My Personal Stories

Engineering Decisions.
Real Outcomes.

These are the programs that shaped my philosophy — where strategy met hardware, and execution separated the winners from the also-rans.

Click to expand story
+

Honda came to TRW with a compressed timeline and a high bar: deliver a quotation and complete technical proposal for a Supplemental Restraint System (SRS) in six weeks. What they were really asking was whether we could execute — not just design, but execute at scale with zero margin for error.

I decided early: the only way to win trust was to compress the entire development cycle into the proposal phase itself. We wouldn't present concepts. We'd present a production-ready design.

Week One
Design Specification Lock

I started by translating Honda's SRS requirements into a hierarchical architecture. Every signal was mapped. Every mechanical interface was defined. This wasn't exploratory work — it was the foundation for manufacturing.

  • ECU specification completed
  • ECU Requirements matrix fully decomposed
  • Complete ECU schematics delivered
  • Mechanical interface requirements finalized
Week Two
Parallel Acceleration

While the schematic locked, I launched parallel workstreams that would typically run sequentially.

Hardware Refinement

  • Optimized component placement for signal integrity and thermal performance
  • Established PCB layout rules — rigorous signal grouping discipline
  • Executed auto-routing with critical signal verification via forward annotation
  • Released Gerber files to manufacturing and our quick-turn PCB fabrication partner
  • Started pick-and-place machine programming

Mechanical & Manufacturing Readiness (Running in Parallel)

  • Finalized ECU housing design per Honda specification
  • Completed secondary housing variant design with PCB fixation points
  • Reserved production capacity at TRW's Marshall, IL facility (including third-shift readiness)
  • Coordinated aluminum machining operations
  • Prepared component magazine carts for assembly
The Parallel Strategy: While the PCB house was running quick-turn fabrication, mechanical design was advancing. Manufacturing schedules were locked. We had manufacturing samples built and validated before the formal design review — eliminating schedule risk.

The Japan Trip: Technical Presentation & Design Review

Three weeks in, I traveled to Honda's engineering center in Japan with fully assembled prototypes, test data, and detailed design documentation. This wasn't a PowerPoint presentation. This was a technical design review grounded in working hardware.

  • Worst-case analysis (WCA) across all functional modes
  • A revised proposal incorporating Honda's feedback
  • Proof-of-concept validation on actual vehicle electrical systems
Walking in with working prototypes while competitors were still in simulation changed the conversation completely.

Manufacturing Readiness: Two Variants

By the end of week two, we had:

  • Quick-turn PCBs fully assembled and tested
  • Aluminum-machined housings delivered (two variants, covering different platform requirements)
  • Both variants integrated and verified
  • Production line at Marshall staffed and ready
We weren't waiting for design freeze to start manufacturing. We were ready to produce on day one.
★ The Outcome · ~3 Months Later

Honda awarded us the ECU platform: 10,000 systems annually (ECUs plus integrated sensors).

Launch vehicles:

Honda Accord
Honda Odyssey
Acura RDX
Acura MDX

Production volume (MY2003): 1.2 million units

Why This Mattered

This program taught me that "agile" in hardware means something specific: parallel workstreams, compressed decision cycles, and manufacturing readiness built into the design phase — not bolted on afterward.

The six-week timeline wasn't a constraint we adapted to. It was leverage. By compressing design, building prototypes, and validating in Japan before formal award, we eliminated the risk that would have haunted a traditional sequential approach.

More importantly, we built Honda's confidence. When you show up with working hardware, manufacturing plans, and supply chain readiness while others are debating topology, you've already won the negotiation.

That 1.2M unit volume validated the approach: disciplined planning + aggressive parallel execution + customer co-development = market dominance.

Click to expand story
+

When I inherited the side airbag sensor program, the logistics were strangling us. Each vehicle had up to 11 sensors. Each location required specific calibration and its own part number. The result: approximately 150 different SKUs in inventory, each with unique documentation, test protocols, and supply chain overhead.

The engineering team was drowning. Manufacturing was drowning. Procurement was drowning. And profit margins were disappearing under the weight of variant management.

The question wasn't "how do we manage 150 variants better?"
It was "why do we have 150 variants at all?"

The Problem Beneath the Problem

The traditional approach treated each sensor location as a unique design problem. Left-front, right-front, left-rear, right-rear — different accelerations, different pressure regimes, different calibration curves. Different part numbers. Different everything.

But here's what I noticed: the physics wasn't actually that different. The variation was in how we configured each sensor, not what the sensor fundamentally did.

That was the insight. Instead of 150 designs, we needed one intelligent platform that could adapt to any position.

The Innovation: Universal Configurable Architecture

I invented a communication protocol and sensor architecture that eliminated the concept of "variant." I called it the Universal Satellite Airbag Sensor (USAT).

Hardware
Two Core Designs
  • USAT accelerometer-based (for primary crash detection)
  • USAT pressure-based (for supplemental sensing)
Software
One Unified Firmware with Configurable Parameters
  • Vehicle calibration curves
  • Position-specific calibration data
  • Sensor ID and protocol stack
  • All programmed at vehicle assembly
System Integration
Assembly-Level Configuration
  • The airbag module stored vehicle-specific calibration data
  • The main ECU communicated with each side sensor at assembly
  • Once all side sensors were configured and programmed, airbag deployment logic enabled
This wasn't a hardware redesign. It was a systems redesign — moving specificity from the part number to the configuration database.

Execution: Engineering Simplicity

The USAT development required 4 engineers instead of the 12+ we'd need to manage 150 variants in parallel. We focused on:

  • One set of accelerometer designs (validated across the full acceleration envelope)
  • One pressure sensor interface (characterized across operating conditions)
  • One communication protocol (robust, low-latency)
  • One calibration framework (flexible across vehicle platforms)
  • Manufacturing readiness for two core designs instead of dozens

The Manufacturing Reality Check

At vehicle assembly plants, the magic happened:

  • Sensor arrives as a generic USAT — no P/N variant needed
  • Assembly line programs the sensor with vehicle calibration curves, position identification, and sensor address on the bus
  • Main ECU scans and configures all side sensors
  • Self-test passes → Airbag system armed
One smart inventory. One purchasing transaction. One logistics footprint. Infinite application flexibility.
★ Results: The Business Impact

Cost Engineering:

  • 30% cheaper per unit — reduced variants, simplified manufacturing, lower testing overhead
  • Reduced development cost: 4 engineers instead of 12+
  • Manufacturing cost reduction: simplified PCB designs, single BOM structure
  • Logistics cost reduction: one inventory location, one supplier relationship, one quality program

Profit Improvement: Direct. Immediate. Compounding with volume.

Production Scale:

  • 100,000 sensors per week for Honda vehicles
  • 5.4 million per year
  • Over 30 million cumulative units shipped across platforms
Why This Mattered

This program taught me that complexity isn't always a product problem — it's often a systems problem. The customer saw 150 variants because we'd encoded locational differences into hardware when they should have been encoded into software and calibration.

By inverting the problem — designing one universal platform and moving specificity to configuration — I eliminated 140+ variants without sacrificing performance or reliability.

The USAT became a reference design for how to scale airbag sensor technology across diverse vehicle architectures. It's the difference between managing complexity and eliminating it.

A 5.4M unit annual program built on the foundation that simpler systems, configured smartly, outcompete complex ones every single time.

Click to expand story
+

After the USAT program, I took on an expanded mandate at Robert Bosch: leading Ford's passive safety business globally while simultaneously owning the North America Passive Safety Technology Roadmap. This was no longer a single program — it was an organization.

Organization Led
Cross-Functional Engineering Leadership
  • Hardware development team
  • Algorithm and calibration development team
  • Crash support team
  • Program management
  • Electronic Lab
  • North America Passive Safety Technology Roadmap ownership

Managing this breadth required building systems — not just leading engineers. Every team had to operate with clarity of ownership, speed of decision, and a shared standard of zero-defect launch execution.

Technical Achievements

Platform Innovation
Single ECU Platform — Reconfigurable Accelerometer (SMB100)

Developed a single ECU platform based on the reconfigurable SMB100 accelerometer — enabling one hardware design to serve multiple vehicle lines and crash sensing configurations. The same architectural principle that drove the USAT, now applied at the ECU level.

Algorithm Innovation
Rollover Algorithm — Minimum Onboard Sensors

Invented a rollover detection algorithm architected around the minimum viable sensor complement. By eliminating sensor redundancy without sacrificing detection accuracy, the approach reduced system cost while maintaining ASIL-level safety performance.

Patents Filed & Granted
Intellectual Property — In Production
  • DE102008040043A1 — Occupant Protection: Seat-back / Occupant Sensing Technology   Google Patents ↗
  • DE102008040043B4 — Rollover: Personal Protection Unit Controlling Method   Google Patents ↗
First-to-Market
Airbag Module with Integrated Active Safety Sensors — Ford Vehicles

Introduced the first-to-market airbag module integrating active safety sensors for Ford vehicles — merging passive restraint control with active safety sensing in a single module. This convergence eliminated redundant hardware and reduced integration complexity at the vehicle level.

SOC & ASIC Strategy
Next-Generation Airbag Silicon Roadmap

Defined and introduced the semiconductor building blocks that would power the next generation of airbag ECUs — a multi-chip strategy built for scalability and integration:

  • Scalable Rollover, Yaw & Accelerometer SOC — architected for next-gen airbag ECU platforms
  • Airbag SOC + Companion ASIC — paired silicon strategy for tighter integration and BOM reduction
  • Scalable, Highly Integrated FET Drivers & Diagnostics — with PSI5 interface, power supply Buck/Boost
The semiconductor roadmap wasn't just about the next product — it was about locking in Bosch's position as the architecture standard for passive safety ECUs across the next decade of vehicle programs.
★ What This Built

A vertically integrated engineering organization capable of owning the full passive safety stack — from sensor silicon and algorithm to ECU hardware, crash calibration, and vehicle integration.

The technology roadmap defined at Bosch North America during this period shaped how passive safety ECUs were architected globally — influencing product strategy, silicon partnerships, and OEM platform decisions across multiple vehicle generations.

The Leadership Lesson

Leading a multi-disciplinary organization at this scale taught me that technical strategy and organizational design are inseparable. The roadmap only works if the team executing it has clear ownership, the right capability, and a shared commitment to zero-defect delivery.

The patents, the first-to-market launches, the SOC roadmap — none of it happens without a team that trusts its leadership to set a vision and then remove every obstacle between them and execution.

Platform thinking + organizational discipline + silicon strategy = sustainable competitive advantage.

Click to expand story
+

At GM, I applied the same platform-thinking philosophy that drove the USAT at Bosch — this time to machine vision cameras for Ultra Cruise. The problem was structurally identical: 7 cameras per vehicle, each defined by field of view and placement location, with evolving and partially undefined requirements. The traditional answer would have been 7 different camera designs. That was the wrong answer.

Situation
7 Cameras · Partially Undefined Requirements · Cost Pressure
  • Ultra Cruise vehicles required 7 cameras per vehicle — forward, surround, and interior positions
  • Each camera position was defined by field of view (FOV) and physical placement
  • Requirements were evolving and partially undefined at program start
  • Conventional approach: one unique design per location = 7 separate development programs
Task
Release a Low-Cost Camera Platform for Ultra Cruise (PSAT) Vehicles

Design and deliver a camera system that serves all 7 positions on the vehicle while minimizing development cost, tooling investment, and piece price — without compromising image quality, safety performance, or future flexibility.

Action: Single Scalable Configurable Camera Architecture

I designed a single scalable configurable camera for all Ultra Cruise PSAT-equipped vehicles. One hardware platform. All positions. All fields of view.

Hardware Configurability
One Design — Multiple Lenses, Multiple FOVs, Multiple Locations
  • Single PCB design supports multiple lens assemblies for different FOV requirements
  • Hardware architecture validated across all 7 vehicle camera positions
  • Mechanical interface designed for multi-location mounting compatibility
Software Configurability
ISP-Driven Flexibility — Up to 32 Configurable Modes
  • Configurable ROI — Region of Interest defined per vehicle position at programming
  • Configurable Resolution — 8MP sensor with software-selectable output resolution
  • Configurable F-Sync — frame synchronization adjustable per location and application
  • Utilized full ISP capability — up to 32 configurable modes per camera
  • Each location independently programmable at vehicle integration
OTA Capability
First OTA-Capable Camera — Configurable & Updatable via ACP4

Designed the first Over-The-Air (OTA) capable camera in GM's PSAT program — fully configurable and firmware-updatable via ACP4. As Ultra Cruise requirements evolved post-launch, cameras could be updated in the field without hardware changes. This was the architectural moat: the camera platform was not frozen at Job 1.

One hardware design. Programmable for each location. Updatable over the air.
The requirements were still evolving — and so could the camera.
★ Results

Development Cost: Reduced GM engineering cost — one validation program instead of seven

Piece Price & Tooling: Reduced PSAT piece price and tooling cost through shared hardware architecture across all camera positions

ISP Utilization: Full exploitation of ISP capabilities delivered a platform with:

  • Up to 32 configurable modes — one camera, 32 different operating configurations
  • Independently programmable per vehicle location at assembly
  • OTA-capable — firmware and configuration updates via ACP4 across vehicle lifetime

First-to-market: First configurable, OTA-capable PSAT camera in a GM production program

The Pattern Repeats

The Honda USAT eliminated 150 sensor variants with one configurable platform. The Ultra Cruise camera eliminated 7 camera designs with one configurable platform. The principle is identical: encode variation in software and configuration, not hardware.

What changes is the domain — sensor silicon to machine vision. What doesn't change is the insight: complexity hidden in part numbers is cost bleeding out of your margin.

The OTA capability was the forward bet. Ultra Cruise was a program with evolving requirements by definition. Designing a camera that could be reconfigured post-launch wasn't over-engineering — it was eliminating the future cost of being wrong today.

Click to expand story
+

When GM decided to bring Adaptive Driving Beam technology to North America, it wasn't just a product launch — it was a regulatory first. No production vehicle had ever been sold in the Canadian market with ADB lighting under the new SAE J3069 standard. I led the program that changed that.

This story has two chapters: the development of GM's smart headlamp foundation through IntelliBeam, and the delivery of the MY2022 Cadillac Escalade ADB system — the first ADB-equipped vehicle launched in the Canadian production market.

Chapter 1 — IntelliBeam: GM's Smart Headlamp Foundation

I led the development of GM's IntelliBeam smart headlamp system — GM's automatic high-beam assist technology. IntelliBeam was the platform foundation: it proved that a camera-based system could reliably detect oncoming and leading vehicles and automatically switch between high and low beam without driver input.

This program established the sensing architecture, vehicle integration strategy, and validation framework that became the stepping stone to full Adaptive Driving Beam. Without IntelliBeam, there is no ADB roadmap.

IntelliBeam · Smart Headlamp System
Automatic High-Beam Management — Camera-Based
  • Led end-to-end development of GM's automatic high-beam control system
  • Camera-based detection of oncoming headlamps and leading tail lamps
  • Automatic high/low beam switching — seamless, reliable, driver-transparent
  • Established vehicle-level integration architecture reused across multiple GM platforms
  • Laid the sensing and validation groundwork for ADB development

Chapter 2 — ADB: First in North America

Adaptive Driving Beam is a fundamentally different technology from automatic high-beam. Instead of switching between two states, ADB continuously shapes the light beam — providing maximum illumination while carving out a shadow zone around other vehicles. It requires precise real-time sensing, complex beam control, and regulatory compliance with standards that had never been tested in North American production before.

Program Achievement
MY2022 Cadillac Escalade — First ADB Vehicle in the Canadian Market
  • Introduced the first GM vehicle with ADB to the Canadian production market
  • MY2022 Cadillac Escalade — flagship vehicle, highest-visibility launch platform
  • Met SAE J3069 — the ADB performance standard for North American markets
  • Met CMVSS108 — Canadian Motor Vehicle Safety Standards for exterior lighting
  • Regulatory first: no other production vehicle had cleared this path before
FMVSS108 ADB — February 2022
Designed & Developed Own Validation Tools — $200K Tooling Savings

FMVSS108 ADB validation required specialized photometric measurement tools and test methodologies. Rather than purchase off-the-shelf test equipment that would add significant cost and delay, I designed and developed GM's own proprietary tools to validate ADB compliance with FMVSS108 February 2022.

  • Designed custom FMVSS108 ADB validation tooling and test methodology from scratch
  • Validated vehicle-level ADB photometric compliance against FMVSS108 Feb 2022 requirements
  • Eliminated dependency on external test house for compliance demonstration
  • Saved GM $200K in tooling and external validation costs
  • Established reusable validation infrastructure for future ADB programs across the GM portfolio
When no commercial tool existed that could do exactly what the standard required, we built our own.
That's not a workaround — that's engineering leadership.
Regulatory & Standards Compliance
Multi-Market, Multi-Standard ADB Certification
  • FMVSS108 — US Federal Motor Vehicle Safety Standard, ADB provisions (Feb 2022)
  • CMVSS108 — Canadian Motor Vehicle Safety Standard, exterior lighting
  • SAE J3069 — SAE Adaptive Driving Beam performance standard
  • Managed FuSa system compliance and cybersecurity requirements to ASIL-D
  • Developed GM's strategy for next-generation Adaptive Driving Beam exterior lighting
★ What This Delivered

Market First: MY2022 Cadillac Escalade — first ADB production vehicle in the Canadian market, meeting SAE J3069 and CMVSS108.

Cost Savings: $200K in tooling and validation costs eliminated through own-developed FMVSS108 test tools.

Infrastructure Built: Reusable ADB validation tools and methodology for future GM programs — a one-time investment that pays forward across every ADB launch that follows.

Regulatory Precedent: Demonstrated FMVSS108 ADB compliance through GM's own internal capability — no external dependency, no schedule risk from third-party test house availability.

The Leadership Lesson

Launching the first ADB vehicle in the Canadian market meant navigating a standard that had never been applied to a production vehicle. There was no playbook. There were no commercial tools. There were no precedents to cite in a supplier meeting.

That's exactly when engineering leadership matters most. You don't wait for someone to hand you the validation methodology. You build it. You validate it. You defend it to the regulator.

IntelliBeam built the foundation. ADB was the summit. The $200K in saved tooling costs was a byproduct — the real value was proving that GM could own its own regulatory destiny on a first-of-kind technology.

Click to expand story
+

Scout Motors was a greenfield OEM within the Volkswagen Group — a rare opportunity to define vehicle electrical architecture from a blank sheet. The mandate was clear: replace the traditional HCP (High-Computing Platform) domain architecture with a modern zonal architecture that could scale across multiple vehicle platforms while enabling the software-defined vehicle vision.

This wasn't a single-company effort. It required cross-brand alignment across Audi, Porsche, Skoda, VW, and Rivian–VW Tech — organizations with different platforms, priorities, and engineering cultures — all converging on a shared architectural foundation.

Cross-Brand Collaboration
Audi · Porsche · Skoda · VW · Rivian–VW Tech

Collaborated across the full VW Group ecosystem to define and develop the zonal architecture framework — balancing brand-specific requirements with the need for a common, scalable platform foundation.

  • Engaged Audi and Porsche engineering leadership to align on zonal ECU compute strategy
  • Coordinated with Skoda and VW teams on software-defined architecture compatibility
  • Collaborated with Rivian–VW Tech on SDV platform integration requirements
  • Drove consensus on shared architecture principles while preserving Scout brand differentiation

Scout Zonal Architecture Leadership

ECU Design & Development
Scalable Solutions Across Multiple Vehicle Platforms
  • Led Scout's zonal ECU designs and development — architected for scalability across vehicle line variants
  • Approved zonal ECU schematics, PCB layouts, and component selections
  • Integrated numerous secondary ECUs into the zonal ECU — reducing vehicle cost, weight, and harness complexity
  • Enabled a single zonal ECU platform to serve multiple vehicle configurations without hardware redesign
12V / 48V System Architecture
eFuse Introduction & Power Distribution Strategy
  • Defined the complete 12V/48V system architecture for Scout vehicles
  • Introduced eFuses (solid-state power distribution) — replacing traditional fusing for enhanced diagnostics, protection, and energy management
  • Designed power distribution architecture supporting both 12V legacy loads and 48V high-power systems
  • Enabled intelligent load management and fault isolation at the zone level
Vehicle Ground Architecture
Multi-Protocol Network Integration

Defined Scout's vehicle ground architecture to support the full communication protocol stack required for a modern SDV platform:

  • 1000BASE-T1 — high-bandwidth Automotive Ethernet for camera and sensor data
  • 100BASE-T1 — Automotive Ethernet for ECU-to-ECU backbone communication
  • FlexRay — deterministic, fault-tolerant communication for safety-critical nodes
  • CAN-FD — high-speed CAN with Flexible Data-rate for powertrain and body systems
Contract Manufacturer & Supplier
CM Selection Framework · Sourcing · Development · Qualification
  • Supported CM selection framework — defined criteria, led evaluation, and drove selection decision
  • Managed supplier sourcing, development agreements, and qualification programs for zonal ECU supply chain
  • Approved zonal ECU validation requirements including EMC and reliability test plans
  • Led SoC and zonal microcontroller selection — benchmarked compute performance across candidate platforms
Power & Thermal Engineering
Worst-Case Power Analysis · Thermal Management · Housing Design
  • Led worst-case vehicle power consumption analysis for 12V systems across all zonal load scenarios
  • Defined thermal management strategy for zonal ECU platforms — junction temperature, derating, and packaging
  • Led zonal ECU housing design optimization — thermal performance, mechanical interface, and manufacturing cost
ADAS L2++ & IVI
Vehicle Architecture, Integration & SoC Selection
  • Led Scout's ADAS L2++ vehicle architecture and integration — sensor fusion strategy, ECU design, system-level development
  • Led IVI vehicle architecture and ECU integration within the zonal framework
  • Led SoC selection for ADAS and IVI compute platforms — benchmarked AI inference, camera ISP, and real-time OS performance
  • Ensured ADAS and IVI compute nodes aligned with zonal power distribution and communication architecture
Connectivity — LEO Satellite Integration
Starlink · Kuiper · Shark Fin Elimination
  • Led LEO satellite connectivity systems integration within the Scout SDV vehicle — including Starlink and Amazon Kuiper
  • Led shark fin antenna elimination activities — consolidating 5G, GNSS, FM1, FM2, and Wi-Fi into the zonal architecture
  • Reduced antenna system complexity and vehicle roof penetrations through integrated antenna design strategy
  • Aligned satellite connectivity with OTA update infrastructure and SDV data pipeline requirements
REEV Integration
Engine & Onboard Generator Integrated Into the SDV Architecture

Led the REEV (Range-Extended Electric Vehicle) design — integrating Scout's internal combustion engine and onboard generator into the SDV architecture. This required bridging traditional powertrain control with modern zonal and software-defined vehicle frameworks:

  • Defined the control architecture for onboard generator integration within the SDV platform
  • Aligned engine management system communication with zonal ECU and 12V/48V power architecture
  • Ensured REEV operating modes were compatible with OTA update capability and zonal power management
A greenfield OEM is a once-in-a-career opportunity — no legacy constraints, no inherited debt, no "that's how we've always done it."
You either define the architecture correctly from the start, or you spend the next decade paying the cost of getting it wrong.
★ What This Built

Zonal ECU Platform: A scalable, configurable zonal ECU architecture applicable across all Scout vehicle variants — reducing secondary ECU count, vehicle weight, and harness complexity.

Cross-Brand Alignment: Architectural consensus achieved across Audi, Porsche, Skoda, VW, and Rivian–VW Tech — a shared foundation with brand-level flexibility built in.

eFuse Architecture: First introduction of solid-state power distribution into Scout's electrical system — enabling intelligent fault isolation, diagnostics, and energy management at the zone level.

Full Protocol Stack: Vehicle ground architecture supporting 1000T/100T Ethernet, FlexRay, and CAN-FD — ready for the bandwidth demands of ADAS L2++, IVI, and OTA update infrastructure.

REEV Integration: Engine and generator successfully integrated into the SDV architecture — extending Scout's range capability without compromising the software-defined vehicle platform.

The Leadership Lesson

Building architecture across five brands is not a technical problem. It's a leadership problem masquerading as a technical one. Every brand has engineers who believe their approach is correct — because often, it is, within their context.

The job isn't to pick a winner. It's to define the principle that makes all the specific solutions compatible, then hold that principle under pressure from every direction simultaneously.

Scout's zonal architecture succeeded because we established non-negotiable architectural principles early — power distribution, network topology, ECU integration boundaries — and gave the brands flexibility within that framework.

Constraints that are well-chosen aren't limitations. They're what make a platform scalable.