چارچوب هم آوایی برای تجزیه و تحلیل استراتژی های جایگزین در تست نهایی IC برای افزایش اثربخشی کلی عملیاتی

Owing to capacity limit, yield demand, and cycle time reduction, determining proper strategy for the final testing of integrated circuits (IC) device is critical. Since none of the tests can perfectly distinguish good devices from bad, alternative testing strategies consisting of various setups and testing procedures affect the testing results and testing cycle time. However, this problem has seldom been addressed in literature. This study aims to construct a decision framework to analyze alternative testing strategies and thus derive the optimal strategy balancing operational efficiency, cost, and risk. This framework has been implemented in a semiconductor-testing firm in Taiwan. The results demonstrate practical viability of the proposed framework.

Semiconductor manufacturing is a capital-intensive industry in which on-time delivery, cycle time reduction, fast ramp-up, yield enhancement, and continual cost reduction are the important ways for operational excellence to maintain competitive advantages (Chien and Wu, 2003). The selection of final testing strategies of IC devices is critical in semiconductor companies owing to capacity limit, yield requirement, and cycle time reduction. Because of lengthy wafer fabrication an IC production processes with various sources of variability being involved in the semiconductor supply chains, the final testing of IC devices as the final step in IC production is critical in order to overcome bullwhip effect. Indeed, an IC-testing strategy is a portfolio (Chien, 2002) of different testing procedures with corresponding setups and recipes. Alternative testing strategies consisting of various tests and setups will affect the sensitivity and specificity of testing results and testing cycle time. On one hand, false-passed products will not only cause customer complaints and product returns, but also impact company reputation. On the other hand, false-failed products will cause loss of IC manufacturer and may also affect normal delivery. In practice, a tradeoff is implicitly made among the decision attributes of cost, quality, and risk to design and select testing strategies. However, little research has been done to deal with such decisions systematically and effectively. The difficulty can be traced at least in part to the lack of a decision framework in which the interrelation among the decision elements can be structured and then the involved uncertainty and tradeoff can be analyzed explicitly and systematically. This study aims to propose a framework for analyzing alternative configurations of testing strategies by considering the tradeoffs among throughput, testing cycle time, and the associated costs and risks to enhance overall operational effectiveness of IC final testing. In particular, the proposed framework integrated statistical decision analysis and cost to model the alternative testing strategies and the involved uncertainties while considering the objectives of increasing testing profit and customer satisfaction. For validation, we conducted an empirical study in a final testing factory of a leading semiconductor company in Taiwan, whose primary products are mask read-only-memory (Mask ROM) and flash memory. The proposed framework can assist the decision makers to design and select testing strategy configuration in light of the condition of testing facility as well as the decision context of semiconductor supply chain. The rest of this paper is organized as follows. Section 2 presents related studies. Section 3 details the proposed framework. Section 4 describes an empirical case study in a semiconductor final testing factory to validate the proposed framework. Section 5 concludes this study with discussions of the contribution of this study and future research directions.

Decisions under uncertainty are frequent and critical in high-tech industries from manufacturing strategy to operational decisions in which the decision makers should trade off among multiple attributes including various risks and cost. The selections of testing strategies largely affect the efficiency of test machine, yet little research has been done to deal with the present problem. This study proposes a UNISON framework based on statistical decision analysis to systematically structure the problem and effectively analyze alternative strategies for IC final testing. The proposed framework provides the means to align the fundamental objectives (e.g., profit and customer satisfaction) of a company with operational decisions of selecting testing strategies. The empirical results showed practical viability of the proposed framework. Indeed, similar approaches can be employed to solve other decision problems that will affect operational effectiveness of high-tech manufacturing. Further research is required to collect further data to demonstrate the usability and validity of this framework in different contexts.