An artificial intelligence (AI) system developed by researchers from Aston University in the U.K could one day reduce the long wait at traffic lights and, ultimately, congestion.

The AI system, according to its developers, reads and analyzes live traffic camera footage and subsequently adjusts the lights to keep traffic flowing, thereby reducing congestion.

To accomplish this, the system relies on deep reinforcement learning wherein the system tries to course-correct when it is not performing well. This reportedly enables the system to improve as it goes.AI traffic light system could make traffic jams a distant memory. Source: Aston UniversityAI traffic light system could make traffic jams a distant memory. Source: Aston University

To help train the system, the researchers developed a photo-realistic traffic simulator dubbed Traffic 3D.

Dr. Maria Chli, reader in Computer Science at Aston University, explained that they "have set this up as a traffic control game. The program gets a 'reward' when it gets a car through a junction. Every time a car has to wait or there's a jam, there's a negative reward. There's actually no input from us; we simply control the reward system."

The researchers suggest that the system can be used to view both real and simulated traffic junctions as a means of teaching itself. They also suggest that the AI system may potentially reduce the time, fuel and income waste associated with inadequate traffic signal timing and congestion.

The team expects to begin testing the AI system on real roads later in 2022.

The research paper, Fully-Autonomous, Vision-based Traffic Signal Control: from Simulation to Reality, will be presented at the Autonomous Agents and Multi-agent Systems Conference 2022 being held virtually May 9 – May 13.

To contact the author of this article, email