Unitree Go2-W is the wheeled variant of the Go2 platform, combining the Go2's proven embodied AI capabilities with wheel-leg hybrid mobility for teams developing and testing navigation, perception, and on-board AI functions at higher speeds and over longer distances. On flat and semi-structured surfaces, the Go2-W rolls efficiently while retaining full stepping capability for obstacles and mixed terrain. Sharing the Go2's 4D LiDAR, LLM integration, autonomous navigation stack, and open SDK, the Go2-W is particularly well-suited for last-mile delivery pilots, campus patrol experiments, and indoor-outdoor autonomy research where movement efficiency is as important as terrain adaptability.
Key Highlights
- Wheel-leg hybrid combining Go2 AI capabilities with rolling efficiency
- 4D LiDAR and LLM integration carried over from standard Go2
- Efficient rolling on flat surfaces; full stepping for obstacles and rough terrain
- Suited for delivery pilots, campus patrol, and indoor-outdoor autonomy research
- Open SDK; ROS2 compatible; supports secondary development
- Ideal for teams prioritising speed and range alongside terrain adaptability
