# Neural-Surrogate Co-Design of Aerodynamic Shape and Control

arXiv: https://arxiv.org/abs/2603.06760 · HTML: https://mlsanity.com/paper/2603.06760/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 4 of its 36 references that are in the map.
- First posted on arXiv 6 Mar 2026.

## Builds on (most cited first, 4 of 4)

- [NeuralFoil: An Airfoil Aerodynamics Analysis Tool Using Physics-Informed Machine Learning](https://mlsanity.com/paper/2503.16323/) (2025)
- [Revisiting Implicit Differentiation for Learning Problems in Optimal Control](https://mlsanity.com/paper/2310.14468/) (2023)
- [Robot Co-design: Beyond the Monotone Case](https://mlsanity.com/paper/1902.05880/) (2019)
- [Co-Design Optimisation of Morphing Topology and Control of Winged Drones](https://mlsanity.com/paper/2309.13948/) (2023)

## Abstract (arXiv)

Control co-design jointly optimizes a robot's physical design and its controller. Existing techniques are limited in the number of model parameters and the complexity of the physical models they can handle. In this paper, we propose a gradient-based bilevel formulation that replaces the dynamics models of earlier techniques with a differentiable neural surrogate and differentiates through the trajectory optimizer using a structure-exploiting approach. We validate the framework on fixed-wing gliders, using NeuralFoil as the aerodynamic surrogate. The tasks are maximum-distance gliding, minimum-distance landing, and perching. Experiments on 2D airfoils and 3D wings show that our method runs about five times faster than a state-of-the-art evolutionary baseline. It also outperforms pipelines that optimize aerodynamic proxy metrics.