A Reinforcement Learning agent training framework with Unity integration. Connects RL agents to Unity simulations via WebSocket for real-time hybrid control.
RLUnity bridges the gap between Reinforcement Learning algorithms and Unity simulations. By establishing a WebSocket connection, RL agents can send actions to and receive observations from a Unity environment in real time, enabling hybrid control architectures for complex simulation scenarios.
- WebSocket Communication — Real-time bidirectional data flow between RL agents and Unity
- Hybrid Control Architecture — Combines autonomous RL decision-making with simulation control
- Modular Design — Easily swap RL algorithms or Unity environments
- Real-Time Training — Train agents directly within the simulation loop
- Observation & Action Spaces — Flexible configuration for different simulation needs
| Component | Technology |
|---|---|
| RL Agent | Python |
| Game Engine | Unity |
| Communication | WebSocket |
| Agent Logic | Reinforcement Learning |