3 ms·

It sounds like an interesting book. Is there a more up to date canonical text on failures of complex systems? The Wikipedia page includes "almost all serious n

by rz2k 9y ago

It sounds like an interesting book. Is there a more up to date canonical text on failures of complex systems? The Wikipedia page includes "almost all serious nuclear accidents have occurred with what was at the time the most recent technology". Chernobyl was trying out a new cooling method, but it wasn't a very new reactor. Fukushima wasn't very new either. However, the book was written before both nuclear disasters. I do think there is a social sciences aspect to the risk of nuclear energy. People who know enough to minimize risks are afraid enough of others' over-reactions to transparency about flaws and mistakes that there is an incentive to cover things up when they could be fixed. The other problem is that Chernobyl and Fukushima are talked about as the worst a nuclear disaster can be, when their impact was greatly contained by heroic efforts of individuals

kevinr 9y ago
Not the parent who mentioned _Normal Accidents_, but this is one of my hobby-horses, so I'm glad you asked! Prof. Nancy Leveson of MIT has done extensive research over the last forty years on the safety engineering of complex systems, and her 2012 book _Engineering a Safer World_[0] (open-access PDF at that link) is partly a response to Perrow's _Normal Accidents_. [0]: https://mitpress.mit.edu/books/engineering-safer-world https://mitpress.mit.edu/books/engineering-safer-world In particular, and unlike Perrow, she thinks we really can design complex sociotechnical systems which aren't disasters waiting to happen, and her book provides a theoretical framework for doing so, plus a fully-worked design analysis technique and accident investigation technique using that framework. These have been and are being used under real-world conditions in a variety of industries including transportation, the military, and even software. There's a conference every year at MIT which I've been going to every year for about four years now where folks present. Every year there are more industries and disciplines represented, and there have also been presentations of analysis techniques applying the theoretical framework in practical contexts like cybersecurity and privacy. You're also very right that there are huge social and human factors components to safety, and her book talks a lot about getting that right and analyzing what went wrong at various levels of abstraction, including cultural and regulatory. I highly recommend it, if you're interested in this subject. Based on what I know of the US Navy, I'm certain that the accident investigators working on this collision are doing so informed by this methodology.
AdamJacobMuller 9y ago
If you're interested in stuff like that amzn.com/B00C5R7F8G is great

by rz2k 9y ago

kevinr 9y ago

Not the parent who mentioned _Normal Accidents_, but this is one of my hobby-horses, so I'm glad you asked! Prof. Nancy Leveson of MIT has done extensive research over the last forty years on the safety engineering of complex systems, and her 2012 book _Engineering a Safer World_[0] (open-access PDF at that link) is partly a response to Perrow's _Normal Accidents_. [0]: https://mitpress.mit.edu/books/engineering-safer-world https://mitpress.mit.edu/books/engineering-safer-world In particular, and unlike Perrow, she thinks we really can design complex sociotechnical systems which aren't disasters waiting to happen, and her book provides a theoretical framework for doing so, plus a fully-worked design analysis technique and accident investigation technique using that framework. These have been and are being used under real-world conditions in a variety of industries including transportation, the military, and even software. There's a conference every year at MIT which I've been going to every year for about four years now where folks present. Every year there are more industries and disciplines represented, and there have also been presentations of analysis techniques applying the theoretical framework in practical contexts like cybersecurity and privacy. You're also very right that there are huge social and human factors components to safety, and her book talks a lot about getting that right and analyzing what went wrong at various levels of abstraction, including cultural and regulatory. I highly recommend it, if you're interested in this subject. Based on what I know of the US Navy, I'm certain that the accident investigators working on this collision are doing so informed by this methodology.

AdamJacobMuller 9y ago

If you're interested in stuff like that amzn.com/B00C5R7F8G is great