An Ontological Approach Supporting the Development of Safety-Critical Software

In several application domains, the development of safety-critical software is subject to certification standards which prescribe to perform activities depending on information relative to different stages of development. Data needed in these activities reflects concepts that pertain to three different perspectives: i) structural elements of design and implementation; ii) functional requirements and quality attributes; iii) organization of the overall process. The integration of these concepts may considerably improve the trade-off between reward and effort spent in verification and quality-driven activities. This dissertation proposes a systematic approach for the efficient management of concepts and data involved in the development process of safety critical systems, illustrating how the activities performed during the life cycle can be integrated in a common framework. This thesis addresses the exploitation of ontological modeling and semantic technologies so as to support cohesion across different stages of the development life cycle, attaching a machine-readable semantics to concepts belonging to structural, functional and process perspectives. The formalized conceptualization enables the implementation of a tool leveraging well established technologies aiding the accomplishment of crucial and effort-expensive activities.