IDFuel, the Industrial Design Weblog - Post details: The Smallest Thing...

The Smallest Thing...

08/29/05

Most often, design problems are proportional to the size of the design mistake. A large design error -- like the complex, laborious, and ugly design and engineering of the Aztec truck -- makes for a large performance disappointment. A small design error -- like an unfortunate screw placement in a VCR remote control housing -- might make things inconvenient every once in a while, but mostly they are small issues. However, when designers begin dealing with complex systems with many interrelated parts, a little design decision can lead to the utter failure of the design.
082905_simpcommain

A perfect example of this tiny flaw that brings down the house exists in my kitchen. The door of my refrigerator does not close.

Actually, it closes easily enough, when you put your mind to it, but mindfulness isn't something we've come to accept in an age where precision manufacturing and magnetized rubber door gaskets virtually ensure a perfect closure. When the door is operated normally, it just doesn't close. After a couple of minutes of investigation, it turns out that the culprit is the inside veggie crisper drawer.

For some reason, the door closing does not have enough oomph to close the drawer all the way, and so it closes it just enough so that it will consistently stop the door just outside the gasket's magnetic grip.

The end result is that the refrigerator is broken. Not kind of junky, or inconvenient, but just plain not cold, which is almost unbelievable, considering the simplicity of the problem. The door would only need to swing less than one more degree to close, so fixing it would mean removing one miniscule piece of plastic from the outer edge of the drawer handle. If the drawers had been made curved like this in the factory, this problem would not exist.

Move on. Same small kind of thing, different problem. In specifying design materials for your flagship lighter-than-air craft, you respond to a need for a weather-proof, heat-reflective coating with an aluminum paint impregnated canvas. It's a great solution, and it works beautifully for a number of flights. The trouble is, you've set up a situation in which your paint is incredibly flammable, your coating can burn, and the stuff that its coating is hydrogen.

One small choice of a different paint might have caused other problems, but it almost certainly would have kept the Hindenburg from being the absolute disaster that it was. In fact, after the Hindenburg, the paint used on later Zeppelins was changed to include a fireproofer and bronze, rather than aluminum, which was much less combustible.

As designers take on more and more complex systems, this problem will become more prevalent. Both space shuttle disasters were attributed to almost trivially small pieces of the assembly (though obviously not trivial, given the situation). Honda's choice to include a molded rubber floor in their Element vehicle resulted in a widespread and disastrous rumor that the car could be "hosed out" when it got dirty -- which destroyed the 500+ dollar airbag system. Countless other disasters are waiting to happen due to poorly placed buttons, improper materials, and unusual usage styles.

So, what can designers do about it? Two basic concepts are a good start: Testing and skepticism.

Definitely test your product as much as you can afford. Test it the way you hope users will use it, and then put it in a focus group and see how they actually will use it. And test it that way too. Maybe even try to imagine how a user might use it in extreme situations -- distracted, talking on the phone, in the shower... You get the idea.

And be a skeptic. Designers need to give up on the idea of Titanics. Every product is sinkable, and your user will find a way to do it. Your job isn't to eliminate the paths to failure, because that's not possible. Your job is to make sure that flaw is only a minor inconvenience; That despite it's failure, astronauts return home, airships stay up, and chicken stays below 40 degrees.

Permalink

Previous post: Technology eats itself

Next post: Which view is correct?

The Smallest Thing...

08/29/05

Search

Categories

Archives