Mohan Trivedi
THE BIGGEST PROJECT now underway in his lab could literally save
lives and potentially billions of dollars. And for Mohan Trivedi and
his dynamic team of students, the importance of the work they are
doing generates an enthusiasm that is positively palpable. "It is
very exciting to be working on projects where we anticipate very
clear benefits for society not so far off in the future," says ECE
professor Trivedi.
Trivedi, who directs UCSD's Computer Vision and Robotics Research
(CVRR) laboratory, is referring to a project sponsored by Caltrans
and the UC Office of the President’s Digital Media Innovation
initiative. The program’s goal: develop a system to detect and
analyze traffic accidents using cameras and other sensors deployed
along roadways, and build a robotic emergency-response team that can
swing into action long before the police or other emergency
personnel can reach the scene. "If you just shave five minutes off
the response time to accidents involving potentially fatal injuries,
you’d save hundreds of lives," says Trivedi. "It would also reduce
the amount of traffic congestion caused by accidents and other
so-called traffic incidents, and Caltrans estimates that those
incidents cost California alone an estimated $10 billion a year."
Trivedi has been at UCSD since 1995, but his work in the
transportation sector dates back to before 1979, when he got his
Ph.D. in electrical engineering from Utah State. In the 1980s he
taught at Louisiana State University and the University of
Tennessee; at both schools, he consulted with Japanese railways on
robotics to detect and repair rails.
Since transportation is one of the core "layers" addressed by the
new California Institute for Telecommunications and Information
Technology, Trivedi is heavily involved as Cal-(IT)²'s layer leader
for UCSD. In that capacity, he's teaming up with counterparts at UC
Irvine to forge a combined research agenda that will speed up and
enhance the new technology each campus might otherwise develop on
its own.
On Oct. 19, Trivedi hosted the first Cal-(IT)² workshop on
intelligent transportation ("smart roads") and telematics ("smart
cars"), with more than 100 attendees from auto companies, other
industrial partners, Caltrans, research funding agencies as well as
UCI, UCSD and affiliated  organizations. He took the opportunity to
introduce workshop participants to MIA (picture at right)—the
"mobile interactive avatar" built in his lab. The mobile robot is
equipped with a two-way video interface able to beam pictures back
to a control room from an accident scene via high-speed wireless
Internet lines deployed on the UCSD campus. Together with other
wireless equipment now under development in Trivedi's lab and
elsewhere, MIA could one day become the cornerstone of an elaborate
network of cameras, sensors and robots along highways to make
"intelligent" roadways a reality.
Meanwhile, with all the research he's now overseeing, not to
mention almost daily contact with companies and organizations
interested in the outcome of his projects, Trivedi still finds time
to teach. Undergraduates get the benefit of his expertise in machine
intelligence, while the outgoing professor gives his graduate
classes insight into computer vision. Says Trivedi: "Teaching and
research are intertwined." | 
