Build-to-order supply chains (BOSCs) have recently received increasing attention due to the shifting focus of manufacturing companies from mass production to mass customization. This shift has generated a growing need for efficient methods to design BOSCs. This research proposes an approach for BOSC design that simultaneously considers multiple performance measures at three stages of a BOSC – Tier I suppliers, the focal manufacturing company and Tier I customers (product delivery couriers). We present a heuristic solution approach that constructs the best BOSC configuration through the selection of suppliers, manufacturing resources at the focal company and delivery couriers. The resulting configuration is the one that yields the best global performance relative to five deterministic performance measures simultaneously, some of which are nonlinear. We compare the heuristic results to those from an exact method, and the results show that the proposed approach yields BOSC configurations with near-optimal performance. The absolute deviation in mean performance across all experiments is consistently less than 4%, with a variance less than 0.5%. We propose a second heuristic approach for the stochastic BOSC environment. Compared to the deterministic BOSC performance, experimental results show that optimizing BOSC performance according to stochastic local performance measures can yield a significantly different supply chain configuration. Local optimization means optimizing according to one performance measure independently of the other four. Using Monte Carlo simulation, we test the impact of local performance variability on the global performance of the BOSC. Experimental results show that, as variability of the local performance increases, the mean global performance decreases, while variation in the global performance increases at steeper levels.

No description available.