Computer Vision and Machine Learning approaches for Fault Detection and Diagnosis in Chemical Pilot Plants

IFP Energies nouvelles (IFPEN) est un acteur majeur de la recherche et de la formation dans les domaines de l’énergie, du transport et de l’environnement. De la recherche à l’industrie, l’innovation technologique est au cœur de son action, articulée autour de quatre priorités stratégiques : Mobilité Durable, Energies Nouvelles, Climat / Environnement / Economie circulaire et Hydrocarbures Responsables.

Dans le cadre de la mission d’intérêt général confiée par les pouvoirs publics, IFPEN concentre ses efforts sur :

• l’apport de solutions aux défis sociétaux de l’énergie et du climat, en favorisant la transition vers une mobilité durable et l’émergence d’un mix énergétique plus diversifié ;
• la création de richesse et d’emplois, en soutenant l’activité économique française et européenne et la compétitivité des filières industrielles associées.

Partie intégrante d’IFPEN, l’école d’ingénieurs IFP School prépare les générations futures à relever ces défis.

Computer Vision and Machine Learning approaches for Fault Detection and Diagnosis in Chemical Pilot Plants

IFPEN is a major player in the triple energy, ecological, and digital transition by offering differentiating technological solutions in response to societal and industrial challenges of energy and climate. The development and deployment of intelligent monitoring systems for chemical pilot plants is crucial for ensuring safe, efficient, and sustainable operations.

Chemical pilot plants at IFPEN are complex systems involving multiple unit operations including reactors, separation units, heat exchangers, and control systems. Early detection of anomalies and accurate fault diagnosis are essential for preventing hazardous situations, equipment damage, minimizing downtime and optimizing performance.

The integration of advanced AI-based anomaly detection and fault detection & diagnosis (FDD) techniques offers significant potential for improving process monitoring, predictive maintenance, and operational decision-making.

Recent advances in machine learning, particularly in time series analysis (GAF, MTF, time series to image conversion), deep learning (GANs, Autoencoders, CNNs), and hybrid AI approaches (Transfer Learning, GAN-Autoencoder combinations, knowledge distillation), provide new opportunities to enhance industrial process monitoring. The challenge lies in adapting these techniques to the specific characteristics of chemical processes: multivariate sensor data, process dynamics, environmental variations, and the need for explainable AI in safety-critical applications.

Internship objective :

• Conduct a literature review of fault detection techniques, focusing on machine learning and computer vision approaches.
• Implement and adapt selected computer vision techniques for time series analysis approaches for anomaly detection and implement them using synthetic pilot plant data (Tennessee Eastman Process).
• Validate developed methods with real plant data.

Profile :

We are seeking a candidate with an engineering degree or pursuing a Master's (M2) in Applied Mathematics, Artificial Intelligence, Data Science, or Computer Science with machine learning background. Chemical engineering students with strong AI background are also encouraged to apply.

• Technical Skills: Strong foundation in machine learning and computer vision, with proficiency in Python and experience with deep learning frameworks (TensorFlow/PyTorch)
• Knowledge Areas: Familiarity with time series analysis, computer vision techniques for signal processing, multivariate data analysis, and real-time systems
• Programming: Experience with scikit-learn, pandas, OpenCV, and data visualization libraries

Soft Skills: Analytical mindset, problem-solving abilities, adaptability to industrial environments, interest in applied AI research and willingness to learn about chemical processes.