Segmentation Models for the Duration of Expected Preventive Maintenance in Semiconductor Fabs

The execution and duration of preventive maintenance (PM) activities have a significant impact on whether effective control is maintained over factory equipment while ensuring that the equipment is healthy and available for production. Although most PMs in the industry have similar structures, the task of generating an automated segmented view of the PM duration and flow to highlight inefficiencies and corresponding opportunities for improvement is complicated by a reliance on manual data entries by human operators. In addition, most optimization and analysis models treat a PM as a single activity while ignoring its parts. In this paper, two models for segmented PM durations are proposed: a parametric model to describe expected PM duration and a practical non-parametric model to generically segment a PM into its parts. Both models provide new insight into PM duration and a clear method to identify inefficiencies in PM duration.