Composition of modules' stock using simulated annealing

An assemble to order policy considers a trade-off between the size of product portfolio and assembly lead time. The concept of modular design is often used to implement the assemble to order policy. Modular design impacts assembly of products and the supply chain. In particular, storage, transportation and production are affected by the selected modular structure. For a fixed number of modules, a stock composition that minimizes the mean final assembly time is difficult to establish. This problem is highly combinatorial: it is not achievable to look for an optimal solution. A first step leads to the elaboration of heuristics. This article proposes to combine the results operated with this heuristics to the use of a simulated annealing algorithm. Computational results showed that this method could be efficient