The prototype is beautiful. It works. The demo went well. Now the client wants 500 units by Q3. What happens next is where most hardware teams hit a wall — not because they're bad engineers, but because prototype engineering and production engineering are genuinely different disciplines.
What changes completely
Tolerances become real constraints. In a prototype, a skilled technician can hand-fit a part that's slightly out of spec. On a production line, that's a reject. Every dimension that matters needs a tolerance that can be reliably achieved by your chosen manufacturing process — and confirmed by measurement.
Assembly must be documentable. Prototypes are built by engineers who designed the product and can work around ambiguity. Production workers follow work instructions. If a step can't be written down clearly, it's a problem waiting to happen.
Components must be sourced reliably. The specialty capacitor you picked because it had great specs? It has a 16-week lead time and a $2,000 MOQ. In production, you need alternatives qualified and stocked.
Every failure must be traceable. If a unit fails in the field, you need to know its serial number, who built it, what batch of components it used, and what test it passed before shipping. This infrastructure doesn't build itself.
What doesn't change
The physics. If the design has a thermal problem, it has a thermal problem at 1 unit and at 1,000. Production doesn't magically fix fundamental design issues — it amplifies them.
This is why we push clients toward DFM review during detailed design, not after. Design decisions made at prototype stage (board stackup, connector choices, housing material, fastener type) become production constraints that are extremely expensive to change later.
The pilot production phase
Between prototype and serial production sits a critical bridge: the pilot run. Typically 10–100 units, built using production tooling and processes. This is where you find out what you don't know — which jigs you need, which steps take longer than estimated, which components have incoming quality issues.
Skipping the pilot run to save time is one of the most expensive shortcuts in hardware. Every problem it would have caught gets discovered instead on a 500-unit run — with proportionally higher cost and impact.
What a good production handoff looks like
- Complete, current BOM with approved vendors and alternates
- Production drawing package with GD&T-defined tolerances
- Work instructions for every assembly step
- Functional test procedure and pass/fail criteria
- Incoming inspection criteria for critical components
- Traceability system — serial numbers, lot codes, test records
Producing this package is part of our scope on every project. If you're at the prototype stage and thinking about what comes next — let's talk.