Programs that ship measurable outcomes.

Problem

GM manufacturing lines were shipping vehicles with electrical defects that escaped traditional EoL testing, dragging plant build performance to 23%.

Approach

Architected an SDV 2.0 toolchain that lets each vehicle on the line act as its own tester. Built APIs bridging vehicle controllers with plant IT, designed a Power BI decision matrix for team leads, and led off-line diagnostic triage workflows.

Impact

• ▸Plant build performance: 23% → 82% in weeks
• ▸89% of held volume recovered for shipment
• ▸50% reduction in test cycle time
• ▸80% improvement in product quality ratings

Problem

HIL programs across Mexico, Europe, and North America had fragmented tooling and no shared view of bench utilization or test outcomes.

Approach

Defined automation strategy and built APIs enabling test scripts, HIL simulators, real-time models, and physical hardware to interoperate. Pipelined SharePoint datasets to Splunk via GitHub Actions for live dashboards.

Impact

• ▸Real-time HIL usage visibility across global programs
• ▸Standardized fault injection and test reporting
• ▸Unified KPI roadmap for vehicle build validation

Problem

Vision-ECU defects were caught late in vehicle integration, threatening launch timelines and requiring nights/weekends to root-cause.

Approach

Designed and built a dedicated automation bench. Wrote scripts that streamlined AVSS execution and surfaced vision-related ECU issues earlier in the cycle.

Impact

• ▸Single HIL bench supporting multiple ADAS products
• ▸Earlier defect detection on critical launch programs
• ▸On-call root-cause support that protected launches

Problem

VeSCoM buy-off across BoR, prep, and Flash pre-ride milestones was manual and a recurring bottleneck.

Approach

Built an automation tool that orchestrated verification steps and surfaced integration issues earlier, with reproducible artifacts for sign-off.

Impact

• ▸30% improvement in project delivery timelines
• ▸Faster resolution of SW integration issues

Problem

Radar, LiDAR, and camera ECUs each had bespoke firmware workflows, slowing prototype iteration.

Approach

Built a Linux/Python tool over XCP that handled firmware updates across the AV sensor suite. Added FPGA-based camera debounce/delay logic and re-init recovery for runtime resilience.

Impact

• ▸Unified flashing across Radar/LiDAR/Camera
• ▸15% improvement in system reliability
• ▸15% reduction in image processing delays

Problem

KERZ engineers needed a Windows-side tool to monitor, alarm, simulate, and calibrate ECU variables live during track and bench testing.

Approach

Developed a C#.NET application speaking XCP over Ethernet — simulator, alarm, graph, and calibration modules wired together for real-time control.

Impact

• ▸8–12% improvement in sub-system efficiency
• ▸40% improvement in system reliability
• ▸20% reduction in downtime