Research and Development (R&D) semiconductor manufacturing environments are characterized by large equipment diversity, heterogeneous process configurations, and limited automation. These characteristics lead to significant variability and limited predictability in processing times, even under identical nominal operating conditions. Unlike industrial-scale facilities, where stable operating conditions facilitate standardized recipes and automated control, automation in R&D fabs must remain flexible and is often bypassed to accommodate novel experimental steps. As a result, Manufacturing Execution Systems (MES) tolerate on-the-fly recipe modifications, and operators frequently intervene manually, further amplifying processing-time variability.

This paper addresses this gap by proposing an approach to derive equipment processing-time profiles by integrating Manufacturing Execution System (MES) traces and equipment-control data. The resulting processing-time approximations are intended to feed proactive scheduling, cycle-time estimation, and the existing simulation-based decision support tool tailored to the specific variability and constraints of R&D environments.