This is my attempt at utilizing a DQN to teach a robot to solve the cart-pole problem.
Original 3d models designed for a pole to be held upright on the cart and balance. Upon completion of building the model, the wiring would keep the pole balanced on its own and did not allow for any balancing. I pivoted, manually recreated the model to train to self-balance on its bottom two wheels.
Short video demonstrating the early attempts at training the model: https://www.youtube.com/watch?v=-TPxhqB77YM