# Sim To Real Transfer

> Bridge simulation to real hardware. Domain randomization, system ID, hardware abstraction.

- Skill: `aselimc/sim-to-real-transfer` (Agent Skill)
- Install (CLI): `npx skillmds@latest add aselimc/sim-to-real-transfer`
- Raw SKILL.md: https://api.skillmd.com/api/skills/aselimc/sim-to-real-transfer/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: aselimc (https://skillmd.com/u/aselimc)
- Updated: 2026-09-10
- Page: https://skillmd.com/skills/aselimc/sim-to-real-transfer

---


# Sim-to-Real Transfer

## Domain Randomization
- **Visual**: texture, lighting, camera pose, background
- **Dynamics**: friction, mass, damping, actuator delay
- **Observation noise**: Gaussian noise on sensors, dropout
- **Action**: smoothing, latency simulation (1-3 step delay)

## Strategies
| Approach | When | Pros | Cons |
|----------|------|------|------|
| Domain randomization | No real data | Simple, scalable | Over-conservative |
| System identification | Some real data | Accurate | Requires instrumentation |
| Fine-tuning in real | After sim training | Best performance | Needs safe exploration |
| Asymmetric actor-critic | Privileged sim info | Uses GT in training only | Complex implementation |

## Hardware Abstraction
Design interfaces that work identically in sim and real:
```python
class RobotInterface(ABC):
    def get_observation(self) -> dict: ...
    def send_action(self, action: np.ndarray) -> None: ...
    def reset(self) -> dict: ...
```

## Key Libraries
Isaac Sim/Lab, MuJoCo, Gazebo, PyBullet