transhuman-priestess:
transhuman-priestess:
transhuman-priestess:
The more I read into reports about industrial and transportation accidents the less I feel like “operator error” actually exists
Ok so “doesn’t exist” may be a slight overstatement. A better way of phrasing it might be “operator error is often used as a way of warding off close examination of how systems fail.”
You read about airlines accidents attributed to pilot error, and almost universally you find overworked, overtired people who have to deal with inadequate training, and poorly maintained equipment. Often investigations uncover a pattern of management ignoring problems that pilots regularly have to deal with. Out-of-date terrain data, false sensor readings, confusing systems presentation, fatigue.
The cargo airline industry fights to keep its pilots exempt from crew rest requirements and a fatigued crew crashes a mile short of the runway. Only the two crew on board die, so really it’s no big deal, right?
Amtrak builds a new bypass to cut 10 minutes off the travel time from Portland to Seattle but doesn’t give the engineers enough training to prepare them for it, nor installs adequate signage to warn of a 30mph curve, so on the inaugural run the engineer hits the curve at 80 mph.
Construction on a nuclear power plant runs into trouble and so to make a key pressure-bearing component fit, they install an S-bend around a pipe, which causes falsely water level readings. Operators open a valve to reduce what they think is excessively high pressure in the reactor and it melts down.
And all of these get simplified, either initially, or in perpetuity, as operator error. Because operators are cheap and easy to replace. Firing someone and laying the blame on them is cheaper than reassessing and restructuring a management culture built on passing the buck.
This is an extremely valuable addition thank you selky ❤️
related pet peeve as someone who used to work on industrial machinery: blaming the technique of the person that fabricated it, specifically (nine times out of ten) blaming the welder. Plane crashes, structural failures, car accidents, pressure vessel explosions, nuclear incidents, and even the loss of entire ships and submarines have all been blamed on “bad welds” (i.e. poor welding technique, or welds not conforming to the print) when that’s simply a bad way to look at it; it’s finding one worker to blame and then not doing anything to fix the problem. In critical applications, there should simply never be a situation where a bad weld causes a catastrophic failure, for three reasons:
- QC should have caught it.
- if QC didn’t/couldn’t catch it, it should have been engineered redundantly so that one bad weld wouldn’t cause total collapse, and it should have been subjected to regular inspections.
- if there is no way to get around a single cracked weld as a failure mode, it should have been designed with the knowledge that eventual failure is effectively inevitable as stress fractures and corrosion weaken the joint over time, i.e., fail-safes should have been in place.
so if that’s the case, if there are supposed to be reduncancies, why do welders keep taking the blame?
a) Welds are most often made by human welders, especially in critical applications like nuclear reactors, aerospace parts, pipelines, bridges and buildings, and repair/retrofitting of existing parts (e.g. automotive repair, though mostly not auto fab anymore) where the use of robots is unfeasible. this means that all the above issues re: “operator error” apply. There’s a human being you can pass the buck to and say “he did it.”
b) Welds (or, more often, the surrounding HAZ) are almost invariably the point of failure when a welded part is subjected to extreme stress. If you find your big important contraption (plane, boat, bridge, nuclear reactor, whatever) in pieces and it’s cracked along the welds, the welder is going to logically be the person you blame. Not the engineer (or lack thereof), not the QC department (or lack thereof), not the boss that didn’t provide adequate time, materials, or conditions to make a cleaner joint, not the fitter who left a huge gap in the fitup nor the project manager who didn’t budget for redoing mis-cut parts, not the malfunctioning machine with dodgy voltage controls that the shop refuses to replace because “it still works,” not the foreman who was rushing the workers to reduce the amount of billable time spent on each task so that his team metrics would look better - when you see a part fail, it’s easiest to blame the person who physically made it, so that’s who gets blamed.
Looking for someone to blame is never a good way to deal with the results of a whole system going wrong, because you will definitely just be pointing fingers at the last guy to touch it.
