Lot output time prediction with a look-ahead hybrid ANN in a wafer fab

Traditional wafer lot output time prediction methods are based on the historical data of the fab. However, the influence of the future release plan has been neglected. In addition, a lot that will be released in the future might appear in front of another lot that currently exists in the fab. For these reasons, to further improve the accuracy of wafer lot output time prediction, the future release plan of the fab has to be considered (look-ahead), and a look-ahead hybrid ANN (SOM+FBPN) is proposed for the same purpose in this study. Production simulation is also applied to generate test examples. According to experimental results, the prediction accuracy of the proposed methodology was significantly better than those of four approaches, FBPN, case-based reasoning (CBR), evolving fuzzy rules (EFR), and the hybrid ANN without look-ahead by achieving a 20%-49% (and an average of 35%) reduction in the root-mean-squared-error (RMSE) over the comparison basis - the FBPN.