Artificial Intelligence To Control Traffic Light System. This new AI-controlled traffic light system could end traffic jams for good. City management could change radically with this automated system.
Those who drive will know all too well, especially on city journeys, how tiresome driving can be as you negotiate vehicles, pedestrians, and the traffic lights that populate the streets. Today, most of the world’s traffic signals are regulated by computer systems, but we still have no way of making artificial intelligence capable of detecting key points at rush hour and clearing traffic congestion. Or perhaps we are already preparing for that?
AI researchers have devised a system that can analyze traffic and manage traffic lights accordingly.
A team from Aston University has published research related to an artificial intelligence system that would allow decongesting city streets. This system is able to read information obtained from live cameras and manage traffic lights to keep traffic flowing. To do this, it uses a type of deep learning called reinforcement, with which the program understands that it is not executing the tasks correctly and tries alternatives.
Traffic management in the city could be radically different with the use of AI systems.
The British university has obtained some enlightening data regarding congestion in urban areas, for example, the British citizen spends, on average, around 115 hours in traffic jams and spends an average of 894 pounds, 1,049 euros at the exchange rate, in lost gasoline during those moments. The researchers decided, given the complication of the matter, to build their own simulator, called Traffic 3D, to begin training their AI. The surprise came when they used what they learned at a real intersection, being able to adapt perfectly to the given traffic situation.
It seems that the era where traffic lights are controlled by computer hardware and rely on quasi-cyclic circuits has come to an end, or so says George Vogiatzis, Professor of Computer Science at Aston University:
The reason we have based this program on learned behaviors is that in this way it can understand situations that it has not experienced explicitly. We have tested the system with a physical obstacle as the cause of the traffic jam, rather than the usual light changes at traffic lights, and the system has passed with flying colors. As long as there is a causal link, the computer will be able to decipher the cause. This is a really powerful system.
¡Stay tuned! Go and visit our blog. We post new articles every week.