Researchers at MIT have demonstrated a simple system that uses virtual tokens tracked over wireless networks to detect traffic congestion and reroute traffic.
I guess it’s appropriate that RoadRunner was developed within spitting distance of downtown Boston, since few places on Earth have as many problems with traffic congestion.
Researchers at MIT’s Future Urban Mobility Group have demonstrated a very simple system that uses virtual tokens tracked over wireless networks to detect traffic congestion and, when congestion is present, to reroute traffic.
Developed by Jason Gao, a graduate student in computer science and electrical engineering, along with Professor Li-Shiuan Peh, the system – called RoadRunner – doles out a token to every vehicle entering a specified zone. When the number of tokens inside the zone exceeds a preset limit, RoadRunner stops issuing tokens and instead sends out audio alerts with alternative routing to vehicles approaching the zone.
Unfortunately, RoadRunner can’t be effectively deployed yet because it depends on the 802.11p wireless protocol, which has not yet been delivered in many commercial products. According to Gao, the RoadRunner team chose 802.11p because it offers higher power transmission than other standards and is more resistant to noise and interference. For the same reasons, 802.11p is also being employed by collision-avoidance technologies under development and by developers of driverless cars.
“We have two scenarios that we are envisioning in the future,” said Gao. “One is that in the near term cars will actually come with 802.11p built-in.” The other scenario, he said, is that 802.11p may in the future be supported in cell phones. “We are working on shrinking the 802.11p components to get them small enough and low-power enough that they could be integrated directly onto phones in the future,” said Gao.
For the RoadRunner demonstration, however, the team installed off-the-shelf 802.11p radios that are roughly the size of a hardcover book.
The key advantage of RoadRunner, besides its being automated, is that it’s inexpensive, since it doesn’t require installing roadway sensor or transponders in vehicles. Once 802.11p is deployed – whether in cars or cellphones – all that is required is installation of software on the client. “One of the primary goals of RoadRunner was to find a way to deploy a system without incurring a huge amount of cost,” said Gao.
The program the MIT team developed for managing RoadRunner, Gao said, can be used by traffic managers to establish congestion zones simply by drawing polygons. And, in addition to setting the number of tokens allowed in a zone, traffic managers can also specify how far in advance of reaching a congested area vehicles are alerted. In the demonstrations, the system was set to alert vehicles two intersections before the congestion.
Unfortunately, there’s nothing to stop drivers from ignoring the rerouting instructions. And my experiences driving in Boston lead me to suspect that RoadRunner’s commands are more likely to be ignored than honored.
With that in mind, future traffic managers deploying RoadRunner will have to have some way of enforcing its directions. One way is to identify cars in congested areas that don’t have tokens and have traffic fines automatically sent to the registered owner, just as many toll bridges currently send out bills for autos that crossed without a token.
The second option is, in my view, more attractive: Get rid of the drivers. Picture driverless cars that are automatically rerouted to their destinations when RoadRunner sends congestion alerts. Some people might feel controlled by such a system. For my part, it would mean a little more time to finish reading the newspaper. That’s a lot better than sitting in frustration behind the wheel stuck in gridlock.