Evolving- Ai-Car-Simulation

Main Car
Map 1
Map 2
Map 3
Map 4
Map 5

Inspiration

Currently we are in gen-4 and in future gen-5 is coming. The world is leaning toward AI more than ever. So, we can assume in the near future AI automation is bound to take over so self-driving car is not a pipe dream anymore, and in order for an AI to drive a car without any accident it must train itself without any known error. So, we aim to make a simple AI simulation where the training can happen.

Description

We simulate self-driving AI cars in Python, where they learn to drive using NEAT, a technique based on reinforcement learning.

How we built it

We made it with python and used the Pygames and neat packages which is neural evolution of augmenting topologies. For the IDE we used Pycharm as it is very user friendly.

Challenges we ran into

Since our project implements several neural links and AI integration, we kind of ran into a problem there.

Accomplishments that we're proud of

We are proud that the project turned out as just as we hoped it to be.

What we learned

We had a blast doing the project as this is our first time joining a hackathon. So, we really appreciate the HackCsb Team, and we learned a lot about AI and neural networks.