🤖 AI-Powered Robotic Arm Control 🦾

A reinforcement learning project where AI agents learn to control a robotic arm.

Read the full story: Teaching Robots to Reach: A Reinforcement Learning Journey.

Overview

This project combines:

• 🏗️ Custom robotic arm environment built with Gymnasium
• 🧠 Deep reinforcement learning implemented via Stable Baselines3
• 🎮 Human-playable interface using PyGame

Features

• 🏭 Custom 2D robotic arm environment with:
  • Joint motors
  • Dynamic target generation
• 🤖 PPO for precise control learning
• ✍️ Blog post documenting the entire journey

Usage

Manual control:

python robot_arm_env.py

Train the agent:

python train.py

Watch trained agent:

python use.py

License

MIT License - See LICENSE for details.

Acknowledgments

• Farama Foundation for Gymnasium
• Stable Baselines3 team for their RL implementations
• All the coffee that powered this project ☕