Adaptive Traffic Light Signalization

Spring 2012

Torstein Stromme
Stephen Smith
Project description

My role: Support the research team in various ways as needed, with focus on the practical implementation of the system in the lab and in a real traffic signal system pilot test in the field. One likely task will be working on software that will provide a real-time flow of traffic detection events from video camera sensors into the system.

About project: As part of the Traffic21 initiative at CMU, we are investigating the design and application of adaptive traffic signal control strategies for urban road networks. Our research has three broad themes: (1) development of signalization strategies that allow real-time response to shifts in traffic conditions (accidents, traffic dispersal at completion of major events), (2) low-cost deployment of advanced signalization concepts (implying strategies that work well with limiting sensing and methodologies for incremental insertion of adaptive signal technology), and (3) principled coordination of transit systems and personal vehicles through adaptive traffic signal control. We have developed and validated a new decentralized traffic signal control approach using a microscopic simulation model of the Pittsburgh downtown road network as an experimental evaluation testbed, and we are now engaged in carrying out a pilot test of the approach on a 9-intersection grid in the East Liberty area of Pittsburgh.

Return to project list