01 · DFMEA

• Ran the seven-step AIAG-VDA process, calibrated to the company's Software DFMEA guidelines and reviewed by senior power management and safety engineers.
• Broke specification content into four categories: ECU roles, NM PDU bit assignments, wake masks, and activation criteria.
• Identified and rated 14 failure modes, 6 of them HIGH priority by RPN.

02 · Detect at Authoring

• Built six automated pre-commit validators, one for each HIGH failure mode. A change that breaks a rule cannot be merged.
• The checks cut each mode's RPN by 50–67%. They went live in June 2026 and have already caught defects in routine updates.

03 · Cross-Validate

• Compared the specification against the deployed AUTOSAR ARXML configuration files and flagged ECU-network pairs where spec changes had not reached the build.
• A new wake-mask validator checked 157 ECUs and found 2 with missing bits. Both were corrected, and the validator then passed on all 157.

04 · Trace End to End

• Building a traceability index that links each requirement to its responsible system function, its software interface, and the test that validates it.
• Requirements with no implementation reference are flagged as implementation debt, and those with no test are flagged as coverage gaps.

01 · Define & Measure

• Conducted a Voice of Customer survey across the specification consumer community.
• Baseline scores for GB-5280 established quantified targets: testability 5.82 / 10 and traceability 5.64 / 10.

02 · Analyze

Four specification format concepts were evaluated against the survey findings:

• Pseudocode
• Graphics
• Excel
• Gherkin / BDD feature files

03 · Improve

• Gherkin syntax for behavioral requirement language, readable by architects and directly testable.
• Excel for the activation criteria sheet (SDV-8280A).
• Word with supporting diagrams for foundational Partial Network content (SDV-8280).

04 · Control

The split specification architecture is now the standard for the power management team.

Verification plan

• Scenario 1: Tracking. A Gazebo simulation with targets of varied size, distance, and acceleration, plus cover and debris to force reacquisition. This is followed by a physical test bench to capture friction, heat, and degradation that simulation misses.
• Scenario 2: Reliability. A design review of structural natural frequencies, then a vibration test harness run to failure to begin tracking MTBF.
• One extended tracking run doubles as the 5-hour endurance test.

Gaps, risks & assumptions

• Gaps: interface control documents for shared requirements, subsystem decomposition, and safety requirements.
• Risks: performance targets may be too strict, real-world use cases may not be covered, and failure modes are not yet addressed.
• Assumptions: values are estimates to be refined with stakeholders; operator skill and training are still unknown.

First week

• Centralize systems engineering documentation and link all tracked engineering work to a system-level requirement.
• Hold an alignment session with the mechanical, electrical, controls, software, and test leads. The session would confirm current compliance, negotiate requirements, and pressure-test the six-week timeline.

Deliberately not formalized yet

• The full calibration sequence depends on operator interfaces and final use cases.
• Heavy process documentation is deferred: systems engineering should bring clarity, not paperwork.
• Tracking decomposition and failure modes depend on solution choices, such as the sensor selection.