The scene is straight out of a black comedy: A janitor working in a lab is driven mad by an incessant beep, so they flip a switch that kills the sound – but also shuts off a storage freezer t…
I mean ooops, big error from that cleaner, but I kinda feel like this is on the lab. The lesson here is that it probably shouldn’t be set up so that it is possible to turn off sample freezers so simply. And if sample storage is so critical perhaps they should have distributed storage so one freezer loss isn’t such a disaster. What happens if that site loses power for a significant period of time? Why is there no alert system of temp rises?
In computing, if a data storage company stored all their data in volatile storage with an easy ‘power off’ switch (that makes an annoying sound) and no backups for their data I don’t think any one of their clients would be very impressed with them when the exceedingly inevitable happens and they lost all their clients data.
This is classic academia. I didn’t realize how janky our entire academic lab infrastructure was until I moved into industry. When the freezer started alarming and couldn’t maintain temperature, the contents should have been moved right then. Instead, they decided to leave the freezer running not quite right and keeping juuuust cold enough for stuff to not thaw.
As a quick story, we had to replace our -80 freezer once when I was in grad school and we found samples frozen in the back of it that were literally older than my advisor that have been handed down over the years. These are tubes with cell lines preserved that, as far as I know, are the only preserved samples.
In fact, during grad school, I wanted to study a strain of bacteria that had a series of papers written about it back in the 70’s and 80’s. When I reached out to the authors, they told me that they lost those cultures due to equipment failure decades ago, and that is why they stopped publishing papers about it.
The lab apparently did have a notice on the thing with the date when they came to repair it, and a guide on how to turn the noise off, plus a “this area doesn’t need cleaning” sign. So it’s at least a shared responsibility. Yes, there should’ve been back-ups, but the cleaner had no reason to even be in the room, let alone to mess with breakers.
A sign telling someone not to do something is not a good enough control measure. It should be locked and access restricted. If this is business critical as they claim they could have done >100 things to prevent this.
I work a lot in safety. If you had a safety critical system just ‘guarded’ by a sign and someone got hurt when the sign was ignored you would be extremely liable for the damages. That sign would be no defence in court. In general nobody reads signs, ever. And that’s if indeed they even can read the signs. Was the cleaner literate? did they speak the language? Do they have comprehension of the sign’s instructions?
To give a vaguely topical example, imagine a submarine with a switch that could open the doors even underwater, that just had a sign saying “don’t press button when submerged”. That would be a truly dreadful design. A better design would have actual control measures e.g. the door motors cannot overcome the pressure from the depth of water preventing opening, depth sensors that lock out the control, the button behind a locked switch cover that only trained, competent staff members have the key for etc. A sign is not a control measure, ever.
Oh, I’m absolutely on board with it not being good enough. But I’ve had personal experience with janitorial staff just pulling the plug on the icp-ms at night as the noise was too loud. Customer kept wondering why the vacuum was so bad each morning and up to regular again right before they left, so we stayed longer and watched it happen. You can’t hide every plug, at some point people are meant to be taught what’s acceptable and what isn’t in the environment they work in.
I feel like actually flipping the breaker is a pretty intentional move. Not a “woops, we opened the door when we weren’t meant to”.
In a different version of the article the outlet and plug were locked out. That’s why the janitor flipped the breaker.
While Lakshmi waited for the freezer’s manufacturer to come perform repairs, her team added a safety lock box around the freezer’s outlet and socket. A warning was posted on the freezer, according to the court filing.
“THIS FREEZER IS BEEPING AS IT IS UNDER REPAIR. PLEASE DO NOT MOVE OR UNPLUG IT. NO CLEANING REQUIRED IN THIS AREA. YOU CAN PRESS THE ALARM/TEST MUTE BUTTON FOR 5-10 SECONDS IF YOU WOULD LIKE TO MUTE THE SOUND,” the warning read, according to the suit.
3:2:1 rule for data. Always have a backup for your backup. I am also surprised something didn’t freak out on those refrigerators. Seems like an oversight.
“backing up” biospecimens and other physical samples is not as easy as baking up data. Freeze/thaw often degrades things, and volumes maybe be quite small and not amenable to aliqoting.
I mean ooops, big error from that cleaner, but I kinda feel like this is on the lab. The lesson here is that it probably shouldn’t be set up so that it is possible to turn off sample freezers so simply. And if sample storage is so critical perhaps they should have distributed storage so one freezer loss isn’t such a disaster. What happens if that site loses power for a significant period of time? Why is there no alert system of temp rises?
In computing, if a data storage company stored all their data in volatile storage with an easy ‘power off’ switch (that makes an annoying sound) and no backups for their data I don’t think any one of their clients would be very impressed with them when the exceedingly inevitable happens and they lost all their clients data.
This is classic academia. I didn’t realize how janky our entire academic lab infrastructure was until I moved into industry. When the freezer started alarming and couldn’t maintain temperature, the contents should have been moved right then. Instead, they decided to leave the freezer running not quite right and keeping juuuust cold enough for stuff to not thaw.
As a quick story, we had to replace our -80 freezer once when I was in grad school and we found samples frozen in the back of it that were literally older than my advisor that have been handed down over the years. These are tubes with cell lines preserved that, as far as I know, are the only preserved samples.
In fact, during grad school, I wanted to study a strain of bacteria that had a series of papers written about it back in the 70’s and 80’s. When I reached out to the authors, they told me that they lost those cultures due to equipment failure decades ago, and that is why they stopped publishing papers about it.
The lab apparently did have a notice on the thing with the date when they came to repair it, and a guide on how to turn the noise off, plus a “this area doesn’t need cleaning” sign. So it’s at least a shared responsibility. Yes, there should’ve been back-ups, but the cleaner had no reason to even be in the room, let alone to mess with breakers.
That’s just not good enough.
A sign telling someone not to do something is not a good enough control measure. It should be locked and access restricted. If this is business critical as they claim they could have done >100 things to prevent this.
I work a lot in safety. If you had a safety critical system just ‘guarded’ by a sign and someone got hurt when the sign was ignored you would be extremely liable for the damages. That sign would be no defence in court. In general nobody reads signs, ever. And that’s if indeed they even can read the signs. Was the cleaner literate? did they speak the language? Do they have comprehension of the sign’s instructions?
To give a vaguely topical example, imagine a submarine with a switch that could open the doors even underwater, that just had a sign saying “don’t press button when submerged”. That would be a truly dreadful design. A better design would have actual control measures e.g. the door motors cannot overcome the pressure from the depth of water preventing opening, depth sensors that lock out the control, the button behind a locked switch cover that only trained, competent staff members have the key for etc. A sign is not a control measure, ever.
Oh, I’m absolutely on board with it not being good enough. But I’ve had personal experience with janitorial staff just pulling the plug on the icp-ms at night as the noise was too loud. Customer kept wondering why the vacuum was so bad each morning and up to regular again right before they left, so we stayed longer and watched it happen. You can’t hide every plug, at some point people are meant to be taught what’s acceptable and what isn’t in the environment they work in.
I feel like actually flipping the breaker is a pretty intentional move. Not a “woops, we opened the door when we weren’t meant to”.
In a different version of the article the outlet and plug were locked out. That’s why the janitor flipped the breaker.
https://www.cnn.com/2023/06/27/us/janitor-alarm-freezer-rensselaer-polytechnic-lawsuit-new-york/index.html
3:2:1 rule for data. Always have a backup for your backup. I am also surprised something didn’t freak out on those refrigerators. Seems like an oversight.
“backing up” biospecimens and other physical samples is not as easy as baking up data. Freeze/thaw often degrades things, and volumes maybe be quite small and not amenable to aliqoting.