Assessing the efficiency of supply chain scheduling algorithms using data envelopment analysis

Where different supply chain planning algorithms are used, generally similar results may pose some challenges on the differentiating powers of evaluating different production schedules because of the increasing complexity of a supply chain network structure. For the comparison purpose, performance evaluation of different supply chain planning algorithms aims to use different supply chains models with different demands, capacities, and commonality through efficiency perspective by using a modified network rational data envelopment analysis (DEA) model. The proposed DEA model has the abilities: (1) to treat only undesirable outputs that exist without normal output, and the situation where input and output are both zero by introducing two new parameters to denote the maximum inventory and amount of delayed demands of a given node in a given time period; and (2) to evaluate the effect of the undesirable outputs/inputs on efficiency with assumption that they leave the system at the end of the current time period and re-enter the system at the beginning of the next time period. To prove the effectiveness of this DEA model, eighteen scenarios with different demands, capacities, and multiple periods are compared. In addition, this study tests the DEA model on a wafer testing/probing operation of a leading global semiconductor manufacturing and testing company in Taiwan by internal supply chain perspective. Results show that the DEA model proposed in this study can be used to assess the efficiency of a real-world operation with undesirable outputs/inputs, such as inventory and delayed demands.