The lack of visibility into high rate (100 Hz.) Improved Auxiliary Power Unit (IAPU) Shuttle data can potentially cause unnecessary launch delays. Current Shuttle day of launch procedures require strip chart identification of IAPU performance issues just prior to liftoff and during on-orbit operations. The current methods are outdated, require a high level of operator attention, and don't support continuous on-line analysis of the data. Depending on fatigue and training level, an operator might miss a significant feature or misinterpret a signature reading. The issue of how to accurately detect and predict space component performance in a streamlined operations environment through automated diagnostics is one of many issues that face the Shuttle Program and the designers of the next generation launch vehicles. A technical approach to automated diagnostics is to develop fast, efficient signature recognition and event detection algorithms that can automate the capture of anomalies, off-nominal, and nominal events and free up the operator to concentrate on the verification and resolution of the event. A generic neural network-based monitoring and analysis prototype system has been developed by the authors and demonstrated for the Shuttle IAPU application. This prototype has run in parallel during IAPU pre-launch operations since STS-66 (late 1994) in the Kennedy Space Center Launch Control Center. The major features of the system include high speed event detection, neural network analysis, event correlation, interprocess communication, and full rate graphical displays. The Shuttle IAPU demonstrations are paving the way for incorporation into the Shuttle countdown procedures and the technologies developed will also be utilized by Rockwell's Reusable Launch Vehicle Design team. Accordingly, the overall architecture has been designed for growth using COTS tools and industry standards.