Semi-Lazy Rearrangement Solver for Accelerating Task Planning Using SPITE Dynamic Roadmaps
Authors: Marta Markowicz, Stav Ashur, James Motes, Nancy M. Amato
Venue: ICRA 2025 Workshop on Future of Construction
DOI:
Link to Publication
Abstract:
In construction, we require collision-free planning to work in complex environments full of materials and debris. We introduce a rearrangement planning approach that aug- ments a task planner leveraging lazy motion validation with a dynamic roadmap that quickly approximates the feasibility of motions after object rearrangement. This approach enables the task planner to produce candidate plans that are more likely to yield valid motion plans without significant overhead, thereby reducing the number of iterations between task planning and motion validation layers and decreasing the overall planning time.
@article{markowicz-slrs-ws-25,
title={Semi-Lazy Rearrangement Solver for Accelerating Task Planning Using SPITE Dynamic Roadmaps},
author={Markowicz, Marta and Ashur, Stav and Motes, James and Amato, Nancy M}
maintitle={International Conference on Robotics and Automation}
booktitle={Workshop on Future of Construction}
year={2025}
}