Title: Strategic Route Choice Behavior

Speaker: Dr. Song Gao (Assistant Professor of Civil Engineering, University of Massachusetts, Amherst）

Venue: Shun De Building, Room 510

Time: 10:00AM-12:00NOON, Wednesday, December 15, 2010

Host: Dr. Hai Jiang

Abstract: In this talk a series of studies of strategic route choice behavior under real-time information are discussed.  Conventional route choice models under real-time information consider only the reactive responses to information on the spot, and ignore the fact that travelers can strategically plan ahead for information that will be available in the future.  Algorithmic studies of finding optimal strategies are abundant in the literature, however empirical research of strategic route choice behavior is a new area.

We present three studies on this topic where human subjects are recruited and their route choices solicited in hypothetical networks with risky travel times and real-time information.   The first study uses a web-based interactive map, the second a full-scale driving simulator, and the third again a web-based interactive map but the network is congestible.  The general conclusion from these studies is that a non-negligible portion of the subjects are strategic, and thus a route choice model under real-time information should be general enough to include both strategic and non-strategic behavior.  It is also found that network complexity adversely affects subjects' strategic thinking ability, and individual strategic route choices collectively could save system-wide travel cost under proper network settings.

A latent-class discrete choice model based on the cumulative prospect theory (CPT) is developed and estimated using the stated preferences (SP) data from the first study.  The two classes are strategic and non-strategic behavior.  Risk attitudes that change with outcome probabilities are well captured by the CPT, which is a more general theory than the expected utility theory to model decisions under risk.

Bio: Dr. Song Gao has been an assistant professor of Civil and Environmental Engineering at the University of Massachusetts Amherst since 2007. Her research focuses on modeling and optimization of transportation systems, including travel behavior models, network optimization algorithms and equilibrium models. Prior to joining the University of Massachusetts Amherst, Dr. Gao worked as a transportation engineer at Caliper Corporation, Newton, MA for three years. She received Honorable Mention (second place) in the INFORMS Transportation Science and Logistics Dissertation Prize Competition in 2005.  Dr. Gao earned her Ph.D. and M.S. in Transportation from MIT in 2005 and 2002 respectively, and her B.S. in Civil Engineering from Tsinghua University in 1999.