Final Lecture Intelligent Transportation Systems ITS and the Impact of Traveler Information Emerging Themes in Transportation Economics and Policy Moshe Ben Akiva 1 201 11 545 ESD 210 Transportation Systems Analysis Demand Economics Fall 2008 Outline ITS Introduction to ITS and its applications Dynamic traffic management The impact of traveler information Modeling and simulation for dynamic traffic management DynaMIT Travel behavior models Data and calibration Evaluating the impact of traveler information Appendices Examples Case Studies Useful reference on ITS http www itsoverview its dot gov 2 Introduction Intelligent Transportation Systems ITS combine advances in information systems communications sensors and advanced modeling and algorithms to improve the performance of surface transportation 3 Benefits of ITS Travel Decreased travel time Improved safety and security Increased reliability Decreased cost Improved trip planning Improved emergency response 4 Benefits of ITS cont Economic Increased productivity On time delivery Environmental Decreased air pollution Fuel savings 5 ITS Applications Advanced Transportation Management Systems ATMS Network management including incident management traffic light control electronic toll collection congestion prediction and congestion ameliorating strategies Advanced Traveler Information Systems ATIS Information provided to travelers pre trip and during the trip in the vehicle Advanced Vehicle Control Systems AVCS Technologies that enhance driver control and vehicle safety 6 ITS Applications cont Commercial Vehicle Operations CVO Technologies that enhance commercial fleet productivity including weight in motion WIM pre clearance procedures electronic log books interstate coordination Advanced Public Transportation Systems APTS Passenger information and technologies to enhance system operations including fare collection intramodal and intermodal transfers scheduling and headway control Advanced Rural Transportation Systems ARTS Mostly safety and security technologies for travel in sparsely settled areas Reference Sussman Joseph M Introduction to Transportation Systems Artech House 2000 7 ITS Applications cont This lecture will focus on dynamic traffic management which integrates ATMS network management and ATIS traveler information applications 8 Dynamic Traffic Management Traffic Management Center Information Surveillance Information Loop Travel Demand Control Control Loop Traffic Conditions Supply 9 Impact of Traveler Information Provide information to road users before and during a trip through in vehicle technologies Public transport users can also benefit from information provided through Vehicle location systems Scheduling systems e g online journey planners Freight operators can use Commercial Vehicle Operations CVO technologies to more effectively manage their fleets The information not only lead to more efficient network flows but can also be used for strategic transportation planning purposes 10 User Response to ATIS Awareness System Decisions Access Usage Trip Decisions Travel Response 11 Example Impact of Traveler Information Images removed due to copyright restrictions Critical component user behavior models 12 Modeling and Simulation for Dynamic Traffic Management Systems Off line evaluation of Dynamic performance stability and robustness Effectiveness of surveillance control system designs Future system and network modifications Development of new concepts and algorithms Real time decision support systems Route guidance Adaptive traffic control Incident management 13 DynaMIT DynaMIT is a simulation based real time system predicting traffic providing travel information 14 Broadcasting Traffic Information Images removed due to copyright restrictions 15 DynaMIT Framework Prediction based guidance Prevents over reaction Database Network Historical information Real time Inputs Traffic surveillance Control State Estimation Supports compliance Demand Simulation Supply Simulation Prediction based Information Generation Demand Simulation Supply Simulation Information Generation Information Dissemination 16 1 4 a Disaggregate 2 Estimated Informed Aggregate OD Historical Demand Simulation in DynaMIT d 5 c b e 3 6 a Disaggregation of historical OD flows b Travel behavior update c Aggregation d OD estimation and prediction e Generation of driver population 17 Traveler Decisions Access Pre Trip Usage Travel Response T ravel or Not Destination Choice Mode Choice Departure Time Choice Route Choice En Route Usage Travel Response Return to Origin Change Destination Change Mode Switch Routes 18 Travel Behavior Models in DynaMIT Route and departure time Multiple driver classes Value of time access to information Path Size Logit1 Path choice set generation Shortest paths Link elimination Random perturbation 1 Ramming M S 2002 Network Knowledge and Route Choice Ph D thesis MIT 19 Travel Behavior Models in DynaMIT cont d Information Instantaneous predictive Descriptive prescriptive Link sub path path Response to Information Pre trip En route Media Variable Message Signs VMS Television traffic websites In vehicle radio cell phone GPS navigation system 20 Travel Behavior Models in DynaMIT cont d Prescriptive Compliance Descriptive Pre trip En route 21 Data and Calibration Disaggregate surveys diaries Detailed individual data Limited Aggregate traffic sensors Easy to collect widespread coverage Special estimation methods Source Balakrishna 2006 See appendix 22 Disaggregate Calibration Example1 2 Pre trip response to unexpected congestion Golden Gate bridge San Francisco CA Home to work trips Conclusions Travel time expected delay congestion level Impact willingness to change travel patterns Alternative types of ATIS Trigger different travel responses compliance rates Experience based factors very significant 1 Khattak A A Polydoropoulou and M Ben Akiva 1996 Modeling Revealed and Stated Preference Pre trip Travel Response to ATIS TRR No 1537 2 Polydoropoulou A 1997 Modeling User Response to Advanced Traveler Information Systems ATIS Ph D thesis MIT 23 Aggregate Calibration with DynaMIT Benefits Jointly adjusts all model parameters Route choice OD flows supply Uses general aggregate traffic data e g counts speeds travel times Applies to any traffic model Updates available parameters with latest traffic data More detail in Appendix 24 Case Study1 Boston 182 nodes 211 links AM peak 7 00 to 9 00 Map of downtown Boston highways removed due to copyright