Disaster’s & Accidents: Human Error by Design

I am a disaster junkie. It started innocently enough with a childhood fascination with volcanoes. I was ecstatic when I felt my first earthquake in San Francisco, devastated when I missed out on F3 tornado that ripped through downtown Fort Worth, TX, my hometown, in 2000. Growing up in gulf (of mexico) state meant great coverage of hurricanes (the Galveston Hurricane of 1900 still has the highest body count of any U.S. natural disaster.) My fascination is not limited to natural disasters alone. I own the “I Love Disasters” issue of Murder Can Be Fun, the “infamous zine of bizarre death, murder, mayhem, crime, disaster,…” (now out of print) which includes an account of the great Boston molasses flood that killed 51 people. I developed an interest in plane crashes in college visiting this site for cockpit transcripts and other info (they list September 11th as the the worst aviation disaster in history but the collision of two 747′s on the Canary Islands is still #1 in my heart.)

So, yes I have a “thing” for the cataclysmic. And guess what? So does HCI, by way of psychology and engineering, specifically human factors (because what’s the number one cause of engineering disasters: human factors!) This doesn’t mean blame the dumb pilot, ship captain, or whomever. On the contrary human factors practitioners (HCI can be considered a sub-discipline or specialization in this field) seek to understand in part the design flaws that led users to err. Take the Canary Islands crash I mentioned earlier (583 dead–yikes!) In The Design of Everyday Things the Don (Donald Norman that is) explains that there were misunderstandings between the pilots and air traffic control that led to the crash (read the cockpit transcripts for yourself.) Surely innocent charts and graphs couldn’t lead to death and destruction? Think again–Edward Tufte blames the Challenger explosion on crummy overhead projections (see The Challenger: An Information Disaster from his book Visual Explanations.)

Can’t get enough of this stuff? Check out the Association for Computing Machinery’s (ACM) RISKS Digest (they even have an RSS feed–sweet!) As for myself I’ll be eagerly awaiting next spring when I will (hopefully) take a class called Human Errors and Complex Systems Failures in the Industrial Engineering & Operations dept. here at UMich.

March 22, 1977 crash of Pan AM 1736 and KLM 4805 in Tenerife, Canary Island