Radar is an interesting direction because it works much better in poor visibility conditions compared to cameras.
If autonomous systems combine radar, lidar and camera data effectively, the redundancy could make perception systems significantly more reliable, especially in fog, rain or low-light environments.
The real challenge is probably not only sensing, but how well the system fuses and interprets all that data in real time.
Rebuilding something yourself often reveals design decisions that aren’t obvious when you only read documentation or use the tool as a black box.
Curious what parts were the most difficult to reproduce compared to the original OpenClaw implementation.