‘‘Managing’’ Learning by Doing: An Empirical Study in Semiconductor Manufacturing

This article examines the contributions of human resource and organizational practices to the development and supply chain management interface. It addresses this issue in the context of the semiconductor industry by highlighting the importance of these practices for learning-based improvement in manufacturing. One of the most important factors for competitiveness in the semiconductor industry is the ability to manufacture new process technologies with high yields and low cycle times. The more effective management of new process technologies within the manufacturing facility aids firms in managing production costs, volumes, and inventories. Efficient management of new process development and introduction translates into enhanced internal supply chain management performance by improving the design of internal workflows, manufacturing performance, and the acquisition and installation of new manufacturing processes. Because much of the knowledge that underpins semiconductor manufacturing is idiosyncratic, however, firm-level differences in human resource and organizational practices are likely to have consequences for performance. The article derives learning curve models of the rate of improvement in manufacturing yield (i.e., the rate of learning) and cycle time (i.e., the speed of production) following the introduction of a new process technology in a manufacturing facility. It then tests the influence of the use by semiconductor manufacturers of teams for problem solving and intrafirm knowledge transfer, the level of internal adoption of information technology (IT), and more extensive and effective workflow and production scheduling systems on manufacturing performance. It finds that the manners in which semiconductor manufacturers allocate engineering resources to problem-solving activities, utilize information technology in the manufacturing facility, schedule production, and control the ‘‘shop floor’’ influence the levels and rates of improvement in both manufacturing performance measures. The article makes several contributions to the literature on product and process development and, accordingly, to research on the product development/supply chain interface. In particular, the model of organizational-based learning provides a better understanding of the determinants of learning-based performance improvement. In particular, better manufacturing performance results not strictly from greater cumulative volume but also from the