Hacker News

points by notenchange 17 hours ago | 0 comments

Reading the text it seems to me this is due to a very widespread issue in Spain and other southern European countries: people don’t care about the warnings because there are too many false positives.

That’s why people here, when they hear the fire alarm they just wait for it to go off: the probability that it is actually a fire is minuscule… but the alarm triggers a lot.

Not sure how to actually solve this… but I’m also not making excuses for what happened. High speed trains is not the same as a fire alarm in some office building, the people in charge should be held responsible somehow.

[–] PaulDavisThe1st 11 hours ago

I'm a volunteer firefighter in rural NM. We have a commercial property close to us that repeatedly has false alarms on its fire alarm/smoke detection system, to the point where even the best of us are suffering from "the boy who cried wolf" feelings.

Because of this, I discovered that (apparently) NY Fire Department charges $5000 per false call on an alarm system directly connected to the dispatch system.

Somethng to think about.

[–] jcrawfordor 8 hours ago

ABQ Fire charges $300/call for false alarms after the first three. It steps up to $500 after number ten. There is, of course, also the $25 annual permit fee for a reporting alarm system. Failure to comply with these fees is a misdemeanor. Just in case some local precedent will help advocate :)

Honestly it's remarkable to me to think that anyone ever gets to that $500 "excessive" tier, I've had one false alarm where the fire department attended in years of having an alarm system and I was mortified. felt like the biggest idiot in the world.

[–] brudgers 7 hours ago

$500 is probably cheaper than a Simplex service call.

[–] jonah 10 hours ago

Pretty common in lots of places - you get charged after n false-alarms and are required to fix your broken sensors or whatever is causing the issue.

[–] Maxion 10 hours ago

Sometimes code gets in the way. E.g. in the school I went to growing up they had to have firealarms in all kind of closets and storage rooms. We got at least once per semester a firealarm when some teacher went in to sort stuff due to the dust.

[–] js2 9 hours ago

There's generally two types of smoke alarms each with pros/cons: photoelectrical which are very susceptible to dust and ionizing which generally are not:

https://www.nfpa.org/news-blogs-and-articles/blogs/2023/03/0...

I wouldn't think code dictated the alarm type. (I just replaced all nine sensors in my home due to age and went with all photoelectric this time and it was less than six months before I had my first false alarm. Never had any false alarms with previous ionizing type.)

[–] femto113 9 hours ago

"False" isn't really binary. In my condo building we used to have in-unit smoke detectors wired directly into the building alarm, which meant whenever anyone in any of 200 units burned their toast sirens were blaring. The smoke was real, it just wasn't a threat to the whole building. This would happen once a month or so, and because it's a highrise that sent the big ladder truck down our little street. The fire department got sick of this so we updated a few years ago to have heat detectors that are wired into the building-wide alarm but the smoke detectors only go off locally. False alarms have fallen to basically zero, but the downside is that for reasons not entirely clear to me the new system has to get inspected every 6 months, which includes making sure every one of 200 heat sensors is working, so we get days with 100s of very short bursts of the alarms twice a year.

[–] Animats 8 hours ago

In my aerospace days, our buildings had both ionization and temperature detectors, plus water flow on the sprinklers. Any detection would sound loud buzzers locally, but two sensors, or water flow, had to trigger before everything powered off. That's common in big installations.

We once got an alarm because the heating element in a computer room air conditioner had turned on for the first time ever. It was defective and started to smoke. Smoke detection tripped, buzzers sounded, lights on a panel showed where the problem was, and alarms went off in the security control center. But this didn't trigger building evacuation or full power down. That took confirmation from a second sensor.

Someone ran into that computer room, saw smoke coming from the A/C unit, and pulled the big red power switch, installed by the safety people, on the air conditioner. Then we had time to shut down normally, the safety people came to clear the alarm, and the HVAC people came to fix the problem. I kept the burnt heating element on my desk for a few days for people to see.

This is typical for big industrial plants. Bad stuff will happen, and you have procedures for it.

[–] askvictor 2 hours ago

A tradie doing some welding in the kitchen I was working in some years back told me this as he disabled the smoke alarm (to prevent a false alert while he was welding): "if you set off a false alert on the fire alarm, quickly set fire to the wastepaper bin - it saves you from paying the false alert fee"

[–] childofhedgehog 6 hours ago

I learned this from my last landlord after I kept causing issues for the whole building by cooking bacon. The detector was installed ABOVE THE STOVE. It was cheaper for him to move the detector than to pay for the repeat call outs over a mildly smoky apartment due to my need for crispy almost burnt bacon…

[–] JumpCrisscross 6 hours ago

Consider investing in powerful window fans. If it's setting off a detector it's causing havoc in your lungs.

[–] infecto 10 hours ago

These are elegant solutions imo.

[–] futureproofd 10 hours ago

For example, my condo in Toronto charges $800 per false call.

[–] munk-a 9 hours ago

I'm surprised the strata was able to offload that to you. Most building corps are in charge of maintaining the fire alarm system especially if it's a tower (rather than just a SFH community under a strata) where unit to unit spread needs to be addressed promptly.

[–] Kinrany 10 hours ago

It should probably be X$ per alarm regardless of it being false or not, minus first N per year.

[–] rtkwe 10 hours ago

Why? The goal is to incentivize property owners to fix broken systems not to pay for the actual cost of responding to a real call. That's covered by taxes.

[–] ralegh 9 hours ago

Depending on who is paying and how much, someone might light a small fire to avoid the fine.

[–] mindslight 9 hours ago

Arson when the fire department is already on the way. Great idea! Firefighters hate this one simple trick.

[–] bell-cot 9 hours ago

I'd interpret the "minus first N per year" to mean that you get a few "free" fires per year - but after that your habits filling your closets with oily rags, overloading electrical extension cords, and smoking in bed will start to cost you.

[–] munk-a 9 hours ago

Fire is one of those things that really sucks for you - but also really sucks for your neighbors.

You never want to supply any reason for someone to demur about reporting a fire.

[–] eastbound 8 hours ago

I'm sure someone somewhere will interpret it a 3 free demos of firefighters at home for their kid, with sirens and light, pretty entertaining (It's a reference to the Freakonomics study that, if you make parents pay by the minutes for being late getting their kid, they basically take the fine as a slightly expensive childcare facility).

[–] uniq7 10 hours ago

That will also decrease the number of true positives

[–] comeonbro 8 hours ago

https://en.wikipedia.org/wiki/Alarm_fatigue

> Healthcare: The constant sounds of alarms and noises from blood pressure machines, ventilators and heart monitors

> Amber Alerts: sent to cell phones as far away as [...] 15 hour drive

> "known to the state of California to cause cancer"

etc

https://www.youtube.com/user/USCSB

See also a million USCSB industrial disaster postmortems

[–] prmph 6 hours ago

I've had the thought that: "known to the state of California to cause cancer"

Should probably be: "known to the state of California to be a class N carcinogen"

[–] eastbound 8 hours ago

I've deactivated my (EU equivalent of) Amber alerts the second they sent ONE irrelevant notification. It took me a few tries but fortunately they also tried their system a few times A WEEK during installation.

I prefer to risk my life every day at not knowing what the government thinks is really important for me to know.

[–] 1718627440 17 hours ago

It just means that a false positive rate is as dangerous as a false negative rate. Just issuing a warning doesn't come for free when you are designing a security system.

[–] karmakurtisaani 17 hours ago

This is very well illustrated in the fable Boy who cried false positive.

[–] mercer 14 hours ago

that sounds like a character from Harry Potter and the Methods of Rationality.

[–] thaumasiotes 8 hours ago

Any lessons in HPMOR are also going to be found in traditional works.

I was interested to find a discussion of the trolley problem in a book of out-of-copyright Norwegian-region folk tales (in English translation) that Amazon recommended.

[–] jryle70 6 hours ago

False positive? How did you know?

[–] coopykins 17 hours ago

Something similar is happening in my region after the flood that happened two years ago: The alarm was raised late; Now, even mild rain is triggering an alert.

Guess whats the consequence of that? Less and less people pay attention to the alerts.

[–] ricardobayes 16 hours ago

What you mention makes me frustrated, as it also signals the infrastructure is so underdeveloped even moderate rain grinds daily life to a halt: kids don't go to school if rain is forecast. We need to elevate our professionalism in Spain and not accept African-level infrastructure in the 21st century. It's really sad within one continent we have countries producing AI and photon computing, and others can't even figure out how to put gutters and rain drains.

[–] cbovis 7 hours ago

This is not what it is. The infrastructure isn't THAT bad.

Government officials are terrified of taking the blame for another Valencia style flooding catastrophe so any hint of rain blasts the mobile phone emergency system. Once the alert has been sent the officials can feel safe in the knowledge that they've done something.

[–] jonny_eh 10 hours ago

To your point, a contributing factor here was the front train couldn't start due to either a minor or bogus "security" alarm. Whatever alarm that was, it wasn't worth triggering such a tragic accident.

[–] jameshart 9 hours ago

To be fair the Spanish word for security is the same as the Spanish word for safety: seguridad. It could as easily be a ‘safety’ alarm and a mistranslation.

[–] 1718627440 4 hours ago

I do not Spanish to assert whether it is true, but it is often said about German and there it is not true. I think the issue, that it is often mixed in daily speech, but that's true for English as well.

[–] jonny_eh 7 hours ago

I'd bet this "safety" alarm was minor or bogus, not worth risking lives over.

[–] piinbinary 10 hours ago

Agreed

https://en.wikipedia.org/wiki/Alarm_fatigue

[–] throwitaway222 8 hours ago

The fix is to do things right every time, and if we're losing money to false alarms, spend more money to make sure the false alarms stop happening. Make it a new persons problem rather than just making someone "used to it" and start ignoring it.

[–] sznio 9 hours ago

> Not sure how to actually solve this…

fix the alarms? I only had one false fire alarm in my life. and thankfully zero real ones. it's possible.

do fire drills? if alarms get ignored, and you can't reduce the false-positive rate, then introduce fake alarms and monitor the response?

[–] whycome 9 hours ago

Then what? In this case, Suspend the train engineer? The company itself is incentivized to ignore the warnings to keep things on track.

(...on track)

[–] eastbound 7 hours ago

A really important strike that Spanish pilots should do, is striking by respecting the red lights and refusing oral overrides. Hours of delays the first day, trains at the wrong place in the evening, we'd see exactly which regions are affected, and either we agree to repair things or we just leave trains to rot at the red lights rather than risking 900 people's life again (x2 trains = 1800).

[–] 1718627440 4 hours ago

Ultimately red signals due not fall from the sky. They are controlled by the same people that issue oral or written movement authorities. So focusing on just the oral movement authorities doesn't address the real problem. In the end it is the same signal in a different medium.

[–] pelagicAustral 17 hours ago

> That’s why people here, when they hear the fire alarm they just wait for it to go off

This literally happened to me last night, I was having dinner at my apartment, and suddenly I start hearing a smoke alarm, but not quite inside the apartment, I just kept having dinner thinking "bah, someone probably burned a toast".

Then it wont stop, I walk to door and hear people talking, still don't bother enough, wait a few minutes, then go out and my neighbor said "there are three fire trucks down there, I think it's a real one" hahhaha

The hallway was kind of hazy and smelled like burnt food, for sure... so I just grabbed my laptop and went out hahaa.

[–] D-Coder 7 hours ago

My apartment has both a local fire detector and a building fire alarm.

The building fire alarm is soooo loud and obnoxious, you _want_ to leave the building at speed. (Which I think is the right way to do it.)

[–] avereveard 16 hours ago

Because they don't get held responsible enough for bypissing procedures. Some system are too complex for fully automated operarions, and the human need to play along.

Safety culture is blameless, but only if everyone plays along.

[–] inglor_cz 8 hours ago

"bypissing procedures"

I know that this was likely a typo, but it is also a genuinely great word for the sort of attitude related to these situations.

[–] tintor 10 hours ago

Issue and enforce fines for every false positive alarm, and problem will solve itself.

[–] kazinator 10 hours ago

Or, how about, issue a mandatory repair order and follow-up inspection for every false positive alarm that appears to be due to equipment failure. The follow-up inspection checks that the alarm has not simply been disabled.

[–] gambiting 6 hours ago

The escalator failure in 2018 in Rome that caused severe injuries to 24 people was found to be partially caused by safety systems being disabled and overriden because they caused too many false alarms.

https://en.wikipedia.org/wiki/2018_Rome_escalator_accident

[–] IAmBroom 8 hours ago

In a modern passenger rail control system[0], ignoring the warnings results in automated responses like dumping all the hydraulic pressure, setting all brakes at once. This will have two effects: (1) the wheels will slide on a fixed spot, grinding a flat spot, and (2) the operator will either require central control to remotely release the emergency brake state, OR the operator will have to continue at the safest (lowest) speed.

What I'm saying is: false positives should not be ignorable, and modern systems take that choice away (at least, just before the situation becomes unavoidable in worst-case analyses).

That being said, previous fatal accidents in Spain's high-speed rail have resulted in "primitive" uncontrolled areas between controlled areas. Those SHOULD NOT exist. You are essentially saying "lives don't matter that much here " ($/life).

[0] I work in rail safety.

[–] esafak 10 hours ago

[flagged]