Supply chain

Applications :
  • Manufacturing
  • Batch processes
  • Energy production and consumption
  • Health

supply chain example

Methods:
  • (Integer)linear programming
  • heuristics and metaheuristics
  • constraint programming
  • stochastic processes
  • game theory

supply chain model

Five major axis are being explored in the supply chain management area.

In a first axis, different management modes have been modeled: distributed, centralized and hybrid structures. A simulation tool relying on a commercial optimization software has been developed in order to assess the properties (robustness and reactivity) related to the proposed approaches when unforeseen events occur and modify data (demand, material supply, resource capacities).

The second axis focuses on the planning process of an entity within a chain and suggests an approach aiming at improving the coordination between supply chain’s partners (customers, suppliers, subcontractors, see figure). A special attention is devoted to the modeling of temporal features that are associated to the different elements of the studied system (cycle times in the production, delivery delays of suppliers, periodic updating of data). A dynamic planning process and a simulation tool are proposed in order to manage firm and flexible demands.

The robustness concept is the central issue of a third axis based on a two-level decisional structure. Temporal aggregation mechanisms are used at the upper level in order to determine a mid-term plan, robust towards features unknown at this level. A “guiding” plan is then derived from this aggregate plan and transmitted to the lower level which deals with short term decisions; a dynamic disaggregation is then performed which enables to react to uncertainties and disturbances. This kind of approach formalizes industrial planning processes that are frequently used.

A fourth original axis relying on queuing theory and on game theory investigates two important issues in an uncertain context: the definition of supplying policies when different suppliers have uncertain delivery delays and the definition of optimal contract-based policies between adjacent entities within a supply chain.

Finally, as a fifth axis, a supply chain can also be considered as a dynamic system in which inputs and outputs correspond to the product flows. These systems are usually subject to constraints due to the internal structure of the chain (stock capacity) or the nature of products. Considering that, we propose some controllers computed using (max,+) algebra.

Contacts : Laurent Houssin, Colette Mercé,

PhD student : Hassen Gharbi

Selected publications (see complete list on LAAS server) :

  • J.FRANCOIS, J.C.DESCHAMPS, G.FONTAN, J.P.BOURRIERES, Collaborative planning for enterprises involved in different supply chains, International Conference on Service Systems & Service Management (ICSSSM'06), Troyes (France), 25-27 October 2006, pp.1466-1471

  • F.GALASSO, C.MERCE, B.GRABOT,  Decision support for supply chain planning under uncertainty, International Journal of Systems Science, 29p., January 2008

  • H.GHARBI, C.MERCE, G.FONTAN, M.MOALLA, Planification réactive d'un partenaire d'une chaîne logistique, Workshop International: Logistique et Transport 2007 (LT'2007), Sousse (Tunisie), 18-20 Juillet 2007, pp.173-18.

  • L.HOUSSIN, S.LAHAYE, J.L.BOIMOND. Just-In-Time Control of Constrained (max,+)-linear Systems. Journal of Discrete Event Dynamic Systems, 17(2):159-178, 2007

  • J.C.HENNET, Y.ARDA, Supply chain coordination; a game theory approach. Engineering Applications of Artificial Intelligence, 2007, DOI:10.1016/j.engappai.2007.10.003