The hell of traffic jams: soon an old memory thanks to AI – Science and life

Traffic jams: how is road traffic regulated?

Currently, road traffic is regulated by a classic signaling system, the “lights” as we call them. However, this technology is based on very sub-optimal operation. Especially during variations in the number of people on the roads. Indeed, the current system is not capable of detecting traffic jams. So he can’t act on it to fix the problem. It is programmed and requires a human hand to do it.

Read also: An artificial intelligence capable of creating and operating a computer! An artificial intelligence that controls another artificial intelligence? We are already there !

Adaptive signaling to avoid the creation of traffic jams

It is from this postulate that the researchers decided to use a precise technology. This is already present in video games, robotics or the optimization of transport infrastructure. This technology is called “Deep Reinforcement Learning” (DRL). It allows the software to “learn”, the different types of situations that may arise so that it can then recognize them. In this case, the authors programmed the software by rewarding it when it made traffic flow and vice versa.

This adaptive signaling method promises to detect and analyze all factors capable of altering circulation. The detection will go so far as to identify the presence of many vehicles. But also that of cyclists or pedestrians who circulate. It will also be able to take into account weather conditions that may affect traffic. All these elements are fluctuating. Their detection will make it possible to vary the signals according to the crowd or the weather.

Read also: Politics: here is what AI still lacks to stand for election It is an illusion for some, a cultural suicide for others, while some still see it as the ultimate solution to make humanity prosper!

The detection of the various elements is done through the use of cameras. The latter capture and then process the information, using an intelligent algorithm. “The process continues to determine an optimal signal in order to evacuate traffic through the intersection in the most efficient way possible” comment the authors of the study.

At this stage, the method has very robust detection capabilities. The authors have the ambition that their system be deployed globally.

Read also: An old mathematical paradox pushes artificial intelligence to question! AI may even come up against some limitations than a 100-year-old mathematical paradox

Source: Garg et al., Fully-Autonomous, Vision-based Traffic Signal Control: from Simulation to Reality, Aston University, 2022,

We wish to give thanks to the author of this write-up for this outstanding web content

The hell of traffic jams: soon an old memory thanks to AI – Science and life

We have our social media pages here and other pages on related topics here.