42 comments

[ 3.9 ms ] story [ 104 ms ] thread
> The cleaner thought they were flipping the breaker on when they actually turned it off, according to a report filed by the RPI public safety staff afterward.

If, when it really matters, the actual & on-hand operating staff for $SYSTEM cannot understand even basics of the controls for $SYSTEM - then you're control UI is a complete failure, and you'll end up f*cked.

(Though given what was at risk - I also wonder why the controls were not behind padlocked plexiglass covers. And there was no remote monitoring system with 24x7 on-call staff. And...)

Not a great idea to place a circuit breaker behind a locked cabinet when a trip on that circuit breaker requires time-sensitive resetting of it.
You could put thin safety wire and a large handle. Some P-51s and other airplanes had a soft limit on the throttle with wire. Shove the throttle hard enough to break the wire and you get more power, but so much that the engine needs to be at least inspected before it flies again.

In this case, it would probably stop someone from making a mistake but you can also kill the power in an emergency.

I think they’re lying. I doubt they tried to silence an alarm by going to the circuit breakers and flipping them, hoping it would restore power and fix the condition causing the alarm.
Obviously. And besides, if they couldn’t differentiate between “on” and “off”, they’re obviously not qualified to control the circuit breakers in the first place.

But I think they’re trying to make it sound like it was an honest mistake, rather than carelessness / bad intent, which I assume matters from a legal point of view.

In a line of circuit breakers it is impossible to not know which one is off.

Unless they are all randomly off and on for some strange reason

The pain isn't so much the $1MM of lost money (though that does matter), it's the enormous emotional cost of losing all your work.

I think I'd go on sabbatical for a year or so if something like that happened. I literally wouldn't be able to handle coming in the next day or month or at least few months.

This. I have been devastated when a project I had been working on for a few months was cancelled. I can't imagine having literal years of your work go up in smoke.
Humans hate repetitive sounds so much it's sometimes used as a form of torture.

The purpose of an alarm is to alert someone. Maintenance had been scheduled, and after that point the beeping served no purpose. The primary fault lies with the person who designed the alarm, there should have been an option to silence it.

did you not read the article before commenting? the instructions for silencing the freezer alarm was attached to the freezer...

I also find it hard to assign any blame to the alarm designer. If there was a way to permanently silence the alarm, that was specifically added to prevent unnoticed loss of temperature. That would be a much worse design. Accounting for human nature.

Yes it silenced the alarm for a few minutes, just long enough to make you notice when it starts back up.

I’m fairly confident momentary silence is part of the aforementioned torture technique. Did you not think about the article when you read it?

> the instructions for silencing the freezer alarm was attached to the freezer...

Reminds me of the "instructions to use a space toilet" in "2001 A space oddysey".

I’ve seen individual researchers ruined due to USB drive shenanigans, but our freezers were remotely alarmed (snmp trap!!) but also routinely patrolled by security who would note the temperature in their log.

The real root cause of this unfortunate and horrible failure is a lack of monitoring and human checking. Lab leadership should have to answer why security wasn’t all over this. At some point, it’s cheaper to make it someone’s job to check the thing than it is to miss the alarm!

> I’ve seen individual researchers ruined due to USB drive shenanigans

Could you elaborate?

Maybe someone unsafely unplugged the drive and corrupted it?
Or was just running around one filled with important irreplaceable data but no backups.

I'm sure some IT person is tut-tutting getting ready chide the hypothetical researcher for not following the 3-2-1 rule, or something, but science researchers aren't IT people and they can be expected to not follow IT best practices.

This is it without a doubt. The number of times I've told lab mates to back up there data... Only to discover it's on random usb drives.
I ran an overhead-funded data storage archive including multiple identical NAS on different campuses and a super fancy offsite tape archive. The researchers would rather buy a USB drive than do the paperwork to get more space because they had to explain why they needed more space. It was just paperwork, and not even mine. It served to justify our own grant requests so they didn’t have to include data storage as part of their grant.

Anyway, their life’s research is on a set of these drives, with plenty of duplication to avoid the disk failure gnome. One day, all the drives disappeared. Tuns out the researcher normally left the drives on a desk, unsecured, in a ground floor office on an urban college campus. No cameras could see the office, so it could have been anyone.

It’s literally a miracle that stuff didn’t walk 24/7.

>A sign on the lab freezer door explains the source of the alarm, and also had instructions on how to silence it, the Times Union reported. “No cleaning required in this area,” the sign said.

Why not lock the door to the room with the freezer? Why was there no backup power? What did they think would happen if the breaker broke?

Obviously, the lab can't foresee every risk. I just hope they don't consider the lawsuit as the last step in solving their problems.

The same thing happened at a lab near me. During lockdown maintenance just decided to turn off the freezer. Maintenance was absolutely not supposed to touch the freezer. Presumably some notice was given to someone to go through the building and turn off anything on, and they thought that overrode the standing notice to not switch off the freezer.

You can't really prevent this kind of thing other than automated battery backup mobile network connected alerts.

Is it a slow news day, or did this really just come to light?

> Byline: today

> when the worker … shut off the circuit breaker Sept. 17, 2020

The lawsuit appears to have just been filed.
If it’s this easy to destroy decades of work, it’s not the janitor’s fault.
yeah, it's pretty petty to destroy this low wage workers life because they couldn't afford to just secure the area completely and make sure the only way to kill it is if you have a key to the room.

A worker can't be expected to read a sign, now if they had thoroughly trained him maybe, but it doesn't sound like that's the case. It's not outside the realm of possibility that some cleaning staff might not be very literate or maybe English isn't their first language, etc...

They're suing the contracting company, not the employee. And that company presumably has insurance, so hopefully no one's life is being destroyed.

Obviously the employee may suffer some blowback, like having a hard time getting their next job, but probably no life-destroying consequences

Weird how the repairs were going to be done 4 days after the incident. So the beeping sound essentially had no purpose?

This is fundamentally a design problem: annoying sounds should be used for emergency situations involving immediate rectification, not “fix this sometime in the next week.”

Imagine if this had happened to whatever refrigerator that contained the blood samples that let researchers trace HIV back to Congo in the late ‘50s. What an incalculable loss. Who knows what else has been lost due to similar accidents and snafus.

(Btw - does anyone know a link to that paper that identified the earliest cultures of HIV? I’d love to read about the fieldwork that led to the discovery of those ancient samples that had somehow remained viable in a, uh, less than stable developing-at-best county.)

Summer intern deleted production database and all backups that were stored on the same server.

Same news. Not the intern's/janitor's fault. It is the lab's director's responsibility

Data stored on the same server, albeit duplicate, is not a backup.
I've always said that maintenance workers are the weakest link in any system.

Not sure why people are letting them off the hook. They're not supposed to use their brains, as they will always do the wrong thing. When they run into something they dim't have explicit orders to deal with, they should stop.

> They're not supposed to use their brains, as they will always do the wrong thing.

(a bit offensive but) maybe the process shall be followed/improved ?

They're not paid enough to be on the hook.

In any case, when workplace accidents happen, the liability is on the business entity, not the individual, who is an agent of the entity.

When we were in the hospital after my wife's C-Section, one of the cleaners that came in told us we could just unplug the machine that my wife was hooked up to and the beeping would stop. Ignorance is dangerous...
This is not ignorance. This is UI/UX design.

Some of us are old enough to remember people carelessly pressing the OK button in a Windows dialogue without reading the mesage in it.

Everybody wants to stop an annoying sound.

I still carelessly press OK buttons.
"Signs, signs, everywhere there's signs."

The ol' "but we put a sign on it" excuse. It's decades of research and they put a sign on it. It's the lab's fault.

exactly. now the university and the insurance companies will spend more on lawyers than purported damages just to find out that they are fault for being just as negligent and that the actual value of the samples was only in $10K - $100K range - you can't just claim what you think they were "worth" to you. everybody looses, except the lawyers in this one.

you don't plug cryo-freezer with a tiny valid temp. range into the standard outlet with no UPS, put up a bunch of signs ("please don't unplug") and no monitoring.

sounds like the samples were worth exactly the same value as the contents of that shared fridge at work, which is also apparently a 20 year science project, and uses the exact same security system.