Automotive Skills Autonomous
c21a84a6ae
Autonomous daily run (Fri DOCS). - CHANGELOG.md [Unreleased] staged with W23 polish (cs-concept, tara, fmeda) and docs entries - New example README stubs: cs-concept-builder, tara-builder, fmeda-builder (coverage 7.9% -> 11.8%) - STATUS.md regen restores canonical domain spread (safety=15, quality=10) from Tue 2026-06-02 rules - Open issues: 12 (was 13 Thu - external delta, no autonomous label/close actions this run) Generated by automotive-skills-daily-standup scheduled task.
1.3 KiB
fmeda-builder — Example
What this skill produces: An ISO 26262-5 Hardware Safety Analysis (FMEDA) xlsx enumerating failure modes per HW element, allocating safety mechanisms from the upstream TSC, applying diagnostic coverage percentages, and deriving formula-driven hardware safety metrics (SPFM, LFM, PMHF) with auto-verification against per-ASIL thresholds.
Typical input shape: Upstream TSC xlsx (from tsc-builder) for safety-mechanism allocations plus a HW Bill of Materials JSON listing components with base failure rates (FIT), package-level grouping, and safety-relevant classification (safety-related vs QM).
Expected output: <item>-fmeda.xlsx — multi-tab workbook (Title, Element Catalog, FMEDA Worksheet with editable DC%, SPFM, LFM, PMHF, Dashboard, ASIL Targets, Traceability) where SUMIF + IF formulas propagate DC% and base-rate edits straight into the SPFM/LFM/PMHF metric tabs and Pass/Fail flags against the ASIL target row.
Sample I/O: Input a TSC stub for an EPB controller + 64-row HW BOM JSON → Output EPB-fmeda.xlsx with 184 failure modes enumerated, SPFM = 99.2% (ASIL D target ≥ 99%, PASS), LFM = 90.4% (ASIL D target ≥ 90%, PASS), PMHF = 8.7 FIT (ASIL D target < 10 FIT, PASS).
Run: Trigger by phrasing, e.g. "Build the FMEDA for the EPB controller from the TSC and BOM".