Integrating optimisation and simulation approaches for daily scheduling of assembly and test operations

Abstract

The purpose of this paper is to show how the results of an optimisation model that can be integrated with the decisions made within a simulation model to schedule back-end operations in a semiconductor assembly and test facility. The problem is defined by a set of resources that includes machines and tooling, process plans for each product and the following four hierarchical objectives: minimise the weighted sum of key device shortages, maximise weighted throughput, minimise the number of machines used and minimise the makespan for a given set of lots in queue. A mixed integer programming model is purposed and first solved with a greedy randomised adaptive search procedure (GRASP). The results associated with the prescribed facility configuration are then fed to the simulation model written in AutoSched AP. However, due to the inadequacy of the options built into AutoSched, three new rules were created: the first two are designed to capture the machine set-up profiles provided by the GRASP and the third to prioritise the processing of hot lots containing key devices. The computational analysis showed that incorporating the set-up from the GRASP in dynamic operations of the simulation greatly improved its performance with respect to the four objectives.

Improving performance of dispatch rules for daily scheduling of assembly and test operations

“…Scheduling thousands of lots a day to minimize key device shortages as well as optimize any number of other objectives is a complex problem for which more advanced analytic techniques are still needed. Further discussion of AT operations can be found in Bard et al (2015).…”

“…After defining hot lots and dynamically redefining their weights, a machine will try to select hot lots first even if changeovers are required. The algorithmic details are presented in Bard et al (2015).…”

“…The enhanced GRASP was implemented in C++, and is abbreviated as GRASP_opt to distinguish it from the GRASP logic implemented in ASAP. Bard et al (2015) developed a deterministic discrete event simulation model to support real-time decision making in the semiconductor backend scheduling area using the simulation software ASAP. In ASAP, much of the information needed to manufacture a part is defined in its route, which is a list of processing steps to be followed along with the machine and tooling required.…”

“…The AT scheduling problem was formulated as a MIP and solved by an enhanced GRASP in Bard et al (2015). Preemptive goal programming logic was used to deal with the hierarchical nature of the four objective function components mentioned above.…”

“…To decide on the best machine-tooling configurations and how to assign lots to machines, a three-phase heuristic was implemented. As a real-time alternative tool to the GRASP, Bard, Jia, Chacon, and Stuber (2015) developed a deterministic discrete event simulation model using AutoSched AP (ASAP) that similarly schedules AT lots over a given planning horizon. The built-in rules in ASAP performed poorly compared to the enhanced GRASP, though, so three new dispatch rules were formulated.…”