Prototypes: A Fast Chance to Fail

The return on a prototype is immediate and enduring. Nearly every time I have plucked a fresh prototype out of the box, I have learned something. Usually something important. Something project-saving: The size was wrong, the finish was critical, the button was inaccessible, the device wasn’t serviceable.

But, prototypes are expensive so we don’t want to pull the trigger too soon. Can we get one last feature in before the PO? Have we vetted the manufacturability? Has the whole team signed off? When is the team free for a robust review?

These are reasonable questions that inevitably waste time. The highly talented team you’ve assembled, wastes much more money spinning their wheels than the actual prototype cost.

That new feature will be tacked on later and will fit better once the device is physical, tactile. The manufacturability becomes clear when the clearance for a tool or finger is tangible. The team will confidently sign off with an object in hand. Leave the thorough review for when the non-mechanically inclined see everything for themselves.

Certainly, excessive prototyping must be avoided. At the same time the return on the time invested in questioning the release is costly, too.

Delaying Prototype Development Costs Too Much

As a CAD monkey I’ll admit to digital myopia, to investing hours behind a patch of unruly surfacing or the interface of several features. Eventually I come up for air and zoom out from the .001” square patch I was working on located on the interior of the housing. Why did this consume me? CAD let me blow it up to twelve inches across. My detail oriented workflow demanded I fix it before moving on. CAD didn’t show me that my tolerance fit was too tight or that when it was corrected it would deform that patch beyond anything I modeled.

Prototypes constantly save me from this digital microscope.

As much as I’ve learned to imagine the things on my screen, they never measure up to reality. I can’t feel a curve on the screen nor heft the device as I spin it. The screen won’t show me how my tolerances actually look with the real materials and processes applied. I won’t know that the interaction of my button material to the housing exceeds the theoretical value I calculated for the spring force.

Even a relatively crude prototype has saved me from each of these problems and more.

This is especially critical in the medical device industry were ergonomics are critical. Doubly so when we cannot customize our product to the user. These devices are used by every kind of person that come in a myriad of shapes and kinematics. Designing those kinds of ergonomic features defy a single designer. A brick of bondo around a 3d printed frame can be placed in everyone’s hands. All the differences can be rapidly vetted.

Now we can fail fast off my first model: a full scale prototype before I figure out the detailed manufacturing requirements, fitments, etc.

Or we delay and iterate so we can capture all those critical engineering features: Wall thickness in place. Molding process and complexity set, tolerances to purchased parts and mounting for durability. Then we make our prototype and we realize I missed that critical ergonomic feature.

What was gained in that delay? I will now redo these engineering features, most of which were decimated when I reached up to the top of my CAD model’s tree and blew the whole thing up. Certainly we will gain some of that time back as many of the answers are already there. But, if I’m doing my job, I would have gotten there, with all the bells and whistles only much sooner.

So please, give your engineers a fast chance to fail so they can work faster for you. They also like to break their toys which is great for morale. For me anyways.