Connected Corridors: New Methods for Urban Traffic Management

January 23, 2017

BIDS will host a tea event on January 24, 2017, with guests Brian Peterson and Anthony Patire of the Institute of Transportation Studies’ Connected Corridors program. They will be discussing a pilot program in partnership with Caltrans to improve urban transportation corridor management on the I-210 corridor in Southern California.

Caltrans and the ITS’ Partners for Advanced Transportation Technology (PATH) developed the Connected Corridors program in 2012. Since then, Connected Corridors has been working with Caltrans as well as Los Angeles County; the cities of Pasadena, Arcadia, Monrovia, and Duarte; and regional mass transit agencies within the metropolitan area to pilot an integrated corridor management system being designed and largely built here at UC Berkeley. The pilot envisions connecting traffic sensing, traffic control, and estimation and predictive traffic modeling techniques in unique ways to assist people and institutions that direct and manage traffic conditions in an effort to improve travel conditions for those living and traveling within the corridor.   

The system being developed by Connected Corridors will use traffic modeling software to provide traffic state estimation in real time. When an incident or event occurs within the corridor boundaries, the system will develop multiple response plans and use the current estimated state along with predictive modeling to recommend one or more response plans for deployment. It will be capable of implementing a response plan, once selected, by sending commands to connected traffic control systems. The response plans will include recommended traffic reroutes, intersection- and ramp-control strategy changes, messaging and traveler communication strategies, and potential multi-modal traveler options to optimize usage of the corridor’s available capacity.

The system will collect real-time traffic data from both freeway and arterial roads as well as probe data from GPS sensing devices to feed the estimation models. It will also collect data from traffic controls, such as intersection signals, ramp meters, and dynamic message signs, to maintain the real-time state within the estimation and predictive models.

This approach presents unique problems and opportunities for data science. Current modeling challenges involve the prediction of transportation network demand, traveler routing, and origin/destination pairs for travelers. Understanding driver behavior and using data to tune and calibrate the models are of particular interest as well.

Never before has so much transportation-related data been available, and this trend of increasing data is expected to continue. This data is critical for understanding not only transportation issues but also issues related to energy demand and usage, vehicle automation, and urban planning.

Caltrans hopes to implement Connected Corridors in multiple corridors throughout the state within the next 10 years.

On the January 24, Brian Peterson, the development manager for the team responsible for designing and developing the system will discuss how the system is being designed as well as opportunities for those interested in the project. Anthony Patire, research and development engineer, will discuss the traffic modeling challenges and how data science may contribute to the solution.

Those looking for more information may visit the Connected Corridors website.

For some information regarding recent challenges on the I-210 corridor, please follow these links: