Native Home Assistant Integration for Wyze Products

Services & Integrations
21

Honestly Wyze is just taking advantage that others (secKatie and Mrtl8) have developed HA integrations/add-on to fill the gap Wyze left. Without their work, I think the brand would have been ditched by a lot of iot afficionados a long time ago. I cannot thank enough SecKatie, Mrtl8, and all the others who managed to keep DWB alive recently, but these integrations still depend on internet. Now it’s time for Wyze to pickup the job and develop it into a local control.

While seeing RTSP back in beta testing is interesting, I didn’t get into it yet. Only a local stream without ability to control the cameras’ functions (especially pan cams, what’s the point of pan cam if we can’t control the pan?) has no interest to me unfortunately. Not to mention all the other wyze devices

22

My thinking is that Wyze doesn’t seem to be aware of the exponential growth of Home Assistant or the weight of an actual “Works with Home Assistant Badge” might mean to new hardware sales.

I think many people would welcome Wyze hardware in Home Assistant, Hubitat etc. If they work fully, not a little bit.

Better, more expensive Wyze cameras? I’d expect to pay more. Still pay a cloud access? Definitely needed for my OGs (V5s) lol, if OGs are not-included in RTSP.

In comparison, and almost all camera brands are made in China, Reolink has a full integration for HAOS. Reolink & Aqara are fully China produced far as I know.

Claims of USA Reo only servers. Don’t know, I don’t use their cloud and won’t be using it.

Reolink doesn’t offer nearly the detection sensors Wyze does. in my test camera model anyhow, I like that Wyze has so many, Like T3 smoke & co alarm sounds listening etc.

These are my personal thoughts only, they have to live with their final business decisions, and are starting RTSP again so very late in the game.

Maybe almost too late for me, TBD. I’m getting up there in age. Waiting a real long time is not in the picture.

I have a Pan V3 RTSP feed via Beta RTSP in Home Assistant. But for me it’s rather useless without detection sensors and operations. I only did one of three Pan V3s, why bother other than testing on one?

I can get RTSP only from Aqara, cheap.

The Reolink test camera I mounted exposes, all detection sensors local in the camera firmware, person, animal and vehicle, does RTSP, Onvif etc. etc. and all local.

Though they offer cloud services also that’s supposedly USA based, :thinking: The main local detections is all I currently need.

I don’t need a Frigate configuration headache, or a Coral TPU to use it. Runs right in the HAOS Platinum rated integration, only uses a couple % of cpu per camera. I could mount many.

4k, runs fine outside through brick walls on 5ghz WiFi.

Dislike the Reo cables and power bricks, odd that choice, L shaped brick and barrel connectors, :thinking: lol, smh.

No other dislikes. The test camera has not made a single false detection of a person trigger since proper setup adjustment and install. 3 months now. It can pick up vehicles at 50 yards and is not a zoom camera.

I’m not plugging Reolink by any means. Just stating obvious differences. Just the likable options that can work for me if necessary. (pressure) lol, like one customers opinion matters.

If Wyze gets there, very soon, exposes sensors, comes up with an HAOS integration or lets Seckatie expand the existing one, that is their call. (API)

I have some 15 Wyze cams, I doubt OGs will be included in RTSP but that’s unknown to me.

I can replace them with V4 or V5 :wink:, regardless I’ll have some Wyze stuck in every window in a window mount.

I’m old, can’t wait very long. So unless Wyze moves fast, I may have to self exclude myself, excepting Wyze window mounted cameras. their little square design beats every one I’ve looked at.

I like my Wyze cams, OGs, V4s, Pan V3s, all have worked fine on my Mesh WiFi. I hope I can continue Wyze cameras purchases as needed, when needed that, Fully, Work With Home Assistant.

23

They actually gave a lot of explanation on that. They told us that they were not able to keep up on multiple firmware branches so they couldn’t guarantee the security on the rtsp Branch, but other people were welcome to build on it if they wanted to.

Also, they never actually removed the beta firmware. They delisted it, but you could still download it directly from their website.

Open this drop down and go to the second to last section of it to see the direct links to the original Wyze rtsp firmware (not the new one in beta) and get it from them directly on their website, Or the last section has the firmware from the wayback machine Internet archive

Common 3rd Party Solutions to get RTSP on Wyze Cameras:

For those who want all of the original and now deprecated Wyze RTSP firmware directly from Wyze, you can still get them here:

[Insecure] RTSP FIRMWARE [missing critical updates] Directly from Wyze:

If needed, here they are forever saved on the internet through the wayback machine:

The problem is that if you remember the CVE issues that came out, the original rtsp firmware also had those cve vulnerabilities in them. Wyze updated their production firmware Branch and delisted the rtsp branch because it wasn’t updated. Personally, I still consider the rtsp Branch fairly safe because the CVE’s basically required a person to be on your local network in order to execute any of them. They weren’t real risks from some remote access. Wyze still allowed people to Make their own choices and continue to use the rtsp firmware if they wanted, they were no longer going to continue to advertise it or imply that it is kept up with updates and patches when it is not. The difference now is that instead of creating a separate firmware Branch, they are going to integrate rtsp into the public production firmware. So now we don’t have to choose one branch or the other and it will continue to be updated indefinitely. This is a big reason many of us are pretty excited about it.

Anyway, point is that they did tell us, and it was a reasonable decision, just a really frustrating one. But having a separate firmware Branch was also frustrating for those of us who want to use their public services and also have rtsp. So the new solution is something many of us have long been hoping for.

Agreed. I think part of the struggle is that they started doing this many years ago and have a lot of design debt to overcome now. It will take some transitioning. I will settle for incremental improvements over time.

I’ve met all of the owners in person, and I find that they’re actually all really awesome people. I have really high respect for all of them. Based on what they have said over and over again, I think it’s less that they care about cloud, than that they care about doing what will benefit the largest number of their users Or at least the largest number of their users would actually care about and use. They have done a lot of analytics and found that it is a very small fraction of their user base that cares about local only connectivity or home assistant or any other integration. It’s not that they are opposed to things that are highly desirable by a small fraction of people, but rather that if they have a limited budget of finances and developer time, etc. they need to prioritize what they put their efforts toward, and logically, it makes sense to prioritize the projects that are going to be used by the greatest number of people.

That’s an oversimplified explanation. Their wish list is determined by a lot more than just the largest number of votes or just how many people will actually use the feature, or how much will it cost, how long will it take chronologically and man hours, ongoing maintenance costs, foundational prerequisites, design debt considerations, and many other things. The roadmap and wish list have at least a couple of dozen considerations for what gets prioritized next. I actually talked to a lot of employees including some on the leadership team that told me all about how they integrated their Wyze stuff i into home assistant and homekit, etc and use that as their primary access. Several of them are highly supportive of doing a full integration with home assistant and really want it to happen at some point. Some told me this in person, and over the last couple of years a couple have even direct messaged me about it. I think it will happen in good time. But there is a lot involved.

I will definitely continue to push and advocate for it. The more the better. :+1:

Wyze adding support for RTSP is enough to hold me off from buying some Reolink cameras on Black Friday like I’d been intending to do.

I think in the long run we will have to wait until the matter camera libraries launch by the end of the year, and then see what Wyze decides to do. I bet sometime within the next couple of years they will either be widely supporting matter or home assistant or both. Failure to make efforts that way could be devastating to keep up with the competition.

1 Like
24

Thanks for your insight on this. When I said owners, I was referring to customers, not Wyze. I do believe Wyze is a cool bunch of wonderful people. My point was that I DO recognize local control & Home Assistant is a minority niche request and, understandably, Wyze needs to prioritize their bread and butter, low-cost cameras with subscription service. In fact, I pay yearly for their subscription, because I believe in Wyze and have probably purchased around 25-30 cameras.

However, as stated above I want to see Wyze (& every IoT company) change so that smart-home devices and cameras are purpose-built to function locally (without the cloud), and the cloud being value-added service. This helps ensure local service during cloud outages & reduces e-waste when products are discontinued or companies end. …Just yesterday, I read that iRobot is likely going out of business, which will likely kill off the iconic Roomba. So many tech companies pretend to be “going green” and caring for the environment, but their business model of being cloud-dependent needlessly contributes to tech waste of perfectly functional products.

What I also don’t understand is how unpaid volunteer developers can create Home Assistant integrations, but Wyze can’t? Maybe offer some official resources to those willing to do the heavy lifting, to minimize extra work Wyze has to put into Home Assistant? Maybe Wyze is just holding out for Matter, so instead of worrying about support for Home Assistant, Apple Homekit, Smarthings, Google Home, etc, they can only worry about conforming to the Matter protocol standard?

1 Like
25

I don’t actually think official Home Assistant support would help increase Wyze business much, otherwise they would have already implemented it. I do want to companies to design smart-home products local-first, with the cloud as a value-added service. This helps respect individuals privacy and control of their own home, while allowing businesses to profit from providing additional features and convenience.

Side-note, I would love to see a practice of discontinued smart-products having all code open-sourced, so independent dev’s can breath new life and help minimize e-waste.

3 Likes
26

My bad! That is definitely a big misunderstanding on my part. Lol.

Firstly, I want to say that I overwhelmingly agree with what you posted in your reply to me. :slight_smile:

Fyi, related to Roomba, there are actually a lot of projects out there to convert many of their models through firmware flashing to be open source local only devices. But that is above the skill level of the majority of users. Most people will just throw them away. They should probably try to sell the devices because there are a good number of hobbyists who would be willing to purchase them and flash them.

Having said that, everything you said was right. That devices should still be designed to work entirely locally if needed and then just have cloud as a bonus value. I am extremely supportive of this standard. That should be the primary business model for all IoT companies, and frankly, they should all be moving toward this standard And the market should begin to reward those who do this.

Some past devs have commented on this and explained that before they give too much access to their API they would need to invest a lot of work in securing the API and then it would also take a commitment of ongoing maintenance for which it would be mostly a liability without an obvious direct connection to revenue to offset the commitment. Their compromise was to allow unofficial access through the cloud and allowing people to do limited reasonable amounts of polling. This is basically what all of the existing projects do. The problem is that polling can quickly overwhelm their server and bandwidth. A few years ago, something like a single digit percentage of their user base using the home assistant integration was tying up well over 50% of their server resources from over-polling things like sensor States and camera states, etc. It was unsustainable. Wyze instigated a rate limit which quickly locked up a large number of users from being able to use their Wyze devices. There was a negotiation between Wyze and the Wyze home assistant integration Dev, and Wyze agreed to compromise to remove the rate limit if the dev promised to at least remove support for sensors and Camera as a sensor which we’re taking up the majority of the polling resources. This would still allow people to get a lot of benefit from the integration without killing off all their operational expenses with no revenue to compensate, making it unsustainable for them. It really sucked to lose support for so many of the things we all liked about the integration, but I did kind of understand where they were coming from. But what they need to do is update the integration so that it doesn’t need to use polling anymore and can simply work the same way the app and Google and Alexa integrations work to where they get a direct push or update from wyze when something changes instead of forcing the devs to do constant polling. Than it wouldn’t be as taxing on their server and would be part of the same update that gets sent out to all the others anyway at the same time.

Yes, some of the employees have actually made almost exactly these comments. The hope hope is that instead of trying to support a dozen different integrations and having to do continual maintenance on individual integrations that will cause them a lot of extra work and maintenance forever with liability losses, it is possible that they may be able to do something like just support matter, and that will automatically give their platform access and support to all the other major platforms. With just supporting matter, than in theory, they would have direct local integration with Apple Homekit, Google, Alexa, home assistant, Samsung SmartThings, hubitat, And countless others all at the same time without having to waste all the man hours, costs, and all the other issues related to custom individual integrations. As soon as the matter standard was announced and starting to be worked on, they updated The wish list for basically every other integration to be a status of “probably not” with the only exception being home assistant, which got a status of “maybe later” instead of probably not. So, in addition to some of the comments from devs about this topic of not wanting to do individual integrations and lifetime commitment to maintenance on those individual custom integrations, their hope was that it might be possible to do something like a single integration with matter and get people all the support they want for everything else. But even with that, they still didn’t rule out having a separate direct custom integration with home assistant and left that one to be the only integration marked as maybe later instead of probably not. That’s actually reassuring to me.

But, I don’t blame them for wanting to wait until matter is fully released with all the camera libraries and capabilities so they can make an actual educated decision about whether that is going to be sufficient and worth their time. It’s hard to commit to something that doesn’t even exist yet and that you don’t know what all the details are. But it makes a lot of sense! That they feel they would have been wasting their time doing individual integrations and causing themselves a lot of problems in the long run. They already learned their lesson about design debt with their app. If they would have been building custom integration beset for other platforms over the last couple of years, and then matter launches, a lot of the time, money and resources would have felt like a waste if matter covers all the needs for all of those integrations which could be leaving extra vulnerabilities or other concerns when they are deprecated in favor of matter later.

IDK, has a business owner myself, I think they’ve made some reasonable decisions in this area to hold off committing to more integrations until matter releases the full details on the devices that make up Wyze’s primary product line. My frustration over the delays has mostly been with the matter consortium politicking and causing delays for the cameras.

Anyway, I very much agreed with what you wrote in your reply to me and I’m very supportive of all of that. :+1: I’m very much looking forward to the matter camera libraries being released by the end of the year (according to their estimated launch). Then I think we will have to wait for Wyze to analyze it and work whatever they conclude into their roadmap timeline, if anything. I will for sure be asking them a lot of questions about it in their next AMA once the matter, camera libraries are finally launched. There will no longer be a reasonable excuse to delay coming to a decision and working towards something in my opinion.

Obviously I don’t speak for, work for represent wyze in any way. I can only do my best to represent the desires of rest of the user community when I get chances to talk to some of the employees. But rest assured I will definitely be pushing them Hard to consider and support matter and home assistant.