GM CEO Rick Wackoner delivered a keynote address at the Consumer Electronics Show. He spoke about Boss, but he also discussed a new GM initiative that is of particular interest to me. GM plans to dramatically expand OnStar with vehicle to vehicle communications(V2V):
One of the next big developments in automotive electronics is connecting cars with other cars.Connecting them electronically, that is… to keep them from connecting physically.
In recent years, advances in electronics have allowed for technologies like anti-lock brakes, traction control, electronic stability control, and obstacle detection. At GM, we’re working our way up this advanced technology ladder to help our drivers avoid accidents and improve traffic flow. In fact, we’re now offering a Lane Departure Warning System and a blind-spot alert system on the 2008 Buick Lucerne and 2008 Cadillac STS and DTS. We’re preparing to take it even further through the use of GPS and advanced transponder technology that we believe will revolutionize the driving experience.
We call it V2V, or vehicle-to-vehicle communications.V2V starts with collision avoidance and builds from there. And the key difference between V2V and sensor-based “vision” systems is in the electronic communications. Today’s vision systems send out a signal that determines the speed and location of the vehicle ahead of you, and directs your car accordingly.
It’s excellent technology.
These next-generation systems promise to be considerably better… because they’ll be significantly less costly, yes… but more importantly, because they’ll use transponders to “talk” with other vehicles within a quarter mile of your vehicle. So, if six cars ahead, somebody in a transponder-equipped vehicle steps on the brakes… in your lane or the lanes on either side of you… your transponder will immediately know that, and start slowing down your car before you’re even aware you may need to stop. This type of technology, unheard of 15 years ago, has the potential to minimize traffic jams and, more importantly, greatly reduce highway accidents and fatalities… with minimal or possibly even no roadway infrastructure required.
I think the key to building a scalable computer controlled transportation system is vehicle to vehicle communication. I don't think it will be possible to centrally control millions of vehicles. Each vehicle will need to decide for itself when to speed up, slow down, or turn. It might be possible for a central controller to coordinate traffic, but even that looks too difficult considering the problems facing the much smaller air traffic control system. In order to support each vehicle's decisions, the vehicle will need to communicate heavily with its peers. As Mr. Wagoner notes, it is much easier to get a message from the car in front of you that it has slowed down then to figure that out with heavy duty computer vision.
In the interest of full disclosure, my research focuses on programming languages for peer to peer communication. I may very well be a victim of the old adage "if you have a hammer everything looks like a nail," but I don't think so.
Links: GM Keynote Transcript
