Lean production and Lean philosophy have been developed, studied and implemented over the world since it was know from Toyota Production System. International manufacturers/OEMs have more or less developed their own Lean production systems. Over the years, arising challenges from new technology and internationalization have force the OEMs to understand and evaluate the design of supply chain to support their lean production system, and to manage the SC and compete as a whole against other supply chains. When apply Lean principles to the level of Supply Chain, coordination is an essential element to success. Today, researches have been done on joint efforts on decision making among supply chain actors, and applying Lean tools such as VSM on an extended level, namely a supply chain. Whereas the methods to improve SC coordination when implementing Lean through SC remain undeveloped. By interviewing Lean supply chain experts, experienced personnel in Supply Chain Management, and Lean experts from various organizations, a methodology to identify and visualize the improvement areas in supply chain coordination was designed. The methodology was then tested in a European OEM’s supply chain. Value Stream Mapping, as the base, was adapted to support identification and visualization of improvement areas in supply chain coordination. Several misalignments in the SC coordination were found in the current state, then visualized in current state map and approved by various SC actors. The methodology was then improved according to the process of testing. The purpose of this thesis is to develop a methodology to identify and visualize the improvement areas in (manufacturing) supply chain coordination within a Lean perspective.

Dans la recherche de performance de chaînes logistiques, la modélisation et la simulation de ces chaînes prenant en compte des aspects organisationnels s‘avèrent nécessaire. Dans cette perspective, nous proposons tout d‘abord un nouveau cadre méthodologique de modélisation et simulation orienté agents de chaînes logistiques, prenant en compte explicitement les aspects organisationnels, tant structurels que dynamiques, ainsi que des observables et indicateurs dans la modélisation de ces chaînes. Ce cadre méthodologique est structuré selon deux niveaux d'abstraction principale qui sont : un niveau conceptuel et un niveau opérationnel. Pour chacun de ces niveaux nous avons différents modèles qui sont proposés et décrit en détail. Ensuite, afin de simuler les modèles de simulation de ces chaines logistiques obtenus par ce cadre méthodologique, nous proposons une architecture logicielle multi-paradigmes spécifique. Cette architecture qui supporte l‘intégration de différentes plates-formes de simulation, est basée sur un médiateur, constituant un système multi-agents, et assurant l‘interaction entre les simulateurs. Les agents de ce médiateur sont chargés de collecter les informations permettant de produire les observables, maintenir la structure organisationnelle lorsque des entités organisationnelles (groupes) sont distribués sur différents simulateurs, l‘interopérabilité entre ces simulateurs, et enfin garantir la validité de la simulation d‘un point de vue temporel. Enfin, la mise en œuvre du cadre méthodologique et de l‘architecture logicielle proposés sont illustrées sur un cas industriel réaliste de chaîne logistique. La présentation des phases de conception est détaillée ainsi que la spécification et l‘implémentation du cas industriel au sein de l‘environnement de modélisation et de simulation proposé. / In the search for performance of supply chains, modeling and simulation of these chains, taking into account the organizational aspects are needed. In this perspective, firstly we propose a new methodology for modeling and simulation of supply chains oriented agents, taking into account explicitly the organizational aspects, both structural and dynamic as well as observable indicators in the modeling of these chains. This methodological framework is structured along two main levels of abstraction, conceptual and operational levels. For each level we have models that are proposed and described in details. Then, to simulate the simulation models of supply chains obtained by this methodological framework, we propose a multi-paradigm software architecture specific. This architecture supports the integration of different platforms simulation is based on a mediator, providing a multi-agent system, and ensuring the interaction between the simulators. The agents of this mediator is responsible for collecting required information to produce observables and maintaining the organizational structure while the organizational entities (groups) are distributed over different simulators and interoperability between these simulators, and finally ensuring the validity of the simulation for a temporal point of view. Finally, the implementation of the methodological framework and the proposed software architecture are shown in a realistic industrial case supply chain. The presentation of the design phase is detailed and the specification and implementation of industrial case in environmental modeling and simulation proposed.