Workspace layout boosts robot-human collaboration with probabilistic guarantees
Physical arrangement of objects reduces goal inference ambiguity by up to 40%...
Shared autonomy systems help humans and robots work together by combining human input with autonomous assistance. Most prior work focuses on improving the algorithms that infer human intent, assuming the workspace is fixed. However, researchers from the University of Colorado Boulder (Tung, Gupta, Kumar, Huang, Hayes, Roncone) flip this assumption: the physical arrangement of objects directly affects how easily the robot can separate candidate goals under noisy user inputs.
They formulate workspace design as an optimization problem and derive a probabilistic correctness guarantee under a bounded noise model. In simulation experiments across multiple tabletop scenarios, optimized layouts improve goal inference reliability and reduce ambiguity compared to baseline arrangements. They also demonstrate a real-world shared autonomy system that integrates the proposed inference framework. This work highlights environment design as a complementary—and often overlooked—axis for improving shared autonomy systems, accepted at ICRA 2026 Workshop on Shared Challenges in Human-Centered and Resilient Robotic Autonomy.
- Optimizing tabletop object layout reduces goal inference ambiguity compared to standard arrangements.
- Probabilistic correctness guarantees derived under bounded noise model ensure reliability.
- Real-world shared autonomy system validates the approach in physical human-robot collaboration.
Why It Matters
Shifts focus from better algorithms to smarter workspace design, making human-robot teamwork more reliable and intuitive.