Olivier Cardin’s research while affiliated with École Centrale de Nantes and other places

Various works propose solutions addressing the sustainability of IoT technologies to reduce their energy consumption, especially in the domain of wireless sensor networks. The diversity of applications, as well as the variability of their long-term constraints, forces them to dynamically adapt the network through time. Accordingly, this study formalizes the SADHoA-WSN framework to tackle the reconfiguration process. This proposal is a dynamic Holonic Control Architecture, linking the physical network evolution to the decisions made by a virtual multi-agent control system. The potential of such an approach is demonstrated by applying this framework to the energy optimization of communicating materials, i.e., materials equipped with inner wireless sensor nodes. The first implemented components of SADHoA-WSN and their related experimental results validate it as a promising energy-efficient dynamic methodology. This work lays the groundwork for optimized energy control in IoT networks.

The pick-and-place task is a common activity performed by robots in industrial settings. Traditional methods rely on fixed holders to maintain object positions, but this approach lacks flexibility and increases installation complexity and costs. To address this limitation, computer vision-based solutions are proposed to enable more flexible pick-and-place applications. Additionally, employing multiple robots in a shared workspace offers advantages such as increased performance and the ability to achieve complex objectives. However, coordinating and allocating tasks among multiple robots while ensuring collision avoidance remains a challenge. This study focuses on task allocation and motion planning, presenting two strategies: a predictive approach based on simulation optimization and a reactive approach based on control by multi-agent systems. The predictive approach utilises simulation optimization to anticipate collisions and optimizes task allocation, while the reactive approach emphasizes real-time decision-making and coordination among autonomous agents. The paper presents a case study for practical implementation on industrial robots.

The fourth industrial revolution aims to achieve greater flexibility and adaptability in manufacturing systems through the use of information and communication technologies. The Digital Twin technology has emerged as a promising solution to support human-centred decision-making in this context. Despite the growing interest in this area, there is still a lack of applications that integrate decision-support functionality and emphasize the relationship between real-time Digital Twin models and what-if simulation models. Hence, this paper discusses the integration of simulation models into a Digital Twin architecture to assist operators in making appropriate decisions. A proof of concept is presented to demonstrate the feasibility of this approach and to open up perspectives for further research in this area.

Reconfigurable Manufacturing Systems (RMS) have emerged as a promising approach to address the dynamic demands and uncertainties in modern manufacturing. However, the lack of a comprehensive and universally accepted definition of the notion of configuration in RMS poses challenges for researchers and practitioners. This research article aims to bridge this gap by suggesting a core reference model for RMS, allowing to exhibit a generic modelling of configurations. The article critically examines existing definitions and approaches to configuration, providing a comprehensive overview of the different dimensions and components of configuration in RMS. By addressing the lack of a consistent understanding of configuration in RMS, this article contributes to the development of a shared knowledge base and paves the way for further advancements in the field. It also highlights the importance of establishing a clear and holistic definition of configuration for effective implementation and utilisation of RMS in practice.

To ensure efficient dynamic manufacturing scheduling within a Cyber-Physical Production System (CPPS), it is essential to develop reactive control architectures. Heuristic-based optimization algorithms can provide this necessary reactivity and agility. In this paper, a hyper-heuristic is proposed. A set of atomic rules for resource selection are combined in a decisional strategy previously developed. The latter results from an optimization-simulation process. A new incremental learning mechanism is introduced in this article; it allows the system to evolve smoothly to integrate new events from the CPPS and its environment. A comparative study with a metaheuristic and heuristics on 56 instances, with family-dependent setup and processing times, demonstrates the interest of the proposed approach.