Scaleable Expert Systems for Adding Crisp Knowledge to Pilot-Vehicle Interfaces

摘要

Vehicle systems operations, from automobiles, to airliners, through craft for seaand space, are heavily dependent on automated uses of data generated by both on-board and off-board sources. Not only are vehicles affected, but so are diverse systems such as process, manufacturing, and power generation plants. Historically, the volume of data in each instance keeps growing because of such factors as task and system complexity. Other factors are environmental, accuracy, safety, cost effectiveness, and customization requirements. A crowded operational domain is also a factor for vehicle operations and communications networks. Tragedies such as that which occurred at the Three-Mile Island nuclear power plant and during the V-22 Osprey Tiltrotor crash are examples that point out the volume and complexity of the data and automation involved. These incidents also point out that it is essential to have a 'scaleable framework for the organization and prioritization of raw data. In this way, the operator gets pertinent information necessary for reaching critical decisions. This framework also should support decision aids that offload operational details so that human resources can be applied more appropriately to goal achievement and other management issues. This paper outlines such a framework and gives examples.