An online integrated operational decision support system with predefined knowledge base for semiconductor industry

We have developed an online predefined knowledge-based decision support system for AMID (Thailand) Ltd. It is a single interactive, computer based decision-making system for a large variety of operations, tasks, and conditions, which don´t require special programming skills for maintenance. AMD is dealing in the manufacture of flash memory with a variety of devices. Each device has several common or device specific rules. Each production lot has to pass through a series of different engineering rules to get released. Each rule is defined by an experienced engineer and shared in a centralized common database. Each device has got a different program as per the type of package and the type of booting. If we combine all variables the scenario will become very complicated. For e.g. Millions of flash memory/week, thousands of lots/week, hundreds of test program, devices, packages and a series of common and device specific rules of each device to be applied on each lot. The system has to make several calculations on production data to validate each rule applied on that device. Secondly, the fault diagnosis and the decisions should come online immediately after such an event. The system has proved to be an error free system and the best decision making tool on production environment; so this paper describes a deep-knowledge base management and use of online decision support system shell for diagnosing faults in manufacturing operations of the semiconductor Industry with real world application.