At VIAME, we are developing new concepts for lightweight design in several sectors (transportation, energy, machining,…). This research project will focus on an optimized data-based framework for Digital Twin (DT) generation of robot systems that can speed up design and evaluation. Machine Learning (ML) techniques will be applied to infer system models based on information obtained either from high-fidelity simulations or measurements. Suitability of ML methods will be assessed for various degrees of nonlinearity, non-smoothness, modelling uncertainty and parametric dimensionality in robotic applications. The thesis will be performed remotely at Ecole Central de Lyon, France.

Ichchou, Mohamed