Reproduce “Intent-Handover: Grounding Language in Human-Usage Regions for Trustworthy Robot-to-Human Handovers” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Hanxin Zhang, Abdulqader Dhafer, Hongbiao Dong, Zhou Daniel Hao
Researchers from the University of Leicester and the University of Birmingham developed "Intent-Handover," a robot-to-human handover system that grounds natural language into explicit spatial constraints to consider human usage intent and hand-gripper clearance. User studies demonstrated increased trust and perceived safety when these human-centric considerations were integrated into the robot's grasping and delivery decisions.
Key insights
- Actively considering the human's intended usage region for an object during handover significantly enhances human trust, specifically in terms of perceived robot competence and reliability.
- Proactively avoiding collisions between the robot's gripper and the human's receiving hand substantially improves perceived safety during object transfer.
- Optimal human-robot interaction comfort in handover scenarios is achieved when both human-usage region awareness and hand-gripper collision avoidance strategies are integrated, highlighting their synergistic effect.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.