Inventory decisions for an assemble-to-order system with two ordering opportunities and multiple transportation modes

We consider inventory decision for an assemble-to-order system with two ordering opportunities and a one-time uncertain demand. The manufacturer can place her first order of the components ahead of demand realization with relatively lower prices. The second emergent order can be placed with relatively higher prices after demand is realized in case of component shortages. There are a number of lead times for each component representing different transportation modes for the second order. The manufacturer has to choose the one that best balance the time and cost tradeoff. We formulate the problem as a stochastic programming problem. We further consider a simpler version with a single component and multiple lead times. We propose several structural properties and a decompose algorithm to solve the problem efficiently.