Rocket Recovery Simulation

🚀 Rocket AI: Takeoff & Landing Simulation

Watch machine learning take flight.

Rocket AI is a Unity-based simulation showcasing a rocket that takes off, hovers, and lands autonomously — all powered by reinforcement learning using Unity ML-Agents.

In this project, an AI agent was trained over thousands of episodes to master the delicate balance of rocket propulsion and gravity. The result? A self-taught lander that can launch, stabilize mid-air, and perform smooth landings on a designated pad — with no human intervention.

🧠 Features:

• ML-Agents Integration: The rocket's brain is a neural network trained through Unity ML-Agents using a custom reward system for stability, control, and precision.

• Physics-Based Realism: Uses Unity's physics engine to simulate real-world challenges like inertia, thrust, and torque.

• Dynamic Feedback: Visual indicators and stats show the AI's decision-making in real-time.

🔧 What's Included:

• The full simulation scene (playable demo)

• Basic stats display

Whether you're into rockets, AI, or both, this simulation is a fun and educational dive into the intersection of machine learning and aerospace dynamics.