When I started this Zero to Maker process, I was inspired by the idea that I could become a self-made industrial designer – that I could bypass an expensive education by learning only what I really needed to get started. Now that I’ve started down the road, I’ve realized that it’s a very long road. The good news is that I remain committed to my initial belief that there’s a less expensive (and faster) way to learn: through a strategic blend of internet resources, access to the right tools, and involvement in a community of like-minded makers. Most important for me, however, has been the commitment to trying to figure out and examine more of what I don’t know.
In a stroke of pure luck, I met someone at the Open Hardware Summit who’s thinking about exactly this. Matt Sinclair, a practicing industrial designer who is also studying for his PhD at The Design School at Loughborough University, was giving a talk on DIY Reverse Engineering during one of the breakout sessions of the Summit. In one of his first slides, he explained his research of how digital fabrication technologies (and Additive Manufacturing technologies in particular) will impact the professional industrial designer’s role and what happens when the consumer takes design into their own hands. Basically, what the self-made industrial designers aren’t thinking about.
Matt Sinclair’s project to reverse engineer a mouse
He went on to list five common issues that new makers and those of us without professional degrees tend to overlook. Matt was kind enough to summarize each of the issues into new-maker speak. By no means are these issues all inclusive or exclusive. In fact, I’m sure every maker, professional or not, runs into these issues in some form or another – the learning by doing (or mistaking) process.
Enter Matt…
Tolerances
As amazing as the computer-aided tools are, no machine or process will make a part that’s 100% accurate to the dimensions of the CAD model – it’s going to be a little off. That’s okay. Design engineers actually plan for this type of variance and set acceptable limits, which is known as tolerance. It’s easy enough for a newbie like me to figure out a tolerance for one part, probably through trial and error. The real challenge comes when you have multiple parts – each with differing materials and tolerances – that need to interact with each other. These “little” amounts of acceptable variations can add up quickly, and trying to sort out where the root problem (or problems) are can prove to be quite a headache. Cue next issue…
Functional Dimensions
Okay, you’ve got your parts back and they don’t fit together like you wanted. How do you know where the problem is? Is Feature X of Part A too big, or is Feature Y of Part B too small? The first thing to realize is that this always happens, so don’t feel bad about it (it’s why engineers use rapid prototyping and soft tooling to check before they commit to the final tools). The second thing is to anticipate problems and generate a list of dimensions, preferably in order of importance with regard to the functionality of the parts. This lets you eliminate certain features and identify others as being the source of the problem.
A rendering of Matt’s design
Post Production Finishing
The quality of surface finish from 3D printed or CNC machined parts are still a long way from what we expect from mass production processes like injection molding. If your parts are purely functional, or not visible, the surface finish may be acceptable. But if the aesthetics of your product are important then you’ll probably need to clean up, or finish, the parts. Basically there’s two ways of doing this, you can add material (by coating, in particular metalizing) or you can remove it (by sanding, polishing, sand-blasting, vapor smoothing etc). If the dimensions of your part are important, you’ll need to know how much material is added or removed by the process you choose, and design the original part accordingly.
[David's Note: This is so important. In my last post, I may have gone a little overboard about my excitement for CNC machines without mentioning the sanding we had to do afterwards because the parts didn't quite fit together. It’s truly amazing what the machines can do, but it seems to me the most experienced CNC operators know just as much about what their machines can’t do.]
Testing
The only real rule for testing your products is that the testing needs to be appropriate to the product’s use. If you’ve designed a product that’s intended to save someone’s life, you’d better be testing it a lot more vigorously than a product that’s designed to sit on someone’s shelf and look beautiful. A product doesn’t only need to perform when it’s new – you need to understand how its performance will degrade in certain conditions, and how it degrades through use. Software designers have it easy! They can release beta products and they can release bug fixes or upgrades. Almost any physical product has the capability to injure someone – saying “we’ll fix it in the next release” is almost never an option.
Redesign
More than anything else, the necessity of redesigning your product because of the points above is something that amateur makers seem to underestimate. In my professional practice, it’s not uncommon to have to redesign a product twice, so it’s the third iteration that actually gets manufactured. A client that doesn’t have this built into the project timeline is one that sets warning bells off. Accept that you’ll have to go through this process, and embrace it as a way of improving the final product, and the final stages of your product’s development will be a lot less soul destroying!
For more information on Matt and his work, please visit his website. Let us know what you think in the comments. What you would add to this list? Any stories to share?
More:
Follow David’s Zero to Maker journey
With the Galaxy Skin, one can have table clock, smart projector, or even a wristwatch. Now we await Apple’s statement regarding this.
