Zhen Shi and Peter Sandborn
University of Maryland
College Park, MD 20742
In this paper, an optimization methodology is used to select the locations and characteristics of test, diagnosis and rework operations in electronic systems assembly processes. Real-coded genetic algorithms are used to perform a multi-variable optimization that minimizes the yielded cost of products resulting from an assembly process that includes test/diagnosis/rework operations characterized by costs, yields fault coverage, and rework attempts. A general complex process flow is analyzed using the algorithms proposed in this paper, and a multichip module assembly process flow is used to demonstrate that the methodology can identify optimum test and rework solutions that result in a reduction in yielded cost.
Complete article is available to CALCE Consortium Members.