First of its Kind AI traffic Light System will bring ease to Road Ragers
The newly developed system will read live camera footage and adapt traffic lights to compensate with the objective of making traffic flow easily...
Do you ever catch yourself screaming at a traffic light for taking too long when nothing is coming from the opposite end? Well, lucky for all of you with road rage, new AI systems developed by Aston University scientists are the first of their kind and read live camera footage and adapt traffic lights to compensate with the objective of making traffic flow easily.
Deep reinforcement learning has been in the works and these traffic lights use just that. What is this said learning? Well, it is when the program understands it is not doing well and tries to differentiate the actions it is taking thus improving to make progress. In the scientist’s testing, their newly developed system outperformed all other methods designed based on phase transitions.
In the United Kingdon, urban areas witnessed estimated congestion of 115 hours of time and wasted £894 in gasoline in 2019. This is a reoccurring thing however as this is reported every single year. One of the leading causes is the inadequate traffic signal timings experienced in the U.K.
Traffic 3D was created to train their developed program. The Traffic 3D taught the system how to handle different traffic and weather scenarios. In real situations, the newly developed system adapted to real traffic intersections despite being trained on simulations.
"We 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."
Dr. Maria Chli, researcher in Computer Science at Aston University, explained.
Currently, the system used for traffic lights is based on a magnetic induction loop in which a wire sits on the road and registers cars passing over it. The program counts the cars and reacts to that data. The AI system sees the traffic firsthand, thus it reacts quicker.
“The reason we have based this program on learned behaviors is so that it can understand situations it hasn't explicitly experienced before. We've tested this with a physical obstacle that is causing congestion, rather than traffic light phasing, and the system still did well. As long as there is a causal link, the computer will ultimately figure out what that link is. It's an intensely powerful system."
Dr. George Vogiatzis, senior lecturer in Computer Science at Aston University, said.
The newly developed system learns autonomously no matter what traffic junction it is set up at. It can also be manipulated to allow for emergency vehicles through quickly but the program is always learning and teaching preferences instead of being programmed.
1The research paper, Fully-Autonomous, Vision-based Traffic Signal Control: from Simulation to Reality, is being presented at the Autonomous Agents and Multi-agent Systems Conference 2022 being held virtually this week.