Thèse Jumeau Numérique pour le Pilotage Court et Moyten Terme dans les Industries de Traitement de Surface H/F - Doctorat.Gouv.Fr

Établissement : IMT Mines Albi École doctorale : SYSTEMES Laboratoire de recherche : CGI - Centre de Génie Industriel Direction de la thèse : Jacques LAMOTHE ORCID 0000000331999876 Début de la thèse : 2026-10-01 Date limite de candidature : 2026-06-07T23:59:59 This PhD thesis focuses on the digital twin for short- and medium-term production management in the surface treatment industry. This industry has several unique characteristics (such as make-to-order production on customer-supplied parts, very short lead times, a highly diverse customer base, and significant product variety) which shift the challenges of planning toward capacity agility to regulate work-in-progress between workshops.
The central objective of the thesis is to propose a digital twin system centered on finite-capacity simulation models to support short- and medium-term planning. It will involve designing complementary services to assist in this planning, including:
· Optimization of key decisions,
· Skills development and Management of multi-skilled operators,
· Order grouping to form batch sizes,
· Detecting changes on critical parameters for the models (e.g., process routes, macro-routes, product mix, skills availability, etc.).
The digital twin will be designed using a systems engineering approach that:
· Formalizes requirements,
· Anticipates the impacts of scalability and process evolution to propose an architecture adapted to these changes,
· Evaluates the added value of the digital twin.
Several scientific challenges have been identified in this research topic:
· Coupling discrete-event simulation and equivalent models (reinforcement learning) for finite-capacity simulation: While this topic has been widely studied in recent years (refs), the challenge lies in proposing an efficient, effective, and frugal coupling tailored to the decisions to be supported and the specificities of surface treatment processes (some highly automated, others highly manual).
· Coupling finite-capacity simulation and optimization in a distributed control context: The coupling of simulation and optimization has long been studied for production systems (refs). Similarly, bi-level planning (short/medium-term) is well-established (refs). However, the models and resolution methods need to be adapted to the underlying industrial context. Here, as the production system consists of workshops with a degree of autonomy, the challenge is to propose local optimization approaches for each workshop across at least two time horizons (a few days and a few weeks), taking into account their specificities (skill management, order grouping), while maintaining global control of priorities and inter-workshop collaboration via the digital twin.
· A systems engineering approach applied to the digital twin, accounting for the impacts of scalability and technological evolution on the digital twin's architecture.
The central objective of the thesis is to propose a digital twin system centered on finite-capacity simulation models to support short- and medium-term planning. It will involve designing complementary services to assist in this planning, including:
· Optimization of key decisions,
· Skills development and Management of multi-skilled operators,
· Order grouping to form batch sizes,
· Detecting changes on critical parameters for the models (e.g., process routes, macro-routes, product mix, skills availability, etc.).
The digital twin will be designed using a systems engineering approach that:
· Formalizes requirements,
· Anticipates the impacts of scalability and process evolution to propose an architecture adapted to these changes,
· Evaluates the added value of the digital twin. The main steps of the PhD include:
· Analysis of requirements and architecture of the digital twin,
· Definition of simulation models,
· Definition of optimization and coordination services,
· Application to a few examples from the industrial case study,
· Evaluation of the digital twin.

Formation
- Master ou ingénieur en informatique ou en génie industriel avec fort intérêt pour la recherche et les applications industrielles.
Compétences techniques
- Ingénierie Système
- Simulation à événements discrets et recherche oéprationnelle
- Compétences en développement informatioque (Java, Python)
Compétences transverses
- Anglais écrit et parlé (minimum level B2) ET Francais (minimum level B2)
- Autonomie et capacité à travailler dans une équipe de recherche.
- Motivation pour contribuer à une application industrielle de la recheche.
Compétences complémentaires (pas indispensable)
- Connaissances en apprentissage par renforcement et analyse de données massives pour la simulation
Documents demandés pour candidater: CV, lettre de candidature, résumé de la thèse de master, copies de publications, relevés de notes, lettres de recommendation (expérience de recherche et industrielle) et tout autre document jugé utile pour appuyer votre candidature.
Site de candidature : Il est demandé de déposere les canditatures sur le lien suivant : recruitee/*****
Application deadline: 7 Juin, 2026, 12:00 PM via le lien : https://institutminestelecom.recruitee.com/o/phd-thesis-proposal-digital-twin-for-short-term-planning-in-surface-treatment-industries
Convocation pour interview: vers le 15 Juin, 2026.
Contacts:
Jacques Lamothe, CGI IMT Mines Albi, ****@****.**
Séverine Durieux, CGI IMT Mines Albi, ****@****.**
Victor Romero, CGI IMT Mines Albi, ****@****.**