Lightning Traffic

A Machine Learning traffic lights simulator made for nwHacks 2019.

Sample run: cars are in blue, streets are in grey and traffic lights are red or green


The project consists of an artificial inteligence that finds an optimal schedule
for the traffic lights of a city. Initially, the AI does not know the position of the traffic lights
nor the schedule of the cars. With time, the AI learns from the simulations and tries to minimize
the total time spent by the cars on traffic by assigning the times when the traffic lights
should be on and off.


The project was written in Java and the GUI of the simulations is done by Processing.
The AI learning process is based on Genetic Algorithms.

Project developers:

Jaden Balogh

Ivan Carvalho

Noah Eibl

Jean-Philippe Abadir

Built With

Share this project: