Been looking on Xiaomi, Roborock and Roomba vacuums, and possibilities of HA integrations, and found out that there are robot vacuums made by ETA and Tesla Smart (different from car manufacturer).
Most of the vacuums look pretty similar when checking the apps. Map, set of commands, wifi 2.4ghz…
What are the chances these will work with HA Mqtt integrations?
Been looking mostly on ETA Tiger.
Edit: Not sure if it might help, the ETA tiger seems to use as a platform Smart Life App and its shared with Nedis and its Tuya based.
Just because the apps show similar (pretty standard) features on the outside, doesn’t mean that the underlying code is the same as the currently supported vacuums.
The vacuum’s software will need some kind of MQTT client installed on it to be able to communicate via the MQTT integration.
Lesser known brands have a smaller chance of getting an integration made for them, simply because people with the right technical skills to develop an integration don’t own them or have never heard of them.
That’s not to say that there will never be an integration for the ETA vacuum you’re interested in. However, it will probably be a custom integration in HACS, provided someone is interested enough in getting one of these vacuums and writing an integration for it.
I suggest you stick to the more popular brands if you require something with a functional integration now.
Thanks. I went for ETA Tiger simply because its a device apparently manufactured in EU, while the manufacturer also claims the service is maintained in EU as well.
Main reasons to jailbreak Xiaomi or Roborock is here not present.
Went throught the documentation for Valetudo and MQTT. Not supported as well.
I guess you are worried more about “Chinese cloud” than “cloud” (which is just an euphemism for “someone else’s computer”), but from my point of view the main problem(s) still exists.
They could shut down the server/app at any time and you will be left with a (very expensive!) paperweight.
Not to mention possibly sending video, photos, map of your house to their servers (and doing who knows what else with it).
But average muppet seems not to care about privacy nor local control any longer (some exceptions in this community, of course) and this is why we can’t have nice things.
IMHO you are going about this backwards. But then again I will only buy a device with local control. Therefore I limited my selection not just to a device which was supported in Valetudo, but rather recommended.
But then again I have long experience with many Single Board Computers (SBC) other than RPi and I know well how difficult it can be to get any kind of decent support on “random Android / ARM” devices.
I am worried about american cloud services as well. Been extensively trained in GDPR in EU for my job, and i am more confident about sharing data within EU.
If the server/app will be unreachable for whatever reason, it appeas that the ETA Tiger is able to do its job on automatic mode. two buttons - On (auto) and go back to dock.
I would most likely not buy a device which does not have a manual override or when the most basic function will not be maintained offline.
OFC i may lose some functionality, and local MQTT and FE would be a great asset.
If it is anything like mine that auto button will just start cleaning the whole entire house, which is not as useful as it may sound.
Before I started using one of these things, I thought that we would be able to automate the cleaning every day (or whatever, depending on the room).
What I learned in practice however is that it is still quite a manual process (again, depending on the room). For example, in the dining room we need to move the chairs out of the way first. And so on. So we essentially clean on a room by room basis. Currently we are using the Valetudo web UI for this (from our phones, as another web page in the HA app), but I am working on putting 4-button keypads in every room so we can clean a particular room at the press of a button (along with controlling lights, media, etc.).
At any rate, no matter what sort of automations/control/UI you have in mind, it will almost certainly be easier to implement via something like Valetudo which exposes pretty much all functionality of the device (vacuum and/or mop, how many pass(es), which room(s), error messages like low water levels, or robot is stuck, etc.) to Home Assistant. Again, this is dependent on the specific model! Which is why I bought a recommended model. Because that’s probably the one the developer(s) use in their own home(s).
Small flat i my case, but I understand it might be an issue for different situation.
Anyway, I got some news.
The vacuum cleaner is Tuya based. There are “Smart Life” branded applications for other appliance manufacturers, and “Eta Smart Life” is just ETA branded Tuya application.
Account created in ETA Smart life is different from base Tuya Smart life, however HA integration is cloud based.
I am aware, but there is an integration which allows to control the device locally + the ETA Smart application isnt using its servers within borders of EU.
The point is, you could just as well by Roborock or something else. I did.
Either get over it being Chinese at its heart and accept it might phone home, or go with real, full European (and thus expensive) brands. The Tuya devices contain Tuya software, so you cannot assume anything about what it does and does not do.
There is a third way, purchase a unit that is supported by Valetudo.
Valetudo replaces the “cloud” component with local control. In fact that is the raison d’être for the project…
Anyway I am not sure “European” necessarily equates to “expensive” (and conversely, “made in China” to “inexpensive”) as it appears more to depend on the features and quality of the unit in question and where it sits in the product lines of the company.
Maybe I got lucky (bought mine some time ago). But I always wondered how long they can keep playing this cat and mouse game with the manufacturers.
I watched some of Dennis Geise’s presentations (which I found fascinating). It seems that every year they disclose new vulnerabilities which the manufacturers then patch. He is just doing it for funsies it appears but in the meantime we have local control of the robots. But the whole business seems quite unsustainable to me in the long term, unfortunately.
EDIT:
They only ever supported a handful of models in the first place, and only a couple recommended ones.
Decided to look at current recommended models, they are still recommending the Dreame L10S Pro Ultra (same as I have) which just so happens to be on a Lightning Deal for $499 (at least in Amazon in US, I followed affiliate link from recommended models) for the next 5.5 hours or so.
I can say it is a very nice machine. If I had the cash I would pop on another one just to have a spare in a heartbeat at that price. I bought on a Black Friday sale and they were not even that cheap.
Just went through the ETA Smart Life application and Tuya app. ETA vacuum can be enrolled in both, in exactly same manner. However only if the vacuum is enrolled in Tuya app, it can be integrated with HA.
Even when it uses the same software platform, it looks like ETA established their own cloud instance, separated from Tuya IOT, and it cannot be linked to the Tuya HA integration (user ID and QR code scan does not work).
I can either went through enrollment on Tuya Smart app, and then link the vacuum robot to the Home Assistant, or try LocalTuya. In this case it will work in HA, without the map, slightly less dependant on the cloud.
LocalTuya will see the device on local network even when paired with ETA Smart Life, but I wonder if there is a way how to obtain local key for it.
The robot has Chinese libraries in its software, and the European Cloud has Chinese Tuya libraries too. Even more likely the Robot is fully Tuya software and its Cloud is a Europese Tuya run server, because Tuya has Europese datacenters. The manufacturer app is just a different outside layer. You proved that because it is 100% compatible. Do you honestly think they wrote any of the base stuff all over again, to do exactly the same as Tuya’s? Tuya probably wrote the manufacturers app too.
I don’t like it, but I accept the fact that the risk is still there that the data ends up in China. I’m not saying you should not buy their things. I think the problem is overstated. The US is equally bad and many accept that. But you keep trying to explain it away, which you can’t. Until you block the robot from internet or completely replace its firmware. So if it is important to you, don’t be fooled.
Ironically, you could have avoided all that hassle if you had gone with the Roborock.
Despite being a Xiaomi (Chinese) brand, their vacuums operate with the standalone Roborock app, which is at least certified by TÜV Rheinland for EU IoT standards.
Even better, the HA Integration is local-first, which meant that it won’t reach out to the cloud unless it has to.
If you’re still in time, I suggest you return the vacuum and get a Roborock.
The ETA EU backend is different instance from Tuya EU, its not just different “outside layer” / frontend. Meaning that when i create account in ETA app, it wont work in Tuya app.
The technology behind it is the same, no argue about that.
