Skip to main content
eScholarship
Open Access Publications from the University of California

The Institute of Transportation Studies at UC Berkeley has supported transportation research at the University of California since 1948. About 50 faculty members, 50 staff researchers and more than 100 graduate students take part in this multidisciplinary program, which receives roughly $40 million in research funding on average each year. Alexandre Bayen, Professor of Civil and Environmental Engineering and Professor of Electrical Engineering and Computer Science, is its director.

Cover page of A Revealed Preference Methodology to Evaluate Regret Minimization with Challenging Choice Sets: A Wildfire Evacuation Case Study

A Revealed Preference Methodology to Evaluate Regret Minimization with Challenging Choice Sets: A Wildfire Evacuation Case Study

(2020)

Regret is often experienced for difficult, important, and accountable choices. Consequently, we hypothesize that random regret minimization (RRM) may better describe evacuation behavior than traditional random utility maximization (RUM). However, in many travel related contexts, such as evacuation departure timing, specifying choice sets can be challenging due to unknown attribute levels and near-endless alternatives, for example. This has implications especially for estimating RRM models, which calculates attribute-level regret via pairwise comparison of attributes across all alternatives in the set. While stated preference (SP) surveys solve such choice set problems, revealed preference (RP) surveys collect actual behavior and incorporate situational and personal constraints, which impact rare choice contexts (e.g., evacuations). Consequently, we designed an RP survey for RRM (and RUM) in an evacuation context, which we distributed from March to July 2018 to individuals impacted by the 2017 December Southern California Wildfires (n=226). While we hypothesized that RRM would outperform RUM for evacuation choices, this hypothesis was not supported by our data. We explain how this is partly the result of insufficient attribute-level variation across alternatives, which leads to difficulties in distinguishing non-linear regret from linear utility. We found weak regret aversion for some attributes, and we identified weak class-specific regret for route and mode choice through a mixed-decision rule latent class choice model, suggesting that RRM for evacuations may yet prove fruitful. We derive methodological implications beyond the present context toward other RP studies involving challenging choice sets and/or limited attribute variability.

Cover page of Micromobility evolution and expansion: Understanding how docked and dockless bikesharing models complement and compete – A case study of San Francisco

Micromobility evolution and expansion: Understanding how docked and dockless bikesharing models complement and compete – A case study of San Francisco

(2020)

Shared micromobility – the shared use of bicycles, scooters, or other low-speed modes – is an innovative transportation strategy growing across the United States that includes various service models such as docked, dockless, and e-bike service models. This research focuses on understanding how docked bikesharing and dockless e-bikesharing models complement and compete with respect to user travel behaviors. To inform our analysis, we used two datasets from February 2018 of Ford GoBike (docked) and JUMP (dockless electric) bikesharing trips in San Francisco. We employed three methodological approaches: 1) travel behavior analysis, 2) discrete choice analysis with a destination choice model, and 3) geospatial suitability analysis based on the Spatial Temporal Economic Physiological Social (STEPS) to Transportation Equity framework. We found that dockless e-bikesharing trips were longer in distance and duration than docked trips. The average JUMP trip was about a third longer in distance and about twice as long in duration than the average GoBike trip. JUMP users were far less sensitive to estimated total elevation gain than were GoBike users, making trips with total elevation gain about three times larger than those of GoBike users, on average. The JUMP system achieved greater usage rates than GoBike, with 0.8 more daily trips per bike and 2.3 more miles traveled on each bike per day, on average. The destination choice model results suggest that JUMP users traveled to lower-density destinations, and GoBike users were largely traveling to dense employment areas. Bike rack density was a significant positive factor for JUMP users. The location of GoBike docking stations may attract users and/or be well-placed to the destination preferences of users. The STEPS-based bikeability analysis revealed opportunities for the expansion of both bikesharing systems in areas of the city where high-job density and bike facility availability converge with older resident populations.

Cover page of City-wide traffic control: modeling impacts of cordon queues

City-wide traffic control: modeling impacts of cordon queues

(2018)

Optimal cordon-metering rates are obtained using Macroscopic Fundamental Diagrams in combination with flow conservation laws. A model-predictive control algorithm is also used so that time-varying metering rates are generated based on their forecasted impacts. Our scalable algorithm can do this for an arbitrary number of cordoned neighborhoods within a city. Unlike its predecessors, the proposed model accounts for the constraining effects that cordon queues impose on a neighborhood's circulating traffic. It does so at every time step by approximating a neighborhood's street space occupied by cordon queues, and re-scaling the MFD downward to describe the state of circulating traffic that results. The model is also unique in that it differentiates between saturated and under-saturated cordon-metering operations. Computer simulations show that these enhancements can substantially improve the predictions of both, the trip completion rates in a neighborhood and the rates that vehicles cross metered cordons. Optimal metering policies generated as a result are similarly shown to do a better job in reducing the Vehicle Hours Traveled in a city. The VHT reductions stemming from the proposed model and from its predecessors differed by as much as 18%.