MIT Portugal Collaboration Successfully Tests Software That Could Improve Traffic Management

Thursday, 22 January 2009

As part of the MIT Portugal Program, researchers from MIT and Portugal have made progress in developing traffic simulation software that could make it easier for traffic managers to analyze road conditions and ease congestion in real time.

DynaMIT, a dynamic simulator based on model-based data fusion software, enables traffic engineers to integrate and analyze data from the dizzying array of information sources that have become available as a result of the proliferation of information and communications technologies—including road sensors, electronic toll collection devices, automatic video processing, global positioning systems, mobile sensor networks and smart phones.

Brisa's Traffic Management Center

The software, which was developed at MIT’s Intelligent Transportation Systems Lab (ITS), recently underwent a successful demonstration at Brisa, Portugal’s largest toll road management company and an industry partner on the project.  The demonstration involved data collected on a 25-km inter-urban section of highway between Lisbon and neighboring Cascais—a stretch that’s often congested with commuters heading to Lisbon on weekday mornings. It was conducted by Jorge Lopes, a PhD student in MIT Portugal’s Transportation Systems focus area and a Brisa employee.

Lopes used the DynaMIT software to predict travel time, speed, traffic intensity and density—information that Brisa’s traffic management center may be able to use in coming years to predict recurrent and incident-based traffic conditions, as well as  to devise real-time traffic management support strategies. In so doing, he demonstrated Dyna-MIT’s usefulness as a data fusion “engine.”

Lopes is currently designing an innovative open framework architecture for the dynamic integration of modeling and simulation services into traffic management and information systems. Translation: he is attempting to create a kind of “transportation time machine”—a means of using real-time traffic data to estimate and predict short- and medium-term traffic conditions.

The data fusion initiative is part of the CityMotion project, which is coordinated through MIT Portugal. CityMotion researchers are developing a knowledge infrastructure, computational models and user applications that draw from and integrate real-time, transportation-related information; one of the project’s aims is to improve the traveler’s experience in urban transportation systems, including those in Portugal.

In addition to the ITS Lab, the MIT team involved in CityMotion includes researchers from the SENSEable City Lab and the Department of Urban Studies and Planning. The Portuguese team includes researchers from the Faculdade de Ciências e Tecnologia-Universidade de Coimbra (FCT/UC), the Faculdade de Engenharia da Universidade do Porto (FEUP), the Instituto Superior Técnico/ Universidade Técnica de Lisboa (IST-UTL), and several industry partners. MIT faculty include Moshe Ben-Akiva, Assaf Biderman, Carlo Ratti and Christopher Zegras. Portugal faculty include João Abreu e Silva (IST), Carlos Lisboa Bento (Coimbra), Francisco C. Pereira (Coimbra) and Teresa Galvão (Porto).