The other FAIL blog
May 31, 2015 11:49 AM   Subscribe

Dr. Yotarou Hatamura, who runs the Association for the Study of Failure, is also supervisor of the Failure Knowledge Database Project. He proposes adopting the "Failure Mandala" to promote the systematic understanding and dissemination of failures. To support this approach, he presents a compilation of 100 case studies of failure events organized by topic (also viewable as a single list of PDF files; and available in Japanese). Dr. Hatamura subsequently chaired the investigative committee into the Fukushima nuclear incident, which he discusses here.
posted by Rumple (11 comments total) 54 users marked this as a favorite
 
What a fascinating organization. Thanks for posting, Rumple.
posted by clockzero at 12:04 PM on May 31, 2015


I spend a lot of time analyzing and critiquing undesired events. This looks like very interesting reading, thanks.
posted by ctmf at 12:23 PM on May 31, 2015


This is extremely interesting! I'd thought about writing a book like that someday, a history of failure, though I know others have also been written. Fascinating!
posted by limeonaire at 12:35 PM on May 31, 2015


Related in the lists of failures category: The Encyclopedia of Ethical Failures (Word doc) (HTML version) from the US Department of Defense. A catalog of the many ways in which government employees can screw up, from false travel claims to bid rigging to campaigning for politicians on duty to throwing out government property instead of cataloging and storing it.
posted by zachlipton at 12:48 PM on May 31, 2015 [4 favorites]


So I've only read the 'Failure Mandala' link so far. He's not the first to try to have standard classifications of cause/action/result. It's nice to have as an investigation tool, to not forget to look at all possibilities. The problem is, most failures are not easily traceable to one thing. He briefly touches on it, but then kind of drops the subject. Most failures are a result of a number of small errors that alone wouldn't have been much of an issue. Often those same errors happen all the time and aren't big deals. The small errors combine in unpredictable ways, though, with an occasional big problem.

I mean, in a simple case of the worker burning himself on a hot surface: was the training insufficiently emphatic about the safety risk? Did he have insufficient experience in the task? Was he distracted or tired? Was he in a position where there was no backup opportunity from a supervisor? Could the procedure be arranged to make the problem less likely?

Yes, all of those, sometimes, and maybe any one of those could have prevented the problem. There's a tricky balance between solving this particular problem quickly and efficiently, and taking on an over-abstract general case of all possible similar problems.
posted by ctmf at 1:27 PM on May 31, 2015 [1 favorite]


One item on the Fail Mandala, "Disrespect For Tradition" struck me as quite interesting. I thought it should have a compliment, though: Blind Respect For Tradition. Maybe that falls under "Poor Value Perception"?
posted by cleroy at 3:33 PM on May 31, 2015 [3 favorites]


That's "FailURE Mandala"... the difference between "FAIL" and "FAILURE" is the difference between "just point, laugh and don't think" and "overthinking a plate of failures" (which that round, pie-ish chart resembles).

And "Blind Respect for Tradition" probably comes into play in all the subcategories in the 'Organization is Responsible' section of the chart, thus absolving the individual with a comfortable "I was only following orders". Or it could involve 'Ignorance/Insufficient Knowledge' of the pre-existing failures of the Organization. Any guide that tries to cover everything will obviously miss something - what seems to be missing here is specific mention of Moral Failures, which Dr. Hatamura probably considers a subject for a whole different institution.
posted by oneswellfoop at 3:55 PM on May 31, 2015 [1 favorite]


The best part about working for the Association for the Study of Failure is that you will never, ever, run out of material to study.
posted by Faint of Butt at 4:54 PM on May 31, 2015 [2 favorites]


This is cool, I have to write a report on some failure of my team about once a month. I might try building up the 3d failure mandala with the material I've got so far.
posted by Joe Chip at 8:48 PM on May 31, 2015


I'll admit that I got all excited at the phrase "failure mandala" but left the site a little disappointed that there wasn't something a bit prettier to hang in the office. Perhaps I'll fire up photoshop.
posted by mcrandello at 11:09 PM on May 31, 2015


I'm worried there may be an issue with the narrative fallacy once you start case studying failures.

Someone mentioed earlier about "cascade failures" in complex systems. Yet these can become hidden as we build a narrative of the failure.
posted by herda05 at 1:27 AM on June 1, 2015 [1 favorite]


« Older 3.3-million-year-old stone tools from Lomekwi 3...   |   “Many atheists are still in the closet,” Newer »


This thread has been archived and is closed to new comments