An environment was modeled, utilizing AgentyPy and machine learning algorithms, that contained agents that represented cars crossing a boulevard. The logic ran in a Flask server so that the car positions over time could be pull into Unity
In Unity the cars’ appearance was made randomly with scriptable objects and a pooling system to manage a large number of cars to optimize the program
A camera system was implemented to change the view during the simulation to a bird’s eye view, default and on clicking the cars see the car’s perspective