The system is divided into two categories: object-centric and robot-centric. The former involves tasks like opening a door, while the latter involves moving the robot around objects. The team says the system can be adapted for different form factors and is a step towards developing fully autonomous robots.
Key takeaways:
- New research from ETH Zurich has developed a model for robots that requires minimal manual guidance. The process involves the user describing the scene and action, the system planning a route, and then refining the route into a minimal viable path.
- The system is divided into two main categories: object-centric and robot-centric. Object-centric tasks involve actions like opening a door, while robot-centric tasks involve moving the robot around objects.
- The system can be adapted for different form factors and has been demonstrated on a quadruped robot, ANYbotics’ ANYmal, which was spun out of ETH Zurich.
- The team believes this work can serve as a stepping stone to developing fully autonomous systems that can perform tasks like opening doors without any human intervention.