Intelligent Test Data Analysis Technology (IT-DAT)

Modern weapon and aircraft control systems are software intensive, containing many complex internal and external interfaces. Accurate testing is vital to ensure that the system is functioning properly and will meet all operational requirements. For themost part, current testing focuses on components and subsystems; it is labor intensive and relies heavily on the subjective judgment of expert personnel for an overall system perspective. There are few automated systems to aid in the collection andinterpretation of software data, and the few systems that do exist still rely on the interpretation of data by system experts. This project will develop a more-thorough test and evaluation process, which will also be more efficient. The project willutilize the latest in artificial intelligence techniques to automatically analyze data collected during testing, thereby reducing reliance on error-prone manual data analysis. The project will incorporate a database archiving all test related data andsystem artifacts, providing a query capability for data analysts to review system requirements, identify trends in data, and compare test collected data against historical data. The project will use proven technologies in the collection and archiving ofdata, leveraging existing tools such as the SEA CORP SEAHAWK, TOWERS, and ADCAS tools. Intelligent Test Data Analysis Technology will improve the fidelity of the testing of complex systems. Through increased automation, this technology will also reducemanpower requirements and shorten the time required for testing resulting in savings and improved operational turnaround times. Concentrating first on complex weapon and aircraft control systems, the application of this technology has great spilloverpotential in the commercial market. As the reliance on software and external and internal interface becomes more pervasive in manufacturing, engineering, financial and other sectors, the demand for expert system test methods will increase as well.