Supply chains are typically exposed to several forms of uncertainties; for example, demand uncertainty in inventory planning, lead time uncertainty in production planning, travel time uncertainty in distribution management. In some cases, failure to incorporate uncertainties in the decision making process could lead to an unacceptable level of loss in demand, service level and loss of goodwill. To deal with these uncertainties, two main solution approaches in operations research, namely stochastic programming and robust optimization, can be applied to formulate the optimization models with uncertain parameters. In many applications, however, the resulting reformulations are much more difficult to solve than their deterministic counterpart and thus become intractable. In this talk, we present several decomposition based techniques that can be applied to speedup the solution processes of several supply chain applications, namely stochastic production/inventory routing, stochastic and robust vehicle routing, stochastic multi-item inventory planning as well as uncertain Markov decision processes.
Bienvenue à tous!