As organizations have increased their dependency on computers, many databases have been developed, and the computers containing the databases have been linked together via networks. The innovative aspect of this project is to extend the capabilities of a formalized, knowledge-representation substrate, which is capable of supporting existing and developing knowledge representation techniques, to handle distributed data storage and information processing. The project objectives are to create algorithms to perform remote procedure calls, to access distributed databases, to maintain consistency of distributed databases, and to balance the load of the processing activities of a computer network. Phase I has shown that the use of data-driven programming techniques is appropriate for distributed processing applications. The proper method for maintaining both the consistency and redundancy of information in a distributed environment is by using demons that encapsulate a "behavioral law" which reacts to changes in the database. There is no single appropriate method for distributing information in a network. The use of a rule-based system is a good way to implement the load-balancing strategy for a distributed database. Note: ISX Corporation has acquired the rights to continue Phase II of this project from Teknowledge Federal Systems, who received the Phase I award. Potential Commercial Application: A general software tool for the management of distributed, fault-tolerant, intelligent systems is in wide demand for applications where reliability is crucial and intelligent behavior is needed, e.g., Space Station and autonomous systems.
