In my opinion this is not Xiaomi into Home Assistant (HA). To me, an integration would mean that I need nothing from Xiaomi, all activities are within HA.
from the Github page[0]:
> Xiaomi Home Integration and the affiliated cloud interface is provided by Xiaomi officially. You need to use your Xiaomi account to login to get your device list. Xiaomi Home Integration implements OAuth 2.0 login process, which does not keep your account password in the Home Assistant application. However, due to the limitation of the Home Assistant platform, the user information (including device information, certificates, tokens, etc.) of your Xiaomi account will be saved in the Home Assistant configuration file in clear text after successful login. You need to ensure that your Home Assistant configuration file is properly stored. The exposure of your configuration file may result in others logging in with your identity.
I think (some correct me) you can group most devices into the following categories:
1. Device requires internet for setup, and for usage
2. Device requires internet for setup, but after that don't need it anymore
3. Device can be fully setup without internet, and used without internet
Personally I aim to be fully within 3 as much as I can, but some devices are really hard to find at a good price point that falls into 3. All my HA devices are within 3, except some real-time cameras which I couldn't find below ~300 EUR if I wanted them in 3, so those are within group 2 and now isolated after the setup.
I'm not buying any device that's not 3, everything else turns into a brick as soon as there's some larger change.
I have some older Google Speakers, and while they seemed to be 2, after being powered off for long enough they can't be set up again, not even with internet access since their firmware was also outdated and the app isn't able to set them up again.
I like the taxonomy you've outlined. It would be great if the Home Assistant org were to formalize something like it into levels with logos manufacturers could use in ads and packaging. It would help clarify products for users and, most importantly, provide an incentive for manufacturers toward more local-first interop.
It might be good to invert the order above and name the levels with something like Platinum, Gold, Silver to clearly signal better and best. Manufacturer's marketing people love having external compliance logos, especially manufacturers of commodity hardware.
Home Assistant itself falls into category 2, with some integrations/utilities in category 1. The amount of stuff it pulls from the mothership and GitHub is crazy. I wish it were local-only by default.
I'm curious about what you're issue was finding cameras as, to me, that's the easiest thing to find cloud free since they have a long history of being used in closed networks as POE, onvif cameras long before smart homes were really a thing.
This is how a lot of home assistant integrations work. Sure, many HA users (self included) try to avoid dependence on cloud services and opt for local only solutions, such as zwave or ZigBee or products that work with local-only wifi. But the nature of the beast is that some devices out there are built to talk only to a cloud service.
Having a company start an upstream project is probably a better sign than not having that, however sure, they could pull the plug on their access to cloud service, people may have privacy and security concerns, etc.
they could pull the plug on their access to cloud service
It's happened to me twice.
The first time was about seven years ago, when Fiet Electric sent out a software update that deliberately bricked all of its home hubs, and consequentially turned all of the connected smart light bulbs into dumb light bulbs. Speculation on IoT forums at the time was that Fiet failed to properly license some piece of code that was critical to its system; but that was all speculation. I seem to recall that Fiet put out an e-mail long after the fact letting people know they could no longer use their "smart" devices.
The second time was earlier this year, when Sylvania ended its cloud service, and turned its smart bulbs into merely clever bulbs. They'll still work with the stand-alone Sylvania app, but new bulbs can no longer be added to Apple HomeKit setups. So you need to use two apps (Home and Sylvania) to control the devices in your home. That is, until the Sylvania app is no longer available in the App Store, or compatible with modern devices.
Avoiding Fiet Electric products was easy. But I thought I'd be safe with a big name like Sylvania.
"Integration" is just the term the HA project uses for code supporting a specific device/brand/platform. Home Assistant shows a label on each integration clarifying whether it's entirely local or cloud-based.
I agree that it totally sucks. If you use home assistant, it is mostly to not give control of everything in your home to the cloud of a company. Especially to a Chinese one...
And there is no legitimate reason why control can't be local only, especially when home assistant is there to provide the gateway feature and remote access is needed!
So this action of Xiaomi is just a marketing way to be able to print "compatible with home assistant" on their box despite still forcing to use their cloud!
That is nice because the xiaomi stuff is all over the place in HA. Some devices (the air cleaner for example) have built in support but many of them need custom not really supported addons from HACS. Like the WiFi fans, weighing scales, rice cooker etc
What is Land Rover the Land Rover of again? Highest cost per repair? Least likely off-road brand to be taken off-road? No.1 brand owned by rich land-owners? I legit don't get what that reference was going for.
Home Assistant is a free and open source way of cross-connecting smart devices. It is incredibly powerful. It can easily save you money (e.g. garage door/motion sensor + thermostat temp adjustments), or allow you to craft bespoke convenience/security features.
It is the central hub of a smart-home. Very reliable in my experience.
Home Assistant is one of the best open source projects I've come across. I've been using it for 5+ years on an older RPi and it's been pretty rock solid. Countless updates and everything just keeps on chugging.
I've landed on a mixture of MQTT and Zigbee communicating devices, the latter being much easier to set up and maintain. There's an integration for just about everything I've wanted, some better than others, but all in all just a great project.
I experimented a bit with ZHA (native Zigbee integration) but soon realised I needed something better, and Z2M was that something.
It had all been working wonderfully, until I had a problem recently that meant I had to redo the entire Zigbee setup.
I run HAOS on a VM, with the Zigbee radio being passed through from the host. Recently I wanted to play with Thread so I added a radio and passed it through to the VM as well. All was fine, though I wasn't having a lot of time to experiment and the Thread dongle was connected in an awkward position so I decided to yank it out. At this point all hell broke lose. Z2M wouldn't start, throwing cryptic error messages. After a lot of trial and error, I removed both USB passthroughs, rebooted the VM, shut it down again, re-plugged the Zigbee dongle and re-added the passthrough. At this point the hardware side of things was fine but my Zigbee network was gone. To make it worse, a new one couldn't be initialised because it was trying to use the same ID as the previous one. I had to manually change the ID in the config YAML, restart everything, then re-pair all devices.
I really feel like this stuff should be more resilient to failures. Otherwise, it's pretty good!
I second Zigbee2mqtt. Koen's work is legedenary, I also been fascinated by Zigbee and been using ever since. No need to 3rd part oem vendor lock-in. 99% of the devices I purchased currently nearly 52 on my zigbee network, were paird hassle free.
I've actually done my migration the other way around. I started with zigbee2mqtt, saw that HA now offers ZHA and switched to that. It just works with all devices I own, so the end result is one less moving part I need to update so the choice was easy.
I've been rolling my own stuff, mostly devices posting to custom python servers, storing data into influxdb and mongodb, triggering other servers on events, and lately also integrating Tasmota devices via MQTT, like the microwave, washing machine, computer monitors, small heating fans and the like. I migrated all Hue devices to zigbee2mqtt and am happy with the flexibility.
Initially (7 years ago?) I refused to use HA because I've all too often had the issue that then projects become stale and I need to migrate to something else.
But lately I've gotten the feeling that HA is really here to stay, with a community big enough for this project not to die and maintaining very good hardware support.
What I'm missing out on is an (Android) app, and I think that this would be a good reason to think about moving over to HA.
> What I'm missing out on is an (Android) app, and I think that this would be a good reason to think about moving over to HA.
There is! The Home assistant companion - it brings you a lot of functionality in terms of location, notifications, sensors and what not into the HA world.
Same story here. Everything goes through MQTT, and a single python script has my automation logic. All redeployable via Docker Compose. I never need to worry about updates breaking things, and there’s much less context for me to try to remember.
Home Assistant never “clicked” with me. It makes some hard things easy, but some easy things hard. I just don’t love YAML enough to write logic in config files…
I also hate that HA pushes you to run their whole OS. The docs usually assume you’re running HAOS.
There is an app that is basically a wrapper arround the mobile version of HA. But it works quite well. The dashboards of HA are responsive and there is no need for a native version.
HA is awesome but I have found that over time entropy kicks in if you don't maintain it properly (which I haven't done for a year) connections fail, switches stop doing what they are supposed to...it's on my Xmas list to spend some time sorting it out!
I include this when considering buying something integrated.
For example Philips Hue is overpriced, but their Home Assistant integration is top tier and ultra-reliable. Contrast that with myQ garage door openers (LiftMaster, Chamberlain and Craftsman) recently breaking Home Assistant support on purpose, to essentially replace it with nothing, and they're dead to me.
So Xiaomi adding support, assuming it is reliable, definitely moves them into a better category.
I installed a pair of these, and haven't looked back. Yes, there is a little bit of a curve with rewiring your opener, but there is great documentation available, and safety in the fact that if you mess up, your door just won't open. If you snap a picture ahead of time, it's easy to undo. And from there, you can hook it in to Home Assistant or HomeKit and do whatever you want, which is amazing. My Home Assistant notifies me when the door has been open for 5, 10, 30, 60 minutes, as well as if the sensor is obstructed for the same intervals.
Second this product. I wanted an Ethernet version so I made my own (it's integrated into esphome and the circuit is documented here https://github.com/Kaldek/rat-ratgdo). Apart from general usage, I use mine to tilt the garage door a small amount if it detects bad air quality in my shop (using an IKEA air quality sensor via home assistant).
I'm aware but I'm one of the like 5 people that bought the overpriced MyQ-HomeKit adapter and it's still working to provide HA control. If it ever dies I'll be going ratgdo or opengarage for sure
It's inexpensive if you discount the cost of your education learning how to wire in something like that.
There's an order of magnitude difference in project size between setting up the old MyQ integration with home assistant and learning how to use.. whatever that thing is.
Sometimes I think clever and educated people forget what it's like to be less intelligent or educated.
1. They want to charge for some integrations. I could see this if they didn't make local only impossible if you want anything beyond the clicker. Why aren't these just bluetooth and or wifi so my car and open when I pull up and close when I leave? Hell, if they just added an 'open if closed' and 'close if open' it would make it way better for the car to controll. IMO they are purposely making the non myQ options suck and stuck in the 80s just so they can upsell to a monthly subscription.
2. Their security is a joke. After moving to new phone their app would refuse to login yet would still show me notifications for door events. The only way to stop the notifications was to uninstall the app.
My newer garage door is lacking wifi just so I can add my own automation without even bothering with theirs.
I wouldn't call them overpriced (at least not all products), their quality is typically great, you get what you pay for. We have had Hue lights for over 10 years (pretty much every light in our house is Hue) and never had any issues. I think over that period one light broke. And like you said, the integration is great. In our house we have it configured to use both through the Hue hub and SmartThings.
> their quality is typically great, you get what you pay for.
You're lucky. I also have almost all lights from Hue, and in almost 6 years I've had 6 or 7 completely dead bulbs (out of ~40). One more lightbulb failed in a weird way - it sort of worked, but in 90% of cases refused to completely turn off and still kept some of the LEDs lit. I haven't bothered to disassemble it to see how that happened. And one more light works normally, but somehow fails to report its status to HomeKit. It can be controlled but always shows up as "updating..." - guess it's a software bug of some sort, since it works in the Hue app.
No Home Assistant here at the moment, just regular Hue+HomeKit setup. I have tried HA a few times, but found no significant additional value over what I already have. It's just as dumb as all other "smart" home solutions, still has very limited diagnostics if something is not working (maybe if one really groks its internals it can be debugged better, but I don't) and requires maintenance. I was thinking about building something with plain simple Zigbee2MQTT and a bunch of DIY scripts to make it a little smart, but haven't yet had time for this.
Yup. I paid like $40CAD for a single Hue lightbulb in 2017 and it's still going. In that time, I've had countless cheap Canadian Tire brand (NOMA) & Walmart brand LED bulbs burn out and need replacing totaling WAY more than $40.
I just looked and they have gotten like 20% more expensive though... that said, their non-smart bulbs are still pretty affordable comparatively.
Look for Zigbee devices and most stuff just works out of the box. And when you’re not upgrading firmware, there’s no way for the manufacturer to break anything.
Zigbee is wonderful, especially alongside things like zigbee2mqtt device support [0]. The downside is that its not uncommon to see non-compliant devices, or buggy implementations which is almost more annoying.
I recently installed a zigbee thermostat in my bathroom, which turned out to be flooding the network and causing the rest of my network to become unstable
If you’ve got a spare garage remote control, a raspberry pi or Arduino board, a soldering iron, an optocoupler, and a sense of adventure, you can easily integrate your existing garage door opener into Home Assistant or what have you.
In fact, the Arduino starter kit comes with a few optocouplers and instructions for basically exactly this project!
Or you can get a tube of a dozen 4N25 optocouplers for like $8 on amazon.
I am looking to install a garage opener and due to meatspace constraints I will probably have to use jackshaft. Jackshafts are predominantly LiftMaster and Chamberlain => the smarts are myQ, which I don't want anywhere near my network. Genie jackshafts seem fine, but Genie's reputation is bad to the point where garage door companies may refuse to work on them.
These motors also usually come with the ability to hook up a hardwired button. There are a couple of pre-made (konnected, ratgo) solutions or one could jury rig a z-wave relay.
Alternative is Overhead Garage door company that have separate SKUs for the unit, the battery and the smarts so one can pick two and use the same relay (my current plan).
There may be _some_ proprietary shenanigans with LM and Chamberlain hardwired buttons but Overhead's one really seems to just work through bridging two contacts
Just use a LiftMaster/Chamberlain jackshaft opener, never connect it to any network install the ESP32-based ratgdo device into the wiring harness, be done with this issue.
Nothing else compares due to the digital integration. ratgdo uses the simple serial protocol that the opener button uses, so it has access to a lot of information.
Not sure how it works with a jackshaft vs. the more traditional residential opener, but ratgdo can speak the myQ protocol and control it from mqtt or home assistant. I have it working with my Chamberlain opener as do many others.
Hue is amazing, the reliability of their products is truly impressive. I've had ~40 lights for years and not a single one has died or ever had connectivity issues.
At the moment you can pair them with any ZigBee controller, which I found to be much simpler. This is one firmware update away from not working (which is why I mostly relied on Z-Wave) so YMMV.
MyQ are such scum. I love how on my smartphone I can just press a button in the MyQ app to open my garage door, but if I want to push the button on my Tesla, they want to charge me a subscription fee of $100 a year or something like that
It's worth clicking through and reading details on each one before you commit. Most of them are quite complete, but some only support a handful of devices or features. You can also get a sense if the control is local (i.e. no internet connection) or cloud based.
Meanwhile MyQ “closed” their API and broke all HA integration because it “cost” the company too much. Where in reality, they just wanted people to use their app since they started serving ads. Say what you want about China and the security implications but a lot of their IoT companies are way more opened than our US counterparts.
low-end Chinese phones, including Xiaomi and Huawei, show ads on system apps like Settings and Contacts. soon, smart home appliances might also display or speak ads. there's nothing stopping these Chinese companies from doing so - both technologically and ethically.
Xiaomi definitely should not lead the way in smart home automation.
do you mean putting Bing on the start menu? Microsoft just made them opt-in by default and hoped that folks won't notice or care. you can disable them.
for Chinese devices, there's no way to disable them.
the difference is, for Windows, having Bing on start was a feature (although a bad one). for Chinese devices, you just get ads - you're stuck with ads while changing your brightness.
You get ads for agreeing to buy a cheap subsidized device, like amazon devices. There's lots of guides out there to disable ads/msa.
>soon, smart home
How? With what screen or speaker? The half inch oled on my smart ricer cooker? Xiaomi has been in the smart home market for ~10 years, with 100s of products, except pushing for storage subscription in the mi home app, they haven't done anything aggregious. My robovacuum, air purifier, air conditioner didn't screamed ads in the middle of the night. There's nothing stopping any tech company that inputs to a screen or speaker from monetizing ads, except for past behaviour, and so far Xiaomi home has been very good.
Samsung, and some other major players put a lot of ads and bloatware in their phones, TVs, laptops and other appliances. Xiaomi doesn't have ads on global versions of TVs and Phones. I know the Chinese version does have ads.
As a homeowner, I don't know what is worse. Using stuff where the local government can watch and spy on every step or some Chinese guy watching your boring life if you are low-level person.
…unlike appliances from non-Chinese companies like, say, Amazon? :-)
There is nothing "Chinese" about enshittification of the world around us. My LG TV insists that my "home screen" is not my property, even though I bought the TV, and invents new ways of showing ads and tracking me invasively. Amazon devices show ads and speak ads. Even Apple devices, even though apple pretends they are above this, show ads in app store search results, and send you ad notifications.
I won't even honorably mention Windows, where my computer and the main UI is basically considered a free-for-all for the Microsoft marketing department.
I love Home Assistant but I now have a pretty strict minimum effort rule after years of configuring integrations and building dashboards that I would forget about after 2 months:
I only do automations (no dashboards at all), and try to keep them as simple as possible. Once I feel I’m reaching diminishing returns territory, I stop.
Only use HA if I need to mix different vendors (e.g. turn on the hue lights if the tuya sensor switches to on) or if the vendor app/service has a limitation that doesn’t allow me to do what I want. For instance, I have some automations for my Mitsubishi airco units cause their app sucks. Otherwise I’ll just use the default app or service.
Only configure an integration if I’m going to use it in an automation; I have a bunch of integrations detected that I don’t configure.
I decided to follow these rules a couple of years back, and since then I could address all my needs with almost 0 maintenance.
How many apps do you have installed to control everything? And how is the stuff integrating if you have equipment from different vendor that need to talk to each other, like AC units with PV inverter to start and shutdown based on electricity net production and real temperature in the rooms (using external thermometers, not the one in the AC)? And how do you consolidate and monitor power consumption in a single place, are you using a different solution for that?
It sounds like I use HA in a similar manner. For me, I make HA available to Google Home and HomeKit. HA is just glue to hold everything together.
I tried do make a dashboard some time ago but it felt rather complicated. Google Home and HomeKit integrate the best into my life and the lives of my family that there is no way HA can compete. Maybe that will change if I find myself in a house that I own… Maybe spending time to make a dashboard will have a better value prop.
I assume then, you don’t use HAOS? What benefit do you see by managing your config this way? Do you also have an operator to reload HA when when you update configs?
Readit News
from the Github page[0]:
> Xiaomi Home Integration and the affiliated cloud interface is provided by Xiaomi officially. You need to use your Xiaomi account to login to get your device list. Xiaomi Home Integration implements OAuth 2.0 login process, which does not keep your account password in the Home Assistant application. However, due to the limitation of the Home Assistant platform, the user information (including device information, certificates, tokens, etc.) of your Xiaomi account will be saved in the Home Assistant configuration file in clear text after successful login. You need to ensure that your Home Assistant configuration file is properly stored. The exposure of your configuration file may result in others logging in with your identity.
[0] https://github.com/XiaoMi/ha_xiaomi_home?tab=readme-ov-file#...
1. Device requires internet for setup, and for usage
2. Device requires internet for setup, but after that don't need it anymore
3. Device can be fully setup without internet, and used without internet
Personally I aim to be fully within 3 as much as I can, but some devices are really hard to find at a good price point that falls into 3. All my HA devices are within 3, except some real-time cameras which I couldn't find below ~300 EUR if I wanted them in 3, so those are within group 2 and now isolated after the setup.
I have some older Google Speakers, and while they seemed to be 2, after being powered off for long enough they can't be set up again, not even with internet access since their firmware was also outdated and the app isn't able to set them up again.
It might be good to invert the order above and name the levels with something like Platinum, Gold, Silver to clearly signal better and best. Manufacturer's marketing people love having external compliance logos, especially manufacturers of commodity hardware.
Having a company start an upstream project is probably a better sign than not having that, however sure, they could pull the plug on their access to cloud service, people may have privacy and security concerns, etc.
It's happened to me twice.
The first time was about seven years ago, when Fiet Electric sent out a software update that deliberately bricked all of its home hubs, and consequentially turned all of the connected smart light bulbs into dumb light bulbs. Speculation on IoT forums at the time was that Fiet failed to properly license some piece of code that was critical to its system; but that was all speculation. I seem to recall that Fiet put out an e-mail long after the fact letting people know they could no longer use their "smart" devices.
The second time was earlier this year, when Sylvania ended its cloud service, and turned its smart bulbs into merely clever bulbs. They'll still work with the stand-alone Sylvania app, but new bulbs can no longer be added to Apple HomeKit setups. So you need to use two apps (Home and Sylvania) to control the devices in your home. That is, until the Sylvania app is no longer available in the App Store, or compatible with modern devices.
Avoiding Fiet Electric products was easy. But I thought I'd be safe with a big name like Sylvania.
The "L" in IoT stands for "Longevity."
And it how the _current_ integration works. So what are we gaining here? I certainly don't yet have any option for a vacuum that isn't Valetudo.
And there is no legitimate reason why control can't be local only, especially when home assistant is there to provide the gateway feature and remote access is needed!
So this action of Xiaomi is just a marketing way to be able to print "compatible with home assistant" on their box despite still forcing to use their cloud!
(Very capable, but also making programmers out of home owners since 2012)
edit: I was referring to a sticker that I’ve seen often on enthusiasts' forum posts 'Land Rover - Proudly turning owners into mechanics since 1948'.
The old school Defender is a very capable off road vehicle, but its need for regular unscheduled maintenance is legendary.
Greetings from a Toyota HZJ80 driver :)
Home Assistant is a free and open source way of cross-connecting smart devices. It is incredibly powerful. It can easily save you money (e.g. garage door/motion sensor + thermostat temp adjustments), or allow you to craft bespoke convenience/security features.
It is the central hub of a smart-home. Very reliable in my experience.
https://www.home-assistant.io/blog/2024/11/18/event-wrapup-g...
I've landed on a mixture of MQTT and Zigbee communicating devices, the latter being much easier to set up and maintain. There's an integration for just about everything I've wanted, some better than others, but all in all just a great project.
Almost everything I use is ZigBee, but at first, I used the built-in Zigbee support for it, not realizing what I was missing out on.
After a move, I setup everything again, but this time via ZigBee2MQTT instead and the compatibility is miles ahead of the built-in integration.
Just a word of advice to others who are using the built-in integration atm, not realizing the big difference between the two :)
It had all been working wonderfully, until I had a problem recently that meant I had to redo the entire Zigbee setup.
I run HAOS on a VM, with the Zigbee radio being passed through from the host. Recently I wanted to play with Thread so I added a radio and passed it through to the VM as well. All was fine, though I wasn't having a lot of time to experiment and the Thread dongle was connected in an awkward position so I decided to yank it out. At this point all hell broke lose. Z2M wouldn't start, throwing cryptic error messages. After a lot of trial and error, I removed both USB passthroughs, rebooted the VM, shut it down again, re-plugged the Zigbee dongle and re-added the passthrough. At this point the hardware side of things was fine but my Zigbee network was gone. To make it worse, a new one couldn't be initialised because it was trying to use the same ID as the previous one. I had to manually change the ID in the config YAML, restart everything, then re-pair all devices.
I really feel like this stuff should be more resilient to failures. Otherwise, it's pretty good!
Initially (7 years ago?) I refused to use HA because I've all too often had the issue that then projects become stale and I need to migrate to something else.
But lately I've gotten the feeling that HA is really here to stay, with a community big enough for this project not to die and maintaining very good hardware support.
What I'm missing out on is an (Android) app, and I think that this would be a good reason to think about moving over to HA.
[0] https://www.openhomefoundation.org/
There is! The Home assistant companion - it brings you a lot of functionality in terms of location, notifications, sensors and what not into the HA world.
https://companion.home-assistant.io/
Home Assistant never “clicked” with me. It makes some hard things easy, but some easy things hard. I just don’t love YAML enough to write logic in config files…
I also hate that HA pushes you to run their whole OS. The docs usually assume you’re running HAOS.
For example Philips Hue is overpriced, but their Home Assistant integration is top tier and ultra-reliable. Contrast that with myQ garage door openers (LiftMaster, Chamberlain and Craftsman) recently breaking Home Assistant support on purpose, to essentially replace it with nothing, and they're dead to me.
So Xiaomi adding support, assuming it is reliable, definitely moves them into a better category.
https://ratcloud.llc/
There's an order of magnitude difference in project size between setting up the old MyQ integration with home assistant and learning how to use.. whatever that thing is.
Sometimes I think clever and educated people forget what it's like to be less intelligent or educated.
I want a solution I can download :(
1. They want to charge for some integrations. I could see this if they didn't make local only impossible if you want anything beyond the clicker. Why aren't these just bluetooth and or wifi so my car and open when I pull up and close when I leave? Hell, if they just added an 'open if closed' and 'close if open' it would make it way better for the car to controll. IMO they are purposely making the non myQ options suck and stuck in the 80s just so they can upsell to a monthly subscription.
2. Their security is a joke. After moving to new phone their app would refuse to login yet would still show me notifications for door events. The only way to stop the notifications was to uninstall the app.
My newer garage door is lacking wifi just so I can add my own automation without even bothering with theirs.
I wouldn't call them overpriced (at least not all products), their quality is typically great, you get what you pay for. We have had Hue lights for over 10 years (pretty much every light in our house is Hue) and never had any issues. I think over that period one light broke. And like you said, the integration is great. In our house we have it configured to use both through the Hue hub and SmartThings.
You're lucky. I also have almost all lights from Hue, and in almost 6 years I've had 6 or 7 completely dead bulbs (out of ~40). One more lightbulb failed in a weird way - it sort of worked, but in 90% of cases refused to completely turn off and still kept some of the LEDs lit. I haven't bothered to disassemble it to see how that happened. And one more light works normally, but somehow fails to report its status to HomeKit. It can be controlled but always shows up as "updating..." - guess it's a software bug of some sort, since it works in the Hue app.
No Home Assistant here at the moment, just regular Hue+HomeKit setup. I have tried HA a few times, but found no significant additional value over what I already have. It's just as dumb as all other "smart" home solutions, still has very limited diagnostics if something is not working (maybe if one really groks its internals it can be debugged better, but I don't) and requires maintenance. I was thinking about building something with plain simple Zigbee2MQTT and a bunch of DIY scripts to make it a little smart, but haven't yet had time for this.
I just looked and they have gotten like 20% more expensive though... that said, their non-smart bulbs are still pretty affordable comparatively.
I recently installed a zigbee thermostat in my bathroom, which turned out to be flooding the network and causing the rest of my network to become unstable
[0]: https://www.zigbee2mqtt.io/supported-devices/
In fact, the Arduino starter kit comes with a few optocouplers and instructions for basically exactly this project!
Or you can get a tube of a dozen 4N25 optocouplers for like $8 on amazon.
https://store.arduino.cc/products/arduino-starter-kit-multi-...
You just excluded 99.9% of smart home product customers.
These motors also usually come with the ability to hook up a hardwired button. There are a couple of pre-made (konnected, ratgo) solutions or one could jury rig a z-wave relay.
Alternative is Overhead Garage door company that have separate SKUs for the unit, the battery and the smarts so one can pick two and use the same relay (my current plan).
There may be _some_ proprietary shenanigans with LM and Chamberlain hardwired buttons but Overhead's one really seems to just work through bridging two contacts
Nothing else compares due to the digital integration. ratgdo uses the simple serial protocol that the opener button uses, so it has access to a lot of information.
That said they’re rent seeking to use their products, eg $100/yr or thereabouts for Tesla integration.
Have 2 of them and work great.
At the moment you can pair them with any ZigBee controller, which I found to be much simpler. This is one firmware update away from not working (which is why I mostly relied on Z-Wave) so YMMV.
It's worth clicking through and reading details on each one before you commit. Most of them are quite complete, but some only support a handful of devices or features. You can also get a sense if the control is local (i.e. no internet connection) or cloud based.
Xiaomi definitely should not lead the way in smart home automation.
for Chinese devices, there's no way to disable them.
the difference is, for Windows, having Bing on start was a feature (although a bad one). for Chinese devices, you just get ads - you're stuck with ads while changing your brightness.
>soon, smart home
How? With what screen or speaker? The half inch oled on my smart ricer cooker? Xiaomi has been in the smart home market for ~10 years, with 100s of products, except pushing for storage subscription in the mi home app, they haven't done anything aggregious. My robovacuum, air purifier, air conditioner didn't screamed ads in the middle of the night. There's nothing stopping any tech company that inputs to a screen or speaker from monetizing ads, except for past behaviour, and so far Xiaomi home has been very good.
As a homeowner, I don't know what is worse. Using stuff where the local government can watch and spy on every step or some Chinese guy watching your boring life if you are low-level person.
Did you mean Oppo and Vivo?
There is nothing "Chinese" about enshittification of the world around us. My LG TV insists that my "home screen" is not my property, even though I bought the TV, and invents new ways of showing ads and tracking me invasively. Amazon devices show ads and speak ads. Even Apple devices, even though apple pretends they are above this, show ads in app store search results, and send you ad notifications.
I won't even honorably mention Windows, where my computer and the main UI is basically considered a free-for-all for the Microsoft marketing department.
Deleted Comment
I only do automations (no dashboards at all), and try to keep them as simple as possible. Once I feel I’m reaching diminishing returns territory, I stop.
Only use HA if I need to mix different vendors (e.g. turn on the hue lights if the tuya sensor switches to on) or if the vendor app/service has a limitation that doesn’t allow me to do what I want. For instance, I have some automations for my Mitsubishi airco units cause their app sucks. Otherwise I’ll just use the default app or service.
Only configure an integration if I’m going to use it in an automation; I have a bunch of integrations detected that I don’t configure.
I decided to follow these rules a couple of years back, and since then I could address all my needs with almost 0 maintenance.
I tried do make a dashboard some time ago but it felt rather complicated. Google Home and HomeKit integrate the best into my life and the lives of my family that there is no way HA can compete. Maybe that will change if I find myself in a house that I own… Maybe spending time to make a dashboard will have a better value prop.
https://github.com/shepherdjerred/homelab/tree/main/cdk8s/co...