IKEA sensors for doors and windows, motion, water leaks

Zigbee2mqtt is the bomb. ZHA can't hold a candle to it in terms of reliability, features and device support. The tiny increase in complexity is so worth it.

All this low power electronics designed to run 3 years on a coin battery bothers me. If it's that low power why not get rid of the battery, slap a tiny solar panel (the kind they used on calculators in the 90s) and a supercapacitor, it should be able to power itself indefinitely on normal levels of room light, never go into deep sleep states, and can even turn the room lights on automatically if it needs more power.

It does, which is a big selling point for me. I’m trying to standardise on rechargable AA/AAA batteries wherever possible. A lot of devices recommend you only use alkaline batteries because of the extra voltage but I haven’t actually found one that doesn’t function on 1.2v eneloops yet.

Also, you wrote matter over thread, but the thread is the same as zigbee. From the rest of your post you mean matter over WiFi.

Knowing that this was happening would not have likely saved the bathroom, or the laundry room, but it might have saved the kitchen.

Point is: for water leak scenarios you should think about both clean water coming from inlets as well as clogs in your outlets.. I should get a couple of these to place behind my lowest toilets, where the water came up last time

The leaks that haunt me though... several under the house in the crawl space. Copper pipe pinholes, caused I don't know how much damage (including several ruined water heater elements, maybe even causing the water heater to fail prematurely). Ruined floors. Until they got bad enough that we had no hot water (which was about the time the floor started warping), there doesn't seem to be anyway that we might have noticed. The water bill wasn't unusually high, or maybe it just frog-boiled its way up where we couldn't. There were several more copper pinholes, but they were noticed as muddy places in the crawlspace while fixing the first, or just spidey-sensed. I am also very nearly too fat to crawl under the house, let alone do emergency plumbing while there.

Another incident involved a squirrel. The idiot owners of this house (before us, not the current idiot owners) decided that the easiest place to put new plumbing was in the ceiling (maybe they were too fat too?). Our attic had a new uninvited guest, and apparently PEX piping is very tasty. Or appears that way, the squirrel lost his appetite after biting enough of a hole that there was a leak from above. This leak though, despite the location which should have made it almost instantly obvious, happens almost within 5 minutes of a torrential downpour that lasted better than an hour. That, plus my unwillingness to believe the previous owners had routed plumbing above everyone's heads, meant we didn't get that taken care of quite as quickly as we might have.

My lessons from this, I think, are:

1. We might have caught the crawlspace leaks quickly if there were a humidity sensor (is that how the leak sensors work?) down there, and we could have compared to baseline. That is, if spraying water didn't just fry them immediately.

2. The ceiling leak probably couldn't have been detected very early, given the coincidence with the rain and how it would have raised humidity. Possibly if we had a water pressure sensor, that would have picked it up as a drop, despite there being no faucets running.

Plumbing sucks. Houses build 50 years ago assumed that it could never fail often enough to make it easily accessible. My next house will take into consideration squirrel mitigation concerns.

Drain pans won't handle a catastrophic failure. I installed this in our house to protect against pipe failure. We had the old grey poly-b pipe (since replaced) that was used in the late 90s.

You can DIY this if you are handy with soldering copper pipes, or get a plumber to install it for a couple hundred bucks.

Apparently water damage claims cost exceeds fire.

[0] https://www.geoarm.com/wv01lfus125-fortrezz-z-wave-automated...

You can do this without connected devices, depending on the configuration of your bathroom!

You can get night lights that plug into an outlet, turn on when it’s dark, and brighten when they sense motion. I have some Casper ones (that also let you disable the motion sensing): https://www.amazon.com/Casper-Sleep-Glow-Night-Light/dp/B09L...

You can also configure you bathroom with multiple lights where some are connected to dumb motion sensors. e.g. Panasonic makes bathroom fans with a primary light and a night light - you can connect the night light to something like a Lutron Maestro motion sensor that goes in the wall where your light switches are. (Or if you want to get fancy, you can get a ceiling-mounted motion sensor and a radio-capable Maestro switch/dimmer that can communicate only locally with the motion sensor.)

https://www.aqara.com/us/product/presence-sensor-fp2

Like other commentators, I have both z-wave and zigbee networks running. Z-wave has better standalone penetration (lower frequency), but if I were to build from the ground up again I would go all-in on zigbee. One of the few reasons I have a z-wave network at all is a smart lock, but these days there are more zigbee offerings there.

All lights in my house are hue bulbs, using a deconz setup controlled by home assistant. My aqara motion/contact sensors get ~2 years of battery life.

Since ZWave is licensed and controlled, I've only had issues where (actual example) a siren supported playing multiple tones but the hub didn't support that particular feature directly.

My experience with Zigbee is slightly different. Being an open standard there are lots of cheap choices however it also means that because there is no licensing to use it, not all of the devices branded "zigbee" are fully compatible with each other. My personal experience leans more towards "Pair the device and cross my fingers that it is compatible". If it's not, nothing works.

Otherwise; They all use the same basic tech (PIR, Resistance for water sensing, magnets and reed switches for door/window open closed) so it really comes down to implementation.

The lights are decent, I've had 2 fail so far, but both were on the same spot so I suspect something's wrong there.

The motion sensor has been pretty reliable, still working with no signs of failure. Delays or "dropped" events (the light not turning on) have been very rare and most likely a result of the ancient server I have than the motion sensor itself.

You could open any Z-Wave device and find exactly the same chip, usually on a daughterboard. It worked well enough on mains power, but battery life sucked. This started to change several years ago, but battery-powered devices are still hit-and-miss.

ZigBee has always had a robust infrastructure of manufacturers, who actually compete on quality and price.

An important part of my setup is that all the switches in my house still work, and won't actually cut power to the bulbs. There are decent options available to add a Zigbee (or wifi) dimmer/relay/remote behind an existing light switch, which is also great for controlling lights that aren't Zigbee and/or multiple devices simultaneously.

This has the added benefit that the stuff it's controlling doesn't have to be on the same circuit, and "rewiring" a light switch is done very quickly in software instead of physically rewiring.

This "LR" stuff you mention is what I think some are calling the Z-Wave 800 series. It more than doubles the radio range of the 500 series stuff and greatly reduces power draw for battery devices. However, I have 500 and 700 series sensors that only need a battery change once every ~2 years or so. I don't find that to be too much, and the signal strength has been no problem for me in an old brick house. I do have a handful of powered zwave devices around though, and these act automatically as repeaters for the other devices in the mesh.

Zigbee theoretically can use this same frequency but in practice none of the devices do.

Make sure your hub has actual external antennas.

The flip side to this, is Zigbee stuff always has a lot of little quirks, problems, or bugs. There's literally an entire module/database for ZHA (Zigbee Home Assistant addon) called ZHA Quirks that's designed to work around all the little issues with various devices.

I'll be happy when/if Thread becomes the norm and I can just run everything through Apple HomeKit. Though for some reason, I still don't see a Thread in-wall dimmer available, which seems crazy to me.

A sensor can alert someone but doesn't stop the water. There are alternatives:

1) An electric valve directly at the tap that the machine actuates when it needs water. Usually comes with hose-in-hose to capture water from a ruptured line and detect and stop releasing water.

2) Quantity limiting devices to limit the quantity that can be used contiguously, and reset when the flow stops. Set to the maximum needed by the machine. Should there be any leak, the possible damage is limited. No electricity needed.

https://www.schlage.com/en/home/smart-locks/connect-zwave.ht...

This would give you the "is locked or not" capability you seek here, but obviously requires a replacement of your door lock (perhaps not trivial if renting, but is a very easy DIY otherwise).

And yet some comments are complaining that it’s more expensive than a Tuya-based solution or whatever. I think this is at the heart of the problem: people want the longevity and cost of a dumb-home product, but the value of these things is rooted in software. Cognitive dissonance ensues—“you want me to pay a subscription for a door lock?” At least here, IKEA is allowing point-to-point communication that could eliminate moving targets of other companies’ software.

The new hub is different, and does not work with the old Ikea system.

I totally agree, but the real problem is people uncritically sharing and promoting articles from low-quality publications. In other words, your wrath is better directed at the HN users who submit and upvote articles from these low-effort sites. There are plenty of publications, such as LWN, that use links to help the reader, rather than to help themselves through Google’s algorithms.

> Ikea ... has “decided to delay this functionality” and will provide an update “when it’s time.” The ongoing delay is understandable given that Ikea’s products already integrate well with other platforms, and the company is focused on keeping things as simple as possible for anyone who delves into the smart home on a whim while shopping for a new bookcase. And when you consider the hurdles required to get everything running on Matter networks during this period of transition, Ikea’s delay is more than justified.

Insurance companies will give you a small discount if you have one installed as well.

There are a bunch of options at Home Depot, Lowes, and your local hardware store.

These have a loud alarm on them and I don't believe they need to be hooked up to the hub to work (just for push notifications).

> detect opening doors and windows, motion indoors and out

If these devices are sending that data back to Ikea, they're surely planning to monetize it somehow.