The plethora of prototyping options available to designers today is staggering. From 3D printing to SLAs, CNCs to laser cutting, quick-turn PCB fabs to Arduino boards, it’s becoming really easy to design up some parts, sit down at a lab bench and build those first prototypes. And using prototypes in an iterative fashion to dial in a design is a great methodology. But… getting products from prototyping to manufacturing is still darn tough.
Manufacturing challenges that cannot be remotely uncovered in prototyping include:
The higher the potential volumes the more work that is needed up front to optimize the design for manufacturing and assembly. This is needed not only for reducing production costs and improving yield but also for end product reliability
Hand assembling 5 prototypes is a million miles away from 3 shift manufacturing. Besides the obvious small sample size, engineers lovingly building those first parts is not the same as an assembly line.
- Production test
Related to yields is testing. Few products can afford engineers hand testing each and every product off the line for defects. And allowing the end customer to test the product in the field almost always ends badly.
- Supply chain
Buying components from Digikey or McMaster Carr (don’t get us wrong, we love both those guys for prototyping and even low volumes) is easy but it’s not cheap, and just because a component is on the shelf in 10 piece quantities today does not mean it will be on the shelf in 100K quantities in three months. And those fun development boards are great for prototyping but expensive and bulky to use in real products, and converting to a robust product with robust firmware often means starting over on the hardware and firmware design.
Using 3D printing or laser cutting is fantastic for prototyping but way too expensive in most instances for volume production. Plastic parts and many metal technologies require tooling, which in turn requires the design to be optimized for tooling and molding. For plastic parts this means adding draft at a minimum as well as taking care with features that may require slides or lifters in the tool. And of course tools have relatively large capital outlays and in many cases substantial lead times.
- Design validation testing
Even in cases where a prototyping technology is suited for doing design validation testing (never try to drop test an SLA for example), it’s important to use production ready parts for doing all manner of validation testing. These include system testing, software testing, environmental testing, mechanical testing (drop tests, shock and vibe, etc) as well as regulatory testing like FCC or CE.
Prototyping is good. Keep doing it. But don’t confuse a good prototype with a manufacturing ready product. That’s where productization comes in, and that’s how we earn our keep.