Optimizing SMT Performance Using Comparisons of Efficiency Between Different Systems Technique in DEA

Abstract

Surface-mount-technology (SMT) enables production of reliable printed circuit board at a reduced weight, volume, and cost. Nevertheless, SMT defects increase quality costs and deteriorate performance. In reality, defects are not equally important and their corresponding quality costs vary. In these regards, this research divides SMT defects into three classes; minor, major, and serious. Eight key factors from SMT processes are investigated concurrently with L18(21 times 37) array. Taguchi's accumulation analysis method is first applied and it provides incomplete optimal levels for some factor. In this research, each experiment is treated as a decision making unit (DMU) with defect counts set as inputs. Comparisons of efficiency between different systems (CEBDS) technique in data envelopment analysis (DEA) is then adopted to optimized SMT performance. The corresponding actual improvements in minor, major, and serious defect counts are found 0.352, 2.153, and 4.154 dB, respectively. Analysis of variance (ANOVA) is finally performed to determine significant factor effects. It is found that conveyor speed and reflow temperature are the most influential factors on SMT performance. In conclusion, the CEBDS technique turns out to be an efficient approach for optimizing process performance with categorical data.