With artificial intelligence, traffic jams should be drastically reduced.
Artificial intelligence can do a lot for traffic, not only in cars, but also in them infrastructure benefits from it. This is the result of a study by researchers from Aston University in Great Britain. It is said to be superior to other methods.
The number of hours people spend in the traffic jam spend is high. Not only is this bad for the nerves, the pollution and costs are high. The constant starting is particularly fuel guzzling. Traffic jams are also a climate factor in the USA and other countries.
Therefore, researchers at Aston University have developed solutions based on a artificial intelligence function. The system can work with live recordings and adapt the traffic lights to the traffic. Traffic light timing is one of the main causes of traffic jams.
The team builds a photorealistic traffic simulator, Traffic 3D, to train the AI. This also included dealing with weather data. At a real intersection, the system continued to learn and was able to reduce traffic jams. The system works with “Deep Reinforcement Learning”. It is rewarded when a car crosses the intersection and receives negative reinforcement when a car has to wait.
The system also understands other obstacles that lead to traffic jams and should also be able to deal with them. It independently learns how to make traffic flow more smoothly. Also the preference of emergency vehicles is integrated.