Guides / Cost
Moving a drone part from three pieces to three hundred
July 23, 2026
The prototype flew. Now you need three hundred of the same part, at a price that works, arriving on a schedule.
Almost everything about how the part is made changes at that point, and the parts of it you control are decided before the order goes out.
Freeze the right things first
Not everything has to be frozen. These do:
- The interfaces. Bolt patterns, bores, mating faces, anything another part depends on.
- The material and temper. A change from 6061 to 7075 changes the finish, the price and sometimes the geometry.
- The critical tolerances. The two or three features that matter.
- The finish and its masking. This drives cost and lead time more than most buyers expect.
These can stay open longer: outside profile, pocket shapes, lightening holes, cosmetic details and anything that does not touch another part.
The expensive pattern is the opposite. Freezing the cosmetics early and leaving the interfaces open means the fixture gets built twice.
What actually changes in the shop
| Prototype | Production | |
|---|---|---|
| Work holding | Vice, soft jaws, whatever fits | Dedicated fixture, repeatable location |
| Programme | Written quickly, conservative feeds | Optimised, proven, timed |
| Setups | More, because nobody wants a fixture yet | Fewer, because relationships matter |
| Inspection | Key dimensions checked | First article plus a sample plan |
| Finishing | Sent in small batches, slow | Batched with a rack designed for the part |
| Price per piece | High, dominated by setup | Lower, dominated by cycle time |
The fixture is the hinge. It costs money once and it buys repeatability, fewer setups and a lower cycle time on every part afterwards. For a small aluminium UAV part, somewhere between 20 and 50 pieces per release is typically where it starts paying.
Where the price actually drops
Three effects, and they arrive at different quantities.
- Setup amortisation. The single biggest step, and it arrives immediately. Ten parts share a setup, so ten parts are much cheaper each than one.
- Fixturing and programme optimisation. Arrives once a fixture exists, typically tens to low hundreds of pieces.
- Material and finishing batching. Buying a full bar rather than an offcut, and filling an anodising rack, both need volume.
Ask for prices at three quantities on every RFQ, for example 10, 100 and 500. The shape of that curve tells you where your programme should sit, and whether the shop has actually thought about your part or just multiplied.
Tolerances usually get looser, not tighter
A prototype drawing often has tight tolerances everywhere, because nobody knew yet which features mattered. Production is the moment to fix that, and it saves real money.
After the first units fly you know which fits were fussy and which were irrelevant. Relax the ones nobody cares about, keep the two or three that decide assembly, and the part gets cheaper without getting worse. The reasoning is in what to put on the drawing.
Documents step up
Production is where the paperwork starts to matter:
- First article inspection against the released revision, before the lot runs. See first article inspection on UAV parts.
- An agreed sample plan for lot inspection.
- Mill certificates per lot and processor certificates for placed finishing.
- Laser marked part number, revision and lot code, so a field failure can be traced. See country of origin papers.
- PPAP level 3 if your programme runs one, asked for at quote stage rather than after.
A realistic schedule for the transition
| Step | Typical |
|---|---|
| Drawing freeze and RFQ | your side |
| Quote | within 24 business hours here |
| Fixture design and build | 3 to 7 days |
| First article machining | 1 to 3 days |
| Finishing on the first article | 3 to 7 days |
| Measurement and report | 1 day |
| Your approval | your side |
| Lot production and finishing | quantity dependent |
The two steps buyers forget are finishing on the first article and their own approval time. Both are days, and both sit on the critical path.
Release schedules beat single big orders
If the aircraft is still evolving, a blanket order with scheduled releases is usually better than one large lot. You get the price of the total quantity, the shop keeps the fixture set up, and a revision only affects the parts not yet made.
Say at RFQ that you want this. It changes how the quote is built rather than being a discount you negotiate afterwards.
Send the model, the three quantities and the likely annual volume through the RFQ page, and the reply comes back within 24 business hours with the fixture decision priced separately so you can see it.
Frequently asked questions
At what quantity does a fixture start paying for itself?
Usually somewhere between 20 and 50 pieces for a small aluminium part, and sooner if the part needs several setups. Below that, the fixture cost is larger than the machining time it saves. Ask for both numbers on the quote and the answer is obvious.
Should the prototype be machined or printed?
Print it for fit and packaging, machine it for anything you intend to fly or measure. A printed part has different stiffness, different tolerances and different failure behaviour, so a printed prototype that works tells you less than people hope.
Why is the production part sometimes better than the prototype?
Because production runs get a proper fixture, a proven programme and an inspection routine. Prototypes are often cut in one off setups where more of the accuracy depends on how the part was held that day.
Need this part quoted?
Send the model and the quantity. Price and lead time come back within 24 business hours.