Hacker News new | past | comments | ask | show | jobs | submit login
Nest Protect is a terrible buggy product (plus.google.com)
295 points by somerandomness on Feb 14, 2015 | hide | past | web | favorite | 191 comments

I saw this last night and admit that I've watched the video a couple times. It's a fascinating view into what users are really doing with your product.

<spoilers follow>

And, I'm really impressed with the filmmaking skill exhibited in this short video. The transition between walking with the camera and setting it down to rip the smoke alarm off the ceiling is perfect, and the framing there is great. As the film proceeds, Brad's breathing becomes more apparent and faster, as the tension builds to a climax of automatons screaming "emergency" and "can't be hushed here, can't be hushed here hushed here". I also appreciate the exposure as Brad walks into the garage to find an improvised grave for his machine overlords; completely black, and out of the blackness comes an insulated water cooler, perfectly sized for the smoke alarms. Finally, I like how the anger, tension, and action escalates progressively through the film. It starts off with some walking around, and gradually becomes more violent. The timing is just perfect.

Perhaps not intentional, but it's just so wonderful. This should be submitted to a film festival. It's the most fascinating "home video" I've seen in ages.

(Edit to add one more thing: I think the real genius is the computerized voice, not quite speaking with casual American English rhythm, telling the user that they can't do the exact action they requested by physically pressing a button. Obviously, Stanley Kubrick beat Brad to the punch by a few years, but it still works. And this is real life, not fiction.)

I concur regarding the film. It almost has an atmosphere of a dystopian sci-fi. Can't say whether that reflects poorly or not on the early stages of the IoT.

I think it's an uncanny valley thing. When your computer says "Compiler Error 42", it's acting like a machine, which is fine. When it starts saying "Emergency: this alarm can't be hushed here," then it starts getting human.

I realize the Nest announcer might be a better GlaDOS than GlaDOS. (But not a better HAL than HAL.)

It's exactly like HAL, calmly telling the human that the thing they want to do is forbidden, without explaining why, how to seek an exemption, or what they should do next.

HAL explains why: "This mission is too important for me to allow you to jeopardize it."

"This alarm tone is too important for me to allow you to hush it."

"Our lawyers refused to accept the liability involved in us shipping a product that actually obeys your wishes, because we must assume that you are a moron who will indicate the wrong thing and then sue us."

Speaking of dystopian sci-fi, the red ring and first-person perpective reminded me of the short film "What's in the Box?".


Its honestly the funniest videos I've seen all week.

Interestingly it was telling him the 'master bedroom' was the one that was detecting smoke and yet it was hard to tell if he ever made it to the master bedroom. I've always wondered what a herd of these would do if you loaned one to someone and their house was burning down, you're remaining ones would scream in sympathy or something.

Almost the whole first minute or so of the video is in his master bedroom.

This is straight out of House of Leaves.

Nest did a great job with their thermostat, but the Protect leaves a lot to be desired. I was vacuuming which triggered the sensor. The alarm went off, and it was really loud.

I looked at how to hush it, and couldn't figure out how. My alarm was too high up to safely climb up and press the button -- I had paid for somebody to install it safely before.

So I called their support, and they told me they couldn't legally add a feature to turn it off. Which is a bit bewildering, considering that wave-to-hush had been a launch feature (albeit removed for apparent reliability issues). So I had to dangerously climb up high and remove the alarm and take out the battery.

But the worst thing? It never alerted my phone.

I have my own theories about why this happened. I had recovered my iPhone and not logged back into the Nest app, which I think is required for notifications to start flowing again.

But the support guy thought Nest engineering would be back in touch with me to discuss this crucial flaw within two weeks. Months later, I've not heard back.

Nest had a ton of options after the thermostat. It feels like they put a smaller B team on the smoke detector, despite it being a critically important safety device. It's really bewildering how the Protect turned out this way.

More generally, the lesson is that the Internet of Things is going to be fraught with complications.

Apparently "dirty" power is a major trigger for the wired Nest Protect's false alarms. It's not surprising that your vacuum cleaner was enough to trigger this.

CloudFlare's CEO was ranting about his ~daily false alarms about 6 months ago: "dirty power common on PG&E SF causes small blips in light emitter. Nest interprets as smoke."

More detail here: https://twitter.com/search?q=%40eastdakota%20nest&src=typd

Actually mine just runs on batteries. I think they have an optical sensor which can be triggered by dust. I was told by support that they used to recommend blowing the debris out, but it wasn't good for customer confidence.

When your troubleshooting procedures remind your customers of their (erroneous but widely practiced) methods of making NES cartridges work, you might have a problem.

I don't know about NES, but at least with the N64, I swear by blowing into the cartridge. Works at least 75% of the time, and the other 25% by wiggling the cartridge a bit in the hopes of getting it to reseat better.

It works -- but what you're really doing is fixing an intermittent electrical connection by introducing moisture. Over the long term that's a recipe for corrosion.

We used q-tips with alcohol in my house to avoid this very issue. Cleans the contacts really good with no corrosion issues.

Do they not have an internal power supply to smooth out voltage spikes?

What I don't understand is "dumb" smoke detectors, whether on mains or 9v battery, have been reliable forever. Yet a device that should be not much more than putting a little compute power/wireless connectivity/etc in the same enclosure as a sensor setup with decades of proven reliability somehow fails on the sensor's reliability.

Is there a patent in this space or something that forced them to reinvent the sensor or should this be a poster child for NIH syndrome?

I don't think it's true that regular smoke detectors are inherently reliable. In newer apartment buildings that I've lived in, all of the units are hooked into a central fire alarm system. If a smoke detector goes off in one unit, the alarm goes off in all of them.

In practice it's often been a nightmare. Some places that I've lived, the fire alarm went off multiple times a day, almost every day. Presumably a lot of this is from people cooking or something, but given the frequency I don't think that could explain all of it.

I think in non-networked smoke alarm situations, people are just less aware of the false positives they generate.

And the power supply issue is well solved, too -- see just about any modern phone charger. I just...don't understand.

Conventional smoke detectors use radioactive isotopes (americium). I imagine Nest wanted to avoid putting radioactive materials in their shiny consumer products.

Some do. Many others also use the photoelectric smoke sensor that Protect does, and some use both. The photoelectric type is better at detecting smoldering fires. The ionization (radioactive) type is better for detecting flames. Perhaps with a temperature sensor in the Protect, they thought they were covering their bases.

That's a shame, since the only place a Nest Protect apparently belongs is a nuclear waste storage facility if that's what it takes to get them to shut up after a false alarm.

I might add that never alerting my phone was the worst thing simply because it was the only reason to buy a Protect. A regular smoke detector continues to detect smoke just fine.

This has been a problem with I think literally every IoE type device I've tried. I'm in Australia so I assume the reason is that everyone buys some AWS service in America, and 500ms of latency hits and server load means stuff times out in the end.

Which seems to suggest some bad things about the ability of the designers of these things to build a robust product and forsee obvious problems.

If the smoke detector function is having false alarms, this should be reported to Underwriters Laboratories, the tech arm of the fire insurance industry. Here's the report form.


The NEXT smoke detector has UL approval number UTGT.S25414. That means UL actually tested the thing. But they may not be testing adequately for electrical noise on the power side, because, until now, smoke detectors were simple electrically and insensitive to power problems.

So report problems to UL. They take fire safety device failures very seriously.

I purchased a Nest thermostat and installed it as per the instructions. Set it to 69°F before going to sleep. I awoke at 3AM with the temperature at 83°F and no way to turn it off.

I had to rip it from the wall. It was one of the most frustrating technology experiences I've had in years, and I'll have to see something REALLY compelling if I'm ever to try home automation again.

You might want to check out a thread on thermostat safety I started over on the Spark Community awhile back: https://community.spark.io/t/spark-powered-thermostats-burni... .

It was a fascinating discussion all around with some interesting lessons:

1) No one seems to consider a thermostat as a device which requires safety controls.

2) The built-in safety controls on a consumer HVAC unit tend to be insufficient & kick in well after temperatures in a home can get dangerously hot, if they kick in at all.

3) Because of the custom nature of every HVAC install, with more air going to certain rooms than others, it's not possible to draw any conclusion about the nature of whether a situation with a rogue thermostat can become deadly.

4) It appears a large majority of digital thermostats are built to a "toy grade" standard.

This is likely a topic space that doesn't get enough attention and I hope I opened a few eyes over at Spark. The fact is, such incidents don't happen often so they simply don't get the attention. This is unfortunate, because with the state of modern safety systems and designs an out of control thermostat is something that should never happen.

Nest pretty much says that you and I were doing it wrong.

It worked just fine for heat -- although I later found out that every time it turned on my heat pump it was activating the emergency heat. I tried finding a way to get the unit to be okay with ramping up the temperature over a longer period of time, but that setting was not exposed in their UI. All I wanted was to say "I don't really care what temperature the house is while I'm gone so long as it never goes below 55F or above 80F, but I want you to make sure that when I arrive home at 6PM the temperature has reached the point I set". But...alas, the unit never figured that out. The "learning mode" never actually did anything, no matter what I tried.

Fortunately I was able to get a full refund on the devices from Amazon.

We had the same issue last year (professional install), and spent a lot of time working with Nest troubleshooting two different units. Nest either had a mystery incompatibility with a brand new Goodman HVAC or a dirty electrical current was frying the circuit on the Nest. We had to get a dumb thermostat instead, it works fine. :(

I used to own a house equipped with a heat pump, and in heating mode the cheap-o thermostat would display "e heat". Maybe it's just common to use the same control circuit for that, that might otherwise be used for some sort of actual emergency heating device?

It just said "heat" when the furnace was on.

You have a point -- I don't know for sure, but my extremely high electricity bills (even during two relatively mild months) are my reason for believing that the Nest was activating the emergency heat.

EDIT: corrected that whole last sentence. It made no sense.

We had the same issue and had a pro service call, in our situation the Nest was calling for emergency heat.

I wasn't willing to pay more money for a professional service call. I was hot and angry and frustrated. I de-installed the units and hooked up the old Honeywells.

e-heat is the manual override to use the resistance heater in the heat pump (it also disables the heat pump part of the heat pump). It shouldn't normally be triggered.

I don't know anything about nest, maybe it displays that in other situations.

Did your nest ever properly call for cooling?

No. The unit saw that it was not reaching its setpoint, so it kept trying to reach the setpoint. But there is apparently no feedback from this.

Instead of A/C, the unit activated heating. So the temperature kept rising even further beyond the setpoint. So, more heat.

Positive feedback -- exactly what I didn't want.

I can see both sides of including detection of improper wiring in the software. It's a nice feature when it works right. On the other hand, it's something new to go wrong that is only helpful in one specific case (that you would rather prevent using some other method).

(I'm presuming enough QC in the manufacturing that improper wiring is the most common trigger of positive feedback)

Have you gone far enough into it to compare how the wires are hooked up on the heat pump side to the nest diagram? Not trying to heckle you about it, it's a puzzle, it must be solved.

"it's a puzzle, it must be solved."

I agree with you. I love solving puzzles, that's why I'm an engineer. But I had to balance solving this puzzle with the ongoing puzzle of maintaining domestic harmony. My then-pregnant-wife insisted I get the system working properly ASAP. The best way to do that was to reinstall the Honeywell -- the system worked properly within 15 minutes. As life circumstances dictated that I was away from home for most of the time -- and we moved a couple of months later, the best solution was to get a refund.

Did I have some wiring wrong? Most likely.

Not to beat my credentials against my chest, or to imply that they mean that I could not have possibly made an error... I have engineering degrees and have done a lot of instrumentation system design work. Figuring out how to wire up strange sensors to an existing system is something I do all the time. If I had this kind of difficulty, what kinds of problems are ordinary people having? Or maybe it just reinforces the aircraft industry joke that "you shouldn't let engineers touch tools, they'll break something".

Other people are posting about problems with professional installs on heat pump systems. The two more likely explanations of that are faulty nests and wiring problems. If it's wiring problems, it suggests there may be something that is easy to overlook or misunderstand.

I had exactly the same experience with my Nest thermostat. I think Nest's assertions about furnace compatibility are a marketing wash. The three support people I spoke to were explicitly disinterested in knowing the make and model of my furnace.

I appreciate the company's consumer electronics mindset, but they need to take the business of running HVAC systems more seriously.

My "dumb" thermostat is back on the wall and successfully running my heating system (to the schedule I've programmed it) as it's done for years.

Most furnaces, perhaps all, in the Canada and the US use an 18V control circuit. Close the circuit, the furnace is on.

In the UK, the control circuit is on the mains; close the circuit, the furnace is on.

I have such a love/hate relationship relationship with the Nest Thermostats. It took two years of tweaking, but I think I finally got them set.

When I first got them, they worked great in the sumemrtime. No complaints. But they were bloody awful for me in the winter:

* Nest would, randomly, refuse to turn on AUX heat. I woke up one morning and my house was in the upper 50s (the set temperature was 68) and Nest was just sitting there trying to run the heat pump instead of turning on AUX when it wasn't keeping up. At first I thought there was something wrong with my HVAC system, but I've had two different people look my system over, on three different occasions, and found everything working properly. Nest was just refusing to turn on AUX.

The first time it happened it was in the single digits outside and Nest let my house get into the upper 50s (temperature was set at 68, and it was still trying to run the heat pump even though it was in the single freaking digits outside). I finally took Nest off the wall and manually shorted the aux heat on - I didn't want the house to get any colder while we "troubleshooted" it, and Nest was just refusing, no matter how I adjusted it, to turn the aux heat on and warm the house back up.

* Other times, it would use AUX when it doesn't need it. It was 37 outside once, and nest was burning AUX like crazy when the heat pump was having no problems keeping up. I had a nearly $400 power bill last January. Now, in fairness, it was very cold that month, but Nest using AUX when it shouldn't was a big contributor.

* For instance, you can't directly set the AUX lockout any lower than 35 degrees (more on this in a minute) when, for my house, it probably needs to be around 26. You can also only set certain things from the thermostat itself.

* They are very difficult to troubleshoot. How do you manually turn on AUX heat to test if it's even working at all? As far as I can tell, you can't! The only way we were able to test it is to take the thing off the wall and manually short the wires together (and, of course, it worked fine).

This is not some weird setup either. It's a standard heat pump with electric heat strips for aux heat, that's only about 4 years old. Where I live, this is the absolute most common form of heating. Because I can take the thermostat off the wall and short the wires together to turn aux on (and because I've had two different HVAC contractors check my system), I'm fairly certain the problem is not my HVAC system.

The biggest problem was, after three times of waking up to a cold house, I had a very difficult time trust them. I had an infant in the house, and for something that I'm not supposed to have to think about, I was spending an inordinate amount of time worrying about.

So at one point I replaced one of them with a Honeywell (not the Lyric, one of the touchscreen models). And it worked fine for the rest of winter and was much less crazy to deal with, but once summer rolled around the Honeywell wouldn't turn my air conditioning on! In troubleshooting it, we found that the Honeywell was defective and wasn't putting enough voltage to activate the reversing valve. So the heat pump was still running.

So I ended up putting the Nest back on for the summer, but spent a lot of time reading the Nest forums. Apparently problems with AUX heat are SUPER common with Nests (just Google "nest aux heat" to see tons of complaints).

There's a hack that that you can go through (detailed here: https://community.nest.com/ideas/1144#comment-7150) that allows you to set the lockout to a lower temperature. Which I did eventually set to 26. And, sidenote, I really hope they don't close this hole, because a lot of people are relying on it.

So far this winter, no problems. The hack seems to be holding. I haven't woken up to a cold house yet, and Nest is only using AUX when it should. And my power bills are running 25% less than the year before thanks to Nest not constantly using AUX heat.

I'm still very wary of them. If I wake up to a cold house again, they are going right in the trash can.

> They are very difficult to troubleshoot. How do you manually turn on AUX heat to test if it's even working at all? As far as I can tell, you can't! The only way we were able to test it is to take the thing off the wall and manually short the wires together (and, of course, it worked fine).

I don't own a Nest thermostat myself, but was browsing their blog out of curiosity after reading this story, and it sounds like the latest software update includes a System Test mode:


I'm wondering how much testing Nest did with the photoelectrics in the smoke detector. A teardown [1] seems to show a rather simple design for the led and photodiode beam path, as well as both of those components having places for light to potentially leak in around the diode bases. Other detectors [2], [3], seem to have much more complicated beam paths, with baffles and such, as well as better shielding to ensure the sensitive components are well-protected. Given this, I can see how all the differences in tolerances can lead to gaps which lead to lots of false alarms.

[1] https://www.ifixit.com/Teardown/Nest+Protect+Teardown/20057 [2] http://www.accuratebuilding.com/images/publications/educatio... [3] http://st-nso-us.resource.bosch.com/media/us_product_test/05...

> A teardown seems to show a rather simple design

Don't you usually want lifesaving devices to have simple designs? For example, scuba breathing regulators are famously reliable in part because there's very little to them. Same goes for the AK-47, though I suppose I'm getting off the "lifesaving" track now...

I'd argue overly simple. The nest, and the other detectors, both have baffles in them to block out light from the outside world. However, the Nest negates some of the benefits from this by having the LED shine against a piece of plastic at an angle fairly close to the photodetector. If that piece of plastic happens to be especially reflective, or if the photodiode and/or LED happen to be misaligned, it would be more likely to trigger false alarms. The baffles in the other detectors, on the other hand, are also used to ensure that the light from the LED gets nowhere near the photodiode unless light's being scattered by smoke. The Nest design works great in the lab, but when assembled en mass, hits a bunch of tolerance issues that leads to a more jittery product.

That's a reasonable possibility. I'm looking forward to seeing a teardown of the next release.

I don't want lifesaving devices to have simple designs, I want them to have designs sufficient and appropriate for the task at hand. "Reduced manufacturing cost by $0.25/unit" in a triple-digit-cost smoke detector is not something that I as a consumer care about.

My point is that sometimes, simplicity isn't about reducing manufacutring costs; it's about creating a better product.

There is a difference between "being as simple as required to reliably function" and "being as simple as possible".

False activations of things like alarms are dangerous because it conditions people to ignore the alarm or react slowly.

But it isn't simple?

Look at that circuit board for gods sake, that's probably 6+ layers and apparently required custom gaskets around crucial components. If you consider the device as a whole, it is not simple at all.

That's kind of the crucial point, too. A smoke alarm can only be considered simple when everything on the path from smoke detector to alarm is simple, and that's not the case here. Which makes this seem like a case of not getting the basics right.

I think he meant naive. We all want the simplest design possible that is correct, fault tolerant, etc etc.

I continually come back to the thought that typical consumer software practices - be it games, apps, cloud services, etc - where you iterate quickly and don't always work out the bugs or design scales well to things like Nest protect, etc. For me, anything that ends up in a consumer home device should be solid from the time it is purchased and not require software updates to work right.

This leads to another concern of mine -- are places like fb, apple, Google, etc. capable of creating stable v1 products for the home.

With Nest, the initial thermostat seemed to work well. Original Dropcam works well. There have been questions since the acquisition of both companies.

I agree ... I was an embedded engineer for many years and we knew that once we released a product to production, it would kill the company if 100k of them in the field had to be replaced.

One thing I've noticed with embedded code bases, some can be quite elegant and tight, have a level of fail safe/resiliency, etc. Some, reboot the device on error is your friend.

It is always hard after spending time in larger systems going back to and doing some code archaeology to see if something that seems off is the right thing or just what it seems like -- needs cleaning up.

"If Nest Protect detects emergency levels of smoke, it will sound an alarm that can’t be hushed, by regulation."


(Yes, I know the smoke detection was a malfunction, but occasional malfunctions like that are a fact of life. My ire is directed at regulators forbidding users from silencing their own smoke detectors without physically incapacitating them -- what were they thinking!?)

Or if you have it installed in your master bedroom, with an en-suite bathroom and a steamy shower. Nothing like having all your smoke alarms go off when you're in the shower! We contacted Nest. Their basic response: "Move the Nest Protect further away from the shower." Or, take the stupid thing off the ceiling and switch back to the previous smoke detector, which would detect smoke but was not fooled by water vapor.

Nest had marketed the device as a smoke alarm that you could silence -- because it would have a period where it would warn you that something was wrong before it triggered the alarm, and it would be possible to manually silence it -- but in my direct experience this period never happened. The Protects always went straight from "normal" to "ALARM" without any warning period. Of course the Protect unit in my kitchen never actually detected kitchen smoke, even when I once burned some food quite severely.

I also had a negative experience with the Nest thermostat. I'm still not sure if I had it wired wrong -- I don't believe so, because I had previously installed cheap-o Honeywell thermostats that worked flawlessly -- but one late spring night I turned on the air conditioner to lower the temperature a few degrees, and then went to bed.

I woke up a few hours later, drenched in sweat. Somehow the unit had commanded "EMER HEAT" on my heat pump instead of air conditioning. My bedroom was 91 degrees Fahrenheit! The target temp was 70F, which the unit was never reaching. But it wasn't smart enough to realize "Hey, my actions are just making the temperature INCREASE, I should stop and flag an error." Nest's response: "that's funny, it shouldn't have done that". I took the things off the wall the next day.

I am now very wary of "smart home" technology. I can't imagine the experience that a non-technically inclined user has...

There's a reason that smoke detectors operate the way they do. That reason is that people die when they don't.

The issue I have with the Nest smoke detector is that it is sold as a luxury consumer product. That puts it in the class of "the customer is always right" sales. The Venn diagram of that and things that are designed to save people's lives regardless of their tendency to statistically evaluate the risks and rewards of rare but deadly events inaccurately is disjoint.

If Nest did not comply with the regulations, they could not market their product as a smoke alarm.

Actually, people die when they incapacitate their smoke detectors because of false alarms that can't be silenced any other way.

Citation: http://www.nfpa.org/research/reports-and-statistics/fire-saf... which says that 60% of people who died in a housefire were in a building that had no operating smoke alarms. A few more quotes:

"Surprisingly, the death rate was much higher in fires in which a smoke alarm was present but did not operate than it was in home fires with no smoke alarms at all."

"Smoke alarm failures usually result from missing, disconnected, or dead batteries. People are most likely to remove or disconnect batteries because of nuisance activations."

> people die when they incapacitate their smoke detectors because of false alarms that can't be silenced any other way.

Not sure nest would win if it was easier to disable. Have you ever had a smoke alarm that goes off every couple hours? You can't sleep. Doesn't matter if you can turn it off with a wave.

And let's say you reach the point where you instinctively wave so it doesn't bother your sleep, now what happens when there's a fire?

False alarms are the problem. From the video and reports, it appears the Nest Protect has real problems with this.

Pulled the battery from my kitchen fire alarm after it went off three nights in a row while making dinner. Nothing was smoking or even boiling, much less actually burning. It's an electric stove to boot.

You are contradicting yourself, if the alarm was triggered when making dinner then it must have generated a smoke, how would the detector even know you are making dinner.

With the nest, the alarm turned by itself for no reason, you don't cook in bedroom, he even said that he got notification when he was at work.

Two different reasons, your situation was that smoke alarm was too sensitive, perhaps you placed it close to the oven, here the alarm fires for no reason, and it is not the kitchen one, but bedroom one where normally you don't have smoke.

The alarm was triggering during a non-alarm situation.

That is, in fact, a false alarm.

The behavior described by cowsandmilk is highly predictable: provide false alarms, and people will almost always disable or bypass the alarm.

Yes. Incapacitated smoke alarms are a contributing factor in fire deaths. Hence smoke alarms are deliberately difficult to incapacitate.

Be aware that battery powered smoke detectors are a retrofit, and do not meet building code requirements. They are better than nothing, but failure due to disconnected or missing batteries is already in the realm of worse than the worst thing that can legally be built today.

You're conflating "incapacitate" with "temporarily silence".

They're two separate concepts. We should make it really easy to temporarily silence a smoke alarm (for, say, 15m or 4h), so that people almost never have to fall back on incapacitation.

And mandating wired ones does not address the problem. People still incapacitate them, by unplugging them. More generally, people will always find a way to silence a smoke alarm, and the regulators' goal should be to ensure that they do it in a temporary way.

Is there a link to the regulation that requires this? I was under the impression that most states require a hush/silence button on new smoke alarms. False alarms are inevitable and a hush button reduces the temptation to permanently disable the device.

Underwriters Labs make you pay to see that.

I can show you the table of contents, though: http://ulstandardsinfonet.ul.com/tocs/tocs.asp?doc=s&fn=0217...

But... every smoke detector I've bought in the last decade has had silence button. So how it is against any regulation?

According to a quick google, all smoke detectors in Maryland are required to have a hush function by 2018. So... pretty sure that isn't prohibited by regulation there. FEMA says the best smoke detectors have a temporary silence button. I guess maybe there are some states that prohibit it? Seems awfully foolish though: temporary hush buttons save lives.

If the device detects a sufficiently-high level of smoke, the button is ignored.

Yes, this is infuriating. It happens on my networked system of smoke alarms. It ignores the hush for nearly a minute while the smoke clears out. Ridiculous. Makes me want to just disconnect all of them.

Again with this statement, accompanied by absolutely no citation.

I recently purchased a non-Nest smoke alarm. The documentation explained that if there was an alarm, the hush button would silence it temporarily, but if there was a high level of smoke, the hush button would not work and the alarm would not be able to be silenced.

Since both Nest and other smoke alarm companies say this, I conclude there is some regulation to this effect.

The Nest device in the video did not appear to be experiencing high levels of smoke.

I posted a link to the citation in the root comment of this thread.

Since I can't actually see the document, does it say that the fire alarm can't be muted even in heavy smoke? Or are you just guessing that it might? I would find that surprising.

I don't know. UL doesn't make their rules openly-available. You have to pay them to read them.

But Nest's website says UL's rules forbid the silencing, and I see no reason to think Nest would lie about it.

I found the relevant citation: https://news.ycombinator.com/item?id=9064731

Sorry; I meant a citation to the actual law people are referring to, the one that says smoke detectors shall not be capable of being manually silenced in the presence of smoke.

Smoke detector user manuals might vaguely allude to such a law to justify crappy engineering, but that's not the same as a citation to specific Federal, state, or municipal codes, or even an insurance underwriting code.

The actual law is "Do what UL says". Or, to use a direct quote:

"The U.S. model codes and installation standards require smoke alarms and smoke detectors to be listed in accordance with the Standard for Single- and Multiple-Station Smoke Alarms, UL 217, and the Standard for Smoke Detectors for Fire Alarm Signaling Systems, UL 268."


I'm missing the part that justifies what we see in that video. He's unable to mute the detectors, but there seems to be a distinct lack of smoke.

Why are people carrying water for this idiocy? Nothing in the document you linked has anything to do with the subject at hand. Fine, whatever, you're not supposed to be able to mute the detector in the presence of heavy smoke. But there is absolutely no prohibition against a mute feature in general.

I found a citation that I think even you will accept:


True, and the 4% business was also mentioned in the summary document earlier. That doesn't change my two underlying points: (1) it's an incredibly stupid regulation that seems to have been designed to cause people to remove the battery or otherwise disable their smoke detectors permanently to avoid nuisance alarms; and (2) there was no smoke at any level in this scenario.

Poor engineering on Nest's part led to a situation where a poorly-thought-out regulation came into play. Neither condition is defensible... yet look at all of the people in this thread trying to defend one or both.

(Yes, I'm moving the goalposts to a certain extent, as I was mistaken about there being no law against silencing a smoke detector in the general case.)

Worse, municipalities are starting to pass stricter code, requiring the battery to be unremovable:


UL is not a lawmaking body, so if they have regulations against being able to hush a smoke detector -- and I highly doubt that -- Nest is technically under no obligation to obey them.

Here's the gotcha though...

A smoke detector company sells a safety product, if that safety product fails they could easily be sued for the lifetime earnings of a dead individual (or even family in some cases).

So that requires that these companies have insurance, and in order for insurance to cover them they must meet certain insurance dictated requirements, which can be anything, but UL specifications are an easy thing to point at (to both protect the company from liability and insurance from payouts).

So while in some locations there might be no government regs, insurance is making regulations all on their own, and these companies can either play along OR risk paying out up to millions of they get sued.

In effect, they are. Many regulations reference their standards, require devices meet a specific UL standard, or recommend that devices be stamped with a UL approval.

My wife has silenced three smoke detectors by physically smashing them from the ceiling with a broom handle. It's a bad idea to persistently disturb her sleep.

That's what happens when you can't easily silence a smoke detector. Needless to say, after this treatment they didn't work so well - the only good thing about such a design is that it sells more smoke detectors. For a long time we simply didn't have any that worked though. We do now, but none of our replacement ones are linked together anymore - too much of a nuisance to replace the wiring she ripped out while silencing them.

Luxury consumer product? Is that way of saying, you must be smart if you bought our product? It does seem to have some pretentiousness built into the marketing.

I prefer KISS when it comes to my life and smoke alarms are not something I am going to fool with. With regards to their thermostats, I rarely am in the rooms my house has thermostats but programmables were cheap and easy to setup.

> Luxury consumer product? Is that way of saying, you must be smart if you bought our product? It does seem to have some pretentiousness built into the marketing.

It says no such thing. You're just putting your own biases on it. Luxury means: "the state of great comfort and extravagant living."

Nest products are premium products. They cost more and you're meant to get a better quality of product out the other end.

Nobody thinks they're "smarter" because they buy luxury, although they might try to justify their purchase on the grounds of its USP.

It can be hushed if you stuff it into a cooler, apparently.

I was curious about this. IANAL (or general contractor) but while it was easy to find lots of regulations about required installation of smoke detectors, I couldn't find any restrictions about silencing them. In particular, I turned out nothing in either California or San Francisco fire code.

Of course, Nest is selling to a national market, so I don't doubt that it's required somewhere in the US, but it sure doesn't seem to be that common.

> (Yes, I know the smoke detection was a malfunction, but occasional malfunctions like that are a fact of life. My ire is directed at regulators forbidding users from silencing their own smoke detectors without physically incapacitating them -- what were they thinking!?)

Well for most it just involves pulling the power cord or yanking the battery.

Right, and then it's far less likely to be there when you need it in an actual fire.

has anyone found where this regulation is? I have searched a few states and cannot find this restriction. I have found mounting and wiring restrictions but none on silencing.

They were thinking about protecting human life.

For example, people with cheapskate landlords who would be happy to leave a nonfunctional smoke detector that was muted installed rather than spend $20.

I'm not saying you should be able to permanently mute them; just for, say, 15 minutes per button press.

Also, cheapskate landlords would just take the battery out.

In case anyone isn't aware, Brad Fitzpatrick isn't just a google employee he also invented and created memcached.

And Go's net/http package.

And LiveJournal.

And Gearman.

And OpenID.

And protobuf.

No, wait, that was someone else.


Now now, I didn't actually invent it, just rewrote and open sourced it... :)

"... didn't actually invent it ..."

Kenton, agreed but close enough. Implementation in code has more value than the specification.

As per twitter discussion problem seems to be with insufficiently filtered input power: https://twitter.com/eastdakota/status/566075097568391168

I now believe that conclusion to be false, after talking to more people.

Can you share what you found?

I'll be sharing more but not yet. Still gathering information.

Okay, fair enough. Thank you.

I don't think it's just Nests, I bought a First Alert smoke alarm a year ago and it just recently started giving me random false alarms. Battery is fine but something is setting it off (not fire or smoke, and yes it has both atomic and photoelectric sensors). I went searching for a good smoke alarm without lots of false alarm complaints and unfortunately every smoke alarm on Amazon has complaints about false alarms.

I think in recent years manufacturers have become paranoid about missing an alarm and being sued so they probably default to fire the alarm if there's absolutely any potential reason, real or imagined, that a fire might be occuring.

Usually the culprit for lots of false alarms is an ionizing in the kitchen and poor placement.

No these were false alarms at completely random times with no cooking, dusting, or other activity going on. A couple were late at night well after 2am and were pretty nerve rattling.

One experience like this would be enough to put me off of home automation for about 10 years. I think its unfortunate that Nest chose a target so fraught with safety implications for their second device. A lot could be done to ameliorate problems like this if the consequences of a false negative were not so catastrophic.

I run five of them, three powered and two battery. One of them "falsed" twice in one week about three months after installation, and not at all in the four months since. So I think it wasn't false, I think there was something triggering it. I've had zero other issues.

Note these were all purchased before Nest turned off the waving, but have been updated with e firmware that ignores the wave.

I also run nine of the original batch of thermostats with no issues, and any minor complaints (e.g. turning off heat when in direct sunlight) have all been addressed over the years of updates.

I was a huge fan of the IDEA of the Nest Protect. The reality? Not so big a fan. I think they oversold it. It's still better than my old smoke detector in that the pre-alarm warning is fairly gentle, but I'm still getting on a chair to tap the "shut up" button. Wave-to-turn-off never worked for me even before they disabled it.

My brother bought one, while I was still in the honeymoon period of mine. His false alarmed constantly, and has since been removed.

There's a general pattern of trying to dumb things down to a simple, inflexible interface, too often with half-baked "AI" bolted on. Good products still need attentive human supervision.

I installed Insteon motion detectors and webcams after a robbery, but the included software was such undependable and inflexible garbage that I replaced it all with a simple Misterhouse-based Perl script that sends texts via email.

If I see another tech product ad aimed at millenials featuring bright easter colors and indiepop music pitched by Steve Jobs wannabes who are unable to get any angry nerds to make their products actually work, I might snap.

By the way, the other HN article about dumb smart homes is a good accompaniment to this. ( https://news.ycombinator.com/item?id=9043524 )

You don't get a second chance with a product like this.

The problem is if you have too many false alarms people will rip the smoke detector down and disable it. There's a fine balance.

I assume he means the product doesn't get a second chance to make a consumer impression because any sane person will abandons the product after a mishap like this.

I suspect 300bps means Nest doesn't get a second chance to produce a good smoke detector--not that the detector shouldn't be failsafe.

My guess is Portal 3 will have these everywhere, with GLaDOS repeatedly uttering "can't be hushed here".

I was looking for a portal reference. Definitely reminds me of glados & stupid turrets.

I wonder if one if his neighbors is messing with his nest. Maybe someone has figured out a way to set off nest alarms.

It seems like the majority of people who had problems with the Nest Thermostat in these comments had it hooked up to a heat pump system.

Mine is a first-gen, bought the instant they were available to the public, and it's been rock-solid for the past 3+ years. However, I only have a standard central air and central heat (gas furnace) setup here in Texas. Nothing complicated.

I've been debating the Protects myself, the battery operated ones.

I've had a couple of battery-operated Nest Protects in the house for the past year or so - no problems or false alarms so far (thankfully no accidental fires either).

"Can't be hushed here"

I've never used Nest Protect so I was curious about this. Why is it location-dependent? The Nest site has some info[0]:

Hushing works differently during an emergency. If Nest Protect detects emergency levels of smoke, it will sound an alarm that can’t be hushed, by regulation. However, you can hush the other Nest Protects in the house by pressing the button on the Nest Protect that detected the problem. So, when you press the Nest button on the Nest Protect in the bedroom, it will continue to alarm, but the Nest Protect in the living room will quiet down. We designed Nest Protect to hush this way so you can more easily communicate with other people in the house during an emergency while still being able to locate and address the source of the problem.

0: https://nest.com/support/article/What-types-of-alarms-can-I-...

The idea is supposed to be that if you accidentally told your basement smoke alarm that it was in the bedroom, you'd wake up hearing "Fire in the BEDROOM" and then you'd look around and see there's no fire in the bedroom and push the hush button and go back to sleep and meanwhile your basement is on fire.

So networked smoke alarms require you to visit the one that's detecting the fire and disable it locally.

Now, why they would ever make the regulations such that you can't temporarily silence the alarm, even when you're in direct physical contact with it, is beyond me.

And despite me silencing the master bedroom one, the other four didn't shut off. That's my major complaint.

There multiple persons under the same roof, sleeping at night. The alarm in the basement sounds. You go in there. There is a thick smoke. You disable the alarm. You inhale too much smoke and fall down. The house burns down and everyone dies.

Of course, why on earth would you disable the alarm if indeed there is a ton of smoke is beyond me. But maybe there was a real-life case of somebody confused that did just that, and in the emotional aftermath regulation was passed to make sure this never happens again, common sense be damned.

Wow, that's a lot of alarms! I've got a 3000 sqft. house and it's only got 3. I guess you can never be too safe.

The current California fire code requires them all over the goddamn place.

The video looked like a preview for a new game where you run around Hogwarts collecting smoke alarms in a bucket.

Do you get extra points for finding the golden NEST?

In other news, talented hacker Brad Fitzpatrick now available for hire.

Sorry, Google doesn't work that way.

You're probably thinking of Apple.

Not sure why you were downvoted as opposed to OP but this thread is full of people who don't realize that Google is truly a place where employees are encouraged to speak up when things are bad. Frankly, organizations would never be able to fix their mistakes if their employees are censured when they speak up.

(Googler here)

You're not wrong, but I think you overstate the case. I wouldn't say employees are encouraged to publicly trash the company's products. Not at all. But the company does respect employees' right to speak their mind in public, and it does encourage thoughtful internal dissent.

I often tread pretty close to the line on what I say in public, and have even been reined in by Google legal counsel in a couple of cases. I found the experience of being told to cool it to be surprisingly affirming and liberating, and a powerful confirmation of the true commitment to openness in Google culture, because of the reasons for which it was done and the way in which it was done. Specifically, in both cases I really had crossed a line which could be potentially troublesome for Google in court, and in both cases the attorney who contacted me was respectful of my opinions and my rights to speak them to the point of being very apologetic about telling me to shut up. It was very clear to me that Google really didn't want to silence me, and did it only because they truly had to. I think that's awesome.

Based on my experience, I have zero concern for Brad's job, and wouldn't be surprised if he gets some mild and unofficial kudos.

Also Googler. If anything, this should reinforce the fact that G is far more tolerant of Googlers' expression than most companies would be.

I have friends at Apple (and formerly at Apple) who can attest that not all companies are that way.

Here is my personal experience with Nest Products overall:

- Own 3 thermostats: 1 at home and 2 at my wife Business

- 6 Nest Protect spread in different part of our home (aka master bedroom, hallway, kitchen,...)

The thermostats work fine and usually the only issues I have encountered were with Comcast or AT&T (wife Business) going down and the thermostat not every time able to reconnect automatically... Not a huge deal and overall happy with those. And checking our bills, over time I see some savings in electricity cost... And the big big plus is really being able to trigger the A/C during the summer hot days when our dogs are staying home, and we are not with them.

The Protect overall have been working as expected with the one in the kitchen kicking in often when we are cooking (previous smoke detectors in the kitchen have been doing the same so no surprise). At installation time one of the unit couldn't register, and after reaching out to Customer support got a replacement unit quickly which worked fine. 3 weeks ago out of the blue the Nest Protect in our master bed room started to do exactly the same thing as shown in the video but the alarm after a minute will end and the app was reporting the level of smoke going down... Except there was never any smoke at all... Since it did happen 3 days in a row and on the third day, I was there to witness it, I tried a few things without luck... Ended up removing the battery and again contacted a Nest support (glad that I know a few folks there so I didn't have to go to any phone diagnosis), and 2 days later got a replacement unit. Installed since then and problem didn't happen and the bad unit was returned.

My conclusion: the thermostat is almost rock solid except for flaky internet and doesn't always reconnect. The Protect are definitely more a v1 with still some glitch but still happy customer. I am not a huge fan that Google did acquire Nest and still have preferred Apple but I don't have too many complaints.

I do have too a bunch of Dropcam and again not really happy that they ended under Google control, but their camera are good.

Your thermostat should not require internet. Fact.

It is only required when I used it remotely... Manual works even with internet down

I can not believe someone went out of their way to put a spoken "Can not be hushed" response into those things.

Just what you need in a panic situation with a fire in your house; your smoke alarm playing the smart-ass, at 85 decibels.

Like others have pointed out, it cannot be hushed by law. So what is it supposed to do? Simply not reacting to the button press seems even worse to me.

I have multiple battery nest protects. I replaced "False alarm" alarms with them when my smoke detector and carbon monoxide detector had multiple 100% false alarms. The protects have not been activated by anything so far. The humidity that would set off my smoke detector before does not set off my protect. I've had no problems with my protects.

I have a Nest thermostat and s Nest Protect. I'm s big fan of both. The Nest Protect has not been buggy for me. It has done the one thing I bought it for--not give us false alarms from the shower down the hall. Out previous smoke alarm would go off all the time from steam, but the Nest Protect hasn't. As the video shows, maybe we are just lucky!

That's some harsh language directed at his coworkers...

They can take it. Or they can build better products.

I don't put on kid gloves when a product is this bad.

I'm nicer when it's a beta or unreleased. But I paid $500 for these, for a final product. This is unacceptable.

I'm going to buy a condo here in the next year, I plan on rolling my own automation with Arduinos and Raspberry Pis. It's the only way to do it right.

You are going to build your own smoke and fire detector and "do it right"?

Apparently Nest can't, so why not try?

If there's a fire, your insurance company will likely be able to escape issuing a payout if they discover you knowingly brought about a fire code violation.

I would actually do this, but use Home Depot smoke alarms that are up to code as well. The custom smoke alarm would only be for the sake of alerting me remotely or something.

If you buy a Dropcam, it'll send you "loud noise detected" alerts when the fire alarm goes off.

Imminent danger of death? I want a Nest (that works, though this video is anecdotal and seems staged-ish) as much as the next person, but I have to silence a standard smoke detector once every never. It's inconvenient when I do, but less inconvenient than waking up to find my house is on fire or not waking up at all because I died in my sleep from carbon monoxide or smoke inhalation. Even if I trust my own build and code, do I trust the build/code quality of an arduino and whatever half assed, 3rd party smoke detecting shield I can find over a device designed and built by people with the life saving responsibility in mind? I'd still trust a Nest over any arduino shield.

Being alive >>>>> rare inconvenience.

What if the home contained both normal per-regulations smoke detectors that alerted humans, and had other smoke detectors to alert your home automation system? We could call them sensors instead, to differentiate.

Ideally, the sensors for your home automation system would be configurable to your hearts content, their point is not to alert humans but to alert your home automation systems. The actual smoke detectors are for alerting humans, and are tightly regulated.

I see a future where the home contains many different types of sensors all over the house intended for data collection, and alarms for humans as well. Temperature sensors and heat sensitive cameras combined with smoke detection sensors would let you do some really interesting things with regards to detecting smoke, flame, or even overloaded extension cables.

One of the reasons I'm interested in my home itself being able to detect events such as fire or flood, is so that it could initiate an emergency offsite backup of any sensitive recently-modified files, and as soon as that completed or failed for whatever reason it can shred the encryption key used to boot the full disk encrypted partitions. It could also unlock doors, trigger a gas and or electrical mains cut-off switch, trigger camera feeds to record to offsite servers, call me or anyone else, or even draw attention to the home by turning on and off lights (or changing their color if you have the right bulbs).

Personal risk aside, have fun trying to get your insurance money if something goes wrong, even if it's not the fault of your home-built fire detectors or Raspberry Pi thermostat or whatever. I'd be surprised if that didn't cause problems, should you have to make a claim they find out about your tinkering.

I'd be inclined to keep the minimum required complement of cheap detectors around, for just that reason.

I didn't even think of that... good point. Is this an issue with nest?

The notion that this is an either-or situation, where a device must necessarily be prone to either false alarms or failure to alert during a true emergency, makes me hope that you (and most of the other posters in this thread) don't actually work for companies that make these things.

I don't... and have tremendous respect for people who build software and systems that are "mission critical."

They're in the same situation as civil engineers and the like. They build a bridge that falls down, they go to jail or are financially ruined. We make software that falls over and people pay us for updates.

Where did you get "smoke and fire detector" from "rolling my own automation"? But I can just buy an off-the-shelf detector that actually gets its primary purpose right and hook into that.

> Where did you get "smoke and fire detector" from "rolling my own automation"?

Most likely from the fact that you made that comment in a thread about a smoke detector...

Apparently HN takes a very dim view on people talking about rolling their own home automation - but as someone who's done it, I say: go for it! It's fun and as a hacker you'll have an easy time making something not achievable with consumer solutions.

Within the context of the article, I think false positive in smoke detectors are always inevitable, and by law they can't be silenced. But what a smart home system could do is allow you to intercede and prevent the defective one from cascade-infecting every other device.

Apparently HN takes a very dim view on people talking about rolling their own home automation

There used to be a very different crowd around here. I've been looking around for a site like what HN used to be, when "H" stood for "Hacker" and not "Helpless," but so far no good suggestions have come to light.

"unhushable pieces of crap" I won't be happy to make a check to an employee that speak of the company work this way.

> In case anyone isn't aware, Brad Fitzpatrick isn't just a google employee he also invented and created memcached.

> And Go's net/http package.

> And LiveJournal.

> And Gearman.

> And OpenID.

Working at Google and having done these things, I don't think he has to worry about it. They weren't comp products either, since he spent his own money on them.

Perhaps it tells you something about how strongly the employee feels about the company's work that he is willing to say something like this in public. For such a big problem, firing the employee won't make it go away.

The (sadly uncommon, but proper) fix for this isn't to silence the employee, but rather to fix the piece of crap you are selling so that it isn't a piece of crap anymore.

Hahaha, oh my god the last 2 minutes of that video is fucking hilarious. Looks like a mega64 skit.

if meme, kill it with fire, wasn't more appropriate I don't when it would.

I do not use the nest protect but have had the joys of interlinked fire alarms where one suffered a bad battery but all of them decided to complain about it.

this was a wonderful video, wonder if its hit reddit yet. Silliness abounds there and its not the exposure you want your product to have. Reminds me of the story earlier today about smart homes

G-bots flagging it off the front page... that's HN for you.

Seems unlikely. The author is a Google employee and mentions it in the post.

Not just any Google employee either!

What is the alternate explanation of the points / time / rank then?

And the upthread point about this being a Google employee seem to miss the actual content which is highly critical of Google. And there's absolutely a history of threads highly critical of Google getting flagged down on HN.

I think there is a small faction of anti-Google readers on HN, and then a larger set of people that don't care one way or the other, taking each story at face value. I know there are quite a few of the anti-Google faction that follow me around on HN and try to drag me into arguments about Google in completely unrelated stories. That's why I don't comment much anymore.

With that in mind, and given that HN's ranking algorithm is secret, it seems unfair to accuse a hivemind of intentionally messing with the algorithm to suppress commentary about their employer. Maybe it's a bug. Maybe it's working as intended.

Could one of the HN admins comment? Do you guys notice a lot of Google-related stories getting flagged?

> Do you guys notice a lot of Google-related stories getting flagged?

Not particularly. Controversial stories reliably get lots of upvotes and flags. As far as I can tell, one could replace "Google" with "*" and the same statements would hold.

I'm going to detach this subthread as off-topic now.

Ooh, that's a nice trick!

Jesus how fucking big is that dude's apartment

I'm gonna blow your mind here: http://en.wikipedia.org/wiki/House

I laughed

Nobody mentioning the fact that he works for Google which owns Nest?

That's the point of his disclaimer:

> "Disclaimer: I am a Google employee. I paid for these myself. So I speak as myself."

If it's this Brad Fitzpatrick [1], he probably doesn't need to give a damn anymore.

1. http://en.wikipedia.org/wiki/Brad_Fitzpatrick

It's a post critical of his employers product, and he clearly mentions he's part of Google -- what is there to mention?

Does it matter?

So this guy pretty much publicly admitted he knew his companies useless smoke alarm was going off at his home, but couldn't be bothered going home to sort it out? Poor neighbours indeed.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact