Reducing cost of high integrity systems through model-based testing

Abstract

This work describes the capabilities of a model-based analysis and automatic test generation system. Model checking ensures that an automatically generated code is free from contradictions that are difficult to identify manually in complex models. It generates test vectors and test drivers that can execute against model simulations or auto-generated code, and produces test sequence vectors for testing dynamic system behavior that support feedback, such as integrators or time delays, which are common in control systems. The paper briefly describes tool qualification support, and processes for using this model-based testing tool with modeling and code coverage tools to produce verification evidence that meets the FAA standards for certification of these systems, while reducing the verification costs by as much as 50%.