Jay Rabe

I see the connections and bridges that product development often misses: between the disciplines, the teams that own them, the customer they are all building for, and the product itself.

A hardware product is built from many disciplines and pieces. Mechanical, electrical, firmware, software, manufacturing, known requirements, open unknowns: each one is its own island with its own depth, its own infrastructure, and its own complexities. The islands are not the problem. When islands or their connections are missing, that is where the issues sprout. What the customer experiences is not any single island but the ability to freely travel the bridges across all of them. The bridges that make that possible are just as necessary as the disciplines and objectives themselves and they need to be planned for and designed with the same intention.

Often the way product development is organized keeps each discipline working in its own island until the work is done and then the bridges get figured out. By then the infrastructure is already built and connection requirements that were missed turn into team frustrations, delays, and expensive fixes. The customer's experience lives in unrestricted flow across the whole network and a gap in the network cuts off functionality at exactly the path the product needs to deliver. When someone understands the whole network and its needs holistically they can plan where the bridges and interconnects need to go early, while the work is still taking shape. The interconnects get designed in while the work is still moving. Each team knows how and what their work connects to. The bridges are part of the design from the beginning, the connections and flow are ready when the product arrives and relies on it.

My design depth is mechanical and electrical. Firmware I hold my own in. Software I architect, direct, and integrate rather than write from scratch. This span of depth allows me to see the connections, shift perspectives across the disciplines, the end user, the product itself, and the unknowns that have not fully resolved yet, and zoom from the high level functionality down to the minute detail and back out again. A failure or troubleshooting that encompasses mechanical, electrical, and firmware cannot get handed to a single discipline to diagnose easily. It gets found by someone fluent enough in all of them to see the overlaps and the real root. I can work the problem across that complexity alone or with the teams.

Unknown and fragmented islands exist in every real product. I come in with my own perspectives and ideas but I ask questions first, deliberately and without leading, to get as much clarity as possible from the people closest to the problem before I share what I am already seeing. Those questions can seem unfocused until the plan arrives and the reasoning behind each one becomes clear. What comes back gets distilled into a plan: known connections designed in from the start, unknowns organized into what gets built now, what gets held for later, and what gets a flexible foundation so the path stays open. Sometimes that plan includes a pause for targeted experiments for directional clarity before building anything or is done in parallel. Understanding a critical unknown is not a delay, it helps set up the plan. Sometimes the right answer to unknowns is a breakout connector with power, SPI, and a few chip selects to leave options open. Other times the hardware unknowns resolve fully and what remains can be handled in code as understanding and testing evolves.

The manufacturing connections and bridges need the same early planning as everything else. Housings, wiring, thermal, repairability, assembly: these are where the product becomes real and where missed connections show up last and cost the most. Running prototypes as part of the experimentation and evolution process verify compatibility and keep everything on track. I work fluently with the groups on and across every island so the collaboration gains clarity on what they are building toward, what the unknowns are to plan for, and how their pieces connect into the whole.

With this fluency comes the foresight to give every team clarity on what they are building toward, what the unknowns are to plan for, and how their pieces connect into the whole. The known connections are clear. The unknowns are given room and the right foundations to resolve and evolve. When that picture is clear from the start the work is better, team confidence follows, and the product proves it.

Below is fifteen years of shipped work, not claims.

LinkedIn GitHub

Jay Rabe

Where the disciplines connect

Three pieces across the disciplines

The same person across every discipline, 0 distance between them. A driveline, a whole machine, and a tool, each one several disciplines in a single object.

The GT40 driveline as one CAD assembly: equal-length headers, the reverse-engineered transaxle, the modular bellhousing adapter, and the weld and sheet-metal fixtures that hold everything to position. Reverse engineering, 3D scanning, mechanical design, fabrication, casting, and fixturing in a single model. Most of these parts show up on their own further down.
A robotic retail vending machine as a full CAD assembly: the bent sheet-metal cabinet, the SCARA robot, the touchscreen compute, the payment systems, and the control boards. Mechanical, electrical, firmware, software, and manufacturing all meeting inside one product.
A rotary weld positioner with its own motion control and optional backstepping, driven by a stepper using 555 timers and no embedded processor. The clip shows it running with the direction-change pulses visible; the board beside it is what runs it. Mechanical, welding, motion, and electrical in one object.

Mechanical and fabrication

Parts and assemblies designed with an understanding of how to make them

CAD allows anything to be drawn up, including parts that cannot actually be made or made easily. I learned fabricating, welding, machining, and electrical before learning CAD. This translates to knowing how to build a part or assembly as it is drawn up. Designing with that clarity on what the welding, machining, and assembly can do makes the transition to real parts a smooth one, because I learned to do all three well before learning CAD.

Equal-length GT40 headers

Equal-length headers and megaphones, designed from a 3D scan and fitted to the car. The challenge is routing and packaging eight equal-length runners, merged by the overlapped and evenly pulsed firing order, around the spark plug wires, leaving room to pull valve covers, and holding everything tight to the engine while preserving clearance to nearby components and the body. 3D scan to CAD to fixtures to finished part, one person across the whole chain.

Electrical and PCB

Boards designed, reverse-engineered, and brought up

A deep understanding of the electrical side is what connects the mechanical and the firmware.

Robotic vending program

A robotic retail vending program, 25 people across three time zones. I course-corrected repeatedly, then took over the robot stepper control PCBAs and drove the robot toward design-for-assembly and design-for-manufacture under a tight timeline, coordinating across the hardware, firmware, and software teams. Delivered within 2 percent of scoped cost and to my quoted timeline.

Machine control and motion

Motion and control work that drives electromechanical systems

The vending machine and weld positioner above are clear examples of mechatronics. Two more below.

PLC and HMI integration. A reverse osmosis system's PLC integrated with an HMI over local wifi, so it reports and is controllable without a walk to the equipment. PLC integration and HMI configuration, wired into the rest of the machine.
Motion control board. Mounted and running on the vending robot, stepper and vacuum pickup in frame. Designed from the ground up.

Software and tools

Engineering software I architect and direct

Suspension Geometry Ninja is a commercial suspension design tool I architect and direct. Built to fill a gap neither the existing tools nor SolidWorks cover: a full suspension design and analysis workflow. suspensiongeometryninja.com

Live 3D view

The live 3D view with Ackermann and instant-center trails. The geometry updates as you move it, which is the part neither the existing suspension tools nor general CAD give you in one place. It imports CAD to reference the space a suspension has to fit, and exports its points as IGES to keep building in CAD. CAD severely bogs down on this geometry and cannot give you the full-travel suspension specs and behaviors, which makes designing suspension in it frustrating and close to impossible. Closing the gap between the tools is less about saving time than seeing the suspension behaviors and letting the design actually flow.

Reverse engineering and 3D scan-to-CAD

From a physical part to a manufacturable model

Measure or scan the part, rebuild it in CAD, then prove the rebuild against the scan or with a printed test fitment. Often to reproduce parts that are no longer available.

Automotive air intake, scan over CAD

An automotive air intake, 3D scanned and reverse-engineered to reproduce a part that is no longer available. The purple and green is the raw 3D scan laid over the finished CAD, so the match can be checked surface by surface.

The team

A design is only as good as the hands that build it

Products do not assemble themselves. People build the assemblies, load the firmware, and run the tests, often at another company in another time zone, and that is where a sound design quietly becomes a field failure. Designing for those hands and leading the people who do that work is the job, not a side skill.

At Trueform I took assembly defects from roughly 90 percent to roughly 3 percent, not by inspecting harder but by designing the product so the floor could build it right the first time. On the 25-person program across three time zones, most of the work was keeping the disciplines and the people aligned across the distance to a launch. I do not hand the floor a list of steps. I write the build book and train the assemblers on the why behind each step, not just the what, so they can catch the problem I did not foresee.

Outsourcing the assembly does not remove any of this. It raises the stakes, because the failure point moves to a floor you do not control. The specification, the fixtures, and the design have to carry the intent the assembler never heard you explain.

The through-line is simple. I try to do the job well enough that people no longer need me: a design the floor can build without a call, a team trained to catch what I would have caught, and the knowledge written down in clear fabrication and assembly drawings and the build books, so it outlives my involvement.

Validation and measurement

The proof, not the promise

A design is only a hypothesis until it is measured and certified. Thermal, dimensional, endurance, IPX4, EMC: the certifications are where the plan proves out. The 250,000-cycle result and the low field RMAs are just the receipts.

250,000-cycle endurance

SCARA robot for automated retail vending. Its bent sheet-metal structure was run through 250,000-cycle endurance testing, and the whole machine then passed EMC, IPX4, and surge-immunity qualification as a complete unit.

This page is built on analogy. Islands, bridges, interconnects, network, flow: none of those are strictly engineering terms but every engineer, executive, and stakeholder can follow them through their own perspective of the work. Finding the right analogy to help everyone see a problem clearly, without making them feel like they needed it explained, is one of the things I do best. The analogy dissolves the understanding roadblocks so the teams can all be on the same page and move forward. It often goes unnoticed in how it resolved. It is about shifting everyone to a common perspective they can all see clearly and build from.

If this page has you thinking about your next build, you likely already know if I am the right fit.