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.
[0] https://github.com/XiaoMi/ha_xiaomi_home?tab=readme-ov-file#...
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 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.
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powered off for long enough they can't be set up again, not even with internet access since their firmware was also outdated and the app isn't able to set them up again.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 am ok with devices that default to 2 but can be converted to 3.
Don't recall if Shelly's default to 2 or 3 but I like that you can flash em to tasmota for a gaurenteed 3.
Most devices are in the category 3, except for BLU that are currently in 2. I have almost 1 of everything they make and I did set up the BLU ones using 2 with their app, I did not test to see if their app works without Internet, but I doubt the BLU can be configured without their app.
That’s why I’m sticking to Zigbee as much as possible. The only place where there’s an internet connection is at the Home Assistant computer which has a Zigbee USB stick.
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.
I can’t speak for OP but usually the issue with #3 for me is that the other criteria it needs to meet is that it works with whatever other integration I want. I remember when I looked the TP-Links were a good option but you just needed internet to set it up. Afterwards it just went on a vlan without internet. The camera needs to support Scrypted/Frigate for my use case but depending on needs for PTZ, Wi-Fi, resolution, night vision, etc, I may or may not be able to find one at a reasonable price that doesn’t require internet access for setup. TP Link makes good cameras at a good price but they require internet access to setup, so any camera that falls into #3 will get compared to a TP Link in #2
Everything can fit in group 3, but manufacturers want to steal your data so they try to pretend otherwise.
Not all manufacturers.
At AirGradient, all our air quality monitors can completely run local. Our official firmware is on GitHub and people could even flash their own (adjusted) version.
We believe this is how IoT devices should be and are very vocal about it. So I think there are a few manufacturers that think different.
If those devices ran off POE they would be divine. Trying to keep walls looking clean in our custom build and usb power is a drag.
And this is exactly why I like to buy AirGradient hardware. Your CO2 sensors are some of the best I've found for the price point, and I love your outdoor monitor as well.
That said; I have started moving towards using SEN5x for a lot of my air monitoring. I've noticed the SHT3x temp sensors pretty consistently flake out after 2-3 years, and the footprint of your current DIY kit is rather larger.
Would you consider an updated version with the SEN6x once it comes out, with perhaps an ESP32c3, rather than the 8266?
You're giving them the benefit of the doubt here. In my experience, not only are they greedy, but they're also inept and lazy.
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.
> This is how a lot of home assistant integrations work.
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 it how the _current_ integration works. So what are we gaining here
An official integration, supported (for whatever that's worth) by Xiaomi instead of random people reverse engineering their API until the next breaking change, or even worse, them deciding "this is DDoS so we'll ban anyone from using our API".
Although I bought my Dreame D10 because its on the Valetudo compat list, I actually never put it on, because it turns out OOB you actually don't need to setup anything for the regular mop function to work. It's entirely unconnected, which I prefer anyway. I miss out on remote access and mopping only parts of the room, but I can live with this trade off.
Once they turn them into true IoT shits, then I'll be worried.
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.
The "L" in IoT stands for "Longevity."
I completely agree with the grandparent. This is all avoided by using Zigbee or Z-Wave devices. All our smart lights are Hue. If they decide to stop supporting it, they can be controlled with the Samsung/Aeotec SmartThings Hub, Home Assistant, Zigbee2MQTT, or whatever you please. Similarly, our smart plugs are also Zigbee and we use a couple of Aeotec Z-Wave temperature/humidity sensors.
Best of all, less worries about yet another IoT device with probably vulnerable software that we have to put on a VLAN/IoT WiFi network. Zigbee and Z-Wave are also much simpler than WiFi/Bluetooth, so less likely that they are a swiss cheese of vulnerabilities.
With hue there is still potential risk of a lockin with their hue app. They allow it still, because their sales are not really good. But otherwise they might have restricted it.
Hue bulbs currently are stock standard Zigbee compatible light bulbs. You can pay them into ZHA/Z2M without any issues and control them without any Hue hub or app.
If Hue were to suddenly switch to something proprietary their existing bulbs will all continue to function without their app or hub.
Hue Bulbs use Zigbee. If hue stops supporting the hub or older devices, you can reset and pair them to anything.
The Hue bridge is IP based but can be controlled entirely over your local network. It’s a slim possibility of something breaking (the mobile app mostly) and then the bulbs are still fine.
Yeah, I ditched the Hue Hub last year and paired them all directly with Home Assistant via Zigbee
what's the state of bulbs that don't use apps? I have some from IKEA that get paired to a Bluetooth remote, and it seems pretty good for now but I'm kinda nervous about relying on a device like that.
IKEA bulbs use mostly Zigbee, can be paired directly with their remote so don't require any hub, and connect directly with any Home Assistant with a Zigbee dongle like Sonoff.
IIRC Ikea bulbs all use Zigbee so should be pretty safe.
Philips Hue uses Zigbee and the quality of Hue lights is really excellent (both the lights themselves and longevity, our oldest lights are over 10 years). A lot of Ikea devices also use Zigbee, though it sometimes takes a while before they are supported properly by SmartThings, Home Assistant, etc.
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
"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.
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've landed on a mixture of MQTT and Zigbee communicating devices
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 :)
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!
> At this point the hardware side of things was fine but my Zigbee network was gone.
How come the hardware was ok but the network was gone? Did some Z2M config files go wrong because of two dongles? Did you try to restore VM from snapshots?
Losing the network and having to re-pair everything would be a nightmare for me given the number of Zigbee devices I run (~35) and that some of them are mounted in switch boxes in the walls.
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.
The nice part about mqtt is other things can subscribe to the same stream. I've also seen people do some really interesting things with hosting a copy of HA in the cloud -- this is what they connect to while outside their network -- and using mqtt to push updates to a local HA.
Yes, I see all the praises about Z2M, but don’t see any details how it’s better. ZHA works just fine for me?
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.
I think the launch of Open Home Foundation [0] is a very good sign for the future of Home Assistant.
> 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.
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.
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.
> Philips Hue
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.
In case anyone using a myQ opener comes across this, I feel the need to mention ratgdo which many have found to be a great inexpensive upgrade.
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.
Can this detect if you’re left your garage door open for a while and notify me?
As long as it can report open/close status, you can create an automation for that. That’s what I did.
I added ratgdo to my HomeAssistant and have HomeAssisteant send notifications if the garage door is open (with a button that closes it)
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.
I want a solution I can download :(
May I suggest OpenGarage? https://opengarage.io/
That's a fair point, but it's still likely less expensive than replacing your existing opener if you include the cost of an electrician doing the wiring for you.
Perhaps an iSmartGate Pro works?
I'll just second to avoid myQ at all costs.
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.
For example Philips Hue is overpriced,
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.
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.
Try Konnected for garage doors. https://konnected.io/collections/shop-now?filter.p.tag=Smart...
Have 2 of them and work great.
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
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
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.
Besides the questionable rent-seeking behavior from myQ:
* I don't want to use an integration that needs a round-trip through the cloud to work. Long-term changes are inevitable (company goes out of business, randomly changes API, etc.)
* I fundamentally do not like Amazon Key integration. It gives someone else control over my security hardware which makes me very uncomfortable. I am not sure if it's opt-in or out, but the point is that a myQ device that is installed _can_ be configured to let arbitrary third party to open the door.
If I have a choice, I'd rather set up a system that I control from the get-go rather than try to lobotomize a system that I can't fully control.
Ratgdo doesn't go through the cloud. It's a separate board. You wire it into the opener the same way would a button, and it speaks the serial protocol that a myQ enabled button or console would use. Then it can speak MQTT over wifi.
I never paired my opener with any app nor do I have it on WiFi.
Why don’t you want myQ anywhere near your network? Last I saw, 3rd party security analysis has actually been shockingly good.
That said they’re rent seeking to use their products, eg $100/yr or thereabouts for Tesla integration.
> Why don’t you want myQ anywhere near your network?
Precisely because they're rent-seeking. I have a wifi-enabled garage door opener that I paid a lot of money for. Why should I have to pay MyQ every month to effectively just let something other than their app or their proprietary switch open the door?
“Near my network” suggested it was a security issue. Are there other companies running a cloud-based api for garage doors that don’t charge? It does cost money to keep the servers going.
I personally just took a cheap remote, attached its button to an esp32 with a relay, put on mqtt, and interfaced it with homebridge. Now it shows up in the home app. Works well! iCloud via Apple TV connects it to the internet securely.
Nope, my concern wasn’t security, it was corporate greed
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.
https://store.arduino.cc/products/arduino-starter-kit-multi-...
> 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 just excluded 99.9% of smart home product customers.
Yes, well, but here at HN, hopefully we’ve got the 0.1% in attendance!
Is there a list I could consult to find such companies?
https://www.home-assistant.io/integrations/
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.
I wish FEIT devices were compatible. I've seen their smart bulbs at costco for as low as $2.50/ea.
There's a reason Feit is so cheap.
People have had success flashing custom firmware in the past, so hardware wise they are compatible.
Feit are the only LED bulbs that I have had straight up die on me inside of a year.
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
I'll be the one to plug Valetudo in this thread I guess. Primitively, it replaces the cloud functionality on-device for robot vacuums (see supported models) and replaces it with local services that run offline and can connect to Home Assistant easily.
I will never buy another robot vacuum without Valetudo support as long as that project lives. It's great.
I've been eyeing Valetudo for a while. I think it's a shame that it doesn't and (according to the author) never will support multiple maps. I don't want to have to buy a separate robot for each floor when I can easily carry it to the next floor when it's done.
Some robots support restoring map snapshots. I wonder what's stopping people from using this as a multi floor implementation?
I use a fork [1] of Valetudo and it lets me do just this. I save one map per floor, then restore when carrying it between floors. One floor gets cleaned much more often, but so far I have preferred this over buying two robots.
If you are thinking of deploying Home Assistant (HA), let me give you a few tips that I should have known when started. HA environment is vast. There are myriads of options, features, and functions. There are some gotchas that can be costly down the road.
All of the following are "for now, as you start out":
Do not gyrate over which version of HA to run. Run the HAOS loaded on RasPi, or a dedicated machine. You can migrate to other solution if that does not end up to be the best choice.
Make sure you get the latest Zigbee radio dongle the HA Forum recommends. I use in all my HA installs a combination dongle for Zigbee & Z-Wave and it works flawlessly (NORTEK Quickstick Combo).
Start with a few cheap Zigbee devices. Stay away from WiFi, LoRaWAN and other solutions. Z-Wave devices are great, but more expensive. You know you do not know if you even want to do this.
Initially, just use the built-in Zigbee (ZHA) integrations. Once you are comfortable deploying other ways, like MQTT can be established easily.
Do not spend too much on the devices now. I suggest you pick up a PIR motion sensors, door & window opening sensors, temperature & humidity sensors, smart plugs, and light bulbs. Anything else is overkill to try out. Sub-tip: all of these are battery operated, so try to consolidate on a standard battery. The ones I listed can all run off of 2032 cell batteries. Ikea sells most of these, Aqara is well known, and so on. If it is Zigbee (not Matter) you will have 96% chance it will work with HA.
Have fun!
I started with HA in a VM and some WiFi devices, added BLU later, never had the need for Zigbee. More than a year later and over 40 devices in 2 locations (with VPN between, using the same HA server), I am more than happy with the results.
Conclusion: people have different needs and go different paths.
You probably meant the same by "dedicated machine" but HAOS runs great in a VM too. I run mine on Proxmox and it's been rock solid.
Newer ikea devices are standardizing on AAA because they can be recharged. The new motion sensors, door sensors, and buttons all use AAA batteries.
Thanks for the update. The ones I have are the Remote Control N2 using AAA, and the motion sensors use CR2032.
I didn't know what it was, so: https://www.home-assistant.io
It's the Land Rover of home automation systems.
(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 :)
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.
Closest to the first. He's joking that you can't own a Land Rover without being a mechanic and you can't use Home Assistant without being a programmer.
I love Home Assistant, and except for the occasional update breaking config files it's been very reliable, but there is no way 95% of people could get it installed, let alone get it set up to do anything useful.
>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.
excepting the recent dross from them with the death of the real defender...they can usually be fixed roadside with a hammer and some grease, they're easily the most ubiquitous off-road vehicle globally (well either landrover or toyota) and with 80% of all made still running.
Not sure why your reaction was so emotional, but I think you're thinking of Range Rovers, again though, the old ones were superb, the new ones are for rappers, footballer, and the other assorted nouveau riche.
Are Land Rovers extra hackable? Will be in need of a new car in a few years and that would help o.0
I think they were referring more to Land Rovers making mechanics out of car owners (Due to their famously bad reliability), but I may have misunderstood the joke.
Definitely, I heard som many bad jokes like you need to buy 2 evoque models to be able to drive at least 1 etc. Really bad reliability despite good design.
He meant shit keeps breaking and youll be constantly fixing it
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.
Just use ZigBee or Z-wave devices and then use a bridge to connect to HA.
Anything else is just WiFi and vendor lockdown
Some of the best home automation stuff is cloud-less wifi. For example Shelly switches with Tasmota are awesome and I much prefer those over zigbee bulbs.
The WiFi and BLU Shelly devices are pretty rock solid, and couldn't be more compatible with HA.
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.
Better late than never, I have 3 of their bulbs in the house, and not being able to integrate with HA prevented me from buying more (at least before Matter was announced).
I've been using them with Google Home, so the lights weren't automated with HA. I'll try the integration out.
I have bulbs from wiz, which already have HA support.
Being new to all this home automation stuff I was quite intrigued how they worked though. They're exposed on the WiFi network over a really simple UDP based protocol which led me down a rabbit hole of writing a little go client to mess about with them, took a few evenings.
Not saying Xiaomi bulbs would be quite as simple to write an integration for, but they might have been. It's kind of fun seeing how people have reverse engineered all these custom protocols.
They sold bulbs under various brands. I’ve got some very old ones branded yeelight that work entirely offline with HA.
Tried to rebuy got a different brand and they don’t work offline properly. Bit of a crapshoot
These days I try to buy preflashed tasmota gear. Things like athom.tech
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!
Put Xiaomi devices on my network?That's a no from me dog.
I have some xiaomi bulbs that used to work until I needed to re-pair them to a new wifi and (unofficial) integration kept asking me for pairing code - which is nowhere to be found. I probably got rid of the boxes where it's supposed to be located. Wonder if there's a workaround for that.
I think WiFi IoT stuff has a reputation for being janky pretty much across the board, the recommendation is usually to use Zigbee stuff instead wherever possible. That also has the security advantage of keeping the devices on their own little network that can't access your LAN or the internet directly.
I've gone with Zwave instead - older house and hoping to keep wifi bands uncongested. It's about $30 per device, with Zooz as a big brand. It's been quite solid and the spousal-acceptance-factor is high.
Yeah - I bought them before I got into this, so thought it'd be nice to make use of them
Defaulting to zigbee for anything new
Not familiar with xiaomi bulbs, do they have QR codes on the base?
A few years ago Xiaomi started enforcing region locks to reduce gray market sales, so a bunch of products stopped pairing when app set to wrong region, without indication why. I had to repair my vacuum, air purifier, security cameras on profile with PRC region, and other stuff on profile with Singapore region. Very annoying.
There is python code floating around in github somewhere to grab the codes
Does that includes Qingping air monitors?
Dont those already work via Bluetooth?
My Air Monitor Lite at least do.
Sidenote: I just installed Home Assistant this weekend for the first time. Insanely polished piece of software!
Got my hopes up but it's a start. Was hoping to see more local support as opposed to Xiaomi cloud
What is MIoT that is mentioned? I know IoT is internet of things.
Meaningless mashup of Xiaomi's "Mi" product branding and "IoT" that they use to refer to arbitrary aspects of their IoT products
Some sort of auto pairing tech with their own routers I believe
It stands for Mquantum-blockchain-self-driving-social-cloud Internet of Things.
Why do people care about Xiaomi specifically? Tons of stuff is supported by HA.
I mean....I guess this is better than not having it, but I'm not personally interested in cloud-only smart integrations. Cloud option? Fantastic! Cloud-only? No thank you.
I don't believe it's completely cloud only, "local control" https://github.com/XiaoMi/ha_xiaomi_home#:~:text=Home%20Inte... sounds quite a bit like non-cloud control?
Could still be behind cloud authentication, though.
Read a little further down:
> Xiaomi central hub gateway is only available in mainland China. In other regions, it is not available.
Nonstarter for many, myself included.
ETA: Yes it does say "partial" local control can be done without the gateway software, but is not recommended and does not work unless you are on the same local network. Better than nothing, but still a nonstarter.