In this lesson, you will learn how to architect a compelling safety case by mapping verifiable engineering evidence to specific safety claims to satisfy certification auditors.
The Architecture of a Safety Case
A safety case is not merely a collection of logs; it is a structured argument. Auditors evaluate your system based on the GSN (Goal Structuring Notation) approach, where high-level safety objectives (e.g., "The braking system shall not engage inadvertently") are broken down into sub-claims, supported by concrete evidence.
To build a robust safety case, you must link your previous technical activities—such as code coverage reports, static analysis results, and formal verification proofs—directly to these claims. If you cannot trace a specific requirement to a test result, the evidence is considered incomplete by certification standards like ISO 26262 or DO-178C.
Organizing the Evidence Chain
Think of your evidence as a tree. The root is your safety objective, the branches are sub-claims (e.g., "The software correctly handles concurrent access"), and the leaves are the raw artifacts.
Consider a snippet of a formal verification argument mapped to a code block:
Claim: The buffer write operation is memory-safe.
└─ Strategy: Use formal model checking to prove no overflow.
└─ Evidence: formal_verification_results_v1.2.xml
└─ Code Snippet:
void secure_write(uint8_t *buffer, size_t len) {
if (len <= MAX_BUFFER_SIZE) { // Checked by model checker
memcpy(buffer, data, len);
}
}
By structuring your documentation this way, you move from "we tested the system" to "the system is demonstrably safe." You must ensure that every unit test execution (from your simulated environment lessons) and every static analysis linter report is indexed with a unique identifier that aligns with your traceability matrix.
If an auditor asks why a system is safe, you don't hand them a folder of PDFs. You present the GSN tree, where clicking a node reveals the specific requirement, the hazard assessment that identified it, the unit test that validated it, and the coverage report that confirms its reach.
Key Takeaway: A safety case is a structured argument where every top-level safety claim is supported by a traceable chain of verifiable evidence, ensuring no safety requirement is left unsubstantiated.