Event
TAMP framework supports long-horizon manipulation with interactive information gathering
1 reports · 1 sources
Linfei Yuan and Rachel Holladay propose a task and motion planning (TAMP) framework that alternates task-directed interactive information gathering with plan execution. The framework uses PDDLStream to optimistically plan interactive actions, verifies results during execution, retains actions that remain usable, and replans only the subsequent connections. The authors report that the framework was tested in simulation and on a real robot on tasks such as determining whether an object is movable, whether a pen has ink, whether a drawer can be opened, and whether a container is empty; interaction-driven state updates and recovery replanning can support continued execution under uncertain outcomes and external disturbances.
Report timeline
- Interactive Information-Gathering TAMP Framework for Long-Horizon Manipulation
The authors propose a task and motion planning (TAMP) framework that alternates task-directed interactive information gathering with planning and execution. It plans optimistically for interactive actions based on PDDLStream, verifies outcomes during execution, retains actions that remain applicable, and replans only the subsequent connections. The framework was tested in simulation and on a real robot on tasks such as determining whether an object is movable, whether a pen has ink, whether a drawer can be opened, and whether a container is empty. The results show that interaction-driven state updates and recovery replanning can support continued execution under uncertain outcomes and external disturbances.
arXiv cs.RO ·