Stochastic Planning in the Textile Supply Chain: How Robust is a Newsboy Model ?

Pierre L. Douillet, Besoa Rabenasolo

École Nationale des Arts et Industries Textiles
Laboratoire Génie et Matériaux Textiles - GEMTEX
9, rue de l'Ermitage 59056 Roubaix, France
{pierre.douillet, besoa.rabenasolo}@ensait.fr

 

Summary:

One of the challenging problem in the area of the textile supply chain management is the identification of the best decision a retailer can make when he has to buy a certain quantity of a given good for the future sales season. Obviously, the answer depends on the knowledge available upon the future trends of the market. When expressing this knowledge by a probability distribution, many assumptions are often introduced that are not founded on an actual knowledge, but most of the time on computational ease. In fact, the knowledge one ever can have upon the probability distribution of the future demand is far lower than the limited knowledge he can have over the future demand itself. Despite the impossibility of an exact knowledge on this future demand, the sourcing decision (here: the order quantity), may be the good one in the sense that its optimality is preserved from any possible demand model errors. It should then depend only on exact and verifiable information over the future demand. Therefore it is interesting to investigate how robust are the hints for an optimal order quantity that can be extracted from what is really known upon this future demand. This robustness will be studied by extending the well known newsboy problem to a family of demand models.