Comparative analysis of vision systems for electroplating surface quality inspection

In industrial practice, a human operator establishes the surface quality of an electroplated deposit through a subjective visual inspection, and this tends to be the immediate indicator of a quality defect. Electroplating in connector manufacture is a continuous process yet the quality inspection is an offline, operator-dependent process. The goal of this research was to automate the inspection of surface quality through the application of inline vision systems. Three vision system types were evaluated: a high-speed greyscale system, a standard colour system and a high-speed colour system. The efficacy of each approach was evaluated using attribute repeatability and reproducibility analysis (ARR); other factors considered were throughput, ease of operation and cost. The results proved that the high-speed colour system achieved the highest resolution reliability output for defect identification. Using key factor analysis and designed experiments, the optimum factor conditions were established for the high-speed colour system. To determine the practical implications such as false-positive and -negative results, the work was applied to a high-volume connector manufacturer. The overall benefit of such an implementation is an improvement in the defect rate and a reduction in risk priority number on the failure, modes and effects analysis of the process.