In an upstream supply chain (USC) dedicated to the mass production of customized products, decentralized management is possible and performing in the steady state, if all the links that precede the final assembly line use periodic replenishment policies. These policies require appropriate safety stocks of alternative or optional components. To achieve such performance in the real world, the supply chain must identify the source of any changes. Unexpected fluctuations in the production of USC plants suggest a bullwhip effect, yet most studies of the bullwhip effect fail to consider build-to-order supply chains. A double transformation of available information, derived from bill of materials explosions and time lags, is required to restore steady-state performance. It then remains to detect and quantify changes and, if a build-to-order strategy of alternative components is possible, use decision rules that are robust to such changes.
