Autonomy and Human-AI Interaction
summary
Describes an implemented mobile service robot that blends motion planning with spoken interaction to decide when to act versus ask for help, using a POMDP-based autonomy layer and supervised perception for object and person recognition. Reports a deployed system on three lab robots and a user study with 60 participants performing fetch and guidance tasks, showing about a 30% drop in required human intervention and quantitative tradeoffs between autonomy level and task success. Demonstrations and failure modes are shown.
AI-written summary · about
related talks
52mWhat If Robots Asked Us for Help?
40mHuman-AI Collaboration Models for Better Decisions and Faster Workflows
20mHuman above AI
29mPairing with the machine: How AI Agents are rewriting the rules of software development
12mInteractive Planning-based Cognitive Assistance on the Edge
27mAgentic Engineering: Working With AI, Not Just Using It
13mHuman-AI Symbiotic Robotics
13mHuman + Machine