"Do Diffusion Models Dream of Electric Planes?" Discrete and Continuous Simulation-Based Inference for Aircraft Design

Aurelien Ghiglino, Daniel Elenius, Anirban Roy, Ramneet Kaur, Manoj Acharya, Colin Samplawski, Brian Matejek, Susmit Jha, Juan Alonso, Adam D. Cobb
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:34698-34727, 2026.

Abstract

In this paper, we generate conceptual engineering designs of electric vertical take-off and landing (eVTOL) aircraft. We follow the paradigm of simulation-based inference (SBI), whereby we look to learn a posterior distribution over the full eVTOL design space. To learn this distribution, we sample over discrete aircraft configurations (topologies) and their corresponding set of continuous parameters. Therefore, we introduce a hierarchical probabilistic model consisting of two diffusion models. The first model leverages recent work on Riemannian Diffusion Language Modeling (RDLM) and Unified World Models (UWMs) to enable us to sample topologies from a discrete and continuous space. For the second model we introduce a masked diffusion approach to sample the corresponding parameters conditioned on the topology. Our approach rediscovers known trends and governing physical laws in aircraft design, while significantly accelerating design generation.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-ghiglino26a, title = {"{D}o Diffusion Models Dream of Electric Planes?" Discrete and Continuous Simulation-Based Inference for Aircraft Design}, author = {Ghiglino, Aurelien and Elenius, Daniel and Roy, Anirban and Kaur, Ramneet and Acharya, Manoj and Samplawski, Colin and Matejek, Brian and Jha, Susmit and Alonso, Juan and Cobb, Adam D.}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {34698--34727}, 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/ghiglino26a/ghiglino26a.pdf}, url = {https://proceedings.mlr.press/v306/ghiglino26a.html}, abstract = {In this paper, we generate conceptual engineering designs of electric vertical take-off and landing (eVTOL) aircraft. We follow the paradigm of simulation-based inference (SBI), whereby we look to learn a posterior distribution over the full eVTOL design space. To learn this distribution, we sample over discrete aircraft configurations (topologies) and their corresponding set of continuous parameters. Therefore, we introduce a hierarchical probabilistic model consisting of two diffusion models. The first model leverages recent work on Riemannian Diffusion Language Modeling (RDLM) and Unified World Models (UWMs) to enable us to sample topologies from a discrete and continuous space. For the second model we introduce a masked diffusion approach to sample the corresponding parameters conditioned on the topology. Our approach rediscovers known trends and governing physical laws in aircraft design, while significantly accelerating design generation.} }
Endnote
%0 Conference Paper %T "Do Diffusion Models Dream of Electric Planes?" Discrete and Continuous Simulation-Based Inference for Aircraft Design %A Aurelien Ghiglino %A Daniel Elenius %A Anirban Roy %A Ramneet Kaur %A Manoj Acharya %A Colin Samplawski %A Brian Matejek %A Susmit Jha %A Juan Alonso %A Adam D. Cobb %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-ghiglino26a %I PMLR %P 34698--34727 %U https://proceedings.mlr.press/v306/ghiglino26a.html %V 306 %X In this paper, we generate conceptual engineering designs of electric vertical take-off and landing (eVTOL) aircraft. We follow the paradigm of simulation-based inference (SBI), whereby we look to learn a posterior distribution over the full eVTOL design space. To learn this distribution, we sample over discrete aircraft configurations (topologies) and their corresponding set of continuous parameters. Therefore, we introduce a hierarchical probabilistic model consisting of two diffusion models. The first model leverages recent work on Riemannian Diffusion Language Modeling (RDLM) and Unified World Models (UWMs) to enable us to sample topologies from a discrete and continuous space. For the second model we introduce a masked diffusion approach to sample the corresponding parameters conditioned on the topology. Our approach rediscovers known trends and governing physical laws in aircraft design, while significantly accelerating design generation.
APA
Ghiglino, A., Elenius, D., Roy, A., Kaur, R., Acharya, M., Samplawski, C., Matejek, B., Jha, S., Alonso, J. & Cobb, A.D.. (2026). "Do Diffusion Models Dream of Electric Planes?" Discrete and Continuous Simulation-Based Inference for Aircraft Design. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:34698-34727 Available from https://proceedings.mlr.press/v306/ghiglino26a.html.

Related Material