Pascal Brault

Centre National de la Recherche Scientifique

Papers

Total Citations

H-Index

About

Pascal Brault is a robotics researcher whose work focuses on the intersection of robust control, trajectory optimization, and vision-based robot manipulation. His primary research areas include robust trajectory planning under parametric uncertainties and visual servoing using learned representations. In his highly cited 2021 paper "Robust Trajectory Planning with Parametric Uncertainties" (18 citations), Brault introduced the novel concept of input sensitivity alongside closed-loop state sensitivity, enabling the generation of optimal reference trajectories that minimize the effects of model uncertainties on robot motion. This work provides a principled framework for designing safer, more reliable robot motions in real-world environments where system parameters are never perfectly known. In his 2022 paper "Visual Servoing in Autoencoder Latent Space" (14 citations), Brault pioneered an approach that replaces traditional hand-crafted visual features with learned latent representations from autoencoders, allowing robots to perform visual servoing without requiring explicit feature extraction or tracking. This work demonstrates how deep learning can simplify classical control architectures. Brault's contributions are particularly valuable for applications requiring both precision and robustness, such as autonomous manufacturing and surgical robotics.