PhyScensis: Physics-Augmented LLM Agents for Complex Physical Scene Arrangement

TL;DR

PhyScensis is a physics-augmented language model framework that generates complex, physically plausible 3D scenes for robotic simulation, integrating an LLM, physics engine, and feedback loop for high realism and controllability.

Contribution

The paper introduces PhyScensis, a novel LLM-based framework that combines language understanding with physics simulation to produce complex, realistic physical scenes for robotics.

Findings

Outperforms prior methods in scene complexity and physical accuracy

Generates scenes with high object density and rich physical relationships

Provides controllable scene generation through probabilistic programming

Abstract

Automatically generating interactive 3D environments is crucial for scaling up robotic data collection in simulation. While prior work has primarily focused on 3D asset placement, it often overlooks the physical relationships between objects (e.g., contact, support, balance, and containment), which are essential for creating complex and realistic manipulation scenarios such as tabletop arrangements, shelf organization, or box packing. Compared to classical 3D layout generation, producing complex physical scenes introduces additional challenges: (a) higher object density and complexity (e.g., a small shelf may hold dozens of books), (b) richer supporting relationships and compact spatial layouts, and (c) the need to accurately model both spatial placement and physical properties. To address these challenges, we propose PhyScensis, an LLM agent-based framework powered by a physics engine, to produce physically plausible scene configurations with high complexity. Specifically, our framework consists of three main components: an LLM agent iteratively proposes assets with spatial and physical predicates; a solver, equipped with a physics engine, realizes these predicates into a 3D scene; and feedback from the solver informs the agent to refine and enrich the configuration. Moreover, our framework preserves strong controllability over fine-grained textual descriptions and numerical parameters (e.g., relative positions, scene stability), enabled through probabilistic programming for stability and a complementary heuristic that jointly regulates stability and spatial relations. Experimental results show that our method outperforms prior approaches in scene complexity, visual quality, and physical accuracy, offering a unified pipeline for generating complex physical scene layouts for robotic manipulation.