Using Reinforcement Learning (PPO) to try and teach a robotic arm to pick and place - eviltoast
Reinforcement Learning with a Pick and Place Robotic Arm
hlfshell.ai
tldr I created a custom robotic arm environment where the arm is rewarded if it sorts shapes a certain way, and then punished if it knocks them off a table. I used OpenAI’s PPO method to be somewhat successful. If you want to skip to the code, go here. The why Reinforcement learning is a topic that’s endlessly fascinating to me. I had great success working through Deep Reinforcement Learning in Action, creating a generic Deep Q solver for several classic OpenAI gym environments.

In my first attempt at a long form technical post, I talk about a project where I had to use deep reinforcement learning to try and solve a robotics application. It worked ok - the post talks about my struggles, solutions, and what I’d change up in the future.