SBIR-STTR Award

Repair of shipboard machinery and equipment requires information and expertise not easily available to shipboard maintenance personnel. What information that is available is generally embedded in technical manuals or maintenance/repair cards that are difficult to decipher, bulky, and sometimes hard to maintain. A system that allowed for improved maintenance aiding and guidance to non-experts would provide a more reliable and broadly available capability, which would enhance safety, reduce costs, and improve operational readiness. Under two phase I SBIR contracts for the USAF (f33615-86-c-5169 and -5170), the feasibility for a portable expert system as a maintenance aid for the field repair technician in the inspection and repair of composites, was demonstrated using an off-the-shelf expert system shell and a PC class laptop portable computer. It is proposed to build on the work and demonstrate a maintenance advisor/trainer for onboard ship use. It will be portable, have voice operation, incorporate the latest technology in data storage, graphics, and interactive displays. It will demonstrate a small selected set of current maintenance aids.

This Phase II effort builds on the Phase I which established the feasibility of combining a modern portable computer with an expert system shell operated by voice, and augmented by graphics, to aid a maintenance technician in the diagnosis and repair of complex Navy equipment. Expert systems rules, representative graphics, and speech vocabulary were developed which lead the technician through the diagnosis and repair procedures of two AN/BBQ-5 sonar system problems without reference to printed documents. A report capability is provided which collects information during the diagnoses and repair process and prints out a report. The Phase II work is to develop a horizontally integrated software and hardware system that serves the needs of the repair technician. It includes analysis of the methods of entering the diagnostic knowledge tree, display graphic requirements, report generation for operator and management needs. The system integrated for use with the AN/BQQ-5 will be further explored to develop a modular input system, zoom and high resolution graphics, alternative input systems, such as advanced speech recognition and synthesis, and touch pads. The initial tool for this research will be the dolch computer and the xi plus expert system.