inicioir al DIIbuscar
agent based simulation of softwood lumber supply chain
We present an agent-based simulator designed to test different supply chain configurations and planning approaches for the lumber industry. The architecture of the simulator will be exposed as well as a proposed simulation framework. We start by discussing the modelling of the specific supply chain components such as the business units, the planning units, their relations and conversation protocols, the products and the processes. Then, we expose a detailed application of the technology and illustrate the impact of different planning approaches. The simulator builds on FIPA standards and integrates optimization modules to support the distributed planning decisions. The application deals with a network of sawmill, drying and finishing facilities, warehouses as well as with customers characterized by different buying behaviours (spot, industrial and contract). An industrial implementation is presented as well as some numerical experimentation.