Rémi Pannequin

Another interpretation of stigmergy for product-driven systems architecture

Product-driven systems (PDS) may enable manufacturing companies to meet business demands more quickly and effectively, by localizing decision as near as possible to the material flows. However, the actual architecture of a PDS is still not well defined. This paper proposes a PDS architecture based on a particular interpretation of the concept of stigmergy, where cooperation between production actors is achieved thanks to attributes (informational pheromones) attached to products. This stigmergic design pattern is presented. Agent-oriented components which implement it are presented, are firstly applied to a laboratory platform, and secondly on an industrial test-case.

Stigmergy: A Design Pattern for Product-Driven Systems

This paper proposes a new interpretation of stigmergy, where cooperation between production actors is achieved thanks to attributes (informational pheromones) attached to products. From this interpretation is introduced a new design pattern that can be used to develop product-driven systems. Agent-oriented components which implement it are presented, and then applied on an experimental platform.

Emulica: an emulation-based benchmarking framework for production control experiments1

Abstract Evaluating a proposed control architecture nearly always requires experiments and applications. This paper propose an emulation-based benchmarking framework to design and run such experiments. The proposition is based on an architecture integrating the control system with the virtual controlled system (emulation model), and on generic constructs to build this emulation model. The main feature of this architecture is presented in this paper : the constructs used to build emulation model and their generic interface with the control system. Then, Emulica, an open-source implementation of the benchmarking environment is presented.

Centralized versus Distributed Decisions, an Industrial Case

Abstract Two tendencies have appeared in manufacturing plants control. On the one hand, there are new management tools, such as APS, which consider the supply chain as a whole, and schedule operations globally and in real time. On the other hand, new control paradigms, such as holonic manufacturing and auto ID seems promising, as enablers of mass-customization manufacturing or agile enterprise. So as to clarify relation between these two systems, we develop two models of an assembly factory, each implementing a different control paradigm. The comparison between them leads us to our future path of research.

COOPERATION BETWEEN BUSINESS AND HOLONIC MANUFACTURING DECISION SYSTEMS

Abstract New developments in production control can be grouped around two modes: distributed and centralised control. Interaction between these control modes can be seen as a B2M issue. We propose in this paper an alternative B2M architecture, using holonic products, in order to enable cooperation between centralised business and distributed manufacturing decision systems. We will first give the requirements of such a system, then outline the structure of the holonic product and finally study the possible interaction protocols between the products and the decision system.