Knowledge Repositories for Multiple Uses

1 Problem In the life cycle o f a complex physical device or part, f o r example, the docking bay door of the space staiion, there are many uses for knowledge about the device or part. The same piece of knowledge might serve several uses. Given the quantity and complexity of the knowledge that must be stored, it is critical to maintain the knowledge in one repository, in one form. At the same time, because of the quantity and complexity of knowledge that must be used in life cycle applications such as cost estimation, re-design and diagnosis, it i s critical to automate such knowledge uses. For each specific use, a knowledge base must be available and must be in a fo rm that promotes the eficient perfomance of that knowledge base. However, without a single source knowledge repository, the cost of maintaining consistent knowledge between multiple knowledge buses increases dramatically; as facts and descriptions change, they must be updated in each individual knowledge base. W e have developed a use-neutral representation of an hydraulic system f o r the F I l l airplane. W e demonstrated the ability to derive portions of four diflerent knowledge bases from this use-neutral representation: one knowledge base is for re-design of the device using a model-based reasoning problem solver; two knowledge bases, at diflerent levels of abstraction, are for diagnosis using a model-based reasoning problem solver; and one knowledge base is for diagnosis using an associational reasoning problem solver. W e have shown how updates issued against the single source use-neutral knowledge reposito y can be propagated to the underlying knowledge bases. In the life cycle of a complex physical device or part (e.g., the docking bay of the space station) there are many uses for knowledge about the device or part, about its structure, function, materials, component parts and so on. Such uses might include product definition (i.e., design and manufacturing process planning), testability analysis and test construction, cost estimation, producibility studies, diagnosis of breakdowns, and re-design of features. In an effort to become more competitive, corporations are automating such knowledge uses. For each particular use, a knowledge base must be available and must be in a form that promotes efficient performance of that system. This has resulted in duplicate knowledge in multiple knowledge bases, where a particular piece of knowledge serves multiple uses. Given the quantity and complexity of the knowledge that must be stored, it is critical to maintain the knowledge in one repository. Without a single source repository, the cost of maintaining the knowledge required by multiple knowledge bases is increased dramatically; as facts and descriptions change, they must be updated in each individual system. It is virtually impossible to maintain synchronicity among the knowledge bases that require the same knowledge. Without a single source repository, the costs of other maintenance tasks (e.g., knowledge verification and validation) are not easily shared across knowledge bases. Finally if there is no common knowledge repository, the development of new knowledge bases