An inverse-distance weighting genetic algorithm for optimizing the wafer exposure pattern for enhancing OWE for smart manufacturing

Wafer exposure pattern will determine the number of gross dies fabricated on the wafer and also affect the yield. Although a number of studies have addressed the wafer exposure pattern problem for maximizing the number of gross dies, little research has considered both the yield and gross dies simultaneously. To fill the gap, this study aims to develop an inverse distance weighting genetic algorithm (IDWGA) that simultaneously maximizes the total number of exposed gross dies and minimizes the deviation of die-estimated measurement from the target for yield enhancement and smart manufacturing. This study developed a novel approach for estimating the die yield from a few measurement points and a three-dimensional (3D) contour plot of die estimates for verifying the measurement pattern among the dies. The proposed IDWGA can detect the die yield pattern during the wafer exposure stage and thus optimize the exposure pattern to maximize the number of gross dies and minimize potential yield loss. On the basis of realistic data, experiments were designed to estimate the validity of the proposed approach. The results have shown practical viability of the proposed approach to optimize overall wafer effectiveness for total resource management.