Why Prototyping is Failing You
Three keys to unlock your prototyping superpowers.
What is a prototype?
Believe it or not, not everyone can agree on the definition and there’s a good reason. Most companies are not using prototypes the right way.
A prototype should be a physical representation of your design, that tests only a specific aspect of that design, only to reduce one risk. That’s right. A prototype should only solve one thing at a time. By shining a spotlight on a specific problem, you make it possible to learn quicker from your mistakes and make better solutions.
Keep prototyping budgets manageable by limiting build scope to a single component or subassembly and placing these shortened builds in series. In addition, more iterations will allow you to learn faster, which you can use to improve your design, which will reduce risk. As the risk of failure approaches zero, you can ramp up the complexity in small increments by integrating these subsystems and adding in more production representative processes. The approach of ‘one task at a time’ is used frequently in lean production to great effect.
This new mindset may not sound like that significant of a paradigm shift, but currently many companies try to limit the number of prototypes they make to save cost. Furthermore, they tend to produce a complete device as quickly as possible in an effort to test function. Instead, complexity adds both time and cost, and confounds root cause, all while incurring expedite fees and decreasing efficiency. I would argue that instead, focusing on high numbers of micro-prototypes will generate more and better data, drastically reducing risks that can have profound impacts on your overall development costs and timelines. Unfortunately, making that subtle shift to think in terms of risk reduction instead of function can be difficult for most organizations.
Below are three keys I use to get into this new mindset and unlock the potential of prototyping.
Key 1: Reduce risk with prototype stages
To make this philosophy easier to digest, it helps to think of prototyping in terms of stages. There is no firm number, but I typically recommend three to five. Too few and your costs can go too high, too many and your design efforts start to slow down. Simpler products will need fewer stages. Conversely, more complexity will need more stages.
Let’s use an example of a new disposable surgical drill. Stage one would be identifying requirements and prioritizing the risks of each requirement being met. This can be done through a variety of methods, from Preliminary Hazard Analyses to FMEA’s, but that is for another article. Group those risks by their overall impact to the project into high, medium, and low.
In Stage two, we will focus one at a time on the high-risk items. In Stage one, the team identified these three items as having the highest risks:
Reliability of the motor over time
Usability and ergonomics of the handle
Sterilization of the device
Using the status quo, if you were to design and build a handful of prototypes to test the above, you would need four to six weeks to get in parts, another week or two to assemble, another week to test, and they would still end up in the trash. Furthermore, you would need to iterate a few times to hone a production-worthy solution. You could be looking at multiple months and many thousands of dollars.
Instead, focus on just the reliability of the motor. Keep it simple and extremely low tech. Go cheap. Really cheap. Perfection is the enemy. I often call these 50% prototypes because the aim is to get the design 50% complete and try it. Don’t worry about cosmetics like color or surface finish at this stage unless those are critical to function. You will learn a lot. A simple Arduino fixture for less than $500 can run multiple motors for accelerated life testing in a thermal chamber and take only a matter of days to build. You will know very quickly which motor works best.
Next, you could build a handful of foam or clay handles to check ergonomics with your customers. How many clay prototypes can you build in an hour versus 3D printing? Versus machining? You can have customers build their ideal handle while you film it and then take measurements or 3D scans of the clay. Data is priceless.
Finally, you can simulate a fully built device with a solid 3D printed handle. Leave the expected channel clearances and then run a preliminary sterilization test to see what failures you might encounter. Some high-temp resins even allow for autoclaving. Running this test early allows for integration of fixes into the design and drastically reduces a major project risk much sooner than normal. You will still need to run sterilization validation on production representative parts, but won’t your confidence level be so much higher?
In Stage three, you will continue to reduce risks, this time focusing on the medium group. During this stage, incorporating the various subsystems may be necessary. Do not go too fast too soon. Remember, baby steps. Likewise, in Stage four you will focus on the low-risk items. In this stage, it is generally safe to develop a full assembly but remember to keep it low tech as long as possible. Run design verification testing to boost your confidence.
Finally, in Stage five you have made it to production representative at long last. Now you can replace those rapid prototype processes with longer-term solutions, limiting risk to only what is introduced by these new processes. Plastic injection molded parts, steel tubing, and more will be produced. Since each system has been thoroughly tested, the team’s confidence is high that the system will pass if parts meet print. In addition, in the unlikely event that something doesn’t work, you have a pool of passing rapid prototype parts you can use to debug. Stage five parts will be used for design verification and validation, as well as manufacturing validation.
Key 2: Users must stay involved
Throughout each stage, users must stay involved. Validation with users ensures that more risk is not being introduced along the way and that you are constantly validating your market. This also has the added benefit of showing design evolution and risk reduction, which regulatory bodies like to see.
Key 3: Just one thing
Unless you have the resources of a large company, and most of us do not, it is ok to focus on one thing at a time. If you use the 50% rule, you can iterate many more concepts than the big guys and in a shorter time. Use this opportunity to quickly rule out bad designs and iterate on promising solutions. Keep a detailed record of what you find out with each prototype.
As such, different prototypes are good for different things. Some need to look pretty to show what the final version of the product might look like to investors, making sure you have enough cash to keep going. Some prototypes might be hand sketched pages to represent different GUI screens in your app, ensuring that your UX is good before spending time coding. Some prototypes test and validate the manufacturing process to ensure good yields before going to high-volume production. The point? The resolution and complexity of your prototypes need to change as your project evolves, but they should still only focus on one thing.
Conclusion
When used well, prototyping is a distinct competitive advantage that can leverage exponential growth to drive down timelines and improve sales. Using a risk-based staged approach to your prototyping strategy will not only improve your product quality, but it will also make you and your team look like superheroes.
About the Author
You May Also Like