Asymptotically Optimal Inventory Control for Assemble-to-Order Systems

We consider assemble-to-order (ATO) inventory systems with a general bill of materials and deterministic lead times. Unsatisfied demands are always backlogged. apply four-step asymptotic framework to develop policies for minimizing the long-run average expected total cost. Our approach features multistage stochastic program (SP) establish lower bound on cost determine parameter values control. replenishment policy deviates from conventional constant base stock accommodate nonidentical component allocation differentiates based backlog costs, materials, availabilities. prove that our is asymptotically optimal diffusion scale, is, as longest time grows, percentage difference between under its converges zero. In developing these results, we formulate broad tracking model convergence results which optimality follows specialized corollaries. Funding: This study work supported by National Science Foundation [Grant CMMI-1363314].