Presented at ASMC, May 2008

ABSTRACT
In today's advanced manufacturing environments, rapid problem recognition is critical to the financial success of the factory. Automated data mining and analysis routines initially flag more and more factory problems. Unless properly tuned, however, these automated systems can generate a mountain of process control signals and create additional work for engineers who must determine which are worthy of further analysis and follow-up. Incorporating spatial patterns1 into these automated analysis routines can bring to engineers' attention real signals that would otherwise be lost in the noise. This enhanced signal clarity enables the engineers to better articulate problems and more quickly drive solutions. This paper will explore how to enhance signals by incorporating spatial patterns into the data mining process and offer practical examples showing how to use these results to better focus fab resources in problem-solving activities.

Read the full paper