Automated Testing of Reusable ADA Software Components

We propose a new model for testing Ada components which use aspects of artificial intelligence and software reliability engineering (specifically operational profiles). Our goal is to produce a tool that is capable of constructing operational profiles for both prior and new testing of an Ada component. Before testing a reuse artifact, engineers construct a test-input graph using Pest Control Studio, a tool based on AI techniques to describe the state space of an application. This graph is then ng prior testing) so that a profile of all prior test experience is obtained. Next, a profile for new use of the component is constructed using a copy of the test-input graph. We envision the development of an intelligent test case generator that is capable of using both profiles to optimize new testing. In addition, we intend to use the operational profiles to compute standard reliability metrics for the component and to derive a new metric to measure a components ed on the completeness of its testing. We call this metric the reusability index. BENEFITS: The test generator and measurement engine will both be integrated into the Pest Control software tool which AET hopes to commercialize. The AET, Inc. engineering and management team along with Florida Tech researchers have the experience to develop and market these products with high probability of success.