Root cause analysis for that one included: * A trading application where it was easy to swap the two values. * Absence of sufficient required confirmations. * Cultural problems at Mizuho which left staff afraid to countermand clear errors. * Same at Tokyo Stock Exchange.
-
-
Show this thread
-
Regulator to stock exchange: "When did you become aware of that order?" "$TIME." "Who became aware first?" "$PERSON." "Why did they not countermand the order?" "Mizuho Securities sent it to us and they felt unable to overrule..." "YOU ARE THE STOCK EXCHANGE. YOU HAVE ONE JOB."
Show this thread -
(A thing that I genuinely like about engineering culture over here, heavily influenced by a certain automobile manufacturer near Nagoya, is that "andon cords" and virtual equivalents are widely, widely deployed. https://en.wikipedia.org/wiki/Andon_(manufacturing) … )
Show this thread -
(It's, conceptually, a big red button which causes an expensive remediation. The important part is not the cord/button, it is the organizational expectation that, if you suspect a quality issue, you are *responsible* for pushing button and your concern *will* be addressed.)
Show this thread -
Anyhow: operator error is a thing; properly engineered applications and processes assume that operators will occasionally make mistakes and should be designed to be robust against them.
Show this thread
End of conversation
New conversation -
-
-
Plus, as Ray Dalio argues, it results in a culture where hiding mistakes whenever possible becomes the rational thing for employees. The only fireable non-chronic mistake should be hiding a mistake.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
In addition to that most(probably all...) these types of errors are the result of system issues and punishing the poor person who happens to be fall victim to them is just a good way to make sure the system remains unimproved.
-
No realize you said basically this in the thread. oh well.
End of conversation
New conversation -
-
-
Fascinating! They told us this story at IBM training about Thomas J Watson, but the analysis didn't go deep: http://mbiconcepts.com/watson-sr-and-thoughtful-mistakes.html …
- 3 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.