Jul 2026· 2026 23rd International Conference on Ubiquitous Robots (UR)· pp. 409-416· 0 citations· 20 references
Abstract
We propose a stack-aware task and motion planning (TAMP) framework that rearranges heterogeneous objects within confined and cluttered environments. These settings involve vertical stacking and variations in object size and shape, which increase planning complexity. However, existing TAMP approaches typically assume uniform object geometries or rely on 2D abstractions that ignore detailed 3D stacking dependencies, leading to inefficient plans and frequent failures in tightly packed environments with densely stacked objects. To overcome these challenges, we introduce a rank-based task planner that assigns each potential relocation a lexicographically ordered score based on four factors: object rank, the number of objects stacked above it, buffer space clearance, and the number of blocking objects. The planner then selects the relocation with the lowest combined score, thereby reducing symbolic distance to the goal. During execution, the planner repairs the plan reactively by preserving valid subplans and applying local adjustments when collisions occur. Experimental results from both simulations and real robot experiments demonstrate higher success rates, reduced planning times, and improved object efficiency under tight spatial and stacking constraints.
Household robots have been a longstanding research topic, but they still lack human-like intelligence, particularly in manipulating open-set objects and navigating large environments efficiently and accurately. To push this boundary, we consider a scalable multi-object collection problem in large scene graphs, where the robot needs to pick up and place multiple objects across multiple locations in a long mission of multiple human commands. This problem is extremely challenging since it requires long-horizon planning in a vast action-state space under high uncertainties. To this end, we propose a novel interleaved LLM and motion planning algorithm Inter-LLM. By designing a multimodal action cost similarity function, our algorithm can both reflect the history and look into the future to keep optimizing mission performance on the fly. Extensive simulation experiments of three scenes demonstrate that compared with related works, our algorithm improves the overall mission performance by 31.2% in terms of fulfilling human commands, maximizing success rates, and minimizing execution costs. Full videos https://sites.google.com/view/inter-llm
Ruochu Yang, Yu Zhou, Fumin Zhang et al.· 2026 23rd International Conf...· 0 citations
This paper introduces Embodied Scene Rearrangement Planning (ESRP), a novel task requiring embodied agents to rearrange furniture in 3D scenes to match a target configuration using only egocentric observations and a top-down target layout. Unlike prior rearrangement tasks, ESRP precludes global state access and introduces mutual object occlusions, reflecting the practical constraints of real-world robotic deployment. These factors make aligning partial egocentric observations with the global target layout particularly challenging for long-horizon planning. To facilitate research, we present ESRP-Bench, a comprehensive benchmark built on OmniGibson featuring over 5,400 scene pairs and 8,200 objects. We define three multi-level metrics to evaluate rearrangement quality and provide four baselines: a hierarchical task-and-motion planning method, a vision-language-model-based method, and two learning-based approaches (IL and RL). Experimental results demonstrate that current methods struggle to complete the task efficiently, highlighting ESRP as a challenging frontier for embodied agents in scene understanding and long-horizon task planning. This work serves as a stepping stone toward deploying intelligent agents in real-world scenarios. Project page: https://pie-lab.cn/ESRP/.
Can-Zhi Chen, Zan Wang, Siqi Zhu et al.· IEEE Robotics and Automation...· 0 citations
The Class-Based Heuristic A* (CBHA*) algorithm, which integrates a General Move Constraint to capture minimum displacement costs when vacant units are scarce, is introduced, demonstrating that class-triggered adaptive heuristics constitute a principled mechanism for efficient spatial planning that generalizes structurally to physical constraint systems.
Sanyar Ahmadi, P. Asadzadeh, Amanj Khorramian· 0 citations
This work introduces tentative object pruning, an approach that leverages this property by constructing multiple simplified tasks with reduced object sets and searching them in parallel until a valid, satisficing solution is found.
Anita de Mello Koch, Naman Shah, Cameron S. Allen et al.· 0 citations
Long-term manipulation planning requires robots to reason not only about immediate task success but also about how current decisions affect future interactions with the environment. In this context, household service robots may need to organize groceries in partially occupied shelves while using limited storage space efficiently and preserving access for subsequent placements. In this paper, we consider an online multi-object shelf-placement setting in which future objects arrivals are unknown. Existing approaches do not jointly address semantic organization, dense space utilization, and manipulator accessibility during sequential shelf filling. To address this gap, we propose Semantic-Dense Placement Planning (SDPP), an accessibility-preserving approach that ranks candidate poses using a semantic-density score combining inter-object semantic similarity with spatial proximity. An Accessibility Map (AM) further filters candidates unlikely to be reachable before motion planning and penalizes placements that reduce the remaining accessible workspace. Simulation experiments show that SDPP significantly improves semantic placement quality over state-of-the-art baselines and achieves the highest average shelf density, while the AM substantially reduces the time required to identify feasible placement poses. A qualitative real-world experiment demonstrates the applicability of our pipeline in a domestic shelf-storage scenario.
Benno Wingender, Nils Dengler, Nicolas Busch et al.· 0 citations
Despite recent advances in general-purpose robotic manipulation, real-world multi-object clutter remains challenging to handle for today's prevalent approaches. The problem scales in complexity due to more objects and collisions, more unpredictable contact physics, distractors, and task ambiguity. Bridging this gap to real-world deployment requires effective scene abstractions; yet today, producing such abstractions requires extensive task-specific manual engineering, which does not scale. These abstractions are costly to generate and difficult to adjust or fine-tune. We instead propose a plug-and-play fix to automatically generate scene-specific, task-specific, adaptively updating abstractions on top of existing planning and control stacks. LLM-guided Environment Simplification (LENS) produces a de-cluttered abstracted scene representation by merging (e.g., stacked objects) or pruning (e.g., distant objects) scene entities in a closed loop in response to task progress. These dynamic, task-relevant abstractions are versatile and easy to use. In our experiments, we show that LENS improves classical planning, model-based control, and a vision-language-action model, across a diverse set of highly cluttered manipulation scenes. Project website: https://lens-2026.github.io/.
Aileen Liao, Rachel Holladay, Dinesh Jayaraman et al.· 0 citations
