As part of the PERSEO project of the PEPR Robotique, a PhD topic is offered under joint supervision between the CRIStAL Laboratory in Lille and LAAS-CNRS.

Title: Dynamical Variational Autoencoders for Resilient Aerial-Ground Robot Fleets

The thesis will focus on developing structured DVAEs for modeling, estimating, and predicting the dynamics of air-ground robotic fleets. The work will notably address recursive Bayesian inference, multi-sensor fusion, and fault tolerance.

The detailed topic is available in English et in French. Strong skills are required in stochastic models, stochastic estimation, machine learning, Python programming.

The advisors are Maan El Badaoui El Najjar, (Prof. CRIStAL & Lille University) and Patrick Danès (LAAS-CNRS & Toulouse University).

Please send by e-mail a complete application (CV, application letter, Master’s transcripts, letters of recommendation...) as a single pdf file to maan.el-badaoui-el-najjar@univ-lille.fr,patrick.danes@laas.fr.