A modeling effort shows intelligent trip planners can optimize multi-modal travel and reduce congestion-induced delay by 14 to 20 percent.

Researchers conducted a traveler survey to model the impacts of intelligent trip planners on mode use and network congestion in Los Angeles, California.

Date Posted
10/15/2019
Identifier
2019-B01407
TwitterLinkedInFacebook

On Influencing Individual Behavior for Reducing Transportation Energy Expenditure in a Large Population

Summary Information

Artificial Intelligence (AI) technology has great, but largely unrealized potential to transform our transportation systems. This research quantifies the potential benefits of AI using high-fidelity simulations and survey data.



Methods



Researchers from the Palo Alto Research Center and the Virginia Tech Transportation Institute developed a novel transportation routing and trip planner, the Collaborative Optimization and Planning for Transportation Energy Reduction (COPTER) application, using AI and machine learning. In theory this trip planner will intelligently and dynamically plan multi-modal trips for users while accounting for a wide variety of factors including travel context, directness of route etc.



To test the potential impact of COPTER on travel behavior researchers conducted a step-two test. First, researchers surveyed regular drivers in Los Angeles County about their willingness to switch from driving to a multi-modal trip. To do this the survey asked users about their daily travel habits and then this information was then fed into the COPTER system to plan a multi-modal route for the travelers. Travelers were then asked how likely they would be to use the multi-modal route compared to their regular driving route on. Second, researchers "[used] the likelihood of adoption from the [survey]" to inform a microsimulation of Los Angeles during "peak traffic periods."

Findings

The analysis suggested that intelligent multi-modal trip planners like COPTER can reduce congestion related traffic delay by 14 to 20 percent during realistic peak period conditions and by up to 30 percent under ideal peak period conditions.

On Influencing Individual Behavior for Reducing Transportation Energy Expenditure in a Large Population

On Influencing Individual Behavior for Reducing Transportation Energy Expenditure in a Large Population
Source Publication Date
01/28/2019
Author
Mohan, Shiwali; Frances Yan; Victoria Bellotti; Ahmed Elbery; Hesham Rakha; and Matthew Klenk
Publisher
Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and Society - AIES 19, 2019.
Goal Areas
Results Type
Deployment Locations