VIDEO - Can Traffic Lights Learn to be Smart?

VideoBaher Abdulhai is a professor at the University of Toronto and the director if the Toronto Intelligent Transportation Systems Centre and Testbed. One aspect of his work is artificial intelligence (AI) based street and freeway traffic control, traveler information systems, and emergency evacuation optimization.

This video is a TEDx Talk titled The Traffic Game - Can Traffic Lights Learn to be Smart?

Highlights of the video include:

  • The Pressing Challenge: Immobility!
  • Congestion: Complex System of Systems
  • Congestion: Exponential Crisis
  • Congestion Solutions
  • Converging Realities: Intelligence and Complex Systems
  • Performance: Average Delay Reduction
  • Multiple MARLIN Agents / Intersections
  • Large Scale Simulation Test in Toronto
  • Taking it to the Streets: Hardware Integration

Excellent presentation.  If this moves effectively from theory to practical applications it could generate very positive benefits for everyone.  Reduced travel times, reduced fuel consumption and pollution/greenhouse gases, fewer collisions, less road rage, etc.   Let's hope their trials in Toronto work out as expected.

However there is also a caveat - the human factor.  As soon as congestion is reduced and traffic opens up there will be some drivers who try to take extra advantage by higher speeds, more frequent lane changes, etc. - because now they see an opportunity to get ahead faster.   so there will undoubtedly some loss of the potential benefits from these drivers.  

In conjunction with intersection cameras and other advanced enforcement these negative tendencies could be reduced.