Keywords: activities pooling, resources sharing, Industry 5.0, multiagent systems, machine learning algorithms, self-organization management, multi-criteria resolution.
Objectives: the PhD objectives are:
1. To model the various stakeholders in the production system using a behavioral and selforganizing approach, highlighting all resources (human or otherwise), including the product/new series to be produced, as well as their links and interactions, and their different needs in terms of services. The idea here is to adopt a product/new series-oriented approach, in a context of recombination/reconfiguration/mutualization of workstations according to the product to be produced;
2. To associate these different stakeholders with an intelligent agent, and design the resulting multi-agent system as well as the complete system architecture enabling the different levels (internal and external to the production system) to be defined. The aim here is to fully integrate robots, cobots and AGVs into activities, so as to improve man-machine interactions and the definition of collaborations between stakeholders. A vision based on calls for tender and/or partnerships could be envisaged, using game theory as a starting point;
3. To identify evolutionary trajectories based on identified behaviors, environmental and social constraints, and possible disturbances or failures in the system under consideration.
4. To propose a decision-support tool enabling the integration and comparison of different control and reconfiguration strategies. This will be defined using dynamic multi-criteria resolution.
