Wrench-Feasible Whole-Body Planning Via Time-Layered DAG Optimization for Omnidirectional Aerial Manipulation
Daum Park, Bohyeong Pak, Sanghyun Kim
AI summary
Problem
Omnidirectional aerial manipulators struggle with dynamically inadmissible motions and disconnected planning graphs in cluttered, long-horizon tasks, making safe whole-body trajectory tracking difficult.
Approach
The method uses GPU-accelerated reverse-chain sampling to generate whole-body candidates, prunes them via collision and rotor-thrust wrench constraints, and restores graph connectivity through guide-based refinement before dynamic programming.
Key results
- 91.7% dynamic-safe runs on drawing and 100% on peg-in-hole tasks
- 100 percentage-point success gain over wrench-unconstrained baseline
- Guide refinement reduces mean broken layers by up to 200.2
- GPU batching enables practical planning for long-horizon cluttered OAM problems
Why it matters
Enables safe, real-time whole-body motion planning for aerial manipulators in complex environments, critical for future autonomous drone manipulation applications.
Abstract
Omnidirectional aerial manipulators (OAMs) must coordinate a floating base and onboard arm to track end-effector trajectories under coupled geometric and dynamic constraints. In cluttered long-horizon tasks, collision-free mo- tions may still be dynamically inadmissible and dense time- layered planning graphs can become disconnected. We present a GPU-accelerated whole-body planning framework that com- bines reverse-chain node sampling, collision and wrench-aware feasibility filtering, guide-based connectivity refinement, and dynamic programming on a time-layered directed acyclic graph. At each timestep, the target end-effector pose is converted to a wrist-anchored representation, from which large batches of whole-body candidates are generated in parallel on the GPU. Sampled nodes are pruned using self/environment colli- sion checks and rotor-allocation-based wrench-feasibility tests. When local disconnections remain, sparse guides trigger tar- geted dense resampling to recover connectivity. On drawing and peg-in-hole tasks, the proposed method achieves 91.7% and 100.0% dynamic-safe runs, versus 0.0% and 8.3% for a wrench-unconstrained baseline. Guide refinement reduces mean broken layers by 200.2 and 49.6, yielding 100 percentage-point success gains over no refinement. GPU batching keeps planning practical for long-horizon cluttered OAM problems.