MO-SeGMan: Rearrangement Planning Framework for Multi-Objective Sequential and Guided Manipulation in Constrained Environments
Cankut Bora Tuncer, Marc Toussaint, Ozgur S. Oguz
AI summary
Problem
Rearrangement planning in highly cluttered environments is hindered by joint-space constraints that restrict robot reachability, making it difficult to move objects without first clearing obstructing movable obstacles. Existing methods often lack scalability, rely on task-specific heuristics, or struggle with non-monotone scenarios.
Approach
The framework generates optimized object placement sequences using lazy evaluation to minimize replanning and travel distance, while a Selective Guided Forward Search efficiently relocates only critical obstacles and a refinement step removes redundant pick-and-place actions.
Key results
- Jointly minimizes per-object replanning and robot travel distance via lazy evaluation
- Scalable Selective Guided Forward Search efficiently relocates only critical obstacles
- Subgoal refinement eliminates unnecessary pick-and-place actions
- Achieves faster solution times and superior quality across 14 constrained benchmarks
Why it matters
Enables reliable robotic rearrangement in complex, cluttered spaces, advancing practical automation for logistics and domestic environments.
Abstract
In this work, we introduce MO-SeGMan, a Multi- Objective Sequential and Guided Manipulation planner for highly constrained rearrangement problems. MO-SeGMan gen- erates object placement sequences that minimize both re- planning per object and robot travel distance while pre- serving critical dependency structures with a lazy evaluation method. To address highly cluttered, non-monotone scenarios, we propose a Selective Guided Forward Search (SGFS) that efficiently relocates only critical obstacles and to feasible re- location points. Furthermore, we adopt a refinement method for adaptive subgoal selection to eliminate unnecessary pick- and-place actions, thereby improving overall solution quality. Extensive evaluations on nine benchmark rearrangement tasks demonstrate that MO-SeGMan generates feasible motion plans in all cases, consistently achieving faster solution times and superior solution quality compared to the baselines. These results highlight the robustness and scalability of the proposed framework for complex rearrangement planning problems. Supplementary videos and code are available at: https: //sites.google.com/view/mo-segman/.