When did we start being concerned with usability? Some will say that such concern is part of being human: cavemen worked their stone tools to get them just right. Interaction design mattered even then. But the field of usability research really came into being when the tools we used started to run up against our cognitive and physical limitations. And to avoid hitting literal, as well as psychological, walls, it was the aviation engineers who started to think about usability seriously.
While cars were becoming ever more sophisticated and trains ever faster, it was the airplanes that were the cause of most usability problems around WWI. Cars were big, but didn’t go very fast or had a lot of roads to travel on at the turn of the century. In the first decade of the 20th century, there were only 8,000 cars total in the U.S. traveling on 10 miles of paved roads. In 1900, there were only 96 deaths caused by the automobile accidents. Planes were more problematic. For one thing, the missing roads weren’t a problem. And a plane falling out of the sky in an urban area caused far more damage than a car ever could.
Planes were costly to make, were made useless after a crash, and caused expensive damage on the ground. Something had to be done to reduce human errors of aviators. By the end of WWI, two human usability testing labs were born, Brooks Airforce Base in Texas and Wright field in Ohio, to conduct studies of pilots: what were the characteristics of a successful pilot? To answer such deep questions, the first flight simulator was developed in early 1930’s by Edwin Link.
The early years of usability focused on finding the “right kind of people” to fill the jobs of pilots, as opposed to fine-tuning the aeronautics equipment to make it more user-friendly. For a decade or so this strategy worked. But when WWII raised the demand to find qualified individuals to fly military missions in Europe, Asia, and at home, that strategy of “finding a few good men” wasn’t good enough. The flying machines had to be designed with users in mind, and the true science of human usability was born.
I find it fascinating that computers had a similar usability history. The original computing machines were programmed with cables, making actual hard-coded connections to create algorithms. This state of affairs was vastly improved by a command line interfaces (CLI). For those who entered this world before 1970’s, the everyday human computer interaction (HCI) looked a lot like UNIX—you had to memorize commands and know when and how to use them to get the computer to do anything. As with aeronautics, we needed a few good men to run these machines.
But similarly to men who had the cognitive and physical characteristics to become good pilots, good programmers were hard to find. We had to train people long and hard to make computers do the simplest of tasks. This was not a workable solution in the long run.
GUI’s came next (Graphical User Interfaces). It was no longer about memorizing commands, but rather about recognizing the right interface element to do the job. Recognition beats cold recall every time. More people could now use the computers.
Not all GUIs are created equal. Making something usable is very, very hard and time consuming. But luckily, we can draw on some of the wisdom gained in century’s worth of thinking and research conducted by the aviation industry.
Pilots still have to work long and hard to learn the interface elements of the machines they work with, cockpits still look like some futuristic torture chambers for median IQ endowed among us. But the “cockpit” of an average SUV is getting better. There is a strong push towards making products more usable and towards stopping blaming the users for everyday failures.
February 2011 addition: