Increasing levels of machinery automation for systems health monitoring are providing operators with larger amounts of raw data. However, transforming massive amounts of data into information useful for effective condition-based maintenance (CBM) remains an arduous task. New technology is needed to continually monitor machinery, to identify impending failures, and to accurately predict its remaining useful life. Prognostic software agents can satisfy this growing need as higher levels of machinery automation raise the cost requirements of continuous monitoring beyond the levels of human and company feasibility. Software agent technologies that can automatically perform useful work as human assistants and can readily be integrated into existing automation system environments, represent viable tools to improve machinery reliability and reduce maintenance costs. Software agents can be used to clone human intelligence, perform human-like reasoning, and interact with human clients. Agents can perform tedious, repetitive, time-consuming, or analytically complex tasks on behalf of people who may not have the time or requisite skills to perform these tasks themselves. Agents can serve as expert assistants in monitoring, troubleshooting, and predicting failures in complex machinery processes. Imparting intelligent processing functions into software agents will allow maintenance organizations to leverage valuable "corporate" knowledge across geographically distributed machinery plants, such as aircraft or ship fleets. Agents can be distributed when and where needed to enhance fleet operations, performance, and readiness. Their intelligence can be upgraded remotely. The human-agent team can provide higher levels of productivity at practically the same cost as that of just the human resource alone. This paper describes intelligent prognostic software agents for real-time machinery monitoring applications. The main functions of the prognostic agent include machinery performance assessment, historical data archiving, automated trending analysis, alarm prediction, fault prediction, and prognostic event logging. The prognostic software agent is a generic tool for maintenance personnel to implement CBM. It predicts future machinery faults and determines when maintenance should be carried out. By predicting machinery problems before they occur, unexpected breakdowns can be avoided. In the absence of significant trends, equipment overhaul periods may be rationally extended, thereby eliminating unnecessary maintenance work. The ability to predict future maintenance requirements leads to improved maintenance planning and cost management. Maintenance and repair decisions can be tied to actual plant operating conditions based on the severity of degrading trends and predicted plant problems.
