Iterative design isn’t design by trial and error. Iterative design is a process of continually improving not just the design, but also the problem your design is trying to solve.

Coming up with a solution is often the most straightforward part of the design process. That isn’t to say that creating the solution is easy, or doesn’t require a deep knowledge and honed skill set. It’s just to say that when you have a set of requirements and a well defined problem, you know where you stand and where you have to get to. It’s mostly straightforward. Much harder is the implicit problem of figuring out exactly what the problem is in the first place. If the problem is vague or ill-defined, the design solution will be too.

For example, the problem statement “Design a better car” isn’t very good: it’s too abstract. Iterating a solution with a problem like this won’t get you anywhere, because the solution won’t be converge to anything in particular. There are too many ways that “better” can be interpreted. Does it mean higher gas millage? A sleeker looking? More accessibility features? Simply put, you don’t know when you are done because “better” is too open ended. A more useful problem statement might be “Design a sports car that has 60 MPG in the $25k price-range.” This is a problem that (although hard to solve) can be attacked and iterated. You know when you are done.

A good problem statement gives a tight set of constraints within which to work. Part of the iterative design process is refining those constraints.

Marissa Mayer, the Google VP for User Experience, said it well:

“When people think about creativity, they think about artistic work — unbridled, unguided effort that leads to beautiful effect. But if you look deeper, you’ll find that some of the most inspiring art forms, such as haikus, sonatas, and religious paintings, are fraught with constraints. They are beautiful because creativity triumphed over the ‘rules.’ Constraints shape and focus problems and provide clear challenges to overcome. Creativity thrives best when constrained.”

My Quest

I’ve been on a quest for the last 16 years. It hasn’t been a constant quest, nor a particularly important one. Just one close to my heart: to design a paper airplane perfectly suited for duration flights. An airplane that you throw hard—nearly straight up–so that with just the right trimming it reaches the top of it’s arc, gracefully roles upright, and begins a flat circling descent. I’ve had planes stay aloft for upwards of a quarter of a minute inside the windless sanctuaries of gyms. Although fifteen seconds may not seem like much, every tenth of a second that goes by on the stopwatch feels like a blissful minute. Until 1998, the indoor world record was 18.8 seconds.

But my quest didn’t start with the duration as its goal. It started much more simply.

Design Goal: Make a paper-plane

Like most kids, my first airplane was the “dart”. It’s the Ford Pinto of paper-planes: ready to fall apart, crash and burn at anytime, but cheap and easy to make.

After playing with the dart for a bit, I was unsatisfied: it flies less like a airplane than a rock with ersatz wings. I began to desire something that would stay up: A plane that lets you revel in the majesty of flight. The dart lets you put other people’s eyes out. Something else bothered me: A number of kids added paper clips or tape to their planes. I thought that that was cheating. They are paper-planes, not paper-and-stuff-planes.

These two requirements led to a new design goal.

Design Goal: Make a pure paper-plane that glides well.

It was my father who taught me how to make the little-known arch-nemeses of the dart. The “glider”.

The glider is the opposite of the dart. The dart is thrown with vicious force, like a lit stick of dynamite; the glider is let go hesitantly, like a parting kiss. The glider’s slender wings have little structural integrity and its lack of a fuselage makes it hyper-sensitive to air currents, but with no wind and a gentle launch it glides beautifully like a warm penny on ice. It barely seems to lose any altitude at all as it skims across a room. But if you throw it, its high-aspect ratio wings flutter and it twiddles uselessly to the ground. The dart and the glider form two ends of the spectrum: The dart lacks finesse and is constantly unfolding itself, but can be thrown hard and is impervious to wind; the glider is all finesse and nearly floats, but it is at the mercy of sneezes in the other room.

The plane I wanted lay someplace between the dart and the glider. It needed large wings to make it float (like the glider) but with the ability to be thrown hard (like the dart).

Design Goal: Make a pure paper-plane that glides well and can be thrown.

Sometime around the 2nd grade I learned how to make a plane that was the first step towards this goal. I call it the “classic”. This was my plane for years. It served me well, yielding serenely long flights. The design is overall well-balanced: it launches well, has a flat glide, and keeps itself together. The classic is easily “good enough” and I stopped searching for for a better plane.

This is a pitfall to be careful of in iterative design. It is easy to iterate, get to a place that is suitable, and stop iterating. It’s similar to the problem of “rising to the first level of incompetence”. That is, if you are good at your job, you’ll get promoted. If you are good at your new job, you’ll get promoted again. The chain of promotion continues until you get to the first job at which you are no longer competent. With iterative design, it’s the problem of “rising to the first level of competence”. The design will get stuck at the good-enough and will never progress to great. When deadlines are right, the pressure to stop iterating becomes even stronger. Fight the pressure; otherwise the design will forever be stuck at mediocrity.

It wasn’t until college that I was bitten by the bug of redesign. I realized that to build a maximal-duration airplane, I needed to improve the launch. Towards this end, I wanted as much weight as possible towards the front of the craft (which gives stability when thrown–it’s like the weight at the front of an arrow). Extreme-duration paper-airplane flying is a rather geekish sport, so it is somewhat ironic that I trial-and-erred my way to the next iteration when I was on a date.

Design Goal: Make a pure paper-plane that glides well, can be launched, and has most of its weight upfront

I was pleased by the clean shapes of this craft–all of its weight is kept forward and it performs better than the classic design–but there was still something wrong. With a really good throw, the plane flattens out under the stress and flutters around a bit. This is exacerbated by the wing folder design choice, which yields a cleaner wing fold for less support at a stress point. What I needed, I realized, was a solution to a problem shared by every paper airplane I’d ever seen.

Most paper-airplanes are based on a central valley fold. The problem is that when launched hard, there is nothing to keep the fold together so it flattens out. By the time the plane reaches the top of its arc, it’s bent out of shape. The perfect plane should keep its fuselage held tightly together.

Design Goal: Make a pure paper-plane that glides well, can be launched, has most of its weight upfront, and has a bound fuselage.

I tried many plane designs and kludges, but a solution eluded me. One recent night, I realized how to do it. I just needed to invert the design. Literally.

And there we go: an airplane that fulfills all of the requirements and the latest victory in a 16 year quest. The plane glides wonderfully, can be launched high with hard throws, and it’s the sturdiest paper plane I’ve had the pleasure to fold. The one oddity is that it flies best in what most people would consider “upside-down”. The inversion fold is, as far as I know, completely unique in the world of paper airplanes.

The Moral

Aside from the fact that you now know how to make 4 awesome paper-planes (and one bad one), what’s the moral of the story? As you design, you aren’t just moving towards a solution, you are learning about what problem your solution is trying to solve. So, don’t stop iterating your design. And more importantly, don’t stop iterating your problem.

RT @aza Iterative Design: Towards the Perfect Paper Plane | Follow @aza on Twitter | All blog posts