Prototype vs Production

A working prototype is a good milestone, but it is not the same thing as a production-ready product.

This distinction matters because many electronics projects get into trouble after the first successful demo, then some harder questions pop up.

Can it be built repeatedly? Can it pass EMC? Can it survive real users? Can the parts be sourced? Can it be tested in production? Can it be made at the target cost?

To answer these questions, we need to first understand the fundamental differences between prototype hardware, and production hardware.

Why make prototypes?

There are a few definitions of "prototype" bandied around, but we have settled on something like:

A prototype is a physical device made with the intention of learning from it.

Notably, a prototype is therefore not made with the intention of selling it.

Before we start on a prototype design, we always first ask 'What do we intend to learn from this?'. This sets the framework, the expectations, and importantly, the scope. Not all prototypes will implement every feature of a product, and rarely will the first functional prototype in a project even be in the same form-factor as the final product. Some people see this design effort as a waste, as so much of these devices will never see the final product. However, the value, and the whole reason for going through the process, is in learning lessons.

These lessons might be technical: Can we achieve the sensor resolution we want? Can we hit our efficiency targets? Does this section of circuitry match the simulations?

On the other hand, they are just as often non-technical: Does the product feel right in your hand? Do we need to move a button to be easier to reach? How does this type of display look in direct sunlight?

Prototypes, by this definition, are all about learning these lessons as quickly as possible, before we go down any design paths that could be expensive to redesign ourselves back out of. For that reason, technical and functional prototypes are often made in form factors and technologies that are easy to test, with cost near the bottom of the priority list. Design or non-technical prototypes will often be made with different techniques or materials than the final product, 3D printed rather than moulded, hand cut rather than CNC. 

So why not just make more of them to sell?

There are two answers here.

1. A prototype needs to work on a bench, to be tested. A product needs to work every time it's used, for its entire lifetime, every time we make a new one.
2. A prototype needs to be made as fast as possible, and as flexible to design change as possible. A product needs to be producible, reliable, and profitable.

Brought together, this means a product going to market needs to consider:

• Component availability

• Unit cost

• PCB manufacturability

• Assembly pass rate

• Test access

• Firmware recovery

• EMC and safety compliance

• Thermal behaviour

• Mechanical integration

• Field reliability

• And more..

These are not things that can be added or bodged onto a minimal viable demo, before it has even become an MVP. Each one requires thought, planning, and care of implementation throughout the design process. Meanwhile, a prototype is just a milestone in that process, often not very close to the end.

For example, a connector that is fine for a prototype may be slow to assemble. A wireless module that saves time early may damage the production margin. A PCB that works on the bench may fail emissions testing. Firmware that handles the happy path may fall over in edge cases in harmful ways. 

The danger in shipping a prototype

Often, we get asked to help "tidy up" a prototype for production. Sometimes, this is possible. Often, this prototype is only a half way point towards go-to-market. "Works on the bench" is all very well, however very few customers live on the bench.

This is especially true if the prototype relies on dev kits, unavailable parts, manual assembly, poor test access or a layout that was never designed with EMC in mind.

Trying to force that into production can look efficient. In theory, you save on engineer time, and start making profit sooner. In practice, the cost often appears later as certification delays, poor yield, field failures or support headaches.

The cheapest redesign is usually the one done before tooling, certification and purchase orders are committed.

Moving from Prototype to Production

Early prototypes do not need to solve every production problem, but they should expose the important ones. As the project continues, prototypes generally get closer to a final product, as more questions are answered. The further along in the process you are, the fewer questions should remain. 

Once all of your design questions are answered, the focus shifts towards any remaining production questions. Sometimes these can be answered without the need for additional prototypes, but sometimes this is unavoidable.

To give you an idea of the questions you should be thinking of, consider:

• Are the key parts available in volume?

• Does the Bill of Materials support the target price?

• Is the design likely to pass compliance?

• Can it be assembled without heroics?

• Can every unit be programmed, tested and calibrated?

• Will the firmware recover safely from faults?

• Is the enclosure manufacturable in volume?

The earlier in your process you can identify and assess these questions, the better your position for moving to production.

The practical takeaway

A prototype can be a valid technical success, but still be a poor foundation for production.

Prototypes exist to help you learn something, but are expensive and difficult to produce.
Production ready designs are reproducible, reliable, and ultimately, profitable.

How we can help

Planning your first MVP?
Need a hand going from a prototype to production?
Schedule a call - no sales deck, no pitch, just support.