Returning to work after an upper-limb injury is often a slow process, especially in environments where collaborative robots and humanoids share space with people. A new preprint on arXiv argues that these robots need to adapt their behavior to the physical capabilities of the human they are working with, rather than assuming a fixed level of ability.
The paper, titled "Active Perception of Upper-Limb Functional Mobility through Human-Robot Handovers," centers on the handover moment as a key point for adaptation. By using active perception, the robot can gather information about the user's upper-limb mobility in real time and adjust its actions accordingly.
This approach could make collaborative robots more useful in rehabilitation and return-to-work settings, where a worker's range of motion or strength may be temporarily limited. The preprint is available on arXiv under identifier 2610.19896.