Product design is a corn maze

There are many analogies I’ve heard for the process of designing and developing hardware. Some compare it to solving a puzzle, others compare it to building a house, and some compare it to Dante traveling through hell. All are valid but fail to factor in how dynamic the process is, and how decisions made during this process have ramifications later on.

As a Vermont resident, I’m legally required to navigate at least one corn maze each fall season. For those of you who haven’t had the pleasure, a corn maze is the devil’s work — a labyrinth of passageways mowed into a field of corn and created mostly to cause arguments between friends and loved ones while burning the top of your scalp. Some are themed after characters from the movies or farm equipment, while others are geometric monstrosities. 

I think this is Robert Downy Jr. corn maze. (Image from Cooper’s Farm.)

I believe the process of creating hardware is most similar to solving a corn maze due to the techniques used to solve the maze, the emotional toll it takes on all involved, and the intense relief that washes over you at the end. However, unlike every godforsaken corn maze that exists on this planet, this maze has infinite endings which change depending on the decisions you make. It’s not a watertight analogy, but it’s better than a puzzle.

Why don’t you just compare this to a regular maze?

This is what a regular maze looks like to a maze solver:

Go ahead and solve it. I’ll wait. (Image from Maze Generator.)

Some mazes can be complicated, and some have multiple solutions. But these mazes all have a known start and end position that is visible to you, a professional maze solver.

This is what a corn maze looks like to a maze solver:

Terrifying image courtesy of Dall-E.

Not so easy now, is it? At each intersection, you must make a choice. If you’re not cheating and using GPS or breadcrumbs (I’m looking at you, Hansel and Gretel), you must also remember these choices and why you made them. Inevitably, you’re going to take some wrong turns and need to double back. Now, here’s another annoying part about a corn maze — you’re surrounded by corn, which tends to look like the other corn around it. Did you take the first left, or was that a right turn? Did you go straight? Did you already try going left on the third turn? The sun is peeking out through the clouds, and you’re almost out of caramel corn. It’s been three days, you’re missing a kid, and your wife is annoyed at you for making corny jokes (get it?). Corn mazes are harder because you need to make decisions on the fly, try various outcomes and remember all of this.

What does any of this have to do with product design?

Every project I’ve worked on over the past 10+ years has started with some rough idea of what we’re going to make, a timeline, and a feature list. However, decisions are still made on the fly as we get deeper into the design process. These  have repercussions and can leave lasting impacts. Let’s use an example product.

The pinnacle of product design.

This engineering marvel has four buttons, four lights, a circuit board, a speaker and some ports. It’s made of aluminum because the client wants it to look like an Apple product or something. When you press a button, the light turns on and a different animal makes a sound. It connects to an app as well to track how many times you pressed the button and compares it to how many times your friend pressed the button. It’s an innovative product and there’s a need for it. 

Since this product connects to the internet, it needs a wireless module and an antenna. Most of the time, we’d use a module that has an antenna attached to it like the ESP32 module with integrated antenna, the core of most IoT products.

It’s pre-certified by the FCC, inexpensive, and reliable. We can follow design rules and know that it should perform well in most environments as long as there is no metal immediately surrounding it. However, we have an aluminum enclosure due to #designinspo. This means wel need to add a more expensive patch antenna somewhere not covered in metal. The decision to make the housing out of metal was the first turn we took in our maze, and it means we can no longer use the wireless module with a built-in antenna.

Let’s fast forward 6 months into the product design process. We’ve ordered CNC-machined samples of the housings, have iterated on the electronics, and the app is finally up and running. We’ve received pricing from our vendors and learned that making each housing out of aluminum is going to be too expensive in the long run. Now, we need to pivot to a molded plastic enclosure. We design a delightful plastic housing, and everyone is happy at last. 

Months later, we revisit the design and begin to wonder why we’re using this more expensive patch antenna. The engineers shrug and say “I think we’ve been using that since the beginning.” After hours of scratching our heads, we remember that the housing used to be metal and it drove the need for a separate antenna, but now we can switch back to the wireless module. Everyone embraces and celebrates. We’ve backtracked through the maze and can take a different, more efficient path.

Re-solving the maze after changing from metal to plastic housings

In a perfect world, the decisions you make would have no repercussions on the rest of your design. In reality, the battery in your product is now too close to the wireless antenna on the module and youl need to modify the layout of components or relocate the battery. Fun, right?

Now, even if you’re the smartest maze solver in the world, you’ll always need to backtrack a little. Things change over the course of a year or so. Parts go out of stock, the scope of the project changes, people have opinions and new opinions. New products launch that may compete with yours, so you need to adapt to stay competitive. Nobody solves the maze in one shot, but you can solve the maze faster and more efficiently.

How do I solve the maze faster?

Unlike professional corn maze solving, it’s highly encouraged to document your decisions and the reasons behind them. Medical products require strict traceable documentation whenever the design is modified, but for most products, it’s fine to make a shared document or add notes in your design program when you create revisions. I tend to name my CAD versions in Onshape and leave a helpful reminder summary of the changes I made and why.

This one simple trick will help you 6 months from now and it only takes 15 seconds.

Another trick to solving the maze faster is to anticipate changes and predict which have the most impact on future designs. I like to ask if a product has specific waterproofing and drop-testing requirements, if there are wireless radios, and if there are specific color material and finish (CMF) requirements early on. It’s far easier to make a product less waterproof later on than it is to make it waterproof later. Similarly, you can always remove a wireless radio with minimal design impact. The CMF decisions impact manufacturing methods, how the housings are split up, and how you design the parts. A part designed for injection molding can be easily CNC machined or 3D printed, but going the other way is much harder. 

Once the product is “feature complete,”I like to zoom out and look at the entire product. We can finally view the maze from the top down and see the entrance and exit, as well as the path we took. Sometimes, an obvious “shortcut” in the maze is revealed, simplifying the design by removing the layers of quick fixes that were applied during the iteration process. Take the Juicero Press as a great example of how not to solve a product maze. In case you lived under a rock during this fiasco, it was quickly discovered that the juice pouches could be squeezed by hand easily, negating the need for an $800 IoT-connected kitchen appliance.

I just saved you $800 (Image from Bloomberg Technology)

The team at Bolt wrote a phenomenal article tearing apart the Juicero (physically and emotionally) and discovered a very expensive machined aluminum mechanism designed to compress their juice pouches. This mechanism was impressively overengineered and could’ve been drastically simplified by squeezing the juice pouch like a toothpaste tube instead of a can crusher.  

None of these parts belong in a juicer. (Image from Bolt.io)

If anyone took a moment to zoom out and ask “what are we doing here?” I think the Juicero would’ve been a cheaper product that could’ve survived a bit longer. My guess is that the team was short on time and ran out of budget to make the necessary changes to cost down this product, and someone uttered the words that sends a chill down the spine of any engineer: “Just ship it.”

Lastly, it always helps to bring in another set of eyes later in the design process. During design reviews, engineers can poke holes in the decisions made in the past and recommend alternative solutions. Having a fresh set of eyes also ensures that engineers aren’t impacted by legacy decisions that no longer are included in the design. I absolutely love this process and helping teams realize that a simpler solution may be possible. Hearing “well, we can’t do that since we need that notch in the PCBA for a mechanism, but the mechanism changed, so I guess we don’t need that notch anymore, so that’s possible” warms my heart.

informal is a freelance collective for the most talented independent professionals in hardware and hardtech. Whether you’re looking for a single contractor, a full-time employee, or an entire team of professionals to work on everything from product development to go-to-market, informal has the perfect collection of people for the job.