An expert system for semiconductor device assembly diagnostics

• Main Author: Co, Celso B.
• Format: text
• Language: English
• Published: Animo Repository 2007
• Subjects: Semiconductors; Expert systems (Computer science); Artificial intelligence; Engineering
• Online Access: https://animorepository.dlsu.edu.ph/etd_doctoral/142; https://animorepository.dlsu.edu.ph/context/etd_doctoral/article/1141/viewcontent/CDTG004147_P__2_.pdf
• Institution: De La Salle University

High volume manufacturing (HVM) is the motivation of this study. HVM is made possible by robust system of automation and manufacturing technology. However, there are vulnerabilities that may be catastrophic. One of the major vulnerabilities is the process and equipment problem. The problem can be categorized as either recurrent or new ones. The recurrent problems have well-defined structured solutions and they constitute the majority of the problems. The new ones ha ve no solution and require mobilization of extra resources to investigate and resolve them. The new ones may not be the topmost concern since experience indicates it seldom occurs or it occurs only a couple of times in a year. Real problems are recurrent problems that are mistaken as new ones and their solutions are repeatedly rediscovered after agonizing trial and error and wasted production time. Expert system is proposed as solution for repeatedly occurring problems and its must be developed to quickly assist human in diagnostic activity. Expert system must be structured to deliver the correct solutions and must be unstructured to deal with human inputs and to establish priority of actions for human to execute. Area of investigation will be diagnostic tree and set theory application and the scope of study are defined within process, equipment, and continuous improvement diagnostics. Expert system is proposed as solution for repeatedly occurring problems and its must be developed to quickly assist human in diagnostic activity. Expert system must be structured to deliver the correct solutions and must be unstructured to deal with human inputs and to establish priority of actions for human to execute. Area of investigation will be diagnostic tree and set theory application and the scope of study are defined within process, equipment, and continuous improvement diagnostics. The benefit is cost effective troubleshooting operation by using first expert system before going to resource mobilization. Only truly new problem is referred to resource mobilization.