They’ve had their fair share of teething issues, but driverless cars will undoubtedly revolutionize the way we travel and the cities of the future. New research has shown yet another benefit of getting a computer to be your personal chauffeur: way fewer traffic jams and less fuel consumption, even for human-driven cars.
Experiments led by the University Of Illinois at Urbana-Champaign have shown that even a small proportion of autonomous cars on the road could reduce congestion. It could also cut the total fuel consumption of all cars by 40 percent.
"Our experiments show that with as few as 5 percent of vehicles being automated and carefully controlled, we can eliminate stop-and-go waves caused by human driving behavior," lead researcher Daniel B Work, assistant professor at the University of Illinois, said in a statement.
The preprint study, published online on arXiv, looked at how the autonomous vehicles affected the stop-and-go waves in traffic, which is known to cause traffic jams. In a series of tests, they put 20 or more human-driver cars on a circular single-lane track along with one autonomous vehicle. They found that just the one single driverless car eliminated most of the stop-and-go driving of the human drivers. Less stopping and starting was equally as good for fuel efficiency.
Egos aside for a moment, humans are not great at driving cars, even on a simple circular track such as this. We get distracted really easily, we can be impatient, and we don’t have amazing reaction times. As a result, humans are very uncoordinated en masse. Just one car braking slightly can cause an “accordion effect” on all the cars behind it. This eventually builds up and causes a “phantom traffic jam", a phenomenon you often see on highways when there doesn’t appear to be a good reason for a traffic jam.
You can learn more about this in the brilliantly concise (and funny) video by CGP Grey below.
On the other hand, driverless cars don't have these problems and can create a kind of super-organized and in-sync system. The surprise is, however, that even just one driverless car among humans also has this effect.
"Before we carried out these experiments, I did not know how straightforward it could be to positively affect the flow of traffic," said computer engineer Jonathan Sprinkle. "I assumed we would need sophisticated control techniques, but what we showed was that controllers which are staples of undergraduate control theory will do the trick."
The research also suggests that as large numbers of autonomous vehicles on the road is still relatively far off, even technology currently available, like adaptive cruise control, could help improve traffic.
The plan is to continue this research by experimenting with denser traffics and more complex roads that could involve even more perils and pitfalls, such as multiple lanes.
3
