Requirements Traceability Framework for Complex Safety Critical Software Systems

Abstract

Safety-critical software systems require complete traceability because requirements, hazards, designs, code, tests, and compliance evidence change throughout development. This study presents a requirements traceability framework for complex safety-critical software systems. The framework combines semantic analysis, artifact metadata, lifecycle rules, bidirectional tracing, version checking, and change-impact analysis. It links functional safety, hazard-control, interface, timing, performance, and regulatory requirements with their related implementation and verification artifacts. The framework was evaluated across six increasing system-complexity levels and compared with document-based manual tracing and keyword-based automated tracing. Results showed that the proposed method maintained traceability coverage above 90% at the highest complexity level, while the comparison methods declined more strongly. It also improved missing-link detection, version consistency, verification planning, and compliance review. The framework can support safer development, clearer certification evidence, faster impact analysis, and more reliable maintenance of large safety-critical software systems during continuous system evolution and coordinated engineering across multiple specialized teams.