, Old Dominion University, mng@odu.edu Despite the widely recognized importance of hurricane evacuation planning for residents with special needs, there is essentially no research available that can be used as guidance, as opposed to the numerous empirical research studies on the evacuation behavior of the general population. In this research, we provide these long-overdue insights using data from a large-scale phone survey (over 7,000 samples) conducted in the aftermath of hurricane Irene in the Hampton Roads region in Virginia. Using aggregate and disaggregate analyses, we start to unravel the behavior of this heavily understudied, and potentially vulnerable population group. Our analysis indicates that key differences exist between the special needs and the general population. th 56 Annual TRF Program, Atlanta, GA, Page 25 U.S. 95 FREIGHT MULTI-MODAL CORRIDOR SUPPLY CHAIN: A PILOT STUDY Jeremy Sage, Freight Policy Transportation Institute, jlsage@wsu.edu Ken Casavant, Freight Policy Transportation Institute, casavantk@wsu.edu Significant gaps in knowledge often persist as to the quantity and diversity of products, their value, and supply chain characteristics of commodities moving on freight corridors. Readily available national level data does a fine job of denoting origin and final destination but does not address easily identify the finer path freight products take, especially when freight moves from one mode to another mode. These gaps constrain system managers’ ability to fully account for the costs and benefits of freight corridor investment and to positively affect freight multimodal connectivity. Constraints such as these ultimately impact the efficiency by which local and regional commerce operates by restricting our ability to identify appropriate multi-modal connections and investments – whether they are public or private investments. This study establishes a consistent framework by which freight economic corridors may be established and evaluated. Using the full extent of US-95 in Idaho, we implements a watershed approach to corridor analysis through the detailed assessment of the upstream and downstream connectivity of the corridor. Resulting from this approach is an enhancement of capacity for managing agencies to better prioritize infrastructure investment under the guise of complete supply chain corridor. A FIELD EXPERIMENT ON CARRIER RESERVATION PRICES FOR TRUCKLOAD CAPACITY IN THE TRANSPORTATION SPOT MARKET Christopher Lindsey, Cambridge Systematics, clindsey@camsys.com Hani Mahmassani, Northwestern University, masmah@northwestern.edu An important element of the U.S. motor carrier industry is the transportation spot market. The transportation spot market consists of shipments handled on a one time load-by-load basis and exists to facilitate urgent or unfulfilled demand. It is characterized by price volatility and uncertainty in the availability capacity. These aspects of the spot market make it very challenging for shippers who rely on it. Because of the severe shortage of spot market capacity and its relatively high and volatile prices, shippers must actively as opposed to passively (i.e. conduct traditional transportation auctions) seek carrier capacity. Often, they employ the help of third parties. This research yields insight into the spot market by presenting some results of a hypothetical field experiment on spot market transactions for truckload shipments. In the experiment, motor carriers were presented with a number of shipment options and asked to select the most preferable option. The choice experiment assessed the effect of price and shipment characteristics such as lead time, empty miles, lane, cost and others on the desirability of a shipment (or group of shipments) by randomly varying these attributes across choice scenarios. The research asserts that a shipper’s efforts to secure truckload capacity could be improved with better information on how carriers value potential transactions differently. It does this by demonstrating that given choice data, it is possible to accurately estimate the minimum amount of compensation various carriers require to provide capacity – the reservation price. Furthermore, it explores the practice of sourcing multiple spot market loads simultaneously – bundling. The insights from this study can lead to improved decision-making and improved business outcomes for shippers and carriers alike. th 56 Annual TRF Program, Atlanta, GA, Page 26 SESSION 2.4: Mobility Issues in Bus and Rapid Transit Systems A SELF-COORDINATING URBAN BUS ROUTE: IMPLEMENTATION AND FIELD EXPERIENCE John J. Bartholdi III, Georgia Institute of Technology, john.bartholdi@gatech.edu Russell J. Clark, Georgia Institute of Technology, russ.clark@gatech.edu Donald D. Eisenstein, University of Chicago, don.eisenstein@boothchicago.edu Loren K. Platzman, independent scholar, loren_platzman@mindspring.com David W. Williamson, Georgia Institute of Technology, david.williamson@pts.gatech.edu We describe our experience with a software system to reduce bus-bunching by strategically delaying buses at selected check points on the route. This scheme is notable for its minimal requirements. It abandons the idea of a target schedule or even target headway and instead focuses on equalizing the headways rather than on trying to achieve any particular target. The system adapts spontaneously to changes in traffic intensity, ridership, bus breakdowns, or even detours on the route. After for more than a year of testing, our system went live on 12 August 2013 and assumed control of the main bus route through the campus of the Georgia Institute of Technology. Control has since been extended to additional routes. The scheme has delivered more reliable headways and with less management oversight than did scheduled service. THE COST STRUCTURE OF US PUBLIC TRANSPORTATION: CONSIDERING THE IMPACTS OF CLIMATE CHANGE David Ripplinger, North Dakota State University, david.ripplinger@ndsu.edu John Bitzan, North Dakota State University, david.ripplinger@ndsu.edu The economic costs of public transportation under current and alternative fleet-fuel configurations are estimated. The analysis builds on a previous analysis of the cost structure of US public transportation agencies by incorporating the costs of greenhouse gas emissions associated with diesel, biodiesel, and compressed natural gas. ACCESSIBLE TRANSIT STOPS AND A LINK TO PASSENGER DEMAND: TOOL TO PRIORITIZE AND COORDINATE ACCESSIBILITY IMPROVEMENTS Todd Hansen, Texas A&M Transportation Institute, t-hansen@ttimail.tamu.edu Matt Killary, Texas A&M Transportation Institute, m-killary@tti.tamu.edu Accessible streets, sidewalks, and other public infrastructure are crucial to viable transportation options for all people—especially accessing public transit. Communities across the United States have inaccessible transit stops lacking functioning sidewalks, improperly constructed or non-existent curb ramps, barriers in the way of poles, trashcans and newspaper stands, and lack of benches and shelters. With limited resources, how can communities work to better prioritize investment to maximize the benefit of accessibility improvements for the community? The methodology presented provides an approach to prioritize needed improvements for bus stops based on both their present condition and the impact to current and potential ridership to determine highest accessibility improvement need. Houston METRO and METROLift, the ADA paratransit service of the agency, applied the methodology as a means to work with stakeholders in targeting resources to improve accessibility where significant ridership demand currently exists. Researchers developed a two-tier methodology using a database inventory of physical bus stop location attributes combined with general public and ADA paratransit passenger location data to weight the locations of the bus stops in the index by importance. Accessibility needs for this study include features such as paved sidewalks in good condition, designs for safety, curb cuts, good signage, separation from auto traffic, and comfort for waiting riders. th 56 Annual TRF Program, Atlanta, GA, Page 27 The first-tier is an index that ranks the features of each bus stop based on categories of accessibility and amenities. Each index data component has a different weight based on the importance of the feature for a person with a disability to access the fixed route. The second tier applies the passenger trip data (both general public and ADA paratransit) around stops to determine the potential ride frequency. The index results prioritize the need for accessibility investment. Researchers further assessed whether the investment is the transit agency responsibility or a city responsibility to help in coordination efforts. The results also identify potential areas around bus stops to focus travel training efforts. SESSION 3.1: Community Issues in Transportation CHALLENGES AND ENABLING MECHANISMS FOR DEVELOPMENT OF HIGHWAY PUBLIC-PRIVATE-PARTNERSHIP PROJECTS IN THE UNITED STATES Kia Mostaan, Georgia Institute of Technology, kiamostaan@gatech.edu Baabak, Ashuri, baabak.ashuri@coa.gatech.edu Involvement of the private sector in financing highway projects in the form of public-private partnerships (P3s) is subject to various limitations and challenges that affect state departments of transportation (DOTs) P3 project planning and development, and limit the expansion of the US highway P3 market. Public and private sector stakeholders need to have a better understanding of opportunities for improvement that can contribute to the US highway P3 market growth and promote excellence in P3 project planning and development. The main objective of this study is to identify and explore opportunities that can help state DOTs facilitate transportation planning and financial structuring of highway P3s. Following interviews with sixteen P3 industry experts, we categorize the enabling solutions and recommendations for development of highway P3s into: (1) Enabling financial mechanisms; and (2) Management and organizational recommendations. The recommended enabling mechanisms are then analyzed in detail. Results indicate that P3 developers and contractors can significantly benefit from financial flexibility offered by off balance sheet financing mechanisms and accounts receivable purchase agreements. Assetbased financing and securitization offers state DOTs financial flexibility and utilizes local interest in P3 projects. Further, authorizing secondary P3 market investments can attract a pool of interested investors to P3 projects. State DOTs can also establish mature and transparent P3 programs in order to attract interested investors and promote the partnership culture between the public and private stakeholders. Finally, it is recommended that state DOTs pursue development and procurement of P3 project portfolios to reduce transaction costs and promote competition. URBAN-POVERTY AND PUBLIC TRANSIT: EVIDENCE FROM THE ATLANTA METROPOLITAN REGION Rahul Pathak, Georgia State University, rpathak3@student.gsu.edu Christopher Wyczalkowski, Georgia Institute of Technology/Georgia State University, cwyczalkowski1@student.gsu.edu Xi Huang, Georgia Institute of Technology/Georgia State University, xhuang67@gatech.edu The factors affecting concentration of the urban poor has been a subject of considerable debate and scholars have argued that transportation-cost, housing-cost, familiarity with a neighborhood, and particularly access to publictransport are some of the important factors influencing location decisions of the poor. With limited resources, the poor would like to minimize their transportation and housing costs and would prefer access to public-transport especially if the cost of owning and operating an automobile is high. The last decade has witnessed significant increase in the cost of owning and operating an automobile in the country and a fluctuation in real estate values and rents. These trends are also accompanied by a fundamental shift in the geography of urban-poverty as more poor now live in suburban areas than in the inner-city, and inner-ring suburbs have experienced an increase in poverty while outer-ring suburbs have become wealthier, vis a vis the urban core. Is the rise in suburban-poverty in the American cities consistent with changes in public-transportation systems or is transit no longer a major factor influencing location decisions of the poor? In this paper, we focus on Atlanta Metropolitan Area and use standardized census tract level Geolytics data from 1970-2010 and Metro Atlanta Rapid Transit Authority (MARTA) bus route data to examine this question using regression and spatial analysis. We analyze at the census tract level, th 56 Annual TRF Program, Atlanta, GA, Page 28 poverty trends in the Atlanta MSA with corresponding changes in commuting behavior and public-transport access, comparing inner-core, inner ring suburbs, and outer ring suburbs. Initial results indicate that transportation is one of the important factors explaining the rise of poverty in the suburbs; an increase in the number of people commuting by mass transit is positively correlated with the number of people in poverty. THE BODA-BODA AS A TOOL FOR FEMALE EMPOWERMENT IN KAMPALA, UGANDA Camille Matonis, Harvey Mudd College, cmatonis@students.pitzer.edu The roadways of many Sub-Saharan African cities are a chaotic blend of wheels and legs, where the lack of effective public transportation has led to the dominance of paratransit modes. In the growing city of Kampala, Uganda, this comes in the form of motorcycle and bicycle taxis called boda bodas that carry multiple passengers and offer the fastest and most convenient transportation option. The aim of this research is to examine the use of boda bodas by women and the potential for more women to become boda boda drivers. Informal transit could lessen the gender gap in Kampala by providing women with a transportation mode more suitable for trips related to their roles in society. And because driving a boda boda is very profitable, the entrance of women into this male-dominated occupation could increase gender equality and encourage other women to ride boda bodas. A FRAMEWORK TO IDENTIFY ECONOMIC AND COMMUNITY BENEFITS OF THE PROPOSED BRT SYSTEM IN SOUTHEAST MICHIGAN Utpal Dutta, University of Detroit Mercy, duttau@udmercy.edu Ramakrishna R. Tadi, State of California Department of Transportation, Ramakrisha.R.Tadi@dot.ca.gov Sawan Dutta, University of Michigan, dsawan@umich.edu In recent years, Bus Rapid Transit (BRT) has generated great interest across the United States. There are more than 20 BRT systems in existence and more are in the planning stage (including Detroit). Within the next few years, BRT will be planned and implemented phase by phase in various parts of Southeast Michigan. The purpose of this paper is to present a framework to identify probable economic and community benefits (quantitative as well as qualitative) of BRT in Southeast Michigan. Taxable Real Estate property, employment sector, population age group, daily vehicle miles traveled, as well as, fatal and injury crash data were reviewed to identify Southeast Michigan’s current and future trends. A Shift-Share analysis using data from Cleveland was performed to determine BRT advantage age group. The authors suggested a number of action items to attract choice riders and excite transit dependent riders to the planned BRT system. Based on the literature review and analysis, the authors identified BRT advantage job sectors and age-groups within the Southeast Michigan region. As BRT will be implemented in phases, it will affect the amount, type and time of investment. Considering this uncertainty in implementation, projected economic benefit as a function of type and amount of investment was presented. It is to be noted that in order to achieve the desired outcome the system must be planned / designed / implemented within the context of Southeast Michigan rather than just copying a successful system from somewhere else. th 56 Annual TRF Program, Atlanta, GA, Page 29 SESSION 3.2: Information Technologies for Collecting Transportation Data COMMUTE WARRIOR: ANDROID APPLICATION FOR COLLECTING LONGITUDINAL TRAVEL SURVEY DATA Alper Akanser, Georgia Institute of Technology, aakanser@gatech.edu Vetri Venthan Elango, Georgia Institute of Technology, vetriventhan.elango@ce.gatech.edu Alice Grossman, Georgia Institute of Technology, agrossman3@gatech.edu Ramik Sadana, Georgia Institute of Technology, rsadana3@gatech.edu Komal Poddar, Georgia Institute of Technology, kpoddar6@gatech.edu Yanzhi “Ann” Xu, Georgia Institute of Technology, yanxhi.xu@ce.gatech.edu Randall Guensler, Georgia Institute of Technology, randall.guensler@ce.gatech.edu Use of Global Positioning Systems (GPS) devices provides significantly enhanced spatial and temporal accuracy in travel surveys, higher survey completion rates, and fewer missed trips than traditional travel surveys. Recent advances in smartphone technology and the widespread market penetration of smartphones in the United States make them excellent devices to use in travel survey data collection. Smartphones also have the advantage that people voluntarily carry them almost all the time. However, travel data collection systems based on smartphones have challenges, such as limited battery power availability, privacy concerns, and a need to minimize interaction with users to avoid survey fatigue. Georgia Tech researchers have developed Commute Warrior, an Android application that passively collects travel data for long periods of time while minimizing battery use and providing privacy options to users. The App also features built-in survey functionality to collect attitudinal data that enhance analysts’ ability to process and interpret the passively collected data. This paper details the development of a smartphone based travel data collection system, explains the features of the system, and presents example data. The Commute Warrior travel data collection system provides a rich dataset that can be used for studying driving behavior, emissions analysis, and travel time variability. DEVELOPING A METHOD TO ASSESS NATIONAL DEMAND RESPONSE LEVEL OF SERVICE Ranjit Godavarthy, Upper Great Plains Transportation Institute, ranjit228@gmail.com Jeremy Mattson, Upper Great Plains Transportation Institute, jeremy.w.mattson@ndsu.edu Del Peterson, Upper Great Plains Transportation Institute, del.peterson@ndsu.edu Patrick Nichols, Upper Great Plains Transportation Institute, patrick.nichols@ndsu.edu Jill Hough, UGPTI, jill.hough@gmail.com Demand response transit service is a major source of mobility for elderly and disabled Americans in urban and rural areas. Federal Transit Administration (FTA) grant programs under sections 5307, 5310, and 5311 all have components designed to increase the availability of paratransit or demand response service. However, there is little information in the National Transit Database (NTD) or elsewhere about the extent of demand response coverage across the country. Also, availability of service data is not uniform for all the agencies and the accuracy of the available service area; service times, etc. are questionable. Therefore, it is challenging to know the gaps in the service coverage and to understand unmet needs. The primary objective of the study is to fill the data gaps to the available NTD database to effectively determine the demand response level of service. To that end, a map questionnaire tool is being developed to gather important information, such as geographic coverage and service frequency, from demand response operators. The tool is being tested in two states: North Dakota and Florida. Data obtained from this tool can be used in GIS with existing census data to identify locations that need service improvements. This study evaluates the effectiveness of this tool for collecting the desired information. Based on the results from the project, the study will provide recommendations regarding data needs and an appropriate method for collecting those data. The recommended framework will provide useful information to th 56 Annual TRF Program, Atlanta, GA, Page 30 transit agencies, MPOs, and state DOTs for identifying deficiencies in service while minimizing reporting burden for transit providers. COMPARATIVE EVALUATION OF TECHNOLOGIES / DATA SOURCES TO CAPTURE TRAVEL TIME ALONG ARTERIAL STREETS Rahul C. Pinnamaneni, The University of North Carolina at Charlotte, rpinnama@uncc.edu Srinivas S. Pulugurtha, The University of North Carolina at Charlotte, sspulugurtha@uncc.edu Venkata R. Duddu, The University of North Carolina at Charlotte, vduddu@uncc.edu RM Zahid Reza, The University of North Carolina at Charlotte, rreza@uncc.edu Accurate travel time information is vital to efficiently plan and effectively manage the transportation network. Technologies and data sources such as Bluetooth detectors and INRIX offer the potential to collect travel time data continuously and use it for long-term transportation planning as well as real-time traffic condition monitoring. However, their ability to accurately collect travel time data on arterial streets is still unclear. This paper focuses on capturing link/section level travel times on arterial streets using Global Positioning System (GPS) floating car method, Bluetooth detectors, and INRIX and comparing it with manual floating car method for each travel run. A filtering technique was incorporated to eliminate any possible outliers from the data collected using Bluetooth detectors for improved travel time estimates. Results obtained indicate that travel time data captured for arterial streets using Bluetooth detectors were observed to be less accurate and not dependable when compared to GPS and INRIX. The number of samples captured using Bluetooth detectors was also observed to vary by time-of-the-day (low during morning peak hours) and data filter range. SIDEWALK SCOUT: CROWDSOURCING ANDROID APPLICATION FOR COLLECTING SIDEWALK CONDITION DATA Alper Akanser, Georgia Institute of Technology, aakanser@gatech.edu Vetri Venthan Elango, Georgia Institute of Technology, vetriventhan.elango@ce.gatech.edu Alice Grossman, Georgia Institute of Technology, agrossman3@gatech.edu Yanzhi “Ann” Xu, Georgia Institute of Technology, yanxhi.xu@ce.gatech.edu Randall Guensler, Georgia Institute of Technology, randall.guensler@ce.gatech.edu Sidewalk condition inventories can help local governments and other agencies allocate maintenance and repair resources efficiently and, if made public, can inform pedestrians as to current sidewalk condition. The ubiquitous mobile phone is a promising platform for collecting data related to sidewalk condition, given its integrated camera and GPS functionalities. Sidewalk Scout™ is a crowdsourcing Android app that allows users to record photos of individual sidewalk elements and report specific defects in sidewalk conditions or pedestrian amenities, including sidewalk obstruction, inadequate sidewalk width, improper slope and cross-slope, pavement discontinuities, broken walk signal heads, etc. The Sidewalk Scout™ app feeds data directly into the Georgia Tech sidewalk inventory system, which includes a publicly available webpage to view mapped data with text and image attributes. This paper details the development of the Sidewalk Scout™ Android application along with the preliminary field testing of the application. SESSION 3.3: Transportation Efficiency and Productivity PRODUCTIVITY AND INFRASTRUCTURE INVESTMENT: THE LINK BETWEEN PRODUCTIVITY AND ECONOMIC DENSITY IN NORWAY Eivind Tveter, Molde Research Institute, eivind.tveter@himolde.no Large infrastructure projects can contribute to increased productivity due to agglomeration effects if there is are increasing return to agglomeration These benefits is “wider” in the sense that they are typically not included in a th 56 Annual TRF Program, Atlanta, GA, Page 31 cost benefit appraisal. The importance of this, proposed, link between productivity and infrastructure depends on the quantitative estimate of return to agglomeration. This paper examines, the link between productivity and agglomeration in Norway, using both “pure” employment density and effective density, where density is weighted by distance between regions, as indicator of agglomeration. A model using data at the Municipality level is estimated. .The inherent problem when examine productivity and density, is the possibility for reversed causation. One source of reversed causation is that employees are attracted to a region with high wages, which entails that density is causing wages, not the other way around as postulated by the model. This is accounted for using an instrument variable technique. The results indicate productivity effects from agglomeration, which is higher than the average result in the literature. The result using the “pure” density is an elasticity of wages with respect to density of 0.08, accounting for reversed causation produces a slightly lower estimate of 0.07; using the effective density result in an elasticity of 0.08, and accounting for reversed causation result in an elasticity of 0.13. When the estimate of 0.07 is used on a fixed link project the agglomeration benefit increases the total benefit by 12 percent. HARD RED SPRING WHEAT MARKETING: EFFECTS OF INCREASED SHUTTLE TRAIN MOVEMENTS ON RAILROAD PRICING IN THE NORTHERN PLAINS Elvis Ndembe, North Dakota State University, elvis.ndembe@ndsu.edu The U.S. grain marketing and transportation sector has witnessed substantial increases in the use of shuttle. Shuttle trains described here as those with 100 or more cars are relatively more efficient in moving commodities relative to other rail services. Likely benefits in the form of lower rates stemming from efficient operations are potentially more important in areas that are highly dependent on rail transportation (e.g. North Dakota) due to captivity; perceived captivity, and related issues. The main aim of this paper is to assess the relative benefits of using shuttle trains in shipping hard red spring wheat from North Dakota compared to other rail services. This is undertaken using a time series technique using data between 1999 and 2012. ESTIMATING THE TECHNICAL AND ALLOCATIVE EFFICIENCY OF U.S CLASS I RAILROAD: A DATA ENVELOPMENT ANALYSIS Elvis Ndembe, North Dakota State University, elvis.ndembe@ndsu.edu Saleem Shaik, North Dakota State University, saleem.shaik@ndsu.edu Since passage of the Staggers Act in 1980, the U.S class I railroad industry has adopted various cost reducing strategies to enhance their efficiency. For example mergers and acquisitions allowed less financially stable firms to be taken over by more stable ones to reduce widespread bankruptcies and excess capacity that plagued the industry. The industry has become the most concentrated than it has ever been following a wave of mergers the latest of which occurred in the late 1990s and early 2000s. This and other related strategies have potentially moved the industry towards the efficiency frontier. In that case, the industry is making optimal use of inputs given prices and outputs and producing optimal outputs given inputs. The main aim of this paper is to evaluate the technical (TE) and allocative (AE) efficiency of U.S class I railroad between 1997 and 2011 using data envelopment analysis (DEA). th 56 Annual TRF Program, Atlanta, GA, Page 32 REVISITING CONCENTRATION IN FOOD AND AGRICULTURAL SUPPLY CHAINS: THE WELFARE IMPLICATIONS OF MARKET POWER IN A COMPLEMENTARY INPUT SECTOR Metin Cakir, U. of Minnesota, mcakir@umn.edu James Nolan, University of Saskatchewan, james.nolan@usask.ca The use of complementary inputs is a key characteristic of the production process in many industries. In this paper we explore how market power in a complementary input sector compares to the exertion of market power in a downstream sector for both producer and consumer welfare, as well as for policy. We develop a model of a homogenous product market that encompasses both bilateral and complementary relationships. Our model focuses on the primary input sector and allows for exertion of market power by both complementary input suppliers and downstream firms. We use comparative statics analysis with numerical simulations to study economic equilibrium under different scenarios of market power exertion. With respect to the welfare of primary input suppliers, our primary finding is that market power exercised by the supplier of a complementary input generates greater negative effects than the same level of market power exercised by the downstream firms. We provide a discussion of the implications of the results for policy in the context of current problems within the Canadian grain handling and transportation system. SESSION 3.4: Energy Applications in Transportation PIEZOELECTRIC-BASED ENERGY HARVESTING TECHNOLOGY FOR GEORGIA HIGHWAY SUSTAINABILITY Seonghoon Kim, Georgia Southern University, shkim@georgiasouthern.edu Junan Shen, Georgia Southern University, jshen@georgiasouthern.edu Mohammad Ahad, Georgia Southern University, mahad@georgiasouthern.edu Dukgeun Yun, Korea Institute of Construction Technology, dkyun@kict.re.kr The benefits of a roadway energy harvesting system are potentially great, given the lane-miles and high traffic volume in specific areas of state highways. A piezoelectric method of energy harvesting has advantages over other alternative sources, such as solar panels and wind power. The primary goal of this research project is to prove that the piezoelectric method is a viable alternative energy source for roadways. The scope of the research project includes investigation of the energy harvesting method, a feasibility study, the framework of the piezoelectric method, preparation of equipment and materials, conduction of lab experiments, and development of potential future research. The research project focuses on conducting lab experiments to identify optimal conditions of energy harvesting with piezoelectric ceramic materials under asphalt pavements. Preliminary research results indicate that the levelized cost of energy (LCOE) is relatively high, but potential energy generation can be improved by several variables. Thus, there is an urgent need to conduct studies regarding this technology in laboratory conditions with available products in the U.S. The results of this research project will contribute to the possibility of highways’ self-supporting energy capacity. The amount of generating capacity will be recorded and compared with other energy harvesting methods to determine economic competitiveness. HIGHWAY INVESTMENT, FREIGHT EMISSIONS, AND TRADE You Zhou, Freight POlicy Transportation Institute, you.zhou@email.wsu.edu Jeremy Sage, Freight Policy Transportation Institute, jlsage@wsu.edu Transport infrastructure investment decisions are increasingly multifaceted. The last decade has witnessed environmental concerns rising to be one of the most important factors. Capital competition between U.S. states is often considered as a factor driving many states' lax environmental standards. In this paper, the impact of investment in highway infrastructure on emissions from freight transportation is investigated. Highway investment can directly increase emissions through providing a better highway infrastructure, which leads to more transportation activities. Additionally, government agencies are likely to favor investment projects that put money th 56 Annual TRF Program, Atlanta, GA, Page 33 in public infrastructure to the benefit of trade. As a result, higher trade volume increases the demand for freight. In this manner, highway investment can indirectly increase emissions. A panel data model is employed to analyze the state-level freight CO2 emissions from 1995 to 2011. The results show that both direct and indirect effects are positive and significant. Additionally, by adding a neighbor's emission variable,a spatial panel data model is used to further investigate the interaction between state's own emission and its neighbor's. Interestingly, the result shows a negative and significant relationship. This may imply that the relatively higher road investment from neighbors can absorb some of freight activities induced by own road investment. Thus, effects of own highway investment on emission are impaired. Based on this implication, decision-makers on highway investment should always consider their neighbors. DEVELOPMENT OF NATURAL GAS DISTRIBUTION PROJECTS IN CHINA Liang, Chieh Cheng, University of Houston, lcheng6@central.uh.edu Ran Shi, University of Houston, serena.shi101@gmail.com Wanxiang Lu, University of Houston, nancy.lu.wx@gmail.com Yixiang Wu, University of Houston, Wyattwu1@gmail.com Chen Cao, University of Houston, cathy.ccao@yahoo.com Since natural gas becomes a dominant choice for city gas. Developing the city gas business depends on dynamic policy, market demand and available resources. Base on the development of shale gas, the production of natural gas sharply grew in last few years. This will increase natural gas supply sharply. However, the lack of knowledge and study on the market side would increase the uncertainty for the downstream market investors. Based on this finding, this team aims to understand how the downstream market of natural gas works currently. Therefore, this paper starts from macro background and SWOT of NG usage. Through literature review, upstream supply keeps on increasing in past 3 decades. At the same time, a lot of facilities like transmission pipeline and LNG ports have been constructed to connect different NG source to the distribution network. As the major market investors and player, natural gas distributors are growing rapidly along with the boosting market needs. We found the review of current distributors is necessary to understand the present competitions. In order to make better understanding, this part also introduces the U.S. distributors to conduct a comprehensive benchmarking between them. The main study focuses on two important factors – gas sales and pipeline length. After data collection, we established metric called “Pipeline Utilization Index” revealing the relation between them. Through tracking the changes of this index for last 5 years, we found it practical when evaluating the performance of distribution projects. Finally, we do hope giving reliable indications for future investment through all the data analysis and integrated information. LOCATING BATTERY SWAPPING STATIONS FOR SMART E-BUS SYSTEM Sang Hwa Song, Incheon National University, sanghwa.song.inu@outlook.com Taesu Cheong, Korea University, taesu.cheong@outlook.com Seungwon Na, Korea Aerospace University, seung1na@kau.ac.kr In this study, we introduce several mixed integer programming models and propose an efficient heuristic algorithm for locating battery swapping bus stations for smart electric buses (e-buses). A case study that focuses on locating of battery swapping bus stations in a metro area demonstrates the applicability of these proposed models. th 56 Annual TRF Program, Atlanta, GA, Page 34 POSTERS EFFECTS OF THE RISE IN OIL PRICE ON ETHANOL FUEL CONSUMPTION IN THE U.S. TRANSPORTATION SECTOR Jaesung Choi, UGPTI, jaesung.choi@ndsu.edu David C. Roberts, NDSU, david.c.roberts@ndsu.edu Over the past decade, increase in ethanol fuel production to substitute motor gasoline consumption in the U.S. transportation sector has become more economically feasible due to world oil prices remaining high. To reveal a relationship between oil price and ethanol fuel consumption, this study reviews the oil price elasticity of ethanol fuel consumption from 2001 to 2012 in the 14 U.S. states which historically show high ethanol fuel consumption. The findings show that the U.S. has an oil price elasticity of fuel ethanol consumption of 1.99, which suggests that if U.S. oil prices increase by 1%, then U.S. fuel ethanol consumption can increase by 1.99%. Furthermore, the elasticity in most of the states studied was much higher than in the U.S. Based on the results, the usage of imported crude oil in the U.S. can significantly decrease if the price of oil remains high in the future and that the U.S. will be able to more closely approach the achievement of energy security. PLUG-IN HYBRID ELECTRIC VEHICLES: OPTIMAL ALL-ELECTRIC DRIVING RANGE FOR MINIMUM SOCIETAL COST Eleftheria Kontou, University of Florida, ekontou@ufl.edu Yafeng Yin, University of Florida, yafeng@ce.ufl.edu The objective of this study is to determine the optimal all-electric driving range of plug-in hybrid electric vehicles (PHEVs) that minimizes the daily cost borne by the society when using the technology. An optimization framework is proposed for the purpose and applied to datasets representing the U.S. PHEV and automobile market. Results indicate that the optimal range is 22 miles with an average social cost of $3.02/day when exclusively charging at home. The optimal range is found to be sensitive to the cost of battery packs and the retail price of gasoline. When public charging is available, the optimal all-electric driving range surprisingly increases from 22 to 24 miles, as larger batteries would allow drivers to better take advantage of charging opportunities to achieve longer electrified travel distances, yielding social cost savings. For this case, the optimal density is to deploy a charger for every 0.3537 vehicle. Moreover, diversification of the size of battery packs, i.e., introducing a pair and triple of all-electric driving ranges to the market, could further decrease the average societal cost per PHEV by 12.31% and 12.85% respectively. POLITICAL IDEOLOGY AFFECTS THE WILLINGNESS TO PURCHASE ELECTRIC VEHICLES Yeong Jae Kim, Georgia Tech, ykim445@gatech.edu Democrats’ increasing support for climate change policies affects consumer's willingness to purchase an electric vehicle over a conventional vehicle. Previous research indicates that the political ideology affects the energyefficient product choice behaviors. In addition, recent research reveals that early adopters of electric cars tend to be more Democrats than late adopters. Taken into consideration of widening partisan gap on climate change issues, I hypothesize that Democrats are more likely to consider electric vehicles than Republicans. The data used in this study come from the Gallup organization which conducted a survey of 1,024 adults from May 12 to 15, 2011. I investigate the impact of political identification on consumer's willingness to purchase electric vehicles using Tobit and OLS models. I show that Democrats are more likely to consider electric vehicles than Republicans. I also discover that gender, age, and education level have the most significant influence on consumers' willingness to consider an electric vehicle. This is consistent finding of the previous study that highlights the significance of the value of sociodemographic characteristics on consumers’ energy-efficiency product consideration. th 56 Annual TRF Program, Atlanta, GA, Page 35 U.S FREIGHT TRANSPORTATION ECO-EFFICIENCY PERFORMANCE MEASURE AND BENCHMARKING STATE BY STATE Fesseha Gebremikael, UGPTI, Fesseha.Gebremikael@ndsu.edu Yong Shin Park, UGPTI, yong.park@ndsu.edu Gokhan Egilmez, North Dakota State University, gokhan.egilmez@ndsu.edu It appears that food manufacturing is one of the major driving forces of the global environmental issues that our planet is facing today. And there is a strong need to focus on sustainable manufacturing toward achieving long-term sustainability of the modal freight of the United States. First, this study will assess the direct and indirect environmental impact of 33 U.S. Food manufacturing sector’s modal freight activity, employing the Economic InputOutput Life Cycle Assessment (EIO-LCA). Secondly, data envelopment analysis will be used to determine the eco efficiency score of modal freight of each food manufacturing sector. Thirdly, a cluster analysis will be incorporated into the aforementioned tool for sustainability benchmarking purpose by grouping homogeneous inputs in order to set an improvement benchmarking target of eco-efficiency for each sector. FUEL FOR NEXT GENERATION REFUSE TRUCKS: ELECTRICITY OR NATURAL GAS? THE CASE OF METROPOLITAN ATLANTA AREA Dong-Yeon Lee, Georgia Institute of Technology, dlee348@gatech.edu Valerie M. Thomas, Georgia Institute of Technology, valerie.thomas@isye.gatech.edu Patrick S. McCarthy, Georgia Institute of Technology, mccarthy@gatech.edu Petroleum diesel has long been the dominant fuel for refuse trucks in the U.S. However, natural gas is gaining momentum as an alternative, mainly owing to the domestic shale gas boom in recent years as well as the heavyduty vehicle natural gas engine technology advances. In addition to the conventional natural gas resources including shale gas, renewable natural gas (or biogas) from landfills seems to be a low-hanging fruit for the refuse truck sector. In fact, the existing nine landfills in the metropolitan Atlanta area can produce sufficient amount of biogas to fuel all the refuse trucks in the area. Another interesting and promising addition to the refuse truck fuel technology portfolio is the electric refuse truck which has recently become commercially available. As these alternative fuel technologies compete to penetrate the market, challenging the incumbent diesel, comprehensive study of comparative private and social benefits of each fuel is to be undertaken to better inform fleet managers and policy makers. Here we analyze both private and social life-cycle costs associated with the fuel technology selection and the resulting climate change and human health impacts. For this, we develop a vehicle energy consumption and emissions prediction model for refuse trucks, based on vehicle dynamic simulation and emissions test data. Our model incorporates route characteristics, refuse collection and hauling operation, and Atlanta area’s local geographic conditions. We also build our own regional hour-by-hour electricity generation model to account for the energy use and air emissions from power plants. Since the environmental impacts from the power generation can vary significantly by time of day and year, the hourly resolution of our model provides more accurate estimation capability. We adopt a life-cycle assessment framework to evaluate overall energy efficiency and air emissions associated with the fuel supply, truck production, vehicle operation and maintenance, and end-of-life. Utilizing discrete choice method coupled with the life-cycle cost analysis, we make market share projections through 2040, based on the total cost of ownership analysis and monetized climate change and public health impacts assessment results. We find that the electric truck provides the best life-cycle energy efficiency; reduces climate change impact by 40 – 80% compared to the competitors; and causes only 10% of the environmental impacts of the others. However, the electric truck doesn’t currently have cost-competitiveness, although the electric’s market share is expected to grow up to 90% in 2040. In the short run, landfill gas seems to be the most cost-effective choice from both private and social costs standpoints. The reasons are twofold: first, landfill gas is even cheaper than the conventional natural gas that is cheaper than diesel fuel. Second, landfill gas can be burned in typical natural gas trucks that have less upfront capital cost penalty than the electric counterpart. The higher vehicle purchase cost than the conventional th 56 Annual TRF Program, Atlanta, GA, Page 36 diesel truck can rapidly be recouped by the cheaper fuel cost. Among the four fuel technologies compared, diesel shows the largest air emissions damage costs. More specifically, our model predicts that the diesel fuel’s overall market share in 2020 is to be reduced to 4% when considering social life-cycle costs, as opposed to the predicted 27% of market share based on private total cost of ownership. This implies that policy interventions might have to be considered to compensate the alternative technologies’ higher cost, so as to lower the climate change and public health impacts from the refuse truck sector. ASSESSING THE ECOSYSTEM GOODS AND SERVICE OF U.S MODAL FREIGHT: SUPPLY CHAIN LINKED CRADLE-TOGATE ECOLOGICAL BASED LIFE CYCLE MODEL Yong Shin Park, UGPTI, yong.park@ndsu.edu Gokhan Egilmez, NDSU, gokhan.egilmez@ndsu.edu Exergy based environmental assessment of transportation across supply chain of industrial sector is new area. None of the existing life cycle assessment models include ecological footprint which account for role of ecosystem goods and services. In this study, we develop a model which assess ecosystem and service of U.S. modal freight of industrial sector using Ecological based life cycle assessment. The impacts on the ecological system in terms of cumulative mass, energy, Industrial Cumulative Exergy Consumption, Ecological Cumulative Exergy Consumption of four major modal freights will be calculated. As for an ecological performance metric, efficiency, loading ratio and renewability index will be quantified in order to provide better understanding of ecological performance of modal freight of industrial sector. INVESTIGATING THE CHANGES OF MARGINAL PAVEMENT DAMAGE COST TO KEY VARIABLES USING SENSITIVITY ANALYSIS Tariq Usman Saeed, Purdue University, tusaeed@purdue.edu Anwaar Ahmed, School of Civil and Environmental Engineering, anwaar5148@gmail.com Samuel Labi, Purdue University, labi@purdue.edu Jackeline Murillo-Hoyos, Purdue University, jmurill@purdue.edu Kumares C. Sinha, Purdue University, sinha@purdue.edu For purposes of establishing road user charges for heavy vehicles, recent past studies have estimated the marginal cost of pavement damage (MPDC) by relating highway agency expenditure to the highway usage level (loading) as a function of certain key input factors. These factors include pavement life-cycle length, discount rate, rest period, effectiveness (service life) of rehabilitation treatments, and the costs of pavement reconstruction and rehabilitation treatments. What is lacking in the past studies is a detailed investigation of the sensitivity of the MPDC to changes in the key factors. Such sensitivity analysis is important to agency decision makers because in actuality, these factors are not fixed, but vary considerably spatially and temporally. Therefore, this paper analyzes the effect of changes in the key factors on the estimated MPDC. The results suggest that changes in the pavement life-cycle length, discount rate, rest period, and treatment effectiveness all have significant impact on the estimated MPDC; therefore, incorrect specification or invalid assumptions regarding these key factors can lead to mis-estimation of the MPDC. For reconstruction and rehabilitation costs, however, their impact on the estimated MPDC is relatively small, and therefore changes in the values of these two factors are not expected to impair significantly the MPDC estimates. th 56 Annual TRF Program, Atlanta, GA, Page 37 A THREE-STAGE LEAST SQUARES ANALYSIS OF POST-REHABILITATION PAVEMENT PERFORMANCE Md Tawfiq Sarwar, University at Buffalo, The State University of New York, mdtawfiq@buffalo.edu Panagiotis Ch. Anastasopoulos, University at Buffalo, The State University of New York, panastas@buffalo.edu Recent studies have provided improvements in the forecasting accuracy of pavement performance modeling, by statistically modeling pavement performance indicators as a system of seemingly unrelated regression equations (SURE). This approach accounts for cross-equation error correlation as a means to control for unobserved factors that lead pavements in poor condition to observe poor performance indicators. In the state of Indiana, the most common pavement performance indicators are the international roughness index (IRI), the rutting depth, and the pavement condition rating (PCR). Even though the first two can be accurately measured, the PCR is based on Engineers’ observations of the pavement surface. Therefore, it is possible that the PCR may be measured as a function of the observable IRI and rutting depth. This paper, explores this possibility by estimating a three-stage least squares (3SLS) model of IRI, rutting depth, and PCR, using data collected in Indiana. All three pavement performance indicators are found to be affected by traffic characteristics (truck volumes, traffic composition, etc.), surface characteristics (treatment and pavement type, drainage performance, etc.), and weather information (temperature and precipitation deviations). In addition, the PCR is found to also be significantly affected by the IRI and rutting depth measurements. The results of the 3SLS and SURE models are counter imposed, with the 3SLS model providing significant improvements in the forecasting accuracy of the pavement performance. ASSESSMENT OF SEISMIC DAMAGE OF A BRIDGE PIER USING A SEEMINGLY UNRELATED NONLINEAR EQUATIONS APPROACH Thomas Androutselis, University at Buffalo, tandrout@buffalo.edu Panagiotis Anastasopoulos, University at Buffalo, panastas@buffalo.edu Ioannis Anastasopoulos, University at Dundee, ianastasopoulos@dundee.ac.uk This paper explores an alternative method for the estimation of seismic damage of a bridge pier through the use of a system of unrelated nonlinear regression equations. Nonlinear dynamic time histories are first analyzed using multiple seismic records as seismic excitation. The results are then used to estimate statistical models in order to express seismic damage as a function of statistically significant intensity measures (IMs). As damage indices, the maximum drift ratio, the residual drift ratio, and the ratio of ductility demand over ductility capacity are utilized. The seemingly unrelated nonlinear regression model equations are evaluated in terms of various goodness-of-fit and forecasting accuracy measures, and with out-of- sample observations. The results show that the seemingly unrelated nonlinear regression equations improve the statistical fit and the forecasting accuracy when compared with independently estimated nonlinear regression, as they account for cross-equation error correlation caused by shared unobserved effects across the damage indices. EMERGING METHODS OF SIGNAL TIMING OPTIMIZATION David Hale, Leidos, Inc., david.k.hale@leidos.com Zong Tian, University of Nevada, Reno, zongt@unr.edu Several new and disruptive technologies are poised to change the way we time signals over the next 10 years. Improvements in computer processing speed may allow a transition to more advanced optimization methods, replacing the outdated algorithms in today's state-of-the-art commercial products. Some states are embracing highresolution signal performance analyzers; which identify when signals require re-timing, and in some cases perform the re-timing themselves. Multi-modal intelligent transportation systems are being developed to exploit connected vehicle technology, and respond better to non-automotive travel modes. Mobile device signal optimization apps can now verify and validate new timing plans, using GPS technology. This presentation will summarize advantages and disadvantages of the new emerging methods, and attempt to forecast the future of signal timing optimization. th 56 Annual TRF Program, Atlanta, GA, Page 38 PUBLIC-PUBLIC AND PUBLIC-PRIVATE PARTNERSHIPS IN TRANSPORTATION Janet Kay Tinoco, Embry Riddle Aeronautical University, tinocoj@erau.edu Students at Embry-Riddle Aeronautical University (ERAU), Daytona Beach, FL partook in a research effort for the National Aeronautics and Space Administration (NASA) Kennedy Space Center (KSC), examining partnerships in real property. Specifically, the students researched relevant public-private partnerships (PPPs) and public-public partnerships (PuPs) in a variety of industries, nationally and globally. While the effort for NASA included regulated industries outside of transportation, this poster will focus on those partnerships that were examined in the transportation sector, specifically space, air, rail, sea, and road. Information on past and current partnerships, the purpose behind their formation, structural arrangements, and reported advantages and disadvantages was gathered where possible. Initial investigation into risks associated with partnerships, particularly risks for the public sector was also accomplished. Student groups were tasked with collecting, compiling and comparing information about current and past PPPs and PuPs. The overall goal was to uncover partnerships of relevant groupings in regional areas of interest: North America, South America, Europe, and Asia/Australia. In general, information was gleaned from publications, government websites, company websites, news/media sources, conference proceedings, and other reputable sites available online, as well as, from paper publications in the library. As data collection effort ensued, it became increasingly clear that partnerships, be they PPPs or PuPs, are highly context specific. Furthermore, PPPs far outnumber PuPs in the areas researched, the latter being a natural outgrowth of the success of PPPs. Regardless, some generalities were gleaned from both of these types of partnerships, regardless of context. These generalities will be presented on the poster, along with the other key information such as advantages and disadvantages, legal arrangements, and risks involved. ENHANCED PREDICTION OF TRANSPORTATION PROJECT COSTS USING A RISK-BASED APPROACH Tariq Usman Saeed, Purdue University, tusaeed@purdue.edu Apichai Issariyanukula, Department of Highways, apichai.ben@gmail.com Anwaar Ahmed, School of Civil and Environmental Engineering, anwaar5148@gmail.com Samuel Labi, Purdue University, labi@purdue.edu Jackeline Murillo-Hoyos, Purdue University, jmurill@purdue.edu Kumares C. Sinha, Purdue University, sinha@purdue.edu Project cost estimation at the planning phase of transportation facilities is typically carried out using deterministic approaches. However, it has been found that deterministic approaches yield cost estimates that have significant error and uncertainty due to the variability associated with project environments and conditions. To address this issue, the present study presents a risk-based probabilistic analysis of highway project costs using Bayesian statistical techniques with limited data. Different stochastic cost models based on Bayesian statistics that are applicable at different levels of data availability were developed; these are able to provide the point estimate, range estimate, and full probability distribution of project costs. Using probability theory, the developed model can simultaneously amalgamate information from historical data and expert opinion into the analysis. The application of the new cost estimation approach is demonstrated using a case study involving pavement rehabilitation data from Indiana. th 56 Annual TRF Program, Atlanta, GA, Page 39 HOW MUCH DOES ABOVE MINIMUM DESIGN IMPROVE HIGHWAY SAFTEY? SeyedAta Nahidi, University at Buffalo, SUNY, seyedata@buffalo.edu Gary A Jordan, University at Buffalo, SUNY, garyjord@buffalo.edu Ugur Eker, University at Buffalo, SUNY, ugureker@buffalo.edu Panagiotis Ch. Anastasopoulos, University at Buffalo, SUNY, panastas@buffalo.edu Abstract: Public highways are designed, built, operated, and maintained in accordance with applicable governmental standards. These standards help ensure and promote safe highways for all motorists. This paper uses accident data collected from highways in Indiana, and estimates a random parameters tobit model to assess whether accident rates vary with respect to road sections in which design standards are either met or exceeded. The results indicate that the impact on highway safety is significant, only when the standards are significantly exceeded. As exceeding standards may entail additional monetary costs, these results have potentially significant policy implications with respect to highway expenditures in terms of benefits and costs. DISCOVERING CAUSALITY IN TRAFFIC SENSOR READINGS FOR ROAD ACCIDENTS IMPACT PREDICTION Liyue Fan, University of Southern California, liyuefan@usc.edu Ugur Demiryurek, University of Southern California, demiryur@usc.edu Cyrus Shahabi, University of Southern California, shahabi@usc.edu Road accidents, a key contributor to traffic congestion, post serious concerns to drivers, law enforcement, and transportation agencies. Reducing the impact of traffic accidents has been one of the primary objectives for transportation policy makers. The wealth of data collected from traffic sensors and accident logs offers an unprecedented opportunity to mine and understand the traffic incidents towards mitigating the consequences. In this poster, we will utilize the real-world datasets in our data warehouse and study to predict and quantify the impact (i.e., backlog and clearance-time) of road accidents on the upstream traffic direction and in the surrounding network (e.g., arterial streets) of the accident. Our data are collected from different transportation authorities in Southern California, and include archived traffic sensor readings and accident reports. We implement our previous work on predicting impact of road accidents in upstream traffic direction, which essentially classifies traffic accidents based on their features and models the impact of each accident class on its upstream traffic by analyzing the archived traffic data. However, in reality traffic accidents may cause surges in traffic demand in their vicinity, such as adjacent arterial streets and freeways, where our previous work cannot be directly applied. To this end, we will present our newly developed methods, which investigate the underlying temporal dependencies among time series of sensor readings at different road segments. As a result, we are able to quantify the temporalcausal effect of traffic speed between sensor locations and make effective predictions by identifying local dependency structures. To mitigate the noise and random fluctuations in sensor readings, we apply a variety of time series preprocessing techniques to discover the inherent dependency between traffic speed at different locations. To eliminate spurious causation induced by unobserved confounders, we adopt Granger graphical models for the sensor time series data collected from Southern California highways. With the discovered dependency network, machine learning techniques will be applied to predict the start time and speed change for impacted road segments at the onset of an accident. th 56 Annual TRF Program, Atlanta, GA, Page 40 DO THE SAME FACTORS AFFECT ACCIDENT FREQUENCIES ON HIGHWAY SEGMENTS WITH DIFFERENT TRAFFIC VOLUMES AND TRAFFIC COMPOSITIONS? Ugur Eker, University at Buffalo, SUNY, ugureker@buffalo.edu Nima Golshani, University at Buffalo, SUNY, nimagols@buffalo.edu Thomas Androutselis, University at Buffalo, SUNY, tandrout@buffalo.edu Gary A Jordan, University at Buffalo, SUNY, garyjord@buffalo.edu Panagiotis Ch. Anastasopoulos, University at Buffalo, SUNY, panastas@buffalo.edu Accident frequencies can be influenced by the level some motorists are alert, or by how some motorists perceive risk, under different traffic conditions. For example, under low traffic volume (overall, or truck) conditions, some motorists may be less alert while driving, which can possibly make them more accident prone. At the same time, some motorists may compensate for the same low traffic volume conditions, and drive faster, which can also make them more accident prone. And the contrary also stands; high traffic volume (overall, or truck) conditions can keep some motorists more alert, or can make them drive slower to compensate for the high-risk conditions, which in both cases make the motorists less accident prone. This paper, seeks to investigate the possibility that different factors can affect accident frequencies, when the latter are observed on highway segments under different traffic volumes and traffic compositions. Random parameters negative binomial models are estimated, and through the use of likelihood ratio tests, the results reveal that different sets of factors affect the accident frequencies for different traffic conditions. A SPATIO-TEMPORAL APPROACH FOR HIGH RESOLUTION TRAFFIC FLOW IMPUTATION Jianjiang Yang, University of Tennessee, jyang27@vols.utk.edu Lee D. Han, University of Tennessee, lhan@utk.edu Shih-miao Chin, Oak Ridge National Laboratory, chins@ornl.gov Ho-ling Hwang, Oak Ridge National Laboratory, hwanghl@ornl.gov Along with the rapid development of Intelligent Transportation Systems (ITS), traffic data collection technologies have been evolving dramatically. The emergence of innovative data collection technologies such as Remote Traffic Microwave Sensor (RTMS), Blue tooth sensor, GPS-based Floating Car method etc., creates the explosion of traffic data, which brings transportation engineering into the new era of big data. However, despite of the advance of technologies, the missing data issue is still inevitable and has posed great challenges for research such as traffic forecasting, incident detection, route guidance, and massive evacuation optimization, since all require complete and accurate traffic data. A thorough literature review shows most current imputation models, if not all, fail to consider the fact and the traffic stream characteristics at a certain location are usually closely related to those at neighboring locations and utilize these correlations for data imputation. To this end, this paper present a kriging based spatial-temporal data imputation approach that is able to fully utilize spatial-temporal information underlying in traffic data. Imputation performance of the proposed approach was tested on simulated scenarios and achieved stable imputation accuracy. Moreover, the proposed kriging imputation model is more flexible compared to current models. th 56 Annual TRF Program, Atlanta, GA, Page 41 A RANDOM PARAMETERS HAZARD-BASED DURATION ANALYSIS OF SENIOR TRAVELERS’ ACTIVITY-BASED TRAVEL TIME AND DISTANCE Gary A Jordan, University at Buffalo, SUNY, garyjord@buffalo.edu Thomas Androutselis, University at Buffalo, SUNY, tandrout@buffalo.edu Paria Negahdarikia, University at Buffalo, SUNY, parianeq@buffalo.edu Panagiotis Ch. Anastasopoulos, University at Buffalo, SUNY, panastas@buffalo.edu Srinivas Peeta, Purdue University, peeta@purdue.edu Sekhar Somenahalli, University of South Australia, sekhar.somenahalli@unisa.edu.au This paper identifies important factors that affect activity-based travel time and travel distance of senior travelers in Adelaide, Australia. To that end, and recognizing the longitudinal nature of the data, the time and the length of the distance from origin to destination of the senior travelers are modeled using hazard-based duration models. And to account for unobserved heterogeneity, random parameters are introduced in the models. Travel time and distance are found to be significantly affected by a number of socio-economic and demographic characteristics, trip characteristics, mode choice, trip frequency, time of day of the trip, and type of activity. The results are expected to identify problematic trip patterns, which if improved, have the potential to improve livability for senior Australians. ADDITIVE MANUFACTURING: THOUGHTS AND OPINIONS ON ITS IMPACT ON FUTURE SUPPLY CHAINS Robert Walton, Emory-Riddle Aeronautical University - Worldwide, walton@erau.edu Carl Scheraga, Fairfield University, cscheraga@fairfield.edu Recently there has been much speculation as to the future impact of Additive Manufacturing, also called 3D printing on supply chain management. Additive Manufacturing technology involves the making of parts and products using a computer- driven, additive process, to build a three dimensional object one layer at a time. Three dimensional printing builds plastic or metal parts directly from CAD drawings that have been cross sectioned into thousands of layers. Additive Manufacturing provides a faster and less costly alternative to machining (cutting, turning, grinding and drilling solid materials). This study utilized an anonymous survey on the perceived future impact of Additive Manufacturing on a preliminary sample that consisted of 42 supply chain professionals. All questions on the survey where paired and asked the respondents to give their opinion on a particular impact in the short term (within the next 20 years) and the long term (more than 20 years) of Additive Manufacturing on the supply chain. In all questions there was a statistically significant difference between their opinions on the impact of Additive Manufacturing in the short term and the long term. This would indicate that supply chain professionals do feel that Additive Manufacturing will impact the supply chain in the future. References Birtchnell, T., Urry, J., Cook, C., & Curry, A. (2012). Freight miles: the impacts of 3D printing on transport and society (ESRC EX/J007455/1). Retrieved November 12, 2014, from http://www.academia.edu/3628536 /Freight_Miles_of_3D_Printing_on_Transport_and_Society Clark, L., Calli, L., & Calli, F. (2014). 3D printing and co-creation of value. 12th International Conference e-Society, 251-254. Holland Herald. (2012 September). Another dimension. Holland Herald, 47(9), 42-44. Nyman, H. J., & Sarlin, P. (2013). From bits to atoms: 3D printing in the context of supply chain strategies. ResearchGate. doi:10.1109/HICSS.2014.518 Oko-Institut. (2013). 3D printing - Risk and opportunities. Retrieved November 12, 2014, from http://www.oeko.de /oekodoc/1888/2013-532-en.pdf Silva, J. V., & Rezenda, R. A. (2013). Additive manufacturing and its future impact in logistics. Management and th 56 Annual TRF Program, Atlanta, GA, Page 42 Control of Production and Logistics, 6(1), 277-282. Walton, R. O. (2014a). Waking up to the threat from a 3D revolution: A real but long-term threat. The International Air Cargo Association, 19. Walton, R. O. (2014b). The 6th mode of transportation. Journal of Transportation Management (In press). A REAL-TIME CONTROL BUS DISPATCHING POLICY TO MINIMIZE HEADWAY VARIANCE Simon Berrebi, Georgia Tech, simon@berrebi.net Kari Watkins, Georgia Tech, kari.watkins@ce.gatech.edu Jorge Laval, Georgia Tech, jorge.laval@ce.gatech.edu One of the greatest problems facing transit agencies that operate high-frequency routes is maintaining stable headways and avoiding bus bunching. In this work, a real-time holding mechanism is proposed to dispatch buses on a loop-shaped route using real-time information; Holds are applied at the terminal station to minimize the expected variance of bus headways at departure. The bus-dispatching problem was formulated as a stochastic decision process. The optimality equations were derived and the optimal holding policy was found by backward induction. A simulation assuming stochastic operating conditions and unstable headway dynamics was performed to assess the expected average waiting time of passengers at stations. The proposed control strategy is found to provide lower passenger waiting time on a wide range of operating conditions and better resiliency than methods recommended in the literature and used in practice. These results indicate that the proposed control method could be implemented on a real bus route to improve the capacity and the quality of service. SESSION 4.1: Transport Safety MODELING SAFETY PERFORMANCE FUNCTIONS FOR URBAN AND SUBURBAN ARTERIAL FOR THE STATE OF ALABAMA Jaehoon Kim, University of Alabama in Huntsville, jhkim5@gmail.com Michael D. Anderson, University of Alabama in Huntsville, andersmd@uah.edu Over the past two decades, considerable studies have been performed to develop statistical models for predicting crash frequencies on the roadway facilities. As one of the contributions of the studies, the Highway Safety Manual 1st edition (HSM) was released in 2010, which provides analytical tools and techniques to estimate expected crash frequencies of various types of roadway facilities. After release the HSM, a large number of the HSM calibration researches have been conducted to adopt on each jurisdiction. Nevertheless, there are still some difficulties to apply the HSM to all of the states because of lack of data. Unlike other SPFs of roadway facilities, the SPFs of urban and suburban arterial contain various formulas. The important functions of the formulas can be emphasized as: (1) the function of multiple-vehicle collision, (2) the function of single- vehicle collision, and (3) the function of driveway collision. For the safety analysis of the roadway facilities, the number of accident with roadway information is required. The required data is obtained from Critical Analysis Reporting Environment (CARE), which is the data analysis software package developed by the Center for Advanced Public Safety at the University of Alabama (CAPS). The CARE data consists of 0.01 mile roadway segment and each segment contains ninety-one variables including AADT, highway functional class, total crash frequencies, frequencies of each type of crash, and many other data of roadway environment. However, CARE data consists of multiple-vehicle collision and single-vehicle collision data, except drive-way collision. Also, the roadway inventory does not include the number of driveway on each segment. Therefore, it is not possible to directly apply HSM SPFs to the state of Alabama. The primary objective of this study is to develop new Alabama-specific SPFs for urban and suburban arterials facilities. We develop two types of Alabama-specific SPFs: (1) the SPF for multiple-vehicle collision, and (2) the SPF th 56 Annual TRF Program, Atlanta, GA, Page 43 for single-vehicle collision. To accomplish the objective, there needs to be twofold approach. Firstly, it is necessary to determine over- or under-dispersion parameter. Generally, Poisson regression is used to model countable data. However, the crash frequency data usually exhibits over-dispersed distribution, and it violates the property of the Poisson distribution of which the mean and variance is equal. Secondly, we choose proper regression models to use in this study based on the first step, and develop Alabama-specific SPFs. The Poisson regression model is the most popular model to estimate count data. However, if the data violates the fundamental assumption of the Poisson distribution, it may lead the Poisson model to result in a biased estimation. The negative binomial (NB) regression is usually used to estimate the crash frequency when the crash data is overdispersed. It is most frequently used model to estimate crash frequencies. Based on the preliminary results of this study, we find the multiple-vehicle crash frequency is overdispersed, so that NB regression model is applied to the development of SPF for multiple-vehicle collision. However, NB regression cannot be applied to the singlevehicle collision SPF because the frequency data contains huge zero values. Therefore, we apply zero-inflated regression models to single-vehicle collision SPF. The reported single-vehicle collisions on CARE data occurred extremely small, but the accident possibly could occur. In the preliminary study, the zero-inflated negative binomial regression model was not applicable to the SPF. This study reexamine the result and apply zero-inflate Poisson regression model to develop the single-vehicle collision SPF. DETERMINANTS OF CREW INJURIES IN CONTAINER VESSEL ACCIDENTS Yishu Zheng, Virginia Port Authority, yzheng@portofvirginia.com Wayne Talley, Old Dominion University, wktalley@odu.edu Di Jin, Woods Hole Oceanographic Institution, djin@whoi.edu ManWo Ng, Old Dominion University, mng@odu.edu This paper investigates determinants of non-fatal and fatal crew injuries in container vessel accidents. The determinants are deduced from an estimated probit regression equation based upon 2001-2008 U.S. Coast Guard container vessel accident data. The estimation results suggest that a crew member is: (1) less likely to have a nonfatal injury in containership and ro-ro container vessel accidents if the vessel has a steel hull and the vessel accident occurs at night time; (2) more likely to have a fatal injury if fire is involved in the containership and ro-ro container vessel accidents; and (3) less likely to have a fatal injury in containership and ro-ro container vessel accidents if the vessel is U.S. flagged, steel hulled and powered by a diesel engine. These results are critical in developing new policies in reducing non-fatal crew injuries in container vessel accidents for ocean carriers, container vessel owners, container vessel registries and the IMO. ANALYSIS OF FATAL TRAIN-PEDESTRIAN COLLISIONS IN METROPOLITAN CHICAGO 2004-2012 Ian Savage, Northwestern University, ipsavage@northwestern.edu This paper analyses the 338 pedestrian fatalities on railroads in the Chicago metropolitan area between 2004 and 2012. Almost half (47%) were apparent suicides, 21% were non-suicidal fatalities at stations and crossings, and the remaining 32% were non-suicidal incidents at other places along the right of way. A spatial analysis shows that while there is a general randomness in incident location, there are some common patterns, and also some notable outliers. The frequency of fatalities at stations and crossings and from trespassing in different municipalities is strongly related to the density of public access points to the right of way. But fatalities of these types do not increase with train volume suggesting that pedestrians may exercise more care around busier lines. The distribution of apparent suicides is less strongly related to the density of public access points suggesting that those intending self-harm will seek out a point of access. Apparent suicides are also more prevalent where there is th 56 Annual TRF Program, Atlanta, GA, Page 44 a higher train frequency and a greater proportion of passenger trains that run to a published schedule. They are also more prevalent in municipalities with higher incomes and lower population density. While most of the apparent suicides (70%) are not associated with any copycat activities, the dataset contain clusters of suicides that are proximate in both time and space. There was also a highly-publicized suicide that led to a 95% increase in apparent suicides throughout the region in the 18 weeks following the incident. SESSION 4.2: Transportation Forecasting COMPARISON BETWEEN TRIP-BASED AND TOUR-BASED TRUCK TRAVEL DEMAND FORECASTING MODELS IN BIRMINGHAM, AL Andrew Sullivan, UAB, asullivan@uab.edu William G. Allen, WG Allen, wgallen@isp.com David Lee, Georgia Tech, david.lee@coa.gatech.edu Ehsan Doustmohammadi, UAB, ehsan12@uab.edu Ozge Cavusoglu, UAB, ozge@uab.edu Virginia Sisiopiku, UAB, vsisiopi@uab.edu Traditional travel demand forecasting models do not model truck trips separately, but rather include them implicitly in the non-home-based (NHB) trip category. Little attention is, thus, paid to truck types, trip patterns, trip lengths, or comparisons with actual truck counts. However, in recent years, with a growing realization of the importance of truck traffic both to the overall economy and on urban congestion and pollution levels, there is a new interest in modeling truck movements with greater accuracy and detail. This study explored the possibility of utilizing a tour-based freight demand model as an alternative to the traditional 4-step travel demand forecasting process for estimating truck trips in the Birmingham, AL region. The tour-based model uses a disaggregate approach to capture the intermediate stops of each truck and reflect the implications of those stops on vehicle miles traveled (VMT). In this study, four months of GPS-based truck movement data from the Birmingham, AL region were used to obtain origin, destination, and stop locations information for trucks moving within the Birmingham region. Such information was then utilized to model truck movements within the study region as individual truck tours. The traditional trip-based planning model used by the Regional Planning Commission of Greater Birmingham and the new tour-based truck model of the Birmingham region were run and trip estimates were obtained for the year 2035. The results from two models were compared across a variety of outputs such as vehicle miles traveled (VMT) by mode and time period, truck volumes generated by each modeling approach and truck VMT by area type. This paper introduces the tour based modeling concept and outlines the methodology for estimating truck tours. Then it summarizes the details of the proof-of-concept case study for the Birmingham, AL region. These include data processing, tour-based model development process, study assumptions, and validation efforts. Finally, results from the comparison between the trip-based and the tour-based truck planning models are presented and discussed. The paper summarizes lessons learned from the Birmingham case study that can have direct application to travel demand forecasting in Birmingham and can guide development and implementation of improved truck travel demand forecasting models at other locations in the future. OPTIMAL REFUELING POLICIES MODELS: APPLICABILITY IN BRAZILIAN ROUTES Luciano Marques Reduzino, Universidade Federal do Espírito Santo, luciano.reduzino@gmail.com Marta Monteiro da Costa Cruz, Universidade Federal do Espírito Santo, mcruz@npd.ufes.br Luiza Rosa Fernandes, Universidade Federal do Espírito Santo - Brazil, fernandes.luizarosa@gmail.com Fuel prices have increased dramatically by more than 180% since 2001 to the present day in Brazil. The raising fuel cost has impacted substantially the cost of motor carriers and, therefore, the efficient management of refueling is a th 56 Annual TRF Program, Atlanta, GA, Page 45 critical issue for companies in the sector. However, many motor carriers are facing difficulties to find the best refueling policy along their route due to the great variability of the diesel prices in the refueling points (truck stops). To address this problem, many models have been developed to determinate the optimal refueling policy. These are decision models that indicates which truck stop(s) to use and how much fuel to buy at the chosen truck stop(s) to minimize the cost of refueling for a given origin-destination (o and d) pair. Some variants of the models work in conjunction with truck-routing models, so that both the route and the refueling policy jointly minimize the fuel cost of operating the vehicle. The use of the of refueling policy decision models may reduce up to 12.98% of fuel cost for motor carriers. This paper examines the applicability of these models in Brazilian routes. In the first phase, a review of the literature is presented and the existing studies were classified by the method applied in the modeling process, the validation process and its contributions. Further on, in the second part, the variability of prices in important routes in Brazil was analyzed and a selected model was applied considering this variability and the local specificities. A comparative study was made and the benefits of the selected model were shown by simulation. The results suggest that the decision support model is applicable to achieve the goal of reduce the fuel cost in the Brazilian motor carriers. PEDESTRIAN TRAVEL DEMAND ESTIMATION USING GEOGRAPHICALLY WEIGHTED REGRESSION MODEL Xu Zhang, University of Kentucky, xuzhang_uk@uky.edu Mei Chen, University of Kentucky, mei.chen@uky.edu Pedestrian demand modeling recently has been increasingly studied by researchers. As one of several legitimate modeling approaches, linear regression model has been proved to be able to effectively explain the relationship between the demand and the explanatory factors. However, such model is built based on the assumption that each independent variable is considered to have a homogeneous impact on pedestrian demand across the study area. Therefore, the linear model only explores the average relationship between the demand and the contributing factors. However, studies have shown that the spatial data usually exhibits heterogeneous and non-stationary properties. Assigning the same coefficient to variables across the whole study region in the linear model may inevitably ignore the existing spatial difference and at some degree degenerate the prediction power of the model. In order to account for such spatial variation, a novel methodology, which is called geographically weighted regression (GWR), are introduced and analyzed in the paper. The new model has been known for its capability of taking the spatial heterogeneity in relationship between dependent variable and independent variables into consideration. Several indicators are also proposed to investigate and compare the performance of both linear model and GWR model. Meanwhile, the coefficients of the selected variables generated by the geographically weighted regression model at the associated census tracts are analyzed. It is observed from the study that GWR can account for the spatial variation of the data and improve the explanatory performance of the demand model. ACCOUNTING FOR NATURAL GAS VEHICLES IN REGIONAL AUTO MARKETS – ESTIMATES FOR THE STATE OF TEXAS, U.S. Chen (Sarah) Xu, University of Houston, sarahxc.hku@gmail.com Liang-Chieh (Victor) Cheng, University of Houston, lcheng6@uh.edu Natural gas vehicle (NGV) technologies have become a sustaining force in the U.S. alternative vehicle markets. In the past years, the counts of NGVs have increased steadily, starting from 23,281 in 1992 to 119,217 in 2010. According to the National Petroleum Council, the light duty NGVs can reach penetration of 50% by 2035. For heavy duty NGVs, the market penetration can potentially become 70%. Other studies also report similar trends for NGVs in the US auto markets. Overall, professional experts are in sync in predicting a promising growth of NGVs in the US transportation sectors. th 56 Annual TRF Program, Atlanta, GA, Page 46 The steady price spread between conventional fuels and natural gas fuels is the economical drivers for NGV adoption in US. For decades, prices of natural gas as a transportation fuels have been only half or even a third of conventional fuels, namely gasoline and diesel. Even though higher market penetration of NGVs may drive the NG fuels prices, the abundant supplies from U.S. domestic shale natural gas production ascertain match for domestic needs for natural gas fuels. In the meantime, the growing prices for conventional fuels are indicative of continuing price spread between conventional fuels and natural gas fuels. The potentials of cost savings of natural gas vis-a-vis conventional fuels are a strong incentive for general public and fleets to adopt NGVs. Environmentally, NGV is a cleaner option for less air pollution and greenhouse gas emission than conventional fuels. Natural gas is a clean burning fuel and will produce much less CO2 as opposed to gasoline and diesel. In addition, natural gas also emits lower levels of NOx and sulfur, the main components of greenhouse gas emission. In a highly populated area, such as metropolitan areas, higher adoption of NGVs can lead to significant improvement in air quality and reduction of air pollution, let alone the resulting reduction of pollution-related diseases and social costs. Growth of NGV also helps U.S. energy sector to reduce the dependence of petroleum-based fuels. Transportation fuels consist of more than half of energy use in the U.S. Using U.S. domestically produced fuels enhances the U.S. economy's independence from major oil and gas producing countries, such as Middle East countries. It also helps avoids the impacts of high energy consumption by large oil and gas consuming countries, such as China and India. Adopting NGVs also can diversify the uses of technologies to power vehicles, such as natural gas, propane, electricity, as well as conventional fuels. There has been a body of qualitative studies that predict optimistic landscape of NGV adoption for the overall U.S. auto market. However, the trajectory of the NGV diffusion has not been quantitatively examined. Little is known regarding the prediction of NGV growth over the time horizon. Even US states leading NGV adoptions also display different paces in term of specific annual market growth. As such, the behaviors of NGV technology diffusion for U.S. overall and state markets need to be quantitatively examined. In addition, the diffusion of NGV technologies is strongly conditioned by the natural gas price and the coverage of natural gas fueling infrastructure. Accordingly, a realistic NGV forecast model requires simultaneous assessment of prices and infrastructure. Interestingly, few studies in the extant literature of alternative vehicle technologies directly probe the price and infrastructure effects in their models. This article intends to develop models to forecast NGV penetration in the U.S. and Texas automobile markets. We ask the following interrelated research questions: 1) How to develop a quantitative models to forecast NGV penetration in the U.S. markets? 2) How to predict year by year NGV diffusion in national and state auto markets in U.S.? 3) How to assess price and infrastructure impacts on NGV diffusion? To address these questions, analytical NGV diffusion models will be developed and specified by using national and state NGV-related data. SESSION 4.3: Topics in Transportation Supply Chains THE EFFECT OF INFORMATION ASYMMETRY OF FREIGHT NETWORK INFORMATION BETWEEN SHIPPERS AND CARRIERS ON THE CARRIER SELECTION IN A PROCUREMENT EXERCISE Manasi Katragadda, College of Business, Iowa State University, manasik@iastate.edu Yoshinori Suzuki, College of Business, Iowa State University, ysuzuki@iastate.edu This paper proposes a theoretical framework of how the information asymmetry of freight network information between shippers (or 3PLs) and carriers in the truckload market can be measured and modeled in the context of carrier choice models within a procurement exercise. Trucking procurement exercises are a bidding type environment where many carriers compete to be selected by the shipper to transport their freight across multiple contract lanes. Recent studies have shown that the network information of both shippers and carriers is important for a successful procurement event because the compatibility of these networks determines the rates, but carrier networks are dynamic and change over time. Studies have also shown that if the networks of carriers are not compatible with those of shippers, the shippers will experience a route guide bleed effect, where carriers reject the th 56 Annual TRF Program, Atlanta, GA, Page 47 contracted loads with shippers and the shippers have to pay higher rates to secure transportation services by using other carriers. However, there have been no studies that have measured the information asymmetry of this network information between shippers (or 3PLs) and carriers. This study will address this research gap. It is important to measure the information asymmetry in this context because shippers have limited knowledge about the carrier networks and it is the carriers who specify transportation costs in the procurement exercise based on their network specifications. Studying the information asymmetry of network information of shippers (or 3PLs) and carriers contributes to the knowledge of the carrier selection process because with more accurate information of carrier networks, shippers may be able to make better carrier selections that will lower the freight transportation costs. We propose that econometric models of carrier choice can be used to study the effect of information asymmetry of the network information between shippers and carriers on shippers’ carrier selection decisions. EVALUATING THE SUPPLY CHAINS OF CELLULOSIC TRANSPORTATION FUEL IN TENNESSEE T. Edward Yu, University of Tennessee, tyu1@utk.edu Burton C. English, University of Tennessee, benglish@utk.edu Lixia He, University of Tennessee, llambert3@utk.edu James A. Larson, University of Tennessee, jlarson2@utk.edu James Calcagno, University of Tennessee, jacal@utk.edu Joshua S. Fu, University of Tennessee, jsfu@utk.edu Brad Wilson, University of Tennessee, wilson.bradly@gmail.com Perennial switchgrass is native to the United States and has been considered as a feedstock for transportation fuel production. This study evaluated the net present value (NPV) of cash flows and greenhouse gas (GHG) emissions of a switchgrass-based biofuel supply chains (BSC) in Tennessee to replace 20% of petroleum use in transportation sector. Two scenarios were analyzed for the switchgrass-based BSC: (1) maximization of NPV over 20 years, and (2) minimization of GHG emissions. A mixed-integer mathematical programming model incorporating high resolution spatial data was applied to determine the optimal location of conversion facilities, feedstock production area, and feedstock and biofuel transportation under each scenario to calculate the NPV and GHG emissions. Results show that the NPV of profit over 20 years of the BSC system was 36% higher in the first scenario when compared to scenario (2). However, the GHG emissions produced in the scenario of NPV maximization was nearly 228 million kg CO2e more per year than GHG minimization scenario. The optimal location of the conversion facilities and feedstock production area were strongly related to current agricultural land coverage. Accordingly, feedstock and biofuel transportation was also affected by the selection of land for feedstock production. The transportation component, which included both feedstock and biofuel transportation, accounted for about 7%-8% of total cost and 3%-4% of the total GHG emissions in BSC under both scenarios. However, the biofuel transportation cost in the NPV maximization scenario was 35% less than that in the GHG minimization scenario. FEEDSTOCK IMPORT OPTIONS REGARDING TRANSPORTATION SYSTEMS FOR ASPHALT PRODUCTION Raj Bridgelall, North Dakota State University, raj.bridgelall@ndsu.edu EunSu Lee, North Dakota State University, eunsu.lee@ndsu.edu This paper examines the logistics and distribution channels for importing heavy crude oil, which is a source of bitumen. This study conducts scenario analysis regarding the cost of available transportation mode and carrier options, and examines import taxes and regulations that will affect the transport costs, schedule, and risks. The direct pipeline option via the existing TransCanada XL facilities represents the smallest percentage transportation cost of all other options analyzed. However, the lack of pipeline capacity on the existing XL pipeline and delays in building additional capacity weighs heavily against this option. th 56 Annual TRF Program, Atlanta, GA, Page 48 ECONOMIC IMPACTS OF FREIGHT DISRUPTIONS DUE TO FLOOD RELATED HIGHWAY CLOSURES USING AN INTEGRATED TRAFFIC NETWORK EQUILIBRIUM AND REGIONAL ECONOMIC IMPACT METHODOLOGY Amlan Mitra, Purdue University Calumet, mitra@purduecal.edu Rodrigo Mesa-Arango, Purdue University, rmesaa@gmail.com Xianyuan Zhan, Purdue University, zhanxianyuan@gmail.com Satish Ukkusuri, Purdue University Calumet, sukkusur@purdue.edu The purpose of this paper is to provide a theoretical framework and a methodology to estimate and analyze the economic impacts of disruptions of movements of commodities due to 2008 flood related highway closures in Northwest Indiana. Using data from an inter-regional commodity flow model and the corresponding transportation network flow model, mean travel time of truck shipments, transportation costs, and the dollar value of 43 2-digit Standard Classification of Transported Goods classes were estimated both before and after the flood. Then, a regional input-output model was applied to measure the regional input–output impacts from the change in the values of these commodities among the various industries in the region. Finally, the model was applied to estimate the direct, indirect, and induced economic impacts on the region. This paper is from a research project funded by NEXTRANS, Region V, Regional University Transportation Center, U.S. Department of Transportation. SESSION 4.4: Systems Approaches to Transportation IS HANDS-FREE TEXTING A BETTER ALTERNATIVE OF TEXT DRIVING? Sanaz Motamedi, University of Rhode Island, sanaz_motamedi@my.uri.edu Jay Hone Wang, University of Rhode Island, jhwang@egr.uri.edu In an increasingly mobile era, the wide availability of technology for texting and the prevalence of hands-free form have introduced a new safety concern for modern drivers. To assess this concern, this study investigated drivers’ opinions regarding hands-free and hands-on text driving and their driving performance when they were texting while driving on a fix-based simulator. A questionnaire was first deployed online to gain an understanding of the drivers’ text driving experiences as well as their demographic information. 175 people partook the questionnaire. The results revealed that younger and less experienced drivers are most likely to text while driving and least likely to understand the risks associated with text driving. Through the use of a fix-based driving simulator, this study examined drivers’ performance while they were engaged in some forms of text driving under different challenging traffic conditions. Through a blocked factorial experiment, drivers would either read a texted message or respond to it with two levels of context complexity and using either hands-on or hands-free texting method. Twenty eight drivers with balanced gender and age groups were recruited to participate in the study. Each participant was tested under all combinations of the three factors. Subjects used their own personal smartphones for hands-on section and a computerized voice enabled audio system for the hands-free section. Their performance was assessed based on the number of driving violations observed while driving in each scenario and the assessment was used as the response in the study. Conclusions regarding the impacts of different forms of texting, text complexity, and response mode on drivers driving performance were drawn. THE IMPACTS OF DISTRACTED DRIVERS ON BICYCLISTS' FATALITIES Sia Macmillan Lyimo, South Carolina State University, macmalyimo@gmail.com In the transportation sector, the common mode of transportation is automobile that requires maximum attention from the driver while driving. Any action or situation that diverge the attention of the driver from the primary task (driving) is termed as distraction and therefore distracted driving. Distracted driving can be as a result of three form th 56 Annual TRF Program, Atlanta, GA, Page 49 of distractions; visual (taking driver’s eye off the road), manual (taking drivers’ hand off the wheel), and cognitive (the focus is not into driving). In recent years, the most common source of distraction has been the use of electronic devices while driving including cell phones, tablets, and PDAs. This study investigates the impact of technological distractions while driving. These distractions include cell phones, computers, fax machine, printers, navigation systems, head-up displays, and two way radios. Specifically, this study wishes to investigate the impact of technological distractions on bicyclists’ safety. Furthermore, this study will investigate the effect of roadway characteristics, crash characteristics, driver characteristics and demographic features on the severity of distracted driving. This paper will quantitatively describe the impacts of distracted driving on bicyclists’ fatalities using data from Fatality Analysis Reporting System (FARS) to analyze the impacts of eight driver related factors (distractions) relatively to the four categories of analysis. Policy makers can use the findings from this paper in the prioritization of different attempts for crash reduction by setting out policies that will directly affect the most critical category that is subjected to distractions. ANALYZING WHAT WORKS BEST FOR TRANSPORTATION EDUCATION Sampson Gholston, The University of Alabama in Huntsville, gholsts@uah.edu MD Sader, University of Southern Mississippi, md.sader@usm.edu Joan Chadde, Michigan Technological Institute, jchadde@mtu.edu Richard Stewart, University of Wisconsin, Superior, rstewart@uwsuper.edu Amit Mokashi, University of Wisconsin, Superior, amokaski@uwsuper.edu Transportation has become one of the last frontiers that still remain to be conquered by most businesses in the twenty first century. Yet this cannot be done unless all and transportation professionals, irrespective of their functional orientation and current job responsibilities, fundamentally understand the dynamics of how products move from one place to another. This is one of the disciplines that is growing at a faster pace. The issue is that the number of graduates in this field is not meeting the current industry demand. Many U.S. institutions have recently developed and planning to develop educational degree programs in this area. This research analyzed the need for best practices and identified best practices in transportation education. Best practices are an inherent part of education that exemplifies the connection and relevance identified in educational research. They interject rigor into the curriculum by developing thinking and problem-solving skills through integration and active learning. Best practices are applicable to all degree levels and provide the building blocks for instruction. This paper will share those effective best practices in transportation education. HSR AS TRANSIT: THE CONTINUING TRANSPORTATION-DRIVEN EVOLUTION OF METROPOLITAN FORM Ryan Westrom, MIT, westrom@alum.mit.edu Joseph Sussman, MIT, sussman@mit.edu With high-speed rail (HSR) now often fulfilling a commuting function within an hour’s travel time from principal metropolitan cities, it becomes the latest in a long line of transportation technologies to elicit change in the metropolitan form of these cities. This paper explores this history, and then the potential for this shifting form in the era of HSR. Via a closer look at four case cities home to potential future HSR systems—Coimbra and Leiria in Portugal and Champaign-Urbana and Kankakee in Illinois within the U.S.A.—that will each move within a principal city’s commuting reach—Lisbon and Chicago, respectively—implications for transportation and land use planning are discussed. The unique discontinuous nature of these new potential metropolitan forms presents fresh opportunities to implement planning best practices, providing increased mobility with sustainability and quality of life returns. These speak well to the potential for HSR to serve this new function, and provide support for the consideration of HSR as a new advanced transportation alternative—and innovative form of transit—for these settings. th 56 Annual TRF Program, Atlanta, GA, Page 50 SESSION 5.1: Highway User Charges IMPACTS OF VIOLATIONS OF HIGH-OCCUPANCY-VEHICLE LANES: A SIMULATION BASED STUDY IN TENNESSEE Jianjiang Yang, University of Tennessee, jyang27@vols.utk.edu Lee Han, University of Tennessee, lhan@utk.edu High occupancy vehicle (HOV) lanes are designed to promote ride-sharing to address increasingly growth of traffic congestion on urban interstates. The basic concept for HOV lanes is to move more people rather than more vehicles, which requires at least two occupants for vehicles traveling an HOV lane. Due to the funding constraints, the most commonly used separation between HOV lane and General Purpose (GP) lanes is wider and broken line, a “mental” barrier instead of physical ones. Such HOV lane separation makes it easy to break the rules and hard to enforce. To curb the abuse of HOV lanes, the conversion of HOV lanes to high-occupancy toll (HOT) lanes are gaining popularity around the country, including Tennessee. The effect of violations rates on the facilities’ operation performance needs to be investigated to facilitate such conversion. However, to the best of authors’ knowledge, such studies are still very limited in previous literature. The relationship between violation rates and facility performance needs to be quantified systematically. To address these issues, a simulation study was conducted along a stretch of I-24 in Nashville, Tennessee, which covered 25.07 miles HOV lanes. Through this study, the travel time and other traffic performance metrics were compared under different violation rate scenarios. The effective range of violation rates, which allow HOV lanes carry additional traffic without degrading performance significantly, were provided as guidance for better HOT pricing scheme. Besides, the abuse of HOV lanes posed a great threat for road users’ safety, which also the fares and pricing solutions for the introduced HOT lanes. BENEFIT-COST ANALYSIS OF BOTTLENECK MITIGATION STRATEGIES David Hale, Leidos, Inc., david.k.hale@leidos.com Jiaqi Ma, Leidos, Inc., jiaqi.ma@leidos.com Michalis Xyntarakis, Cambridge Systematics, mxyntarakis@camsys.com Recent research from the Federal Highway Administration involved micro-simulation analysis, for five specific bottleneck mitigation strategies. These strategies included dynamic lane grouping at signalized intersections, dynamic merge control at freeway on-ramps, increasing freeway acceleration lanes from 500 to 1500 feet, hard shoulder running between freeway interchanges, and decreasing freeway lane widths so a new lane could be added. The micro-simulation analyses revealed mobility benefits, but a comprehensive benefit-cost analysis would require assessing safety and environmental impacts. This paper describes the mobility, safety, and environmental analyses that were necessary to compute benefit-cost ratios, and reports on those findings. CHARACTERISTICS OF CAR-LESS HOUSEHOLDS IN CALIFORNIA: EVIDENCE FROM THE 2012 CALIFORNIA HOUSEHOLD TRAVEL SURVEY Suman Kumar Mitra, University of California, Irvine, skmitra@uci.edu Jean-Daniel Saphores, University of California, Irvine, saphores@uci.edu In the United States there are now more motor vehicles than drivers and 23% of households have more cars than adults. However, approximately 10.5 million US households, or 9%, do not own cars. Unfortunately, our knowledge of car-less households is lacking, as is our research on their predicaments. The objective of this paper is to understand the characteristics and the travel patterns of car-less households in California based on the 2012 California Household Travel Survey (CHTS). These households, which seem often forgotten in transportation policy discussions, can be organized in two groups: involuntary and voluntary car-less households. This paper examines the characteristics of voluntary and involuntary car-less households and how their travel characteristics differ. Discrete choice models are used to explain the characteristics of voluntary and involuntary car-less households. th 56 Annual TRF Program, Atlanta, GA, Page 51 Results from this study, will help inform policies that encourage low or zero car ownership while avoiding the potentially negative consequences of not owing a car in an automobile-oriented society. SESSION 5.2: Transport Capital, Funding, and Economic Growth PORT PLANNING BENCHMARKING - BEST PRACTICES IN FUNDING PROGRAMS Libby Ogard, Prime Focus LLC, logard@new.rr.com Richard Stewart, University of Wisconsin Superior, rstewart@uwsuper.edu Ports are a vital part of our global transportation network. Port planning must consider coastal management areas, army corps of engineer’s jurisdictional boundaries and Department of Natural Resource requirements. Commercial port planning varies by state and Canadian province. In Pennsylvania Ports were recently moved from the state department of economic development to the Department of Transportation. In Florida, a grass roots approach to port development helps streamline the port planning and funding processes. In Wisconsin, all the Regional Plan Commissions and MPO’s were surveyed to identify their port planning information needs and understanding of how landside transportation infrastructure must support our marine gateways. This paper will compare Texas, Florida, Pennsylvania and Wisconsin port planning efforts and will highlight state port infrastructure funding programs. CAPITAL INVESTMENT NEEDS OF THE SMALL RAILROADS Anne Campbell, North Dakota State University, eileen.campbell@ndsu.edu Rodney Traub, North Dakota State University, rodney.traub@ndsu.edu This paper examines the capital investment needs of the Class II and Class III railroads in the United States. An electronic survey was administered to 470 small railroads and yielded a 31% response rate. The American Association of Railroads facts of 2013 state there are 560 small railroads operating in the United States and Canada. These railroads have investment needs, for instance track and bridge maintenance, reconstruction, purchase/lease of rail cars and locomotives. Their financial needs are important aspect of the future of small railroads in US, as well as the services they provide for the Class I railroads. This survey asked questions about the past, present and near future capital investment needs of the Class II and Class III railroads. DYNAMIC RELATIONSHIPS AMONG TRANSPORT INFRASTRUCTURE INVESTMENT, NON-TRANSPORT PUBLIC INFRASTRUCTURE, AND ECONOMIC GROWTH IN THE UNITED STATES: A NEW LOOK Junwook Chi, University of Hawaii, Manoa, jwchi@hawaii.edu Jungho Baek, University of Alaska Fairbanks, jbaek3@alaska.edu In this paper, we examine the dynamics of transport infrastructure, non-transport public infrastructure, private capital, labor hours, GDP, and exports in the United States. Using annual data from 1960 to 2012, we employ a modern ARDL approach to identify cointergration vectors and explore the direction of causation among the variables. The central focus of this paper is on the short- and long-run impacts of public investment in transport infrastructure on economic growth that have drawn mixed conclusions in existing literature. We find empirical evidence that a bidirectional relationship exists between transport infrastructure and GDP in the long-run, indicating that expanding transport infrastructure improves aggregated economic output and enhanced economic output increases public investment in transport infrastructure. However, the magnitude of impact of transport infrastructure on GDP is smaller than that of non-transport public infrastructure, implying that expanding transport infrastructure can be a less effective long-term fiscal stimulus, compared to expanding non-transport infrastructure. In the short-run, public transport infrastructure is found to have an insignificant effect on economic output. th 56 Annual TRF Program, Atlanta, GA, Page 52 SESSION 5.3: Operating and Management Performance in the Airline Industry AN ANALYSIS OF THE EFFECTS OF OPERATING AND FINANCIAL LEVERAGE ON THE MAJOR INTERNATIONAL CARRIERS: 1990-2013 Carl Scheraga, Fairfield University, cscheraga@fairfield.edu Richard Gritta, University of Portland, gritta@up.edu Introduction: The global airline industry has always been highly cyclical and somewhat fixed cost driven. The carriers are thus high in what financial analysts refer to as operating leverage. In addition, the majority of the airlines have followed aggressive debt strategies; that is, they have chosen to use large amounts of long-term debt finance to purchase assets. This results in a high degree of financial leverage. In the past, the resulting combined leverage has created severe financial problems for the industry. The purpose of this paper is to examining the effects of this leverage during the years in which the carriers saw unprecedented growth and a return to profitability. The sample will consist of the major international airlines. Methodology: Leverage will be defined using elasticity measures borrowed from economic theory. The degree of operating leverage (DOL) will be defined as the percent change in operating profits given a percent change in operating revenues, while the degree of financial leverage (DFL) will be expressed as the percent change in profits after taxes given a percent change in operating profits. The degree of combined (DCL) is the percent change in net profits given a percent change in operating revenues. These measures will be computer for the years 1990-2013. In addition, the volatility in both the carriers’ pre-tax return on assets (ROA) and the return on stockholder’s equity (ROE) will be measured in order to document the effects of the leverage inherent in this industry. Implications for public policy and for the survival of some industry members will be discussed. Selected Bibliography: Moyer, McGuigan and Kretlow, Contemorary Financial Management, 12 edition, 2012, Ch. 14. “Operating and Financial Leverage on the Major U.S. Air Carriers: 1990-2004,” Journal of the Transportation Research Forum, XLV No.2 (Summer 2006). (Richard Gritta, Bahram Adrangi, and Brian Adams). “Business, Financial and Total Risk in Air Transport: A Comparison to Other Industry Groups Prior to September 11, 2001,” Journal of the Transportation Research Forum, LVII No.4 (Fall 2003), 149-156. (Richard Gritta, Garland Chow, and Ned Freed). “The Effects of Operating and Financial Leverage on the Stability of Airline Returns Over Time: The Contrast Between Southwest, Delta, and USAir,” Transportation Quarterly, LIV No.4 (Fall 2000), 7-22. (Richard Gritta, and Ned Freed). “Measuring the Degrees of Operating, Financial, and Combined Leverage Facing the U.S. Air Carriers: 1979-1995,” Transportation Law Journal,, XXVI No.1 (Fall 1998), 51-71. (Richard Gritta and Ned Freed). "The Causes and Effects of Business and Financial Risk in Air Transportation: Operating and Financial Leverage and the Volatility in Carrier Rates of Return," Journal of Transportation Management, VI No.1 (Spring 1994), 127-149. (Richard Gritta and Garland Chow). "Financial Leverage and Optimal Air Carrier Financing Patterns: An Indifference Analysis," in Distribution Research in the 1980s, (Ohio State University: Transportation and Logistics Research Fund, October 1978), 78-86. (Richard Gritta). th 56 Annual TRF Program, Atlanta, GA, Page 53 PROACTIVE PASSENGER MANAGEMENT WITH A TOTAL AIRPORT MANAGEMENT PERSPECTIVE Axel B. Classen, DLR, axel.classen@dlr.de Florian Rudolph, DLR, florian.rudolph@dlr.de This paper describes the approach of DLR research into proactive passenger management of airport landside and passenger operations. It will be shown, how a dynamic operational forecast can facilitate such proactive passenger management and how this can be implemented. The paper will also explain how a modern and proactive passenger management will fit in the Total Airport Management approach and finally show the potentials for improvement in terms of punctuality and efficiency. WHY REVENUE MANAGEMENT IS A GOOD THING Emmanuel Carrier, Delta Airlines, emmanuel.carrier@delta.com From its origins in the airline industry, revenue management has expanded first into the rest of the travel industry and later into many other industries such as retail and B2B. While they were becoming ever more present affecting a greater and greater number of B2C and B2B transactions, revenue management practices have become increasingly controversial with consumers. In this paper, we look at long series of empirical data to show that revenue management is actually a win-win for both consumers and producers, increasing utilization rates and decreasing waste. We discuss how to keep this legacy alive given the emergence of "big data" techniques. SESSION 5.4: Regulatory, Legal and Competitive Issues in Transportation ASSESSING THE OBJECTIVITY OF TRANSPORTATION RELATED REGULATORY DECISIONS IN CANADA James Nolan, U. of Saskatchewan, james.nolan@usask.ca Savannah Gleim, U. of Saskatchewan, savannah.gleim@usask.ca Like the Surface Transportation Board in the U.S., the Canadian Transportation Agency (or CTA) acts as the regulatory body for transportation issues in Canada. As described on their website, the following is their legal mandate; “The Canadian Transportation Agency is an independent, quasi-judicial tribunal and economic regulator. It makes decisions and determinations on a wide range of matters involving air, rail and marine modes of transportation under the authority of Parliament, as set out in the Canada Transportation Act and other legislation. Our mandate includes:    Economic regulation, to provide approvals, issue licenses, permits and certificates of fitness, and make decisions on a wide range of matters involving federal air, rail and marine transportation. Dispute resolution, to resolve complaints about federal transportation services, rates, fees and charges. Accessibility, to ensure Canada's national transportation system is accessible to all persons, particularly those with disabilities.” (CTA, 2014) Given several controversial and high profile recent decisions made by the CTA on transportation disputes, we want to examine if some form of regulatory capture or bias is driving decisions by the Agency. Building on prior work of Annand and Nolan (2003) questioning the objectivity of the CTA in making certain specific regulatory decisions, in this paper we examine a time series of related regulatory/dispute resolution decisions made by the CTA since 2000. All else equal, if agency decisions are fair in a legal sense, we offer that they should resemble a random draw over time. If so, we would expect to observe some “runs” in the data (possibly a lengthy series of wins or losses) for plaintiffs (typically shippers) or defendants (typically carriers). However, casual observation of recent decisions th 56 Annual TRF Program, Atlanta, GA, Page 54 indicates that they seem to resemble a simple “balancing act”, with decisions alternating over time between plaintiff and defendant. In effect, we suspect that the CTA decision process is dominated by efforts to appear neutral, characterized by a careful balancing of decisions over a fiscal year between shipper and carrier. We rely on non-parametric statistical analysis to examine the likelihood that the decisions made by the CTA are indeed biased in a legal sense by identifying patterns in the data that would be indicative of directed (i.e. non-random) agency decision-making over a given time interval. DOES COMPETITION IMPROVE TECHNICAL EFFICIENCY OF TRANSPORTATION SERVICES? A DEA APPLICATION IN THE CASE OF NORWEGIAN CAR FERRY SECTOR James Odeck, Molde University College, james.odeck@vegvesen.no Edvard Sandvik, The Norwegian Public Roads Administration, edvard.sandvik@vegvesen.no Svein Bråthen, Molde University College, svein.brathen@himolde.no Economic theory suggests that competition enhances efficiency by which transportation services are provided. We test the validity of this theory using the Norwegian car ferry links whereas ferry links connect trunk road segments by transporting cars from one road and to the other. Traditionally, the ferry link services were provided by monopolists subsidized by the government based on long term cost norms. Currently, ferry link services are gradually being exposed to competitive tendering (CT). We use Data Envelopment Analysis (DEA) to measure how ferry links perform and test whether the links that have been exposed to CT perform better than others. The results reveal that: (1) Exposure to competitive tendering enhances efficiency; (2) there are large potentials for technical efficiency improvement in the Norwegian of about 23-33% and; (3) exogenous factors outside the control of the ferry link management impact efficiency significantly. We urge the government to speed up competitive tendering. BICYCLE POLICY IMPLICATIONS OF UNEVEN DEVELOPMENT Caroline Appleton, Georgia Tech School of Public Policy, cappleton7@gatech.edu Bicycling as a mode of transportation is gaining recognition as city leaders, transportation engineers, and citizens alike strive to make urban areas more mixed modal and the most commonly implemented policy solution for improving bicyclist safety, ridership rates, and access is the conventional bike lane. But caution is advised. I motivate an analysis of the bike lane policy implications that follow from the uneven developments in the bicycling research literatures. To do so, I begin by using the three categories of policy scholarship – policy evaluation, policy analysis, and policy process – to describe this area of research. This reveals three conclusions from the literature; 1) the current state of the bicycle research and policy literatures is predominately focused on evaluation and analysis with little attention paid to process, 2) research topics covered generally do not account for the legal dimensions of bicycling, and 3) the range of relevant actors and their actual interests are inadequately accounted for the research. Taking this, I ask what a critical analysis of the bike lane as a designed technology and policy tool reveals about the differences between intended use and alternative use. Recognizing the importance of context, I limit my study to Atlanta, GA. Recognizing the importance of focus, I limit my study to the practice of cars parking in bike lanes. I incorporate state law and municipal code, relevant actor roles, AASHTO’s standard setting, and NACTO’s standard challenging into the study. I conclude with a set of policy proposals that address the unsafe practice of parking in bike lanes highlighting the need for integrative urban planning policies and areas of future research. th 56 Annual TRF Program, Atlanta, GA, Page 55 ECO-EFFICIENCY OF U.S. CONTAINER PORTS USING DATA ENVELOPMENT ANALYSIS (DEA) N. Muhammad Aslaam Mohamed Abdul Ghani, North Dakota State University, nmuhammadaslaam.moha@ndsu.edu Eunsu Lee, North Dakota State University, eunsu.lee@ndsu.edu Yong Shin Park, North Dakota State University, yong.park@ndsu.edu Container ports play a critical role as an international trade gate for every country in the world, especially in regards to the U.S. economy. Intermodalism is more effective and efficient by reducing the time of transferring goods from ship to shore and on-dock transportation. The environmental aspect of container ports is crucial in order to maintain the sustainable, green logistics at ports since maritime activities have been identified to be major pollutant emitters in the U.S. Those port activities include waiting to port, loading and unloading processes using cranes, and transporting containers from ships to inland depots by truck and rail. The purpose of this paper is to present the analysis of the eco-efficiency of U.S. container ports. There are two stages involved in the methodology of this research. In the first stage, Data Envelopment Analysis (DEA) is used to analyze port efficiency without CO2 emission consideration. The result of the efficiency is used in the second stage to determine eco-DEA, which is used to analyze the amount of CO2 produced from crane activities. SESSION 5.5: Global Trends in Transportation and Logistics A NELT-TRF Industrial session with the following presentations: TECHNOLOGY TRENDS IN LOGISTICS AND SUPPLY CHAIN MANAGEMENT Dr. Anthony Pagano, University of Illinois, Chicago DESIGN OF NEW CONCEPT CONTAINER TERMINAL Dr. Moohong Kang, KMI ADVANCED MATERIAL-HANDLING EQUIPMENT FOR MANUAL-DEPENDENT DISTRIBUTION Dr. Yongjang Kwon, KRRI DEVELOPMENT OF CARGO BATCH LOADING AND UNLOADING SYSTEMS Dr. Sukk Lee, KRRI SUPPLY CHAIN ISSUES IN AGRICULTURAL PRODUCTS AND TECHNOLOGIES Dr. Sangyoung Moon, Hankyong National University SESSION 6.1: Systems Performance and Safety ASSESSMENT OF OPERATIONAL PERFORMANCE OF BUS RAPID TRANSIT (BRT) IN LAGOS STATE NIGERIA Olayemi Funmilayo Dickson, National Centre for Technology Management, olayemidickson@yahoo.com Lagos state is one of the worlds’ largest urban agglomeration with an estimated population of 20,000,000 on a land mass of 3,577sqkm. Road traffic, urban mobility, congestion and environmental pollution in the State have been major challenges for many years; The Bus Rapid Transit (BRT) scheme was introduced to help reduce these problems. In the light of this, the impact of Bus Rapid Transit (BRT) system on commuters’ satisfaction in Lagos State was assessed. Data were collected through the use of questionnaire that was administered using simple random sampling technique. Also, secondary data were used. Results showed that majority of the BRT commuters were satisfied with BRT operations in Lagos State in the following areas: (safety/security (50%), speed (53.6%), identity and image (44.2%), fare structure (36.7%), comfortability (50.5%), boarding/boarding platform (38.6%), travel time (32.6%), and capacity (34%) but however express their dissatisfaction in some other areas of operation which are reliability (40.1%) and waiting time (62.2%). In terms of severity of the problems passengers encounter in the use of the BRT in Lagos state, the study revealed that majority (50.3%) of the respondents considers th 56 Annual TRF Program, Atlanta, GA, Page 56 overloading, delay in the arrival of buses/inadequate number of buses (46.9%) and inadequate maintenance of buses (36.1%) to be serious problems. Under the ticketing system, about 79.5% of the users expressed satisfaction with the use of paper ticketing system and 65% of them approved the introduction of an electronic ticketing system following the suggestion of an alternative ticketing system. The paper concluded that there is need for further improvement on the operational performance of BRT system in Lagos State in order to compare well with what is obtainable internationally and suggests possible policy recommendations. ASSESSING THE VULNERABILITY OF EVACUATION PLANS VIA CRITICAL ELEMENT DETECTION Chrysafis Vogiatzis, University of Florida, chvogiat@ufl.edu Jose Luis Walteros, University at Buffalo, josewalt@buffalo.edu Panos M. Pardalos, University of Florida, pardalos@ise.ufl.edu Successful evacuation planning and disaster management need to be robust with respect to disruptions over the infrastructure network. However, in most cases, when considering evacuation plans the capacity of the infrastructure is put to its limits, and a possible failure could have fatal consequences. In this study, we first propose an islanding technique to decompose large-scale evacuation problems into smaller, more manageable ones. The results of this decomposition are then fed into a mathematical formulation that aims to detect the set of critical elements (roads, bridges, intersections) of the evacuation plan. Identifying these elements can lead to the creation of balanced plans and can be reinforced to prevent disruptions during evacuation. Further, they can be used to analyze alternate routes and contingency plans in case of emergency. We present some preliminary results that provide us with evacuation plan robustness assessments. BENEFITS OF SYSTEM SAFETY MANAGEMENT Alireza Edraki, Gannett Fleming, aedraki@gnet.com System safety management has some costs associated with it when you try to implement it in your organization. Often, organizations do not recognize the long run benefits of establishing system safety management in return of upfront cost associated with developing the system. The benefits are all across the organization and project life cycle. It starts from design engineering steps through system requirements, implementation, manufacturing, installation, testing and commissioning. System safety can reduce cost of redesign, reengineering, retest and associated impact in organization and project. At the same time there are major risks related to not running the system safety management in organization which can be summarized in to “Health, economic and legal”. Every single of these risks can have significant consequences in Organization which will be presented in detail. These benefits are mostly generic across any type of project and organization and can be easily translated to financial benefits. SESSION 6.2: Highway Infrastructure and Public Sector Decisions EVALUATING THE APPLICATION OF DIVERGING DIAMOND INTERCHANGE IN ATHENS, ALABAMA Tahmina Khan, UAH, tk0002@uah.edu Michael Anderson, UAH, tk0002@uah.edu A diverging diamond interchange (DDI), a freeway and arterial interchange design, uses crossovers at the ramp intersections to eliminate conflicts between arterial through traffic and the left turning traffic from both the arterial and the ramps. The conventional diamond interchange (CDI) compared with the DDI has similar on-ramps and offramps, but there is a change in the number of lanes on the arterial. A case study has been conducted for an interchange in Athens, Alabama. The existing interchange is a conventional diamond interchange which is compared with a DDI built at the same location by using Synchro/Simtraffic as a simulation tool. The analysis th 56 Annual TRF Program, Atlanta, GA, Page 57 focused on determining level of service and vehicle delay for the two interchange types. In this project effort was made to know if a DDI is the right solution for all interchange locations through the examination of a single location and adjustments to traffic volumes under numerous combination of turning movement scenarios where variation of capacity and the influence of proximity of adjacent intersections were also included. Finally, it was concluded that DDI cannot be an appropriate measure to improve the current CDI network for the study area since only 4 special cases it performs better than CDI. And DDI cannot be the effective traffic calming measure if it is associated with upstream or downstream intersections. HEAVY VEHICLE IMPACT ON RURAL TWO LANE HIGHWAY SEGMENTS OPERATING UNDER VARIOUS LEVELS OF SERVICE CONDITIONS Zijian Zheng, North Dakota State University, zijian.zheng@ndsu.edu Pan Lu, Upper Great Plains Transportation Institute, pan.lu@ndsu.edu Oil boom in Western North Dakota State triggers a surge of oil trucks shipping oil products. The distinct characteristics of oil trucks, such as low speed, large size, and slow accelerate and decelerate have unusual effect on following traffic and result in inaccuracy in traffic capacity forecasting. To estimate effect of oil trucks on two-lane rural highway, in this research, a new set of passenger-car-equivalent (PCE) factor is calculated based on an improved analytical model using two criteria: headway and delay. Several factors are considered in the developed model: vehicle speed, safety passing time, headway distribution, flow rate, and delay to downstream traffic. It is demonstrated that PCE value is related with flow rate of both lanes, and the new set of PCE values reveal fluctuation of effect of trucks on following traffic under various traffic conditions. FREIGHT MOVEMENT BY HEAVY DUTY DIESEL TRUCKS EMISSION IN PORT OF HOUSTON AREA Ehson Khademi, Texas Southern University, ehson.khademi@gmail.com Fengxiang Qiao, Texas Southern University, qiao_fg@tsu.edu Mehdi Azimi, Texas Southern University, azimim@tsu.edu lie Yu, Texas Southern University, yu_lx@tsu.edu Xiaobing Wang, Texas Southern University, xiaobingwang2010@gmail.com The purpose of this study was to characterize the VSP distribution of trucks in port of Houston area in order to give recommendations how to mitigate pollution. The port of Houston handles about 70 percent of the containerized cargo in US Gulf of Mexico. There are 8 main terminals, six of which are general cargo and two are container terminals named Barbours cut and Bayport. The Bayport terminal was selected for the study, the most modern and environmentally sensitive container terminal on the U.S. Gulf. Port truck activities on Port of Long Beach, Port of Los Angeles and Port of Oakland generate approximately 7,075 tons per year (TPY) of NOx and 564 tons per year of diesel PM in 2005. These emissions represent 23 percent of all port-related NOx emissions and nine percent of all port-related diesel PM emissions. With respect to the considerable amount of emission created by the trucks and significant air pollution impacts on living in communities, it causes adverse health effects, particularly for children, the elderly, and those with compromised health. Since the easiest and fastest way to make changes to decrease the amount of emission is possible through making changes in truck circle, so the purpose of this study is designed to investigate the amount of emission created by heavy duty diesel trucks. In order to do so, the problems and constraints were identified by reviewing the literature so that most executable suggestions could be evaluated and offered in the least amount of time and in fastest way. There were 5 heavy duty diesel trucks chosen for 6 weeks, 5days a week. The area of the study was limited to a specified area with the areas covering 11 zip codes. Instantaneous speeds of vehicles were collected on a secondby-second basis using Global Position System(GPS) device. Then, for each second-by-second data, acceleration rate and Vehicle Specific Power (VSP) value were calculated. VSP were used to obtain the operating mode distribution bins according to the standard provided by the Motor Vehicle Emission Simulator(MOVES). The vehicle emissions th 56 Annual TRF Program, Atlanta, GA, Page 58 were calculated based on the operating mode binning approach. Emission factors analyzed in this study are Carbon Dioxide (CO2), Oxides of Nitrogen (NOx), Hydrocarbons (HC) and particulate matter (PM). As a result, the frequency of bin 1 (idling mode) has the highest approximate percentage equal to 59% that represents the amount of time that trucks have been moving with less than one mph, which shows the amount of time the truck have been in long line for entering and leaving the gates, inspection of the vehicle or inspection of the truck load or any other reason that truck is stopped but the driver do not turn off the vehicle. With respect to the information given in the introduction and the review of literature (Background), the highest percentage of operating mode ID and as a result the highest emission distributed, is related to the idling mode. To figure out about unexpected amount of bin 1 (idling mode), it’s suggested preparing a survey about drivers behavior. Also, it’s recommended checking the location of truck when it’s stopped for more than 2 minutes. Also as a future research, it’s suggested improving the efficiency of the current application of Port of Houston Authority and current truck circle in order to avoid idling by managing time as much as possible. SESSION 6.3: Railroad Technology and Development SHORTLINE RAIL IN WASHINGTON STATE: CHALLENGES AND OPPORTUNITIES ON THE ROAD AHEAD Jeremy Sage, Freight Policy Transportation Institute, jlsage@wsu.edu Ken Casavant, Freight Policy Transportation Institute, casavantk@wsu.edu J. Bradley Eustice, Freight Policy Transportation Institute, jbradley.eustice@email.wsu.edu The recently completed State Rail Plan for the state of Washington identified, via a needs assessment, several key issues facing the state’s rail system. Tops among these issues are abandonment, access and competitive needs of the ports, as well as intermodal connectivity. Nearly 2000 miles of rail line had been abandoned in Washington before the late 1990’s, and another 70 have been abandoned since. Many of these miles were a result of Class I railroads stepping away from their less profitable lines. These actions generated opportunity for the creation of many of the state’s shortline railroads on branch and light density lines. However, as the larger rail system contracted and developed, many of the state’s shortlines have been left behind and no longer meet the standards and conditioning required for today’s load limits (286,000 pounds per car), thus requiring further investment should the state or owner seek new or improved operation. The vital question is whether these upgrades will promote and induce further use of these lines. This study inventories the state’s shortline operations and identifies the perceived needs of operators to maintain and grow their operations in order to fill a vital niche in Washington’s multimodal transportation system. Further, the Economic Impact of shortline rail is investigated through a series of regional case studies to demonstrate the breadth of roles played by shortline in Washington State and the potential implications of funding shortfalls that fail to allow them to maintain relevancy. PASSENGER RAIL ENERGY TECHNOLOGIES FOR THE NEAR-TO-MID-TERM: A SUMMARY REVIEW AND CASE STUDY Raphael Isaac, UC Davis Institute of Transportation Studies, risaac@ucdavis.edu Lewis Fulton, UC Davis Institute of Transportation Studies, risaac@ucdavis.edu Projections for the Amtrak Capitol Corridor line, which serves the corridor between San Jose and Auburn, suggest close to a doubling of ridership by 2040 (CA State Rail Plan). The Bay Area’s Caltrain service and other California rail lines also expect dramatic growth. In response to this trend, which mirrors similar trends for passenger (i.e. both commuter and intercity) rail that can be seen across the country, the state has ambitious goals to increase service levels – for example increased top speed on the Capitol Corridor line, from 79 mph to 110 mph, to add track along the current right-of-way, and to extend the service beyond its current endpoints and add new branch lines. Given such major changes being considered, including infrastructure expansion, an increased number of vehicles, greatly increased fuel usage, and, an aspect not to be minimized, a further entrenchment of energy/fueling infrastructure, it seems a good time to evaluate the costs and energy/environmental implications of different rail energy technologies, addressing performance/implementation issues that pertain to both freight and passenger rail. th 56 Annual TRF Program, Atlanta, GA, Page 59 Diesel fuel and its associated locomotive technologies (including diesel-electric) currently provide the power source for approximately 87% of domestic rail (NREL, Transportation Energy Futures), while electricity and its associated infrastructure and locomotive technologies comprise the remaining 13% (with very limited exceptions). Some experimentation with biodiesel-operated rail has occurred here in the U.S., and, as a “drop-in” fuel, this is a technology that seemingly could provide environmental benefits with very little additional cost (and associated risk), at least at the “tailpipe.” Europe has embraced electricity as a rail fuel (at least seven European countries have 50% or more of their track electrified, and Britain has in recent years greatly increased its electrified track), while, with the low natural gas prices of recent years, significant piloting of natural gas has also begun, in particular in North America. Using H2, or fuel cells, to power rail is still in early experimental stages, with features such as refueling needing to be explored further; however, H2 has the potential to provide many of the benefits of electrification without the high costs (particularly up front) of additional infrastructure (and, potentially, related maintenance). Lastly, various hybrid concepts that attempt to work in conjunction with traditional diesel technologies may offer some environmental benefits at a lower cost and perhaps risk level. In addition to reviewing the current literature on these technologies, we will present a case study examining likely characteristics and impacts of each on California’s Amtrak-Capitol Corridor line. We will provide a quantified comparison of up to 5 different fuels for rail systems (diesel, biodiesel, LNG, electricity and hydrogen). We will also assess the potential, costs and benefits of efficiency improvements such as hybridization. Our hypothesis is that strong efficiency improvements will generate net cost savings over just a few years, whereas most alternative fuel projects will take many more years to pay back, though will be needed to achieve very low carbon rail travel. ANALYSIS OF CAPACITY PRICING AND ALLOCATION MECHANISMS IN SHARED RAILWAY SYSTEMS: LESSONS FOR THE NORTHEAST CORRIDOR Maite Pena-Alcaraz, MIT, maitepa@mit.edu Joseph Sussman, MIT, sussman@mit.edu Recently, governments have started promoting the use of shared railway systems as a way to take advantage of the existing capital-intensive railway infrastructure. Up until 15 years ago, all major railways both managed the infrastructure and operated the trains, i.e., they were vertically integrated. In contrast, in shared railway systems, multiple train operators utilize the same infrastructure, i.e., there is some level of vertical separation between infrastructure management and train operations. Examples of shared railway systems are the Northeast Corridor in the U.S. and the railway system in Italy. Such systems can achieve high utilization, but also require coordination between the infrastructure manager and the train operators. Such coordination, in turn, requires capacity planning regulation that determines which trains can access the infrastructure at each time, capacity allocation, and the access price they need to pay, capacity pricing. The need to establish capacity pricing and allocation mechanisms in the railway system is relatively new. There was no need to pay for access or to allocate capacity under traditional vertically integrated railway systems. As a result, the literature in this area is nascent, and there is limited understanding of the trade-offs associated with alternative capacity pricing and allocation mechanisms and their comparative performance. The objective of this paper is to identify some of the trade-offs involved in the choice among alternative capacity pricing and allocation mechanisms for shared railway systems in the context of the Northeast Corridor in the US. In this case, the Federal Railroad Administration requires Amtrak and the rest of the commuters and freight railway companies to agree on a capacity pricing and allocation mechanism in 2015. To analyze these trade-offs, we develop a framework to evaluate the performance of shared railway systems under generic capacity pricing and allocation mechanisms considering both technical and institutional aspects. This framework integrates two modules. The train operator module simulates the behavior of the train operators to determine their demand to use the infrastructure, their access charges willingness to pay, and the fares they would charge to the users. The infrastructure manager model optimizes the timetable design considering the demand from train operators and all the technical constraints from the infrastructure. The results obtained are the demand to schedule trains, the access charges (capacity pricing), and the final train timetable (capacity allocation, which train services are finally scheduled and when). With this information the performance of two alternative capacity pricing and allocation mechanisms is analyzed from the perspective of the infrastructure manager (cost recovery, th 56 Annual TRF Program, Atlanta, GA, Page 60 infrastructure capacity use, etc.), the train operators (access charges, trains scheduled, barriers to entry, etc.), and the users (level of service, fares, etc.). FREIGHT RAIL ECONOMIC DEVELOPMENT Libby Ogard, Prime Focus LLC, logard@new.rr.com The State of Minnesota undertook a study to identify opportunities to grow freight on short line railroads in the state. All short lines in the state were interviewed to identify needs, opportunities and barriers. An outreach effort which involved all the Economic Development Agencies within the state was undertaken to identify what rail projects they felt would be most beneficial for their regions. An inventory of projects was prepared. This paper examines the partnership between the State Department of Transportation and the Department of Economic and Employment Development in Minnesota, and through their joint effort, how to best approach freight rail economic development. SESSION 7.1: Speed, Time, and Delays in Transportation A SPATIO-TEMPORAL BASED ALGORITHM FOR SHORT-TERM SPEED PREDICTION Jianjiang Yang, University of Tennessee, jyang27@vols.utk.edu Lee Han, University of Tennessee, lhan@utk.edu As traffic flows from upstream to downstream, the traffic stream characteristics at a certain location are usually closely related to those at neighboring locations. The incorporation of traffic data at adjacent locations as input variables has been demonstrated to improve traffic forecasting accuracy in recent literature. However, the fashion of how neighboring traffic data are utilized is somewhat arbitrary in most, if not all, of the recent studies where spatial-temporal correlations are pre-specified by the researcher rather than properly identified based on the nature of the sites and behaviors of traffic. The purpose of this study is to propose a model that could determine the most desirable extent of temporal and spatial traffic data to be used from neighboring locations for traffic variable forecasting at a given location. To this end, the kriging model with a moving time window is proposed and developed using high resolution freeway speed data in this paper. The spatial-temporal correlations incorporated in the proposed model are scientifically identified using cross-correlation function analysis. As a data-driven approach, the proposed kriging model can also make responsive prediction by adapting to unexpected changes through unsupervised self-learning. To evaluate the effectiveness of the proposed model, it is compared against three benchmark models (i.e., ARIMA, simple exponential smoothing, and double exponential smoothing) for three scenarios: morning peak, mid-day off peak, and evening peak. It is found the proposed kriging model outperforms all other benchmark models under all scenarios. Even more remarkably, the results suggest that the modified kriging model performs exceedingly well in congested conditions (morning peak and evening peak). This interesting finding indicates that spatial correlations are critical for better forecasting when congestion is present. EFFICIENT FRONTIER OF ROUTE CHOICE FOR EQUILIBRIUM UNDER TRAVEL TIME VARIABILITY WITH HETEROGENEOUS TRAVEL PREFERENCES Mahyar Amirgholy, University of Massachusetts Eric J. Gonzales Abstract not available. th 56 Annual TRF Program, Atlanta, GA, Page 61 REROUTE OR WAIT IT OUT? ESTIMATING THE TIME UNTIL SPOILAGE IN THE PRESENCE OF UNEXPECTED DELAYS J Bradley Eustice, Freight Policy Transportation Institute, jbradley.eustice@email.wsu.edu Jeremy Sage, Freight Policy Transportation Institute, jlsage@wsu.edu Ken Casavant, Freight Policy Transportation Institute, casavantk@wsu.edu On rural roads, unexpected delays significantly impact freight as there is rarely an alternate route. Firms must make decisions in real time to prevent freight from “spoiling.” In the case of agriculture, the actual freight may spoil. For all other products, important deadlines may be missed. The research question is: “In the presence of an unexpected delay, which industries are most time-sensitive?” This paper develops an empirical model using a recent origindestination survey of 4600 truckers conducted by Washington State University on behalf of the Washington Department of Transportation. To expand the dataset, route and reroute options are calculated using the Google Maps API, which is based on a shortest duration algorithm. The model is tested using linear regression analysis. Of the 42 SCTG commodity classifications, only a handful has significant results at the 5% level. The time until spoilage is between 3 and 7.5 hours above the scheduled delivery time. As is expected, the agriculture industry is the most time-sensitive. SESSION 7.2: Factors Affecting Highway Safety WEIGHTS FROM SAFETY PERSPECTIVE FOR INTERCHANGE LIGHTING PRIORITIZATION TOOL Srinivas S. Pulugurtha, The University of North Carolina at Charlotte, sspulugurtha@uncc.edu Ravishankar P. Narayanan, The University of North Carolina at Charlotte, rpottyna@uncc.edu Public agencies rely on the “Total Design Process” (TDP) of the National Highway Cooperative Research Program (NCHRP) Report 152 “Warrants for Highway Lighting” and the updates that followed to identify and prioritize lighting needs along roadway sections. However, this more than 40 year-old guidebook does not account for various factors that range from crash severity to traffic composition and other design criteria. The objectives of this manuscript are to research, explore, and develop an updated mechanism to prioritize interchange locations that require lighting. New factors such as 1) acceleration lane length, 2) deceleration lane length, 3) distance of signboard placement from the interchange, 4) crashes by severity at night, 5) illumination level, 6) the percent of heavy vehicles at night, and, 7) ramp volume ratio were identified and included in the updated tool. Data collected at 80 interchanges (38 with lighting system and 42 without lighting system) along nine corridors was used to estimate unlighted and lighted weights, compare with currently used weights, and update the TDP prioritization tool. While most of the weights computed for factors were similar to the currently used weights, differences were observed for factors such as freeway median width, freeway number of lanes and night-time traffic volume per lane. Warranting points were computed using the current tool as well as the updated tool with new factors and computed weights. The inclusion of new factors other than crash severity and updated weights only had a marginal effect on the warranting points. A comparison of computed warranting points using both the current and updated tools indicate a decrease in warranting points at the interchanges with fewer numbers of night-time crashes or less severe crashes using the updated tool. In general, warranting points increased based on the updated tool for interchanges with more severe night-time crashes. It is, therefore, recommended to consider the number of nighttime crashes by severity instead of night-to-day crash rate ratio for prioritization of interchange lighting system installation or maintenance. th 56 Annual TRF Program, Atlanta, GA, Page 62 CORRELATION BETWEEN CRASH RATES AND INTERNATIONAL ROUGHNESS INDEX ON IN-SERVICE U.S. HIGHWAYS Ahmed Elghriany, The University of Akron, afe3@zips.uakron.edu Ping Yi, The University of Akron, pyi@uakron.edu Peng Liu, The University of Akron, pl40@zips.uakron.edu Quan Yu, Beijing University of Technology, yuquan@bjut.edu.cn Pavement roughness may have a significant impact on traffic safety, and its effect on the overall safety of motorists needs to be studied and quantified. This paper investigates the relationship between the International Roughness Index (IRI) and crash rates by examining traffic safety performance over time and under changing pavement conditions. Specifically, the proposed model provides an insight into the effect of the roadway surface type conditions on traffic safety in order to develop a better understanding of the problem. Data on GIS platform and extracted by SAS software from Oracle servers was obtained from different sources across North America and was compiled and analyzed using a statistical analysis approach. For both Portland cement concrete (PCC) pavement and asphalt concrete (AC) pavement, different model formulations were examined, and a quadratic relationship was found to be most effective to link crash rates with pavement roughness. The obtained regression models reveal that PCC pavement with an IRI value of around 1.50 m/km and AC pavement with an IRI around 0.75 m/km are indicative of a safer roadway, while IRI values of greater than 2.25 m/km for PCC pavement and a value greater than 1.25 m/km for AC pavement indicates a roadway that may be susceptible to much higher crash rates. The findings of this study may serve as a good reference for state and local transportation agencies to use in facilitating roadway maintenance decision making. THE USE OF CHAID DECISION TREE MODELS IN INJURY SEVERITY ANALYSIS OF LARGE TRUCK CRASHES Jill M. Bernard, University of Missouri - St. Louis, jmbhw9@mail.umsl.edu Christopher Mondy, University of Missouri - St. Louis, cmmd6c@mail.umsl.edu Large truck transport is vital for freight shipping in the United States; yet, it can prove to be a dangerous mode of transportation. From 2002 to 2012, 91,145 crashes involving large trucks occurred in Missouri, resulting in 1,156 fatalities and 18,457 injuries. Many factors contribute to large truck crash severity, and it is theorized that these factors and their effect on injury severity vary as a function of gender. Missouri traffic, personal and vehicle crash data from 2002-2012 are used to analyze circumstances that increase the probability of injuries and fatalities, given a large truck crash occurs. CHAID decision tree models are developed for each gender to predict values of injury severity as impacted by crash contributing circumstances to better understand predictor importance and uncover interactions among factors. Results suggest that the major contributory predictors for crash severity for Missouri female drivers to be driving too fast for conditions, driving on the wrong side of the road, improper backing, speeding and improper turning, while major contributing predictors for Missouri male CDL drivers are driving too fast for conditions, improper backing, violation of stop sign or signal, improper turning, and failing to secure loads. Additional results suggest that for Missouri female CDL drivers, when speeding contributes to a crash occurrence, the probability of a fatality is 14.29% and an injury is 28.57%. For Missouri male CDL drivers, when driving too fast for conditions on the wrong side of the road, an 8.70% and 48.91% probability for fatal and injury outcomes exist respectively. As a result, it is recommended that truck driver training programs focus educational efforts on gender specific behaviors that impact crash injury severity in order to enhance road safety measures. th 56 Annual TRF Program, Atlanta, GA, Page 63 SESSION 7.3: Organizational Issues in Transport NORWEGIAN BUS TRANSPORT - DOES INTERNAL OR EXTERNAL ORGANIZATION OF PUBLIC AUTHORITIES AFFECT EFFICIENCY Hilde Johanne Svendsen, Møreforsking Molde AS, hilde.j.svendsen@himolde.no Arild Hervik, Møreforsking Molde AS, arild.hervik@gmail.com James Odeck, james.odeck@ntnu.no This topic addresses how to organize the processes of competitive tendering of bus transport in public sector. The question of internal or external organization will be discussed in all Norwegian counties when the procurement process for bus transport goes from negotiations to competitive tendering. In internal organization the competitive tendering process is performed by a department at the county authorities, while external organization is when the county authorities establish an external unit to take care of the competitive tendering process. The necessary data was collected from the public transport authorities. This study was made with all Norwegian counties as population, and Data Envelop Analysis (DEA) and 2nd stage regression was applied. Passenger kilometer was used as main output. The number of passengers was also tested as output. Labour, capital and energy were used as input variables. Population density, degree of competitive tendering and internal/external organization were the 2nd stage explanatory variables. At the outset, the result tends to explain that the lower degree of competitive tendering that is applied in a county, the more efficient is the county performing. When the variables for Density, number of inhabitants per kilometer of public road are included, the result turns out as opposite and the model tends to explain that higher density gives higher efficiency in the counties in question. For the main variable Organization, the model explains that external organization is more efficient than internal organization within a 10 per cent significance level. As use of DEA is sensitive to outlayers in the population, there were made an analysis where Oslo, the capital in Norway, and Akershus were not a part of the population. By use of this sensitivity analysis the findings suggest that a differentiated practice between rural and urban areas could be implemented. THE RELATIONSHIP BETWEEN LEADERSHIP STRENGTHS AND PERFORMANCE OF FLEET MANAGERS Ahren Johnston, Missouri State University, ahrenjohnston@missouristate.edu John L Kent, University of Arkansas, jkent@walton.uark.edu In this study, the Clifton StrengthsFinder® Assessment and data on 2013 fleet manager performance at a large midwestern motor carrier is used to see if any correlation exists between leadership strengths and on the job performance of fleet managers, as measured average revenue per unit per week and average employee turnover. While personality tests can lead to valuable insights with regards to the potential of a job candidate, the use of such testing may likely violate the privacy rights of potential employees and could potentially violate ADA and EEOC laws. The use of Clifton StrengthsFinder® as an alternative could potentially avoid the legal uncertainty associated with the pre-employment use of personality tests. The Clifton StrengthsFinder® Assessment determines the five top leadership traits of fleet managers, and preliminary results reveal that, after controlling for number of trucks managed and whether company employees or owner/operators are managed, fleet managers with Arranger and Positivity as top traits have lower revenue, and those with top traits including Competition and Analytical have higher revenue. Further results reveal that Responsibility, Individualization, and Relator traits are associated with lower driver turnover rates, and Restorative is associated with higher driver turnover rates. th 56 Annual TRF Program, Atlanta, GA, Page 64 ASSESSING THE EFFECT OF COMPRESSED WORK WEEK STRATEGY ON TRANSPORTATION SYSTEM PERFORMANCE MEASURES Venkata R. Duddu, The University of North Carolina at Charlotte, vduddu@uncc.edu Srinivas S. Pulugurtha, The University of North Carolina at Charlotte, sspulugurtha@uncc.edu This research focuses on evaluating and assessing the effect of compressed work week strategy on transportation network performance measures such as link-level traffic speed, travel time, and volume-to-capacity ratio using TransCAD software. Data gathered for the Charlotte metropolitan area, located in the state of North Carolina, United States, during the morning peak hours was used to illustrate the effect of compressed work week strategy and identify appropriate percentage of employees that need to be part of the strategy. The changes in traffic patterns were observed for every 5% increase in the reduction of trips (say, due to select employees working from home or not working a day each week) to downtown area (with high employment density). The results obtained indicate that reducing 15% to 20% of work commute during the morning peak hours using compressed work week strategy would increase traffic speeds by 5 mph (8 kmph) on 66% of center-lane miles and yield maximum benefits. It could also decrease the average travel time on 64% of center-lane miles. th 56 Annual TRF Program, Atlanta, GA, Page 65

Prelim Program (v6).xlsx - Transportation Research Forum

Related documents

Add this document to collection(s)

Add this document to saved

Suggest us how to improve StudyLib