# SIMART: Decomposing Monolithic Meshes into Sim-ready Articulated Assets via MLLM

Source: [SIMART: Decomposing Monolithic Meshes into Sim-ready Articulated Assets via MLLM](<https://www.sota2.com/research/paper/69c36c260636e5bf4afc65ac>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-28T06:00:01.368Z

Authors: Chuanrui Zhang, Minghan Qin, Yuang Wang, Baifeng Xie, Hang Li, Ziwei Wang

Published: 2026-03-24T16:16:52.000Z

## Abstract

High-quality articulated 3D assets are indispensable for embodied AI and physical simulation, yet 3D generation still focuses on static meshes, leaving a gap in "sim-ready" interactive objects. Most recent articulated object creation methods rely on multi-stage pipelines that accumulate errors across decoupled modules. Alternatively, unified MLLMs offer a single-stage path to joint static asset understanding and sim-ready asset generation. However dense voxel-based 3D tokenization yields long 3D token sequences and high memory overhead, limiting scalability to complex articulated objects. To address this, we propose SIMART, a unified MLLM framework that jointly performs part-level decomposition and kinematic prediction. By introducing a Sparse 3D VQ-VAE, SIMART reduces token counts by 70% vs. dense voxel tokens, enabling high-fidelity multi-part assemblies. SIMART achieves state-of-the-art performance on PartNet-Mobility and in-the-wild AIGC datasets, and enables physics-based robotic simulation.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [3D Articulation and Geometry Estimation on SIMART-Bench (In-Domain Items)](<https://www.sota2.com/research/sota/3d-articulation-and-geometry-estimation-on-simart-bench-in-domain-items.md>): 3D Articulation and Geometry Estimation; SIMART-Bench (In-Domain Items); Type Score: 92.8.

- [3D Articulation and Geometry Estimation on SIMART-Bench AI-generated Items (Out-of-Distribution)](<https://www.sota2.com/research/sota/3d-articulation-and-geometry-estimation-on-simart-bench-ai-generated-items-out-of-distribution.md>): 3D Articulation and Geometry Estimation; SIMART-Bench AI-generated Items (Out-of-Distribution); Type Accuracy: 83.1.

- [3D Part Grounding on AI-generated Items](<https://www.sota2.com/research/sota/3d-part-grounding-on-ai-generated-items.md>): 3D Part Grounding; AI-generated Items; IOU: 80.7.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2603.23386>)

[PDF](<https://arxiv.org/pdf/2603.23386.pdf>)
