Abdul R. Pinjari. Associate Professor Department of Civil & Environmental Engineering University of South Florida

Transcription

1 Toward a Sustainable Program for Measuring, Monitoring, and Managing Transportation Systems in Indian Cities Perspectives of a Transportation Researcher from the United States Abdul R. Pinjari Associate Professor Department of Civil & Environmental Engineering University of South Florida Visiting Professor, CiSTUP, IISc Bangalore

2 Overview of the Talk Why this workshop? Insights from the US-DOT s University Transportation Center (UTC) programs on the role of academic institutions of higher learning in creating more sustainable transportation systems Data, Planning Tools, Technologies, and Research needs Some examples from my research and other colleagues research How can academic institutions and stakeholders work together toward a sustainable management of our transportation systems? 2

3 Why and what is this workshop about? Not about immediate solutions to address Bengaluru s traffic problems Need a long-term and sustained program to measure, monitor, analyze, understand, predict, and manage our transportation systems How different stakeholders can support in building such a program? What are the emerging decision-support needs of the stakeholders? What is the role of research programs at academic institutions? How emerging technology-enabled data sources can help? How the science and practice of transportation planning and operations change in the face of emerging decision-support needs and data sources? Some directions for the transportation research program at CiSTUP 3

4 United States Department of Transportation (US-DOT) s University Transportation Centers (UTC) program Started in 1987 Purpose of UTCs DOT invests in the future of transportation through its University Transportation Centers Program, which awards grants to universities across the United States to advance the state-of-the-art in transportation research and develop the next generation of transportation professionals. [Emphasis added] Mission Internationally recognized centers of excellence, fully integrated within institutions of higher learning, that serve as a vital source of leaders who are prepared to meet the nation s need for safe, efficient and environmentally sound movement of people and goods. Vision To advance U.S. technology and expertise in the many disciplines comprising transportation through the mechanisms of education, research and technology transfer at university-based centers of excellence. Source: 4

5 The US-DOT s UTC program Goals of US-DOT s UTC program Education: a multi disciplinary program of course work and experiential learning that reinforces the transportation theme of the Center. Human Resources: an increased number of students, faculty and staff who are attracted to and substantively involved in the undergraduate, graduate and professional programs of the Center. Diversity: students, faculty and staff who reflect the growing diversity of the U.S. workforce and are substantively involved in the undergraduate, graduate and professional programs of the Center. Research Selection: an objective process for selecting and reviewing research that balances multiple objectives of the program. Research Performance: an ongoing program of basic and applied research, the products of which are judged by peers or other experts in the field to advance the body of knowledge in transportation. Technology Transfer: availability of research results to potential users in a form that can be directly implemented, utilized or otherwise applied. Source: 5

6 The role of transportation research programs at institutions of higher learning (in my view) Conduct basic research to Enhance our understanding of transportation systems Create knowledge base Educate future workforce, innovators, and thought leaders through Interdisciplinary academic curricula Involvement of students in research and knowledge creation process Attracting the best and the brightest into the field Origin of new ideas and innovations in transportation Contribute to solving society's transportation problems Through all the above (slow but steady and sustainable impact) Applied research relevant to society and technology transfer Most successful university transportation research programs are those that were allowed to pursue research in directions they were best at Success = Extent of innovation Success = Reputation in the academic, stakeholder, and student circles Success = Impact on practice and society in the long-term 6

7 The role of transportation research programs at institutions of higher learning (views of the more experienced) Prof. Clyde Lee Founding Director of the University of Texas Transportation Research Center Pioneer in weigh-in-motion systems I always insisted that the purpose of research was to support the academic program of the university. Robert C. Johns Associate Administrator and Director of US-DOT s Volpe National Transportation Systems Center Oversees $265 Million in annual research and innovation projects Former Director of Center for Transportation Studies at Univ. of Minnesota Excerpts from his paper titled Transportation Research Contributions to Society by University Transportation Centers IATSS Research, 27(1), it would be rare for a center and a university to develop successful innovations alone. The university focuses primarily on the function of creating knowledge and the labor pool. While the university does play a role at times in developing and testing innovations and creating technology standards, external organizations usually provide the other functions needed for successful innovations. 7

8 Views of Robert C. Johns (continued ) the university is best at divergent behavior. University research is about discovery and exploration, and resources are needed to fuel this dynamic process. However, the customers of centers, public or private, are more likely to succeed at convergent behavior, focusing on implementing innovations that meet their internal and external demands. Challenges: Knowledge can be a difficult product to sell. Expenditures on basic research can be difficult to defend,... There can be expectations that the university should come up with the one right answer to a complex problem Gap between academic and practitioner cultures If, however, University research is perceived as consultant services to only solve specific problems and provide short-term answers will suffer in comparison to the performance of professional consultants, and it may be perceived as losing its objectivity. 8

9 Data needs (or Measurement needs) Travel Demand Travel behavior Why, when, where-to-where, how, and how much of travel (people and goods) Usage of transportation facilities and services Traffic volumes (on roadways and at signals), transit ridership, etc. Determinants of travel behavior and drivers of travel demand Demographics, socioeconomics, culture, land-use, technology, information, Transportation System Performance Travel time and travel speeds (congestion patterns) Usage (under/over) of different services and facilities Reliability and resilience Infrastructure condition User benefits Short-term management strategies require high resolution data Temporal resolution, Spatial resolution, and Socio-demographic resolution A continuous measurement and monitoring program is paramount 9

10 Some research directions of potential value 1. Next-generation travel forecasting models for Indian cities (Bangalore) 2. Communication and sensing enabled transportation operations and planning 3. Research program to improve the understanding and control of heterogeneous and in-disciplined traffic streams 4. Air quality measurement, monitoring, and modeling 5. Freight and logistics (urban and inter-city) 6. Understand and promote megaregional development (i.e., development of second tier cities around major metropolitan regions) 7. Pavements and Infrastructure 8. Transportation finance and policy research 10

11 Next-Gen travel forecasting models for Indian cities Decision-making at various agencies needs tools to predict urban activity and travel under alternative scenarios What happens to transit ridership, revenue, and traffic congestion under the following scenarios that attempt to make public transit attractive? Changes in fare structure Increase in transit service along a route or a feeder line Expansion of the Metro network and connectivity What is the influence of emerging travel modes and technology? Shared mobility services Information and communication technology (ICT) How does travel pricing or rationing change travel patterns? Peak-period congestion pricing, cordon pricing Odd-even and other rationing schemes, parking regulations How best to formulate travel demand management schemes Telecommuting, Flexible work hours, 11

12 Travel Demand Model Agent-based travel demand forecasting models Study region s socio-demographic and economic characteristics, land use patterns, transportation system characteristics Simulation of daily activities and travel of all agents/individuals in the region Complete (simulated) daily activity-travel pattern for all individuals in the study area Transportation planning and engineering applications System performance measures Traffic volumes & congestion Travel speeds & travel times Transit ridership & revenue Vehicle kilometers traveled Vehicular emissions 12

13 Features required of such models for Indian cities Strong public transportation model components Traveler choices in public transportation systems Static and flexible transit network structures Intermodal connectivity The ability to provide forecasts relevant to the strategies of Responsive to short-term, operational improvements Prediction of network performance (e.g., congestion) at a fine resolution of time and space Other travel modes Non-motorized travel and intermediate modes Emerging modes of travel (e.g., shared mobility) Influence of information and communication technology (ICT) Influence of traveler information and internet connectivity in vehicles Consideration of travel demand management strategies Pricing and rationing Telecommuting and flexible work policies 13

14 Features desired of such models for Indian cities Multi-scale approach for connectivity with other model systems Long-term residential and employment (land-use) choice models Microscopic traffic flow simulators Traffic pollutant emission and dispersion models Energy consumption and urban metabolism models 14

15 Application-1 Spatial-temporal distribution of population activity 12 am 1 am 2 am 3 am 4 am 5 am 6 am 7 am 8 am 9 am 10 am 11 am 12 pm 1 pm 2 pm 3 pm 4 pm 5 pm 6 pm 7 pm 8 pm 9 pm 10 pm 11 pm MINSQMI Diurnal cycle of blockgroup-level population-time density in Tampa, Florida Source: Gurram,S., A. Stuart, A.R. Pinjari (2014). Air Quality, Atmosphere & Health, 8(1), ,000 92, , , , , ,000 2,410,000 person-minutes/sq. mile 15

16 Application-2 Spatial-temporal distribution of network performance Bi-hourly varying a) passenger car volumes and b) travel speeds in Tampa. Source: Gurram, S, A. Pinjari, and A. Stuart, High Resolution Simulation of Urban Travel and Network Performance for Estimating Mobile Source Emissions: Application to Tampa, Florida. The Multi-Agent Transport Simulation (MATSim). Chapter 74, pp , eds. Andreas Horni, Kai Nagel, and Kay Axhausen. 16

17 Application-3 Spatial-temporal distribution of air pollution 12 am 1 am 2 am 3 am 4 am 5 am 6 am 7 am 8 am 9 am 10 am 11 am 12 pm 1 pm 2 pm 3 pm 4 pm 5 pm 6 pm 7 pm 8 pm 9 pm 10 pm 11 pm µg/m 3 Diurnal cycle of ambient NO X concentrations in Tampa, Florida Source: Gurram,S., A. Stuart, A.R. Pinjari (2014). Air Quality, Atmosphere & Health, 8(1),

18 Application-4 San Francisco congestion pricing study Which trips to toll and where? Congested Transit Segment (travel speed below 8 mph) Congested Auto Segment (highway speed below 30 mph road speed below 10 mph) Source: San Francisco County Transportation Authority (SFCTA)

19 Agent-based Travel Forecasting Models in the U.S. and other countries (as of 2011) Seattle Portland Oregon Sacramento Lake Tahoe San Francisco Bay Area San Joaquin Valley Denver Chicago Columbus Ohio Burlington NY Philadelphia Other Cities/Countries Singapore London Netherlands Belgium Australia Los Angeles Phoenix San Diego In application / Model development complete Under development Dallas Houston Atlanta Jacksonville Tampa Miami Ongoing research in India Bengaluru Calicut Chennai Mumbai

20 R&D needed for developing such tools for our cities Improved understanding of travel behavior in our cities with High diversity of population characteristics A multitude of travel modes (including demand-responsive services) Technology and information-intensive environments A combination of long-term infrastructure investments and short-term demand management strategies Development of credible modeling tools in a data-poor environment Infer mobility patterns and travel demand using emerging data sources (sensors, mobile phones, GPS, ) Urban mobility sensing (data collection) using technology Alternative predictive modeling approaches (Machine learning, etc.) Computational and software development Evaluation of uncertainty in data, models, and predictions 20

21 Communication and sensing enabled transportation operation and planning Wireless communication Vehicle-to-vehicle Vehicle-to-other users In-vehicle communication Vehicle-to-Infrastructure Source : Mobility sensing using technology Wireless sensor networks Mobile internet and IoT Cameras High-precision location devices Citizen participation Source: U.S. Department of Transportation 21

22 Communication and sensing enabled transportation operation and planning Planning, operation, and safety of transportation systems can be improved significantly using Sensor networks Communication technology Computational and Data sciences (Urban computing) A multitude of smart-city applications Safety: Collision avoidance and alert systems Real-time control of traffic facilities (actuated and coordinated traffic signals) High accuracy location and mapping (for transit information systems) In-vehicle connectivity (for encouraging transit usage by choice riders) Positive modification of driver behavior (to reduce aggressive driving) Encouragement of citizen participation in system sensing and control Research needed Mobile networks and control in dense and diverse traffic settings Automated sensing of mobility and system performance Analytics of the mega data streams 22

23 Research questions in the context of emerging data New data sources Data from travelers, vehicles, and infrastructure with millions of sensors Social networks, crowdsourcing, web-crawlers, admin records, Research questions Fusion of alternative data sources (with partial information) Construction of complete travel patterns with data from multiple sources Linking with traditional survey data and administrative records (eg. census) Data quality, missing data, self-selection and endogeneity issues Statistical inference The role of theory in a data-driven environment 23

24 Research to improve the understanding and control of traffic networks with heterogeneous traffic streams Improved representation of traffic flows in Indian cities Heterogeneity in traffic streams Multiple vehicle types Longitudinal and lateral movements Absence of lane-discipline User behavior At mid-blocks At intersections and other conflict points Pedestrian facilities Area-wide and network-wide behaviors Actuated and real-time traffic control (signal timing/optimization,...) Single location Corridor Network wide (or Area-wide Traffic Control System) Development of analytical models and micro-simulation tools 24

25 Interactions with the stakeholders of CiSTUP Inputs on intended uses of the modeling tools Measurement of travel demand and system performance (data) Model testing and evaluation Urban simulation software tools Technical-training and technology-transfer 25

26 Thank You 26