PSI - Industrial Systems Management
Team topics
Expertise in flow simulation
The team has an industrial licence for several software products on the market, allowing it to make its expertise available to companies in need of advice. In addition, the team is actively involved in the integration of this type of tool directly into the system management process to provide these systems with short-term behavioural forecasting functionalities and in the native integration of energy consumption into the performance evaluation of these systems.
Optimal control of industrial systems
Currently, most industrial systems are controlled by scheduling tools based on Operational Research theories. This type of tool has proven its effectiveness, but many heaviness and rigidity can be observed in the case of disrupted systems. Based on the laboratory's expertise in Control Theory, we propose to use the theories of Automatics to propose new tools for optimal control of industrial systems, such as AGV fleet routing algorithms for example.
Taking into account the human factor in scheduling policies
In general, an effective Man-Machine relationship established during the scheduling process makes it possible to improve the results obtained by taking into account the human expertise in the calculation. In the case of online system rescheduling, this expertise is all the more important in order to ensure that the proposed solution and the disruption causing the rescheduling are in line. In this sense, we propose the use of tools and associated performance evaluations to optimize this relationship and improve results.
Design of Cyber-Physical Production Systems
The traditional frontier between computer network and industrial network will disappear in the coming years in order to integrate all the elements of the industrial system into the management performance. In addition, the rigid hierarchization currently observed in most systems will tend to be replaced by decentralized decision-making, a characteristic of heterarchy that allows flexibility and robustness in response to the risks encountered. We propose the use of Multi-Agent tools for real-time system control, allowing the integration of a complete control system, from the cloud to sensors and actuators in the field. This theme is expressed through the implementation of industrial-scale demonstrators in the workshops of the IUT in Nantes.
