Before I became an industrial designer there was a dark time when I was studying ….. mechanical engineering! Dark times indeed! However, between those two moment in time, I had the amazing opportunity to be the “Internet of Things Intern” at the FabLab in Amsterdam. I look back on this time fondly, and it sparked a love of electronics which I still cary to this day (and use in my personal and professional life).
Today, I want to share one of the key ideas that I learned at the FabLab that I have repeated to countless fellow industrial designers when advising them on how to work on electronics for school projects, for their jobs, or even for startups. I call it “the onion model”.
Why an onion?
Many reasons, besides Shrek references! Namely:
- Like an onion, electronics projects work in layers.
- You don’t bite into them unless you really know what you’re doing (or really don’t)
- It partners really well with garlic (which I guess is software?) and if you are going to make it right, you use butter (which is good design).
The Model
Concept
This stage is the crunch outside of the onion, when you’re in the phase you still don’t know what the onion actually looks like. It’s also the least clearly defined. All you need to do in this stage is squeeze the onion and figure out what you want the electronics to do. This is the stage of actions, reactions, and context.
Paper Diagram
Once you know what you want the electronics to do, it’s time to whip up a paper diagram. Now, you can do this on the computer, but I like using actual paper.
This isn’t the moment for full on schematics yet, you just want to divide your project up into functional blocks. In that way you can understand the project better and search for components on a block by block basis.
Breadboard Prototype (We can make it work)
Once you have selected your components, go ahead and order the breakout boards for them and pop them all into a breadboard. This is the moment for the Minimum Viable Test! Check and see if the electronics work together, and then you can begin to validate your concept (user interactions, the electronics’ performance).
This is usually a frustrating exercise in squeezing everything onto a breadboard and constantly running out of the right sized jumpers. However, this is an essential step to see if you made the correct choices and assumptions in the earlier stages.
Protoboard Prototype (We can make this)
If you have made a few PCBs in the past, you might get away with skipping this step. However, using breadboard is the quickest way to validate that you can make your electronics for (twice the) cost, make them work repeatedly, and honestly, depending on your requirements this can be your MVP. If you’re making 40 a year, you have some overhead, and there’s not a lot of space/mechanical requirements a Protoboard will do you just fine.
PCB prototype (We can get other people to make this)
Once you have a protoboard version, getting to a breadboard is not too difficult. If you have little experience or patients to solder QFN chips, then you can really get away with making a more professional version of your breadboard prototype (which comes at a cost). If you don’t know what I am talking about then I would definitely get someone to do this for you.
PCB Optimization (We can get a robot to do this)
Look how tiny the center of that onion is! There’s almost nothing left! Well, turns out there can be a great deal. From setting up automated assembly to haggling over the price of resistors because you are going to buy a million, PCB optimization is a whole onion on its own. But for designers, this is just a little bit of the inside of the onion you should hand off to other people.
Okay, lots of words, what does this mean for me?
Want some quick take-aways ehy? Alright!
Know what you want, what you need
Know your recipe, know how much onion you need. Everyone wants shiny, sexy electronics, but if you only need a proof of concept or if you are making 10 products a year, you don’t need to get it assembled by robots overseas.
Set expectations on people and the process
I once talked with a designer lead company, they wanted someone who could speak their language and do design research (so a designer) but also produce a flawless PCB for them….. in 5 months.
Basically, find a person who can go from the part of the onion you are at, and then knows how to get to where you need to! (Or is some amazing designer who has experience working with Electrical Engineers…)
Thanks for reading
This was a bit of a disjointed thing I wrote in the train while working at Mobgen | Accenture Interactive on my graduation!
And now that I am done, I am on the hunt for more things to work on!
