Mechanical Neural Network - OrionHack23

I took part in OrionHack, a space-themed deeptech hackathon that took place over a weekend. We made a unique mechanical neural network design to train a physical and simulated model to learn an XOR gate. This could potentially be exciting for space to avoid issues of radiation affecting calculations. It was a really nice collaboration of Design Engineering, Mechanical Engineering and Mathematics.

Our final project was selected as the winner.

Simulation I helped develop:

Video of presentation:

Presentation shown: