SESHA 2017 Symposium Abstract

Multivariate Analysis Technology in Process Safety Monitoring Application

Tuan, Andy*; Huang, Jack
(Flagship International Ltd, Taipei, Taiwan)

As consumers demand more functionality at lower price for their Information and Communication Technology (ICT) products, high tech device manufacturing industry has widely adopted factory automation and integrated process design to increase factory efficiency and reduce production cost. However better process safety performance is becoming more and more difficult to achieve as both complexity to control manufacturing process and number of hazardous process materials consumed in the factory increase. Traditional uni-variate statistics process control (SPC) charts may not be able to disclose enough insights for plant management to make correct and timely decision for the production line. This paper will illustrate how Multivariate Analysis (MVA) technology can be applied to design and build effective process safety monitoring system and achieve early detection for process excursion and identify critical safety-related parameters which could also often cause abnormal operation that leads to quality and efficiency problems at the same time. A more holistic approach by applying MVA technology not only to monitor process safety factors but also to improve the yield of the process is suggested in this paper.