LabBuilder: Protocol-Grounded 3D Layout Generation for Interactable and Safe Laboratory

Jianbao Cao, Zhangrui Zhao, Bohan Feng, Zixuan Hu, Rui Li, Haiyuan Wan, Chenxi Li, Jingyuan Li, Wenzhe Cai, Lei Bai, Wanli Ouyang, Lingyu Duan, Di Huang, Minting Pan, Sha Zhang, Xinzhu Ma, Shixiang Tang, Dongzhan Zhou
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:11786-11814, 2026.

Abstract

Automated laboratories hold the promise of accelerating scientific discovery, yet their deployment is bottlenecked by the difficulty of designing safe and executable environments. While simulator-based design offers scalability, existing 3D scene generation methods are primarily tailored for household settings, optimizing for visual plausibility while neglecting the protocol grounding and layout-level safety constraints essential for scientific experimentation. We present LabBuilder, an end-to-end system that generates and verifies 3D laboratory layouts from concise textual specifications. It operates through three tightly coupled components: LabForge first curates a meta-dataset of annotated assets and chemical knowledge, translating natural language specifications into structured protocols; building on these protocols, LabGen synthesizes laboratory layouts via an iterative, constraint-aware optimization strategy; finally, LabTouchstone evaluates the resulting layouts as a unified benchmark. Extensive experiments demonstrate that LabBuilder significantly outperforms existing state-of-the-art methods, producing laboratory environments that are realistic and valid under modeled geometric, chemical-safety, and navigation constraints.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-cao26ad, title = {{L}ab{B}uilder: Protocol-Grounded 3{D} Layout Generation for Interactable and Safe Laboratory}, author = {Cao, Jianbao and Zhao, Zhangrui and Feng, Bohan and Hu, Zixuan and Li, Rui and Wan, Haiyuan and Li, Chenxi and Li, Jingyuan and Cai, Wenzhe and Bai, Lei and Ouyang, Wanli and Duan, Lingyu and Huang, Di and Pan, Minting and Zhang, Sha and Ma, Xinzhu and Tang, Shixiang and Zhou, Dongzhan}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {11786--11814}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/cao26ad/cao26ad.pdf}, url = {https://proceedings.mlr.press/v306/cao26ad.html}, abstract = {Automated laboratories hold the promise of accelerating scientific discovery, yet their deployment is bottlenecked by the difficulty of designing safe and executable environments. While simulator-based design offers scalability, existing 3D scene generation methods are primarily tailored for household settings, optimizing for visual plausibility while neglecting the protocol grounding and layout-level safety constraints essential for scientific experimentation. We present LabBuilder, an end-to-end system that generates and verifies 3D laboratory layouts from concise textual specifications. It operates through three tightly coupled components: LabForge first curates a meta-dataset of annotated assets and chemical knowledge, translating natural language specifications into structured protocols; building on these protocols, LabGen synthesizes laboratory layouts via an iterative, constraint-aware optimization strategy; finally, LabTouchstone evaluates the resulting layouts as a unified benchmark. Extensive experiments demonstrate that LabBuilder significantly outperforms existing state-of-the-art methods, producing laboratory environments that are realistic and valid under modeled geometric, chemical-safety, and navigation constraints.} }
Endnote
%0 Conference Paper %T LabBuilder: Protocol-Grounded 3D Layout Generation for Interactable and Safe Laboratory %A Jianbao Cao %A Zhangrui Zhao %A Bohan Feng %A Zixuan Hu %A Rui Li %A Haiyuan Wan %A Chenxi Li %A Jingyuan Li %A Wenzhe Cai %A Lei Bai %A Wanli Ouyang %A Lingyu Duan %A Di Huang %A Minting Pan %A Sha Zhang %A Xinzhu Ma %A Shixiang Tang %A Dongzhan Zhou %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-cao26ad %I PMLR %P 11786--11814 %U https://proceedings.mlr.press/v306/cao26ad.html %V 306 %X Automated laboratories hold the promise of accelerating scientific discovery, yet their deployment is bottlenecked by the difficulty of designing safe and executable environments. While simulator-based design offers scalability, existing 3D scene generation methods are primarily tailored for household settings, optimizing for visual plausibility while neglecting the protocol grounding and layout-level safety constraints essential for scientific experimentation. We present LabBuilder, an end-to-end system that generates and verifies 3D laboratory layouts from concise textual specifications. It operates through three tightly coupled components: LabForge first curates a meta-dataset of annotated assets and chemical knowledge, translating natural language specifications into structured protocols; building on these protocols, LabGen synthesizes laboratory layouts via an iterative, constraint-aware optimization strategy; finally, LabTouchstone evaluates the resulting layouts as a unified benchmark. Extensive experiments demonstrate that LabBuilder significantly outperforms existing state-of-the-art methods, producing laboratory environments that are realistic and valid under modeled geometric, chemical-safety, and navigation constraints.
APA
Cao, J., Zhao, Z., Feng, B., Hu, Z., Li, R., Wan, H., Li, C., Li, J., Cai, W., Bai, L., Ouyang, W., Duan, L., Huang, D., Pan, M., Zhang, S., Ma, X., Tang, S. & Zhou, D.. (2026). LabBuilder: Protocol-Grounded 3D Layout Generation for Interactable and Safe Laboratory. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:11786-11814 Available from https://proceedings.mlr.press/v306/cao26ad.html.

Related Material