A Priority-Driven Finite Capacity Planning System with Enhanced Shifting Bottleneck Algorithm

Abstract

This paper presents the details of a priority-driven finite capacity planning system to address the capacity aggregation issue of the traditional rough-cut capacity planning (RCCP) approach. The system overcomes the limitation of infinite capacity consideration in the traditional RCCP approach through detailed system modeling. The capacity planning process starts by establishing capacity availability through the user-defined production calendars and machine unavailability time periods such as planned machine preventive maintenance schedule. The available capacity information is represented by building machine time lines with finite time buckets down to an increment of minute. Machine loading preferences and standard processing times are then specified in the form of capacity matrices. With the detailed capacity modeling approach, all information needed to model the capacity constraints could be precisely stated. The enhanced priority-driven finite capacity engine can be configured to consider weighted product and machine priorities; product forecast ratio, linkages to critical tooling and fixture constraint, as well as, the ability to cater for shifting bottleneck during the dynamic capacity allocation process. Through the intelligent capacity planning algorithms, the demands are assigned to the available capacity on a level-by-level approach. The priority-driven finite capacity planning system has been implemented in a few companies in the semiconductor backend assembly environment and it has proven to be a practical and effective capacity planning solution based on the encouraging feedback from the end users.