A new study on the potential benefits of autonomous vehicles describes how cooperative robotic cars could increase the capacity of our highways by 273 percent, according to NBC News:
The paper is being presented this week at an Institute of Electrical and Electronics Engineers (IEEE) conference on vehicular technology. Its author, Columbia University’s Patcharinee Tientrakool, wrote her dissertation on a method for cars to communicate safely and reliably that she calls “reliable neighborcast protocol,” or RNP.
Research in self-driving vehicles has naturally focused on how to make the car imitate an intelligent driver: recognizing and navigating obstacles, reading signs and performing other common tasks. If there were only going to be a single such vehicle on the road, surrounded by human-guided cars, then that’s the most important thing to perfect. But what if nearly every car on the road is a robo-car?
Tientrakool’s paper looks at the difference in efficiency between when autonomous vehicles don’t communicate and when they act as a team. She concludes that cars simply managing their own speed would increase efficiency by an appreciable 43 percent, but if they were working together, that number jumps to a staggering 273 percent.
Cars would achieve this by banding together into groups, driving much closer to one another than humans do, and working out the best possible solution to things like merging and changes in traffic. Another study (PDF) by Steven Shladover at UC Berkeley’s Institute of Transportation Studies makes similar predictions based on this “platooning” of vehicles.
Shladover found that people tend to trail about a second and a half behind the car they’re following — a reasonable distance that reflects the speed at which we can react to things like sudden braking. Using what they call cooperative automatic cruise control (essentially the autonomous cars collaborating on speed and formation), they can reduce the time between cars to less than half that. That means, of course, that twice as many cars could conceivably fit in the same space.
It’s still very speculative, since robo-cars like Google’s aren’t yet road-legal — although they are approaching that status in Nevada and California, among other places. And before entire highways are populated with them, it’s likely there will be single lanes dedicated to platooning vehicles.
However far off it may be, transit officials are probably impatient for that day. Doubling or tripling the capacity of the highways without spending billions on new lanes and other considerations is a city planner’s dream — to say nothing of the commuters, whose trip to work would be shorter, safer, and easier.
Five strategies to facilitate the paradigm shift in transportation
Stop investing in roads to build new neighborhoods that cause other neighborhoods to flood
Houston's mean streets: Our city's road design is killing people