Designing with Conviction

Designers create order out of chaos, but design is not science. We need more than science and engineering in our quest to manipulate our environment.

Conviction versus Doubt

Doubting is easy, conviction takes work. This is true in most parts of your life, from religious faith to writing fiction. Doubt is nurtured by laziness and cynicism. Conviction about what you believe and your course of action takes guts, work, and humility. It may require going against the cultural currents and will cause grief, frustration, and a lot of work. What do you really believe? What do you think is the right approach? These are questions which are easy to ask but are difficult to answer.

Running around being cynical and doubtful is not a virtuous state. Once you are convicted of something, you need to work it out in your own mind so it is not a conviction that is contrary to evidence. How do I know if I have a good idea? Part of the inquiry can be mental, thinking through the design carefully, but part of validating your conviction is experimenting and testing. You have to build the thing you are designing and test it. It doesn’t have to be a high-quality prototype. It just has to be something physical that you can work with. Many designers are reluctant to quickly move to a physical model—it is more peaceful to push a pencil or mouse.

One of the problems with 3D printing is you can tend to use existing designs because it is easier to throw those digital files at your printer than it is to create a time-consuming CAD model. You may be very reluctant to change a design after you have spent hundreds of hours drawing a model. That is human nature.

There are ways to avoid this problem of near plagiarism or overinvestment in CAD. Build your idea in cardboard, foam, clay, or wood, to name a few approaches.

Build your idea quickly. Test out dimensions, geometry, and mechanisms. Does it fit in your hand? Does it look like something will easily break? Does your design have a chance of doing what you think it should? These issues will be very clear with a simple model. Check on control actions with hard wired motors or actuators. You can couple your simple prototypes with a microcontroller, PLC, or circuit board taken from a device similar to what you are working on. Don’t requisition materials and issue purchase orders—just scavenge stuff and use some of your coffee money and buy what you need. Cardboard, tape, hot glue, and string aren’t expensive.

You can be creative with how you test things. I have tied thin strings to the discharge hose of a Shop Vac and moved it around a model to see air flow behavior. This system makes a great poor man’s wind tunnel. I have tapped into a self-service car wash to provide high pressure water for testing. You can throw things off cliffs, hit them with hammers, immerse them in water, and perform all sorts of other home testing before you move forward with your design. I learned this early in my career when I was trying to optimize a pneumatic starting system and had our machine shop make several air constricting orifices. These didn’t work very well so I just started experimenting with squeezing a flexible hose with a vice grip. This was a cheap and easy approach, and I was able to get an idea of what would work. I have done many other simple, embarrassingly low technology tests, even with medical devices (using screws and wood as a human analog before moving to cadavers.)

However, humility is required. Complex or high-speed devices will only operate reliability in a narrow band of optimized dimensions, tolerances, surface finishes, material specifications, etc. Complex machines require much more than hunches to make them work reliably. They involve solving a myriad of subtle issues that may not be identified in a lovingly guided working prototype.

This is an excerpt from my new book, "Intense Design:Product Design Lessons From Cold War Era Skunk Works"

Check out my YouTube Channel too!