Submarine Towed Array Systems (TAS) need a process for automating the collection of system specific data related to fleet-wide operational performance. The purpose of implementing such a process is to realize the benefits of applying Condition Based Maintenance (CBM) to these remotely operated, externally located systems. This process should be an integration of elements which automatically collects real-time on-line Reliability, Maintainability and Availability (RMA) data, conducts diagnostic/prognostic analysis, and provides reports and access to users throughout the community. Analysis will include total and sub-system TAS health assessment and prediction, trending of input and output data and other selected maintenance related functions. The development of an integrated hardware and embedded software system with support hardware is required to construct a process accomplishing these functions.The LCE/ARET? TEAM proposes the design and development of a COTS-based automated data collection, centralized repository, and analysis system for TAS RMA related data. This effort leverages previously demonstrated individual ship system health monitoring results. The database will be integrated with new design and existing modeling and simulation tools to predict the Fleet TAS health. The use of a complex Bayesian Belief Network (BBN) will provide the basis for the implementation of an integrated predictive tool with process diagnostic technology. The application of an integrated BBN-based predictive reliability assessment capability to commercial equipment and process plants can be highly valuable in minimizing the down-time of production facilities. This approach can immediately improve the maintenance planning process in many industries, by scheduling outages at the most economical points in the production runs, by taking advantage of just-in-time procurement of repair parts and maintenance-related materials, and by achieving the optimum utilization of plant maintenance personnel, funding and ancillary support facilities. The true value of condition based maintenance is realized only when proactive maintenance is scheduled to avert costly non-random system failures. The potential cost savings in preventing emergent repairs and operational delays is extremely significant when the consequences of failure are ruinous. This occurs when downtime is particularly costly to the operation or when the system fails often. With the current trend of fielding more sophisticated system designs with higher performance requirements, the increased complexity decreases overall reliability. The proposed CBM predictive approach is particularly beneficial to improve these processes.
Keywords: Reliability, Availability, Prediction, Intelligent Code, Maintainability, Monitoring , Automat
