Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces LLM-guided Environment Simplification (LENS), a novel approach that automates the generation of scene-specific abstractions for robotic manipulation in cluttered environments. By dynamically merging and pruning scene entities based on task progress, LENS enables more efficient planning and control without the need for extensive manual engineering. Experimental results demonstrate that LENS significantly enhances the performance of classical planning, model-based control, and vision-language-action models in complex manipulation tasks.
Automating scene abstraction with LENS leads to substantial improvements in robotic manipulation performance, even in the messiest environments.
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/.