USU Computer Scientists Receive Grant to Decipher the Morning Commute

Getting from A to B should be a straight line but every commuter knows a host of factors can derail the best laid itinerary in the blink of an eye. Utah State University computer scientists are employing the power of observation, probability and computation to give transit managers a leg up on heading off ugly traffic snarls.

Omar U. Florez, a doctoral student in USU’s Department of Computer Science, and his faculty mentor Curtis Dyreson are recipients of a $20,000 IBM Scalable Data Analysis for a Smarter Planet Innovations Award. IBM initiated the university competition fall 2010 to foster design and development of systems and solutions to address issues in such areas as transportation, energy, buildings, water, security and urban life.

“The credit for this award goes to Omar, who is a very smart and independent student,” says Dyreson, assistant professor. “He’s already made several contributions to the state of the art and I anticipate he’ll be a leader in this field in the future.”

The team’s proposal involves streamlining ways of searching and analyzing data from video traffic cameras to determine and predict patterns of commuter behavior. The aim of the project is to provide transit managers with reliable data to guide transportation management decisions, as well as access to real-time information to respond to problem situations.

“We’re examining causal relationships — for example, drivers stop at intersections and pedestrians cross the street,” says Florez, who completed undergraduate studies in systems engineering in his home country of Peru. “We can’t assume that participants in the transportation system will always follow the rules of the road. Transportation systems are very complex and influenced by rapidly changing conditions.”

Because video cameras are already in place in many urban areas, they provide a cost-effective and convenient method of data collection, he says.

“The drawback is that most video streams are massive, so special algorithms are needed to distill usable data,” Dyreson adds. “Omar is interested in solving difficult problems in searching and analyzing video data.”

To do this, Florez is developing a computer program using a probability distribution known as a hierarchical Dirichlet Process to analyze each ‘transaction’ in a given transportation system. He’s developing computer models to discern frequent actions of moving objects and to explain how, with rules, they interact with each other.

“Having this information could be valuable to city leaders, particularly in crisis situations, including the heavy snowfall we’ve recently seen in such cities as Chicago, New York City and Boston,” he says. “An urban center’s transportation system is its lifeblood.”

Florez completed a four-month internship in service science research this past summer at IBM’s Almaden Research Center in California and hopes to pursue a career in the field. During school breaks, he returns to Peru to organize conferences to enable fellow Peruvian graduate students studying throughout the world to present their research to undergraduates.

“It’s a great way to share experiences and encourage undergraduates to pursue similar opportunities in graduate study,” Florez says. “We’re trying to foster a research environment. The conferences literally change people’s lives.”

Little computer science research takes place in Peru, he says.

“We basically consume technology but do not produce it,” Florez says. “I want to eventually return to Peru and change that.”

Related links:

USU Department of Computer Science

USU College of Science

Contact: Omar Florez, omar.florez@usu.edu

Contact: Curtis Dyreson, 435-797-0742, curtis.dyreson@usu.edu

Writer: Mary-Ann Muffoletto, 435-797-3517, maryann.muffoletto@usu.edu