SPLTRAK Abstract Submission
Data Mining of Solar Cells Production Data Using Factorial Analysis
Johnson Wong, Dinica Li, Gordon Deans
Aurora Solar Technologies, North Vancouver, BC, Canada

For a large solar cell factory floor, factorial Analysis is a technique which makes use of the entirety of production data and the path information, to delineate performance differences in the process tools of every process step.  This is a data mining approach that has a number of important advantages over single factor experiments using test batches to compare tool performance: 1) there is no experimental overhead, no disruptions brought to the production that could lower overall throughput, 2) being a surveillance technique, it yields simultaneous information about all factors (all process tools of each step) as opposed to one factor at a time, 3) drawing conclusions on 100% of production data, it has inherently much greater statistical resolution and accuracy over sampling experiments.  Moreover, factorial analysis is compatible with all production floors that have end-of-line I-V measurements, and does not require midstream process quality control measurement data to be effective, and is therefore positioned as a very cost effective method to monitor the production.