DESCRIPTION :
Complex cyber-physical systems are evolving at an accelerating pace, operating in increasingly dynamic environments and contending with ever-increasing uncertainty. This requires a high level of adaptability, through a continuous engineering of complex cyber-physical, socio-technical, ecosystems. Digital twins are key enablers, and leverage on both model simulation and data science. Modeling & Simulation is a time-honored activity consisting in building complex analytical models to be simulated to evaluate natural or engineered phenomena. Conversely, data science relies on the availability of data to build complex predictive AI-based learning models. While both could be confused or even opposed, we argue they better complement each other to enhance the ability to best engineer complex systems continuously.
The sound hybridization of model simulation and data science enables a coordinated use of both techniques in complex scenarios (e.g., analytical models for explanation, and data model for recurrent pattern retrieval). Moreover, the hybridization also opens the door to adaptive modeling, where one model is inferred or refined by the others, and vice-versa (e.g., inferring or refining an analytical model from a learning model, and better tuning and explaining a learning model thanks to an analytical model).
Challenges are related to the identification of relevant patterns, and their proper implementations with well-defined interfaces for each model and the required protocols and operators to support the proposed scenarios. We aim to establish the first unifying theory for both model simulation and learning models, and demonstrate its applicability in practice within digital twins for mechanical engineering.
References
* R. Eramo, F. Bordeleau, B. Combemale, M. v. d. Brand, M. Wimmer and A. Wortmann, "Conceptualizing Digital Twins," in IEEE Software, vol. 39, no. 2, pp. 39-46, March-April 2022,
* R. Verdecchia, L. Cruz, J. Sallou, M. Lin, J. Wickenden and E. Hotellier, "Data-Centric Green AI An Exploratory Empirical Study," 2022 International Conference on ICT for Sustainability (ICT4S), Plovdiv, Bulgaria, 2022, pp. 35-45
* Narciso, Diogo AC, and F. G. Martins. "Application of machine learning tools for energy efficiency in industry: A review." Energy Reports 6 (2020): 1181-1199.
* Ahmad, Muhammad Waseem, Monjur Mourshed, and Yacine Rezgui. "Trees vs Neurons: Comparison between random forest and ANN for high-resolution prediction of building energy consumption." Energy and buildings 147 (2017): 77-89.
* Zendehboudi S, Rezaei N, Lohi A. Applications of hybrid models in chemical, petroleum, and energy systems: A systematic review. Applied Energy. 2018;228:2539-2566. doi:10.1016/j.apenergy.2018.06.051.
* Slater L, Arnal L, Boucher MA, Chang AYY, Moulds S, Murphy C, et al. Hybrid forecasting: using statistics and machine learning to integrate predictions from dynamical models. Hydrology and Earth System Sciences Discussions. 2022;2022:1-35.
* Syauqi A, Pavian Eldi G, Andika R, Lim H. Reducing data requirement for accurate photovoltaic power prediction using hybrid machine learning-physical model on diverse dataset. Solar Energy. 2024;279:112814. doi:https://doi.org/10.1016/j.solener.2024.112814.
* Mayer MJ. Benefits of physical and machine learning hybridization for photovoltaic power forecasting. Renewable and Sustainable Energy Reviews. 2022;168:112772. doi:https://doi.org/10.1016/j.rser.2022.112772.
* Rudolph M, Kurz S, Rakitsch B. Hybrid modeling design patterns. Journal of Mathematics in Industry. 2024;14(1):3.
* von Rueden L, Mayer S, Sifa R, Bauckhage C, Garcke J. Combining machine learning and simulation to a hybrid modelling approach: Current and future directions. In: Advances in Intelligent Data Analysis XVIII: 18th International Symposium on Intelligent Data Analysis. Springer; 2020. p. 548-560.
* von Rueden L, Mayer S, Beckh K, Georgiev B, Giesselbach S, Heese R, et al. Informed machine learning-a taxonomy and survey of integrating prior knowledge into learning systems. IEEE Transactions on Knowledge and Data Engineering. 2021;35(1):614-633.
* Thummerer T, Mikelsons L. Learnable & Interpretable Model Combination in Dynamical Systems Modeling; 2025. Available from: https://arxiv.org/abs/2406.08093.
* Hajirahimi Z, Khashei M. Hybridization of hybrid structures for time series forecasting: a review. Artificial Intelligence Review. 2022;56(2):1201-1261.
* Schweidtmann AM, Zhang D, von Stosch M. A review and perspective on hybrid modeling methodologies. Digital Chemical Engineering. 2024;10:100136.
* Yang S, Navarathna P, Ghosh S, Bequette BW. Hybrid modeling in the era of smart manufacturing. Computers & Chemical Engineering. 2020;140:106874.
* Wang J, Li Y, Gao RX, Zhang F. Hybrid physics-based and data-driven models for smart manufacturing: Modelling, simulation, and explainability. Journal of Manufacturing Systems. 2022;63:381-391., The application domain of this work involves Mechanical equipments or systems made of several mechanical equipments. Several potential industrial applications in the field of Process equipment (fluid systems, specific components as valves,…), Mobile (off-road) working machines (as forklift or parts of it) and Production machines (welding robot, machining,…) are targeted. Therefore the developped pattern should be generic enough to encompass the aforementioned applications. Nevertheless, the existing thermal-hydraulic loop (JNEM) available at the Cetim facility may be use as a support for the involved developpment.
The JNEM loop is representative of an industrial process loop. It's a closed, instrumented hydraulic loop. It is equipped with a pump, a heat exchanger, a tank, a regulation valve and three piping sections. Its function is simply to provide the flow, pressure and/or temperature requested by the operator. Several control devices have been added to generate some defects artificially in the future.
This digital twin is designed to serve several purposes: predictive maintenance, optimization of process loop settings and decision support. The main objectives are to:
* Detect, localize, and estimate variations in process parameters (pipe clogging, heat exchanger performance degradation, valve dynamic behavior changes, etc.) through comparison with process parameters measurements (flow, pressure, temperature) at several locations of the physical system
* Optimization of the process loop settings (pump speed, valve opening) to reach target process parameters (flow, pressure, temperature) according to operator requirements (minimization of the time to reach the target, minimization of the energy consumption…) thanks to the simulation of different scenario using the digital twin. The "best" scenario is then automatically applied on the physical system through the driving of the involved actuators or through operator validation.
* Provide monitoring and prediction capabilities: use of virtual sensor to estimate and predict process parameters, such as flow rate (in the event of a flow meter failure), allowing for real-time monitoring and control of the physical system
Environment
This PhD is funded by the CETIM (the French Technical Center for Mechanical Industries) in the context of a collaboration with Inria (the national center for research in computer science). The main advisors of the PhD thesis will be Prof. Benoit Combemale (Inria, DiverSE team), Prof. Julien Deantoni (Université Cote d'Azur, I3S/Inria Kairos team), and Dr. Yoann JUS / Hubert LEJEUNE (CETIM).
The candidate will be involved either in the Inria DiverSE team (Rennes) or in the Inria Kairos team (Sophia-Antipolis), and will register either to the doctoral school in computer science of the University of Rennes or to the one from the University Côte d'Azur accordingly.
The DiverSE team is located in Rennes, France. DiverSE's research is in the area of software engineering. The team is actively involved in European, French and industrial projects and is composed of 13 faculty members, 25 PhD students, 4 post-docs and 8 engineers.
The Kairos team is located in Sophia Antipolis, France. Kairos's research focuses on continuous engineering, promoting synergies between heterogeneous artifacts throughout the whole development lifecycle. The team is actively involved in several projects at the regional, national, and European levels. It is composed of 6 lecturer-researchers, 4 PhD students, and 2 postdocs.
Code d'emploi : Ingénieur Mécanique (h/f)
Domaine professionnel actuel : Ingénieurs, Projeteurs et Techniciens Ponts et Chaussées
Temps partiel / Temps plein : Plein temps
Type de contrat : Contrat à durée indéterminée (CDI)
Compétences : Modélisation 3D, Intelligence Artificielle, Analyse des Données, Programmation Informatique, Ingénierie de l'Information, Système d'Aide à la Décision, Informatique de Bureau, Systèmes Dynamiques, Machine Learning, Langages de Modélisation, Cyber-physical Systems, Sap Business Information Warehouse, Informatique Scientifique, Conception et Développement de Logiciel, Random Forest, Technologies Informatiques, Anglais, Français, Adaptabilité, Implication et Investissement, Travaux de Soudure, Génie Chimique, Modélisation des Données, Méthodes de Conception, Technologie Énergétique, Organisation d'Événements, Mesure de Débit, Débit Volumétrique, Mécanique des Fluides, Elaboration des Prévisions, Conduite de Chariot Élévateur, Énergies Renouvelables, Échangeur de Chaleur, Systèmes Hydrauliques, Hydrologie, Génie Mécanique, Conduite de Machine, Mathématiques, Gestion de Projet, Neurones, Analyse Numérique, Solaire Photovoltaïque, Sciences Physiques, Maintenance Prédictive, Analyses Prédictives, Travaux sur Vannes, Programmation de Systèmes de Production, Pompes, Moule, Simulations, Etudes et Statistiques, Durabilité, Séries Chronologiques, Industrie Pétrolière, Compétences de Modélisation, Marque (Branding), Science des Données, Écosystèmes, Performance Energétique, Actionneurs
Courriel :
Julien.Deantoni@inria.fr
Téléphone :
0299847100
Type d'annonceur : Employeur direct