Even a few self-driving cars significantly speed up traffic in traffic jams

A team of researchers at the University of Illinois found that adding just a couple of self-driving cars to a traffic jam markedly increases the average flow rate. In an experiment carried out on a test site, 5% of cars with an autopilot were "introduced" into the conventional ring traffic. As a result, this gave a decrease in the number of constant braking-accelerations of the participants in the flow and saving up to 40% of fuel.

Unfortunately, traffic jams have become a part of modern urban life. Their main reason is, of course, too many cars on the roads, as well as accidents, outdated road infrastructure, road repairs. The human factor also plays a role, when each driver brakes and accelerates in his own manner, which generates a jerky wave-like movement.

Scientists decided to approach the problem using the latest advances in traffic monitoring. If earlier cameras were mainly used for this, now the trend is GPS data coming from drivers' mobile devices. By processing data on the location, speed and direction of movement of vehicles, the corresponding algorithm calculates the optimal constant speed, allowing you to move without sudden changes in speed.

The next step, a team of scientists from Illinois is going to test the effectiveness of self-driving cars in tighter traffic jams and difficult conditions, including unexpected lane changes between the rows of other road users. In the future, experiments with the autopilot are planned to be carried out on the city streets.