Principal Investigator

Siyuan Liu

Status

Completed

Start Date

Jan. 1, 2015

End Date

Dec. 31, 2015

Project Type

Research Advanced

Grant Program

MAP-21 TSET National (2013 - 2018)

Grant Cycle

2015 TSET UTC - National

Visibility

Public

Abstract

Unexpected traffic incidents and the consequent congestion have significant negative impact on the economy and the quality of people’s lives. Transportation systems can be improved by technologies of autonomous and connected vehicles, which allow vehicular control and communications, and enable effective information dissemination among vehicles. Adaptive vehicle routing is one of the most promising applications of such technologies to achieve better safety and efficiency of transportation systems. Processing and learning real-time traffic information from probe vehicles and social media (e.g., Twitter), vehicles can be routed safely and efficiently in a non-myopic way. Adaptive routing of autonomous connected vehicles has two main goals, 1) at the individual vehicle level, how to provide a routing strategy for each individual vehicle to minimize accident risk, crash rate, and travel time; 2) at the system management level, how to provide globally optimal routes for all the vehicles over time so that the total system crash rate and travel cost can be minimized. We first develop anomaly detection methods to capture the abnormal speed changes, estimate the travel risk and broadcast the warning to neighbor vehicles. Second we propose a Gaussian Process Dynamic Congestion Model that can effectively characterize both the dynamics and the uncertainty of congestion conditions and incidents in traffic. Using this congestion model, we develop efficient algorithms for non-myopic adaptive routing to minimize the total risk and cost of all vehicles in the system. Our approach will be validated by traffic data from two major US cities (Philadelphia and Washington D.C.) and two large Asian cities (Shenzhen and Shanghai in China). The output will bring immediate impact and benefit to Microsoft CityNext and GeoLife, Google driverless car, community-based traffic and navigation applications (e.g., Waze), ridesharing service (e.g., Uber). We will actively engage public agencies (e.g., PennDOT, NSF) for future research support.    

Description

-

Timeline

-

Strategic Description / RD&T

Deployment Plan

-

Expected Outcomes/Impacts

-

Expected Outputs

TRID

Individuals Involved

Email	Name	Affiliation	Role	Position
siyuan@cmu.edu	Liu, Siyuan	Carnegie Mellon University	PI	Faculty - Research/Systems

Budget

Amount of UTC Funds Awarded

$80518.00

Total Project Budget (from all funding sources)

$80518.00

Documents

Type	Name	Uploaded

Match Sources

No match sources!

Partners

No partners!