It seems to me like a team sat for a while, looked at all the possible ways they could get people to mount a package like this in every room and settled on calling it a smoke alarm.
At the most optimistic they started with a smoke alarm and gradually realised they could build a general purpose platform based on the hardware being deployed in lots of rooms and many types of sensors being dirt cheap now.
Decided they could enable new capabilities (and data goldmines!) in software later.
It’s a pity there’s not an actually customer controlled version of it.
Are there any best practices for using a device like this but not having it communicate with the wider world? I.e. It can communicate with you via a homekit hub, but can't connect outside your LAN
https://hackaday.com/2014/07/29/a-cheap-diy-smoke-detector-t...
I then paired it with https://www.home-assistant.io/ to send alerts to my phone, google home, etc.
Haven't had to actually test that the carbon monoxide alarm works yet because testing that is hard/expensive, but for smoke it works just fine.
Given that these devices are battery powered and meant to last for years on a single charge you can imagine how often they actually connect to the wireless network. And how much traffic they send.
And again, they work perfectly fine without Internet access. Or network access for that matter. I love them, and as someone with an irrational fear of being in a fire they have helped a lot. They're much more sensitive (without being an issue, due to pre-alarm) than the alternatives.
