Whatever happened to IoT smoke alarms?

by @edent | , , | 18 comments | Read ~1,803 times.

I've had a Nest smoke alarm for about 7 years. It connects to my WiFi network and occasionally pings a message to my phone that I've burnt my toast. Nifty!

But, due to planned obsolescence regulatory requirements, it needs to be replaced.

In app warning that my Nest smoke alarm needs replacing.

Back in 2014, the Nest cost £100.

In the exciting world of 2021, it costs... £100!

WTAF? Surely a combination of market capitalism and Moore's Law means these devices should have dropped in price significantly? But I can't find any other Wi-Fi smoke/co alarms for sale in the UK. Why is that?

(There's a Netamo WiFi smoke detector for £90 - but it doesn't do carbon monoxide.)

A basic smoke & carbon monoxide detector is about £20 retail. There are hundreds of models at that price point.

A basic ESP32 WiFi board is a fiver. Call it a tenner if you want something like a Raspberry Pi Zero W.

Add in the cost of running some servers, and you're barely scraping £50. Surely there's someone out there who wants to sell a million of these things?

But, I think I can understand why there aren't loads of these for sale:

  • Low market desire. While not exactly Internet Fridge territory - there isn't a huge desire for mostly passive IoT devices. Things like flood sensors and open-window detectors just don't seem to interest people.
  • Compliance and insurance. I assume there's a high cost of certification for these things. And if someone's house burns down, but the detector didn't warn them, then insurance companies start to fight.
  • Software is hard. Nest had a troubled history. Apps need constant updating. Which either means a high cost for the device, or tying people into a subscription plan.
  • Patents. I'm sure Google has a bunch of these and it probably scares off most competitors.
  • Wi-Fi complexity. Wi-Fi is power hungry - which makes it hard to run on in-built batteries. It is also difficult to configure for the user - lots of faffing around with apps and having to reconfigure the device when your wireless password changes.

Alternatives

ZigBee is the low power, low cost, short range radio network for IoT. Found in lots of lightbulbs and other things. The downside is that it needs a central hub in your home in order to connect to the network. In theory, any hub should connect to any ZigBee devices - I use a bunch of generic lightbulbs with my Hue hub.

But the reality is slightly different from the theory. I can't find any ZigBee smoke/co detectors for sale in the UK. There are a bunch which claim to work, which can be shipped direct from China. But they rarely have the requisite safety certification, and they all seem to need their own proprietary hub.

The Roost Battery is a WiFi enabled battery suitable for placing in smoke alarms. When it detects a power draw that indicates the alarm is sounding, it connects to WiFi. But, again, it isn't sold in the UK any more.

Now what?

Do I need to be alerted that my house is burning down when I leave lockdown? No, probably not. But it's fun!
Am I aware of the privacy risks? Yes - and I understand how to mitigate them.
Could I build my own out of a Pi and some scavenged components? Sure - but I can't be bothered.

So, unless one of you clever lot can recommend a smart smoke & carbon mono detector which is available in the UK - it looks like I'm paying the Google Tax for the Nest. Bah!

18 thoughts on “Whatever happened to IoT smoke alarms?

  1. There’s other things I like about nest, if the CO2 alarm goes off it shuts down the boiler, things like that are quite nice.


  2. @Edent To be fair, planned obsolescence is probably appropriate when an essential part of the product relies on a radioisotope with a known half life, and which is too dangerous to be user-serviceable..

    1. David says:

      Interesting theory, but not correct. Smoke detectors are based on americium-241 with a half-life of 432 years. It is also a weak alpha source, which is why it isn't able to penetrate smoke and this is what triggers the alarm.

  3. Beko Pharm says:

    I love reading about your IoT journeys @Edent whilst completely ignoring my own. Like that washing machine I have, that in theory is able to connect over the wifi to… yes to what? I guess to some vendor API and I assume it does so over WEP only too. Or our central heater, that used to have it’s own gateway but since the company never got it running they patched it out so now it sits on my regular subnet and sends an occasional mail unsecured to some SMTP gateway. Don’t think it even uses hardcoded credentials for this. Oh the possibilities (and dangers)…
    I know. And weep. And I ignore it. I’m too busy with other things than “fixing” or getting this zoo under control. It’s a mess. It’s the Internet of Shit.

  4. David Sheldon says:

    Also, is there anything one can do with an embedded computer with a bunch of arm cores and wi-fi etc. that used to be an over-engineered smoke alarm? It seems a shame to add it to the WEEE pile.

  5. I assume the answer is: when they turn the servers off and your IoT kettle stops working, all you can do is shrug. But if they turn the servers off and your house burns down, you'll sue, and (ideally) win.


  6. Rob says:

    For me the benefits of the nest detectors isn’t the internetiness of them - it’s the usability. The alarm speaks to me so I know what to do - “there’s smoke in the hall” - and the regular self tests and reporting these to me give me a sense of safety knowing they are working. A set of 3 interconnected smoke alarms is about £60, whereas a pack of 3 nests is £250. Yes, I’m paying more but it’s less than £20 a year more. Is that worth paying for a better user experience? In my view, yes.

  7. Steve Dee says:

    I have a feeling I've seen smart screens / home assistants that listen out for the audible sound of alarms, and then trigger notifications. It might even have been a Google Home/Nest feature?

  8. For the last 5 years I've wonder about this as well. I ended up doing my #iot smoke alarms with #ESP32 , #arduino, #nodered, #tasmota shkspr.mobi/blog/2021/04/w…

    1. Alex S says:

      @Spencer did you paste the wrong link there? That is just a link to this page. I'd be interested to know how to made a smoke alarm out of an ESP32 specifically how you make it low power.

      1. Tom says:

        You're looking at a tweet, that's why it has a link to this article.

        It seems that there's an integration with Twitter so if someone posts the link, it drops a 'comment' in here. Click the dat under the person's name and it'll take you to the tweet.

  9. Eric says:

    These kids at Google invent new stuff while ignoring what exists in the market. No wonder they project don't go anywhere.
    The big problem is that they don't know one of the disciplined of Project Management called Change Management. You cannot steer a battleship on a dime. You must do it slowly. With and industry that saves life's, you also have to earn trust and do it slowly.
    When the Nest smoke detector was launched there was an old industry of wired home alarm panels installed by professionals (in addition to the stand-alone units).
    https://www.doityourself.com/stry/series-of-hardwired-smoke-detectors-wiring-explained

    They ignored this market segment. Big error. They should have added connectors to insert the Nest as part of an existing system or at least creat a base unit to receive the WiFi signals from all the the Nest units into the panel. Then, by selling thru the network of installers, they could have a constant revenue, sales and most important, the support from the experts of the industry.
    You should also consider that most industry professionals don't consider a component without battery backup (most home WiFi networks) as reliable enough for being part of a life-critical system, but that is another topic.

    The guy in the video below is presenting a product that aims to bridge old wired security and fire alarm panels (critical systems) to the new home automation systems (convenient, non-critical systems). This is an interesting solution, not only because most Home Alarm panels suck at home automation, but also because HA market is evolving too fast as for baking it as part of your safety and security system.
    https://youtu.be/M2uYjqooGnE

  10. James H says:

    I've just come accross the Fire Angel series of smart fire alarms.
    https://www.fireangel.co.uk/home/connected-home/

    Part of the issue might be that fire alarms are targetting M&E consultants rather than consumers and the usual IoT crowd?

    1. @edent says:

      Oh! Those look interesting. Although the app reviews are... not great https://play.google.com/store/apps/details?id=com.fireangel.b2c

    2. Robert Wigley says:

      I have spent hours researching and the FireAngel Connected Pro models (which I believe are using Z-Wave with their proprietary gateway) appear to be the best on the market currently (in the UK anyway). So I bought and installed them yesterday as they are also one of very few systems that are compliant with the upcoming new Scottish regulations (https://www.gov.scot/publications/fire-and-smoke-alarms-in-scottish-homes). First impressions are that they are pretty good. The app has been working flawlessly for me so far (admittedly only a day) and I have setup an Alexa Routine using the Alexa Skill (https://www.amazon.co.uk/Sprue-Safety-Products-FireAngel-Wi-Safe/dp/B07VV8JWMX) to test them weekly. What is really lacking for me is proper smart home integration with Home Assistant so that I can turn on the lights, unlock the front door and have a voice announcement tell me which alarm is sounding in which room. IFTTT support is supposed to be on it's way according to their FAQ and I have emailed them to ask if there is any API that can be accessed to integrate directly into Home Assistant. Currently waiting on a response....

  11. jc says:

    While looking for the same thing I've found https://www.elro.eu/en/. I bought some units ($20 for a smoke alarm and $40 for the connector) and after some small problems getting the devices connected it does seem to work quite well. I'm currently testing using their app which requires the connection device to have an internet connection but I've also found the https://github.com/dib0/elro_connects project. I'm guessing that if that is working it will no longer require that internet connection cause the device can only handle a single connection.

    The system is by far perfect because the connector just won't work if you have an 5ghz signal with the same ssid. Considering the features you get for the prize I can work around those problems by disabling the 5ghz band for my IOT network because almost none of those devices support that anyway.

Leave a Reply

Your email address will not be published. Required fields are marked *