Research Analyzer
← Back ICRA 2026

High-Performance Dual-Arm Task and Motion Planning for Tabletop Rearrangement

Duo Zhang, Junshan Huang, Jingjin Yu

PDF

AI summary

Key figure (auto-extracted from paper)
SDAR achieves a 100% success rate and near-real-time planning for complex dual-arm tabletop rearrangement tasks, outperforming prior state-of-the-art planners.
Dual-arm robotics Task and Motion Planning Tabletop Rearrangement GPU-accelerated Planning Dependency Graphs Real-time Control

Problem

Dual-arm task and motion planning (TAMP) for tabletop rearrangement faces a computational bottleneck due to combinatorial task search explosion and high-dimensional motion planning in close-proximity systems with entangled object configurations.

Approach

SDAR integrates a dependency-driven task planner that decomposes object constraints into parallelizable sub-tasks with a GPU-accelerated motion planner that samples grasp poses and optimizes trajectories in real-time, supported by a multi-stage failure recovery mechanism.

Key results

  • Dependency-driven task planner reduces sub-task counts and enables parallel dual-arm execution
  • GPU-accelerated motion planner achieves 100% success rate on complex, non-monotone tasks
  • Near threefold reduction in task execution time compared to baseline planners
  • Successful real-world transfer to UR-5e robot hardware with reliable performance

Why it matters

Enables reliable, real-time coordination for dual-arm robots in cluttered environments, advancing practical automation for manufacturing, logistics, and domestic assistance.

Abstract

We propose Synchronous Dual-Arm Rearrange- ment Planner (SDAR), a task and motion planning (TAMP) framework for tabletop rearrangement, where two robot arms equipped with 2-finger grippers must work together in close proximity to rearrange objects whose start and goal config- urations are strongly entangled. To tackle such challenges, SDAR tightly knit together its dependency-driven task planner (SDAR-T) and synchronous dual-arm motion planner (SDAR- M), to intelligently sift through a large number of possible task and motion plans. Specifically, SDAR-T applies a simple yet effective strategy to decompose the global object dependency graph induced by the rearrangement task, to produce more optimal dual-arm task plans than solutions derived from optimal task plans for a single arm. Leveraging state-of-the-art GPU SIMD-based motion planning tools, SDAR-M employs a layered motion planning strategy to sift through many task plans for the best synchronous dual-arm motion plan while ensuring high levels of success rate. Comprehensive evaluation demonstrates that SDAR delivers a 100% success rate in solving complex, non-monotone, long-horizon tabletop rearrangement tasks with solution quality far exceeding the previous state- of-the-art. Experiments on two UR-5e arms further confirm SDAR directly and reliably transfers to robot hardware. Source code and supplementary materials are available at https://github.com/arc-l/dual-arm.

Index terms

Task and Motion Planning Dual Arm Manipulation Manipulation Planning

Related papers