Sourish GHOSH will defend his thesis on Monday, March 23, 2026, at the Arts et Métiers campus Arts et Métiers . His work at the MSMP laboratory focuses on "Artificial intelligence-assisted surface metrology for machining: an integrated approach to evaluating surface quality in machines."
Title
“AI-Assisted Surface Metrology for Machining: An Integrated Approach to In-Machine Surface Quality Assessment”
Jury
• Ms. Caroline RICHARD, Professor, University of Tours – Rapporteur
• Mr. Abdelali OUDRISS, Professor, Mines Paris PSL - Rapporteur
• Mr. Arnaud GOTLIEB, Professor, Simula Research Laboratory – Examiner
• Mr. Mohamed EL MANSORI, University Professor, ENSAM – Examiner
• Mr. Ricardo KNOBLAUCH, Associate Professor, ENSAM – Examiner
Thesis summary
Surface quality is a key factor in the performance and reliability of machined components. As production systems evolve toward “zero-defect” strategies and in-process control, it becomes essential to link machining behavior, surface formation, and measurement reliability within a coherent framework. This doctoral thesis proposes an integrated methodology combining cutting mechanics, signal analysis, and artificial intelligence to predict and control surface quality during machining.
Interpretable AI models are being developed to establish the link between signals, machining parameters, and surface roughness, thereby forming the basis of a proactive quality control framework. At the same time, multiscale analysis techniques are being introduced to characterize texture evolution beyond conventional roughness indicators, while AI-based approaches are being explored to improve the robustness of non-contact optical measurements.
The objective of this thesis is to contribute to the development of intelligent and reliable manufacturing systems by enabling earlier detection of surface defects and supporting data-driven decision-making in machining processes.
24th International Conference on Metrology and Properties of Surfaces (Met&Props) - May 27-29, 2024 in Marrakech (Morocco)
