Louis Berthier

I am a CIFRE PhD student in Applied Mathematics & Machine Learning at École Polytechnique, affiliated with the Centre de Mathématiques Appliquées (CMAP) and the SIMPAS team.
My research is conducted in collaboration Michelin, under the supervision of Pr. Eric Moulines, Dr. Ahmed Shokry, Dr. Maxime Moreaud, and Guillaume Ramelet.

My research focuses on the development of data-driven soft sensors for real-time quality monitoring in rubber production lines.
The objective is to overcome the limitations of expensive and offline quality measurements by building adaptive predictive systems that provide reliable, online estimations of product quality.
This involves the integration of machine learning, uncertainty quantification -notably via conformal prediction-, and explainability to ensure robust deployment in complex industrial environments.

Prior to that, my research focused on machine/deep learning and uncertainty quantification applied to:

• Robotics - Adaptive & Intelligent Robotics Lab, Imperial College London: model-based Quality-Diversity optimization
• Neuroscience - CNRS CerCo: CNN-based oscillation detection in neurophysiological data
• Healthcare - EuroMov DHM: image processing and biomedical signal analysis

Industrial Collaboration

I am actively engaged in industrial collaborations as part of my PhD, having worked with two startups, one of which I am still collaborating with. My goal is to translate research outcomes into practical, industry-ready solutions.

Possible collaboration formats include:

• Targeted research internships
• Part-time collaborations (20 hours/week)
• Joint research projects with defined milestones and deliverables

Domains of interest include Neuroscience, BioAI, Retail, Marketing, Manufacturing, Finance, Robotics and Supply Chain.

If you are interested in collaboration or wish to discuss potential applications of my research, please contact me.