Bridge Production: Shipping Real Units Before the Mold Is Cut
Content in this Article
A smart-home camera team needed 1,200 units in customer hands before their Series A metrics review, and their injection tooling wasn't cutting steel for another nine weeks. They printed the enclosures on MJF instead, dyed them to match the final color, and shipped. Three weeks in, returns started coming back with a specific fault: the SD-card door hinge was cracking on one in every forty units — a failure mode nobody had seen in bench testing, because bench testing doesn't involve someone's toddler opening the door forty times a day. They fixed the hinge in CAD and rolled it into the mold design before a single cavity was cut.
Why the First Few Thousand Units Are the Riskiest Ones
Every consumer device carries failure modes that only real users find: a hinge opened at an angle nobody tested, a button pressed with a thumbnail instead of a fingertip, a charging port that collects lint in someone's actual pocket instead of a lab drawer. Injection tooling locks that geometry in hardened steel for the life of the product — which makes the window before steel is cut the cheapest place to still be wrong.
The traditional answer was to wait: finish DFM, cut the mold, then ship. The gap that answer leaves is real-world exposure. A team that waits eight to twelve weeks for tooling before shipping anything is also waiting eight to twelve weeks to find out what their actual customers do differently than QA.
Where the Unit Economics Actually Cross Over
Additive manufacturing carries no tooling cost and a per-part price that barely moves with volume; injection molding carries a large upfront tooling cost and a per-part price that drops sharply once that tooling is amortized. Where those two lines cross depends on part complexity, but industry cost modeling puts MJF's break-even against injection molding at roughly 1,025 units for a quality-optimized part costing around $22–24 each with zero tooling. SLS crosses over earlier for simple geometries — around 667 units — and later, near 1,087 units, for more complex parts.
Why the volume band matters: that crossover band — roughly 500 to 5,000 units — is exactly the volume most first production runs need, which is why bridge production isn't a compromise so much as the economically correct choice for that specific quantity. An aluminum bridge tool for the same part might run $2,000–$8,000 and be ready in two to three weeks, versus $30,000–$150,000-plus and six to ten weeks for production steel. Below the crossover volume, printing wins on total cost even before you count the value of shipping weeks earlier.
Once a program is confirmed and volume climbs past that crossover, injection-molded per-part cost for a comparable part typically settles into a $0.50–$5.00 range — which is why nobody runs bridge production past the point where real steel tooling pays for itself. The point of a bridge program was never to avoid tooling. It was to buy time to make one tooling decision instead of three.
Making Bridge Parts Look and Feel Like Production
Print texture is fine on parts nobody touches or sees — internal brackets, structural ribs, a battery retainer hidden behind a final assembly. It's a harder sell on a customer-facing enclosure surface, which is where two post-processes earn their cost: dyeing and vapor smoothing.
| Process | What it does | Typical effect on surface | Best used for |
|---|---|---|---|
| Dyeing | Adds color — commonly deep black or a custom shade to match final CMF | No change to texture, only color | Any customer-facing part where the print's raw grey/off-white finish doesn't match the launch color |
| Vapor smoothing | Exposes the part to a controlled solvent vapor that softens and reflows the outer 0.05–0.1 mm of surface | Removes the powder-grain texture, seals surface porosity, produces a near-injection-molded finish | Enclosure shells, anything handled directly, parts that need moisture or bacteria resistance |
| As-printed (no post-process) | None | Visible layer/powder texture, higher surface porosity | Internal structure, hidden brackets, parts fully covered by final assembly |
Vapor smoothing does double duty on a bridge part: it makes the enclosure presentable enough to survive unboxing photos and a retail shelf, and the sealed surface it leaves behind resists the sweat, hand oils, and moisture that a printed-but-untreated part would otherwise absorb over weeks of real handling — which matters because the whole point of shipping bridge units is finding out how the product survives exactly that kind of contact.
Running a Bridge Program on Your Project
A Taiwan-based hardware team is usually one phone call away from both sides of this decision — an MJF or SLS service bureau for the bridge run, and a mold shop for the aluminum or steel tool running in parallel. The two don't have to be sequential. Cutting the aluminum bridge tool at the same time you're printing the first 500 units means the tool's own trial shots become a second data point before steel gets ordered.
| Order volume | Recommended approach | Why |
|---|---|---|
| Under ~500 units | MJF/SLS, no tooling | Below where any tooling — even aluminum — pays for itself |
| ~500–5,000 units | MJF/SLS bridge run while an aluminum or steel tool is cut in parallel | The crossover band — printing wins on unit cost and speed, tooling isn't ready yet anyway |
| Above ~5,000–10,000 units | Injection molding, tooling already amortizing | Per-part cost drops to $0.50–$5.00 range; printing no longer competes on cost |
The field data a bridge run produces is worth treating as a formal input to the mold design, not a side effect. A hinge that cracks in week three, a port that collects debris, a button that a specific hand size can't reach comfortably — each of those is a change that's nearly free to make in CAD and expensive to make in steel after the fact.
Bridge production doesn't replace tooling — it buys the weeks needed to make the tooling decision once, with real field data instead of a guess. For a first run in the few-hundred-to-few-thousand-unit range, that trade is usually worth making before you ever commit to 3D printing as a permanent process instead of a bridge to one.