Research

My research activities focus on diagnosis, monitoring and supervision of dynamic systems. The problems I consider are diagnosis, multi-model diagnosis, distributed diagnosis - diagnosability, predictability - health monitoring - anomaly detection.  I am interested in model-based methods as well as data-based methods based on machine learning. The synergy between these two families of methods is also one of my main interests. My research topics are integrated in the scientific strategy of INSA, at the level of the digital society issue by approaching the diagnosis from large masses of data of various and complex natures, at the level of the Mobility and infrastructures issue by proposing methods adapted for the embedded, but also at the level of the energy transition issue by developing diagnosis methods aiming at reinforcing the energy efficiency of the installations. My work developed in the context of automotive diagnosis, industry 4.0 and photovoltaic installations is in line with the strategic axes of LAAS, Transport&Mobility (member of the board), Industry of the Future and Energy.

Here some of my current research subjects:

  • Diagnosis and Diagnosability analysis of patterns with Time Perti nets
  • Certifed diagnosis of fault patterns based on Time Petri nets
  • Predictability and prediction of fault patterns based on Time Petri nets
  • Knowledge extraction for production optmization
  • Anomaly detection in space systems
  • Diagnosis of photovoltaic installations
  • Diagnosability analysis based on chronicles
  • Diagnosis and distributed diganosis based on chronicle recognition
  • Leaning for diagnosis and repair in garage