Adaptive Gait Generation for Hexapod

This project was inspired by natural phenomenon of adaptation witnessed in nature. Genetic Algorithm is used to generate gaits for a robot without manually inputting the gait in program source code. This means, the hexapod learns to walk on its own and is never taught walking by its human masters!

We also used GA to introduce the ability to adapt automatically in case of damage, just like animals do in nature! For example, whenever a dog gets hurt, it is quickly able to adapt to any damage and escape to safety. Our project tries to emulate the same, on a much smaller and slower scale. It is very hard to challenge nature in beauty and complexity!

This project won out of 50 other ideas presented in technical fest APOGEE 2016.

Public Artifacts

Research Paper:

Arduino controller code:

MATLAB simulation for Genetic Algorithm: