Tool-feeder partitions for module assignment in PCB assembly. (February 2017)

• Record Type: Journal Article
• Authors: Raduly-Baka, Csaba; Johnsson, Mika; Nevalainen, Olli S.
• Is Part Of: Computers & operations research. Volume 78(2017)
• Page Start: 108
• Page End: 116
• Publication Date: 2017-02
• Subjects: Complexity theory -- Flexible manufacturing systems -- Tool loading -- Set-up optimization -- Control of assembly process -- Printed Circuit Board -- Job Grouping; Operations research -- Periodicals; Electronic digital computers -- Periodicals; 004.05
• DOI: 10.1016/j.cor.2016.08.017
• Languages: English
• ISSNs: 0305-0548

Abstract

Abstract: The assembly of electronic components on printed circuit boards is performed in a line of automated placement machines, each equipped with a feeder unit that contains the component reels required for a product type. Assembling multiple product types by a single machine requires the replacement of certain components in the feeder unit. Increasingly, assembly machines are equipped with modular feeder units, enabling the replacement of a larger group of component reels. To minimize the machine setup time, component reel modules can be preassigned with the necessary component types, thus requiring only the loading of a suitable module set for each product. This work introduces and analyzes the problem of constructing the minimum number of component reel modules, with the objective that a set of jobs can be manufactured without individual component reel changes. We show that the problem is NP-hard and provide an integer programming formulation for it. We also introduce two heuristics based on job grouping, and analyze the results. Abstract : Highlights: This paper introduces the problem of tool partitioning for modular feeder units. In a traditional formulation, the problem reduces to the job grouping problem. With modular feeder units, the problem is generalized in a nontrivial way. We show that the TPMA problem is NP-hard also when multiple modules are allowed. We apply two algorithms to create module assignments and study their performance.