Modeling resources in research and development semiconductor manufacturing shop floors

Research and Development (R&D) semiconductor manufacturing systems differ from production facilities in various aspects, including the level of characterization and control over engineering processes, the scales of production control, and the range of automatable decision-making processes. As part of an ultimate goal to expand process capabilities and activity volume, this paper focuses on the explicit modeling of resources and their associated states in an R&D shop floor. Three types of resources are characterized and modeled, and their capabilities are analyzed: knowledge about engineering processes (resp., device team holds), workforce (resp., the number of equivalent operators), and machinery (resp., time to repair, time between failures). Particular attention is paid to the variability and granularity, under and out of control, of key aspects characterizing the resources of interest and their impact on operations management decisions. Numerical experiments are conducted and analyzed based on real-life instances.