Skip to content(if available)orjump to list(if available)

Matter – Protocol to connect compatible devices and systems with one another


Comments moved to

Edit: I've been told that this article is the better one so am going to merge the thread hither.


So you went hence thither and then thence hither again whence you came.


My biggest issue with Thread is (AFAICT) it gives each device access to the internet. That's the absolute last thing I want and the reason I went 100% Z-wave (with a handful of legacy zigbee devices). I want a single point of local hardware that my devices talk to and that hub can talk to the internet as needed but I don't want to have to vet every device or deal with companies going out of business.

As it stands now I don't care at all who made my Z-wave switches/bulbs/plugs, they could already be out of business for all I care, that's not the case with 90% of "smart home" shit I see online. "No hub" means "absolutely not" for me because that's a wifi device with full access to your network (unless you segment) and the internet.


Yeah. I was very excited about this until i got to "Bluetooth Low Energy for provisioning" and I was like "oh.." and then "Wi-Fi for high-data rate connectivity" was a "oh. no thanks." That's a clear path to the internet and also using multiple radios so that if I lock down one thing I have to constantly worry about others.

As it stands I dont see this being a thing that benefits me. It is more likely that this pushes smart home manufacturers away from Zwave, and makes my life worse because now I can't buy IoT devices without granting them internet access. Then I have to go through and start setting up IoT VLANs to quarantine devices, and in some really shitty cases, Ive seen WiFi devices where the app sends a message to a server, and the server sends a message to the device... so I can't quarantine the device without modifying it or losing functionality (see MyQ garage door openers). The industry looks bleak.

Agree on all points of preferring Z-wave though and it's also what I use. However, Linus (of Linus Tech Tips) recently redid all the switches in his home with Zwave switches... and they had a firmware issue... And the company refused to publish firmware updates in any way except through "official" Zwave hubs, and not Home Assistant. It eventually got sorted out, but Zwave isn't necessarily intrinsically failproof either.


> Agree on all points of preferring Z-wave though and it's also what I use. However, Linus (of Linus Tech Tips) recently redid all the switches in his home with Zwave switches... and they had a firmware issue... And the company refused to publish firmware updates in any way except through "official" Zwave hubs, and not Home Assistant. It eventually got sorted out, but Zwave isn't necessarily intrinsically failproof either.

Yes, I followed that drama as well and the did eventually make the newer firmware available to HA and similar though that doesn't exactly negate anything I said. For new features/fixes you are dependent on the manufacturer but that's the same everywhere. The big difference is that if Linus' switch company went out of business then his switches would continue to work the same way they always did (bugs and all). But I do agree with you Z-wave isn't perfect and does sometimes require a little more knowledge to work with.

Thankfully I'm all set currently (aside from the SmartThings rug-pull that I have to deal with and finally move everything remaining to HA instead of using my ST hub as just a controller from HA) and I'm just going to sit tight over the next few years and see how things shake out before I upgrade/replace any of my gear.


I was just referring to your comment:

> As it stands now I don't care at all who made my Z-wave switches/bulbs/plugs, they could already be out of business for all I care

If the company was out of business and there was no way to get the newer firmware, LTT might have had issues. Also, buying a switch because it is advertised as having feature X, only to find out that feature X was added in a newer firmware than you have, and you can't get that firmware. That means it isn't just a "I dont care at all who made my product" scenario. It's much lower risk (especially if the feature you want is basic functionality rather than automation), but it's not 0 risk.

I agree with you that Z-wave is going to be 100% better than WiFi devices that call home and join a botnet, but I wanted to make sure it was clear that there is still room for issues.




WiFi is only for devices that need it, like I'm thinking a "matter"-provisioned wireless security camera.

It definitely confuses the user story though, idk why they included wifi at all.


Based on my IoT devices I already have, any device that supports Matter will probably use Wi-Fi to ping back home whenever they can.


I suspect a big part is to allow bridges like the new IKEA hub. Lots of folks put those in wifi.


> Zwave isn't necessarily intrinsically failproof either.

The one nice thing about it, above Zigbee, is that you can expect devices to work together. With Zigbee you frequently have to pair a device with a hub from the same manufacturer.


I have a lot of Zigbee devices, and I've never experienced a pairing issue like this. I use Zigbee2MQTT with Home Assistant, with a electrolama ZZH USB stick -

During my first attempt at home automation, I used to mess around with flashing custom firmware onto WiFi devices (Tasmota, ESPHome.) It was a huge pain whenever the hacking process stopped working (tuya-convert), and I had to open up the device, solder some wires, and flash the chip via USB serial.

Nowadays I can just buy any Zigbee product I want (usually from AliExpress), and get it set up in zigbee2mqtt without any extra steps. And it's all local within my home, and no hubs or cloud services required.

I think Zigbee is better for me because I have a lot of battery-powered devices, such as sensors for motion, door contacts, and temperature/humidity. I'm really impressed that the button batteries usually last for 1-2 years.


Although you can easily circumvent this with a cheap Zigbee USB controller, Raspberry Pi (or other computer) and something like Zigbee2MQTT. My IKEA bulbs, Xiaomi sensors and Philips remotes all work great as a mesh.


You would think so...

On the other hand, Zwave radio spectrum varies depending on region, so you need to be very careful that you don't spend hours troubleshooting something that doesn't work because it's a device intended for the EU market instead of US.

IIRC, Zigbee is better here.


I have more zigbee than i do zwave, never had a problem with either pairing to my Nortek USB Hub


The whole idea of Matter is local control over IPv6. From what I understand, it is basically Zigbee's cluster library re-platformed on top of new encrypted mesh layer that runs on IPV6. This does technically mean that all such devices can send out arbitrary IPv6 packets, and for Ethernet and Wifi can even send out ipv4.

When operating via thread, those ipv6 packets are just using it as low powered wifi alternative for those packets. Both Wifi and Thread connections thus require two levels of joining. One for connection to the AP/thread mesh network, and one for the matter encrypted network. Ethernet based devices on the other hand, can simply join directly to the matter encrypted network.

BLE for joining can simplify this, as a controller can provide the needed info about thread or wifi, as well as the info needed to join the matter network.


Now I very much support local control. Most of my smart home devices right now are either Zigbee or Z-Wave. I have 9 z-wave devices (plus the home-assistant controller), and 4 Zigbee ones at the moment, although only 2 are in use. (These current ones will never upgeade to support matter over thread). I am waiting on he new inovelli switches which will initially run on zigbee, but I will probably later upgrade them to run on matter. (After I buy a new zigbee radio dongle that supports dual stack zigbee/thread operation, and home-assistant officially releases their support for doing that).

But those of us running home-assistant, or the like are not the real target market of Matter. For more typical users, they end up with a bunch of smart devices that are wither wifi based, or have their own hub. Each one has a seperate app. Some might be homekit compatible, some might be hue compatible (work via the hue hub and its app), some might be compatible with amazon Alexa via the cloud, some might work for Alexa via zigbee (but only if you have an echo model that supports zigbee). It is not easy for consumers to even be certain what works with what.

With Matter, every Matter device will be compatible with homekit, will be compatible with Alexa, will be compatible with google home, will be compatible with SmartThings. The only requirements are that your controller device have support for the general type of device in question, and if it is a Thread device you have at least one Thread "border router" (ip gateway, equivalent to wifi AP) in your matter network. Homepods for example include such a border router, and likely some future Amazon Echo models will too, and it has been suggested that some Wifi Routers will include that functionality too.

That is a huge win for the average consumer!


Your first paragraph describes my reaction as well. The article's overture was a roller coaster.


Where is the developer docs? The site leads to some document under solutions that needs to be downloaded to read. It is also not clear why a protocol specs or implementation should be categorized by solutions. Overall the site documentation lacks transparency.


Wikipedia article links to the key websites,

I had less luck finding Matter with DuckDuckGo. Evidently a competitive search term (which is to be expected).


    As it stands now I don't care at all who made my Z-wave switches/bulbs/plugs
This is maybe an optimistic view of Z-wave in my mind... early on quality and reliability problems were nearly universal with Z-wave devices, and today I have a fairly short list of brands I trust (e.g. Zooz) after having been burned by a series of devices with various firmware bugs or just very poor lifespans. The best story I have is the lightbulb that would turn on every time it received an explorer packet, and thus every time my hub at the time ran a heal, which was of course scheduled at 2AM every day and not easy to disable due to the hub's poor firmware. This was years ago, but then there's still a surprising number of Z-wave devices on the market that are just Z-wave, not the poorly named Z-Wave Plus or Z-Wave Plus V2, and thus have significantly inferior network reliability if there are any changes in the environment. In any modestly challenging environment (e.g. battery powered device a few walls away from the nearest powered device) I have gotten used to having to try out multiple products before finding one that was reliable, and that was even after learning not to buy anything that wasn't at least "500 series" (Silicon Labs part number for Z-Wave Plus).

Even Enbrighten, a former GE brand and so ostensibly reputable (this of course depends on how familiar you are with the GE of today) is shipping some total garbage that they haven't even bothered to develop real documentation for. Curiously, I've found that some Enbrighten products and Zooz products are physically identical to the degree that I think they are both sourcing some of their hardware from the same manufacturer... but Zooz has very noticeably superior firmware.


I've got about 20 zwave devices by all of the manufacturers you mention on my network and they have all been very reliable after a year of use. I had a lot of reliability issues at the start, but they all ended up being due to two issues:

* Signal strength. Getting more repeaters and devices fixed this * Flaky service implementation. Uprading ZWave JS on Home Assistant and adding a monitor that got ZWave JS to ping devices when they were unavailable as made the system very reliable.


Thanks for the tip about Zooz, they do look nearly identical may have to pick up a few for additional switches.

I have had no need for any additional documentation for all of the GE / Enbrighten switches I have installed in my house. Other than how to wire it and how to get it in pairing mode, what else is there?


no, thread is a rebranding (and evolution) of the zigbee protocol, which is a low-power, local mesh network with no inherent internet connectivity. zigbee/thread devices work offline first, with switches that are an integral part of the local mesh network. these devices can also be connected to the internet through a hub/border router, if you so choose. matter is the umbrella brand name for the combo of BLE, wifi, and thread that provides a complete IoT ecosystem.

i have a bunch of zigbee devices (mostly ikea tradfri) that i'm eager to upgrade in the hopes that i get more reliability and speed out of my connected devices (mostly lighting).


Why the HELL do companies and whatnot keep rebranding from something INSANELY EASY to search like Zigbee to something ENTIRELY GENERIC like thread?



Because they didn't. The only thing that Thread shares with Zigbee is the radio. It's a completely different standard.


because they do not want you searching for protocol compatiblity, they want you to lock into a vendor

they want you to look for "Alexa" or "Google Home" or "Apple HomeKit" labels, not "Zigbee" or "thread" or "zwave"


I suspect it is desirable as one is supposed to search for the marketed products. The magic smoke that binds the device to the app or blinky lights is of little importance.


At least it wasn't Nintendo in charge of the naming, then it would be New ZigBee.


Drop formerly MassDrop, Wise formerly TransferWise


Stating that Thread is a rebranding of Zigbee is incorrect. Both are based on 802.15.4 for their physical layer but the two are entirely separate and managed by different standards bodies (Zigbee under the Zigbee Alliance, now CSA - Thread under the Thread Group). Where some confusion may come is that Matter is now also under the CSA (just like Zigbee).


while you're technically correct, it's a rebranding from the (zigbee) consumer's perspective, via the CSA being a rebranding of the zigbee alliance. within that context, thread is the most direct analog of the zigbee protocol, as they're both implementations of 802.15.4 for low-power local mesh. consumers won't need to (and most won't want to) understand it deeper than that.


Matter/thread is nothing like zigbee, except the MAC layer is the same. For one, Thread uses IP(v6), and Zigbee does not use IP at all. This will make managing thread networks much more like managing normal networks (hopefully).

I don't know why you think your zigbee devices will be upgradeable to thread - that strikes me as quite unlikely, although perhaps physically possible depending on how much hardware offload the devices use.


> Matter/thread is nothing like zigbee, except the MAC layer is the same.

Matter actually uses an iteration of the Zigbee device model - from an application level they are actually somewhat similar

> I don't know why you think your zigbee devices will be upgradeable to thread - that strikes me as quite unlikely

Most Zigbee radios are dual Zigbee/Thread since both protocols are build on 802.15.4. I suspect that the limitation is going to be on manufacturers rather than hardware limitations


If all they do is connect to the hub and then the hub controls all external access then that's fine. I'm just unclear as to why there is so much talk about IPv6/cloud/TCP/UDP for something that is only talking to a local hub. I mean I totally get you can use all of that (save for "cloud") 100% locally but if that were the case I'd expect more mention of that. I'm ok with using wifi-adjacent for communication, internally but I do not all my "thread" devices to have internet access (local or external).


The idea of thread is that it is basically IPv6 over Zigbee, in the form of 6LoWPAN which is an IPv6 stack optimized for low-power devices with an addressing, discovery, and routing mechanism designed to work well on mesh networks like Zigbee.

To some extent, Thread and Matter are direct replacements for Z-wave at different levels of the stack. There are some different pros/cons between the two though. The major advantage of 6LoWPAN is that it shares a lot of the design and implementation with existing network stacks and can be carried directly over IP networks. This is expected to make more complex 6LoWPAN topologies much easier to implement (e.g. 6LoWPAN traffic can be easily forwarded over the internet by a gateway). None of this is really anything that can't be done with Z-wave, but 6LoWPAN makes it easier by having a lot of common design and implementation with ubiquitous IP stacks. Matter itself has the major advantage of being a newer and higher-level design than Z-Wave which should result in more consistent interoperability of a wider range of devices.

Z-wave will probably remain superior for battery-powered sensors into the future, because Thread doesn't allow for the extremely aggressive sleep schedules (e.g. sleep mode for 18 hours at a time between supervisions for security sensors) that Z-Wave does... although we can of course debate how wise it is to only perform supervision every 18 hours, even if UL allows it for burglar alarms.


> Bluetooth Low Energy for provisioning via a QR code, while it relies on Wi-Fi for high-data rate connectivity and Thread for low-data-rate communications.

They will only use "thread" for low data rate communications. They will all have Wi-Fi. thus the ipv6/cloud/tcp/udp talk.

I want _just_ thread out of these things. I want my devices to talk to my zwave/zigbee/thread network and to not talk to the cloud without my permission.


> My biggest issue with Thread is (AFAICT) it gives each device access to the internet.

Thread is a protocol, this is only as true as the statement "WiFi gives each device access to the internet".

It _can_ give devices access to the internet, but only if you plug your (border) router into an internet connection. (Or otherwise bridge the networks. Your border router could still have internet access itself without bridging the thread network out to the internet.)


Given that most people will use their one and only router for this, it means most of these devices will have access to the internet. Sure, you _might_ be able to prevent it if you jump through hoops (because it really is annoying to setup a second wifi network, at least in my experience) and assuming the devices don't fail if they can't phone home (and we can be sure some will). But everyone is much safer overall if the "common standard used by most people" did not use wifi.


> But everyone is much safer overall if the "common standard used by most people" did not use wifi.

To be clear, Thread is not Wifi.


> But everyone is much safer overall if the "common standard used by most people" did not use wifi.

What wireless protocol has the range to cover an entire house or lot? Bluetooth is sketchy beyond the same room, and I don't want to deploy a dozen hubs just to cover all my stuff...


> It _can_ give devices access to the internet

My, maybe not-so-naive, assumption is that manufacturers will require an internet connection for some/all functionality because they can.


That's already the case right now today.


In the 80s my dad controlled the lights in our home using the Radio Shack plug 'n power system. Plug-in switches and the control box would talk to each other by sending signals through the home electrical wiring.

There was no need for wifi or Bluetooth or rc or anything funky like that. It was about as dumb as a smart home could get.


Good ole' X10 that later became Insteon. I actually have Insteon for all the lights and fans in my house and am quite happy with it's performance. I've been considering a migration, but I'm holding out until I see how Matter/Thread do as an ecosystem first.


I miss Insteon. I know it's kinda-sorta-maybe-alive again but we'll see.

Insteon was simple and mostly just worked. Even my wife saw the benefit of the thing.


Being able to work without internet access is a requirement of the standard AFAICT.


>My biggest issue with Thread is (AFAICT) it gives each device access to the internet.

Uh what? This is definitely not the case with my nanoleaf bulbs..


Thread is an IPv6 bearing mesh network; and if you have a border router that has the appropriate IP routing rules, and your firewall allows it, your Nanoleaf bulbs can most certainly reach out to the internet (granted, they only have the concept of IPv6, so a majority of the internet is inaccessible if only because it is IPv4)


"If you allow the devices to access the internet they can access the internet" isn't really as doom and gloom as the parent is suggesting.


Any ideas if the HomePod Mini (probably the most popular Thread border router) grant broader internet access to Thread devices on the mesh network? I can't find any clear info about how this might work or how to test.


My understanding is it gives them an IPv6 address, but that does not mean they have internet access unless there is a hub that connects between the two.

There's no WiFi, they use Zigbee behind the scenes, what changed is that they are addressed via IP, but it's still a strictly local network.


Thread and Zigbee are entirely distinct protocols; Matter does not use Zigbee in any way, at any layer of the stack, with the exception of some influence on the high-layer device/data modeling


> There's no WiFi

> Matter uses Bluetooth Low Energy for provisioning via a QR code, while it relies on Wi-Fi for high-data rate connectivity and Thread for low-data-rate communications.

So you say no WiFi but the post specific says WiFi. And if there's wifi then I have to set up my network to ensure they don't have internet access, thats not something Im going to leave up to the honor system.


The WiFi is in the hub. You should read it more carefully.



It’s 900 pages long, I don’t have a good point of reference so I’ll ask. Is this a small or large reference doc compared to alternatives?

Or phrased another way. Will no name, cheap devices ever properly implement this spec?


They will because there is an open source reference implementation that they can use:


For reference, the IEEE 802.15.4 spec is ~800 pages long. 900 pages does sound like a lot considering that Matter (AFAIK?) doesn't directly spec any hardware or transport details - those being covered in 802.15.4 and Thread.

Granted, we should remember that those 900 pages include base details that, probably, CSA are not planning to change in the foreseeable future. They need to be very thorough.

To answer your real question: device manufacturers will likely use the Matter SDK. It would be a huge undertaking for a smart-light manufacturer to re-write all of that code from scratch!


I haven’t read the spec, but I believe that some backwards compatibility is built in, and there’s a degree of complexity involved in making a robust framework that isn’t general purpose (like Wifi), but instead offers compartmentalisation of different device types and use cases.

I agree with the manufacturer concerns, but many were in a bit of limbo while Matter and Thread languished in draft RFC hell. The spec was always ‘coming by the end of the year’, and obviously impacted by the pandemic.

At least now, there is some certainty and a path forward.


What I want it the IP level architecture and data interchange formats, etc, not the entirety of the low level spec or hardware/radio specs.


Look at the bluetooth spec it’s several thousand pages.


> Matter lets devices communicate with bridges and controllers locally, which means that your smart home will still work when the internet goes down, and some devices may still have basic functionality even if they lose their cloud connection because the device maker goes out of business.

While I'm glad perfect did not become the enemy of the good and we actually have a standard now, I hope the consortium doesn't stop here, and keeps moving toward more vendor-agnostic, consumer-friendly standards.

> Originally, Matter was supposed to handle enough elements of provisioning and functionality so users wouldn’t have to download an app. In most cases, users will still need to do so.

While I also hope Matter doesn't turn into USB-IF ("Matter 3.1 Gen 2 SuperSpeed"), some clear evolutions or optional certifications (especially something for long-term, in-case-they-fold functionality) would be good to see them working toward.


A co-worker and I were just talking about Matter yesterday. Underneath the hood this runs on a networking protocol called "Thread"[1], which itself uses IEEE 802.15.4. This is pretty similar to ZigBee. Thread runs on 2.4ghz.

Thread uses ipv6 and UDP (TCP optional), so it should integrate well with existing network infrastructure.



> Underneath the hood this runs on a networking protocol called "Thread"

This is correct, but potentially leaves out a level of abstraction [1]. Matter is designed to abstract away multiple networking protocols, including Thread and Wi-Fi.

Philips Hue, for example, plans to support Matter but not Thread [2].


[1] A bit like saying "HTTP runs on TCP". It usually does, but it's not necessary to the standard.



> The Hue Bridge is not just a Wi-Fi to Zigbee translator. It has never been. It’s also the local intelligence of the system. It coordinates that your switches work well with the lights, it is what makes sure that the network is well managed. Entertainment streams, cloud services, automations, schedules … all that and more is managed by the bridge. Such a small, reliable “edge computer“ in the home is essential. matter does not offer a solution for it. Building all the functionalities directly into lamps would make things much more complicated, and we don’t want to move them to the cloud. That is, even if we used Thread, we would still need a bridge.

Second: Mesh networking with Zigbee in an efficient and reliable way is really hard. We have been working on getting that to work well for a decade. And Thread at scale in the complexity of users homes is not yet proven. I’m a bit afraid of how open Thread will be. There will be many, many companies with many different implementations in the same network. So, I see a lot of risk, and we don’t have plans to build Thread light bulbs. I don’t say never, but we have very much a “wait and see“ approach towards it.

Last not least: We have a very happy and large installed base. A key benefit of Philips Hue is future-proof and always up to date. Many people have made large investments in a Hue system as part of their home infrastructure, not as a gadget. And they expect that these products remain relevant for at least ten years. So far, we have never asked consumers to buy new lights.


What's hue's deal with not supporting thread? Seems far superior to their current setup.


You can always read what the GP linked. It details why


And HTTP/3 doesn’t run on TCP.


I was curious how Thread compares with ZigBee in power usage.

One white paper I found suggests that Thread uses slightly less power. E.g., for a device on a CR2032 battery sending a packet every minute, with ZigBee they estimate the battery lasting 1.38 years vs 1.49 years with Thread.

But another white paper at has them the other way around, by about the same amount.

Either way I guess it's close.


I’m always a bit confused what does thread add on top of 6LoWPAN?


A lot. Mesh routing, security, onboarding, and more. 6LoWPAN pretty much only describes how to compress IPv6 headers into 802.15.5 packet sizes


Is Mesh routing RPL or something else?




I'll be closely following the development of matter, especially when it comes to compatible devices. Cloud dependency of many devices for the most basic functionality offered is an absolute dealbreaker for me. While I doubt that there will be meaningful change when it comes to the app mess retrofit smart home systems tend to create, I really do think matter will offer significant improvements. Knowing that I will still be able to dim my smart lights even if company X goes bankrupt makes is slightly easier to justify spending the premium on smart bulbs.

Also, I really hope that having one universal application level protocol for a variety of devices will further improve the home assistant experience. While the device support is great for many big brands, smaller brands (especially those outside the US) are sometimes lacking integrations.


What I don't like about Matter is that the participants in the standard itself don't seem to be proceeding with the ideal of having an abstract, unified control device. From a further link [1] from the posted article:

> Consumers will still need a lot of apps: One of the initial promises of Matter was that consumers would be able to add a device — like Amazon’s Echo — to their smart home controller but wouldn’t have to download a special app for every outlet or light switch they bring into the home. But at launch, and likely for a couple of years as the standard gets more robust, consumers will still need apps for anything beyond the basics, including installation. Even my panelists realized that this was the case.

I hope these companies realize that they are basically excluding themselves from a lot of potential customers by pulling this crap. I would have put smart bulbs as upscale stocking stuffers every holiday season if it wasn't for the fact that I have no idea what product / version / "ecosystem" the potential recipients use. "Just install a new app" is not a reasonable solution.




Give it some time. Since the software for devices is all available as open source, I expect a lot of commodity products (lights, switches) to be made available as pure Matter, without any apps, because it reduces the software development costs. The Matter early adopter companies are all companies with more expensive product lines.


Right - there was no way that the giants would partner together unless they could create their own wedges. It seems that there will be a baseline standard that allows end users to use their hardware on any one of the platforms, but each platform will have their own requirements for developers, starting with proprietary integration for the "best customer experience".


You’re not wrong. But I’m excited anyway.

The fact I can (soon) stop having to check every device I’m interested in to see if it works with my ecosystem is great. I won’t have to settle for a worse device because the good one is Alexa (or whatever) only.

Also? Thread rules. WiFi has been (mostly) reliable but needs constant power. Every Bluetooth home automation device I’ve tried was a total mess of pairing/connection issues. Thread devices have worked great so far so I’m happy to see it continue to get popular.


HomeKit doesn't require an app for each device - sounds like Matter doesn't need one either, it's just that many devices expose controls Matter doesn't support.

My HomeKit light switch doesn't have an app on my phone at all, never installed one - just scanned the HomeKit code and set it up.


It depends. A lot of HomeKit devices have basic functionality through just the Home app, but if you want anything advanced, you need the device app. My Philips Hue bulbs only do "gradually brighten at sunrise" through the Philips app, and Nanoleaf panels need their app to do anything more complex than on/off.


>Cloud dependency of many devices for the most basic functionality offered is an absolute dealbreaker for me.

Story time: My internet got taken out by a dude with a backhoe, took a week to get repaired. I then discovered that my $4k 8Sleep mattress is entirely incapable of _any_ form of local control.

I have to bounce a packet off AWS to change the temperate of my bed, on a $4k piece of hardware. (I'm not mentioning the price to brag, but to point out the absurdity).

That and they couldn't even be bothered to put an RJ-45 on the back, so it completely failed when I migrated my main wifi to WPA3.


There's a lot of VC blood on the floor that bought us this common standard. Standardization is tough or impossible in a market that is minting unicorns -- we just had to wait for things to cool off before we could agree.

This is one of the gotchas of our current tech world: money drives progress, but that progress is almost always at the expense of cooperation.


This is great news, the Matter protocol sounds like it will make things a lot easier in the future.

If you're looking to get started with home automation, I would recommend using Home Assistant as the central software to control everything. Matter seems to solve some issues that I don't really experience as a Home Assistant user. You can use Home Assistant as the gateway that connects all your hubs and devices together.

I would also highly recommend buying Zigbee products from AliExpress. These are far more affordable than anything that you find in a retail store or from big brands. If you're very wealthy then you could afford to go with Philips Hue products for everything, but I've had a lot of success with some very affordable Zigbee and Bluetooth devices.

Here's the products I would recommend (all prices in USD):

* XIAOMI Mijia Bluetooth Temp/Humidity sensors: $4.86 -

  * You can flash these with a great custom firmware:
* Sonoff Zigbee products:

  * Motion sensor: $9.49

  * Door/window sensor: $8.49

  * Temp/humidity sensor: $8.49 (If you prefer Zigbee to Bluetooth)
* Moes Zigbee switch and dimmer modules for wall switches and ceiling lights:

  * 1 gang switch: $11.19

  * 2 gang switch: $13.06

  * 1 gang dimmer for LED lights: $13.06

  * 2 gang dimmer: $14.92
* Tuya Zigbee Garage Door module (with relay + reed switch sensor): $18.39 -

I've been really impressed with these battery-powered devices. I was worried that I'd be constantly changing batteries, but some of my motion and door sensors have lasted for over 2 years on a single CR2032 battery.


What are you using as your Zigbee hub?


I'm using a zzh (CC2652R Stick) from electrolama [1]. For the software, I use Zigbee2MQTT (z2m) [2] running as a Home Assistant [3] add-on. Here's a full list of supported USB Zigbee adapters for z2m [4].






Thank you!


What is the coolest thing you can do with a smarthome? The single biggest selling point?

Seems like you can basically just control appliances from your smartphone. Seems like such a small benefit (frankly I wouldn't even use it) for such a large effort, not to mention the security and data integrity problems.


One thing I've learned over the years is that this 100% depends on you and nobody's answer will really satisfy you, because the needs and routines of every individual can be radically different.

Do you care about energy consumption? If you do, you can pull in all that information into a dashboard, evaluate your usage patterns, and craft automations that ensure you're being as efficient as you want. Turn your lights/AC/music on and off when you go from room to room, etc.

Do you care about micro-optimizing your time usage? Create automations that support your daily routine. Tie your coffee maker's smart plug to the next upcoming alarm on your smartphone so you have breakfast ready by the time you get up. Have your garage door open automatically when it's time to go to work, and it's not a holiday, and your phone is at home, and your car is in the garage.

Do you care about eye strain? Have your smart lights change colour to match the light temperature of sunlight throughout the day.

Do you care about silly things that make you feel all powerful? Have your lights turn off, some white noise start, and all the smartphones in the house switch to Do Not Disturb when you plug in your phone to charge next to your bed after 10PM.

The sky is the limit. Over the years I've ping-ponged between "eh, this is not really worth it" and "this is actually incredibly convenient" depending on how my routine has changed.


My kids were scared of the dark, and letting them control the bathroom and hallway lights by voice (from their cozy beds) saved me from many late-night wakeup calls.

A smart home is just a problem solver, and everyone has different problems.


There are cool things you can do with a smarthome, like let a friend into your house remotely, but those aren't the selling features.

The selling features are the boring things. Like connecting a light switch to a lightbulb without having to pull wires through external walls. Or instead of waking up to an annoying alarm clock wake up to your blinds opening.


Agree, there's no one life-changing killer feature, but there's lots of small ones.

For example my thermostat goes up to 85 when I'm not at home, and I set it to a more comfortable temperature before I head home (it'll automatically get more comfortable when I arrive, but on hot days it's nice to pre-cool). You could accomplish something that works maybe 90% as well with lots of effort and a programmable thermostat, but this is more or less automatic.

My entryway light turns on automatically when I get home

I can turn off the lamp upstairs while I'm laying in bed, without having to get up.


I'd like to chime in and agree. It's the little things.

Through HomeAssistant (HA) I have a spoken notification to remind me to run the dishwasher right when the electricity time-of-use peak pricing ends each day. It also flashes a light a specific color when it says it so even if the tv is on loud, it'll get my attention.

HA automatically turns on my desk light when my webcam turns on.

HA allows me to remote-start my car with a Siri shortcut, instead of having to dig through the terrible manufactures app.

HA lets me know if I forgot to lock the door to my house.


The most recent thing I setup in HomeAssistant + NodeRed was an automation for my front porch lights, which are connected to a Lutron Caseta switch. When my Ring doorbell detects motion at night, the lights will turn on at 31%, and stay lit for 20 minutes. However this automation won't run if the lights are already turned on at some level other than 31%. Any subsequent detected motion increases the duration so that they stay on for 20 minutes after the last detected motion.

I also use Alexa to fake lights being turned on throughout the house when I'm away for an extended period of time, so the house doesn't seem as empty. There's a HomeAssistant extension that will record normal behavior in your house, and replay it on demand, which I'm thinking about using to randomize the lights even better.


Agreed about the little things. The best thing I've done is put a floor lamp on a timer... which sounds really lame but was so nice since there wasn't a light switch or outlet switch in that room.

I had it come on in the morning before I get up, then turn off after I leave for the day. Turn on in the evening, and off at bedtime. You can get mechanical timers for that, but they don't adjust themselves for daylight savings or changing sunrise/sunset times. I was going to set it up to stay on longer when it is dark and overcast out, but I moved to a place with better switch placement, so never did.


I plugged my dumb washing machine, dishwasher and tumble dryer into some zwave plugs which can report on energy draw. Hooked it up to Home Assistant, which is hooked up to Alexa. Now whenever any of my dumb washing machine, dishwasher or tumble dryer finish a run, it announces through my echo dot that it has finished. This is pretty useful to us as the washing machine and tumble dryer are in the garage, and the dishwasher finished beep is pretty quiet.


Honestly, there are no good applications beyond the most boring ones like turning your lights on automatically when the sun goes down.

Some people can figure out how to setup more complex things like "star my coffee when I wake up while gradually bringing the lights up" etc., but the whole experience is abysmal, and worse than Linux in the 90s: you need separate apps (or code stuff with Homebridge), all the devices have extremely limited capabilities (esp. when trying to control them centrally through e.g. Homekit), tgey are extremely brittle and disappear from your network if you look at them funny etc.

It's like Apple's Siri: good in commercials, but only good enough to set a timer. Same here: good in commercials, only good enough to turn on and off based on a simple timer.

Edit: I have a lighting setup with IKEA smart bulbs, Philips Hue lights, a smart plug on the balcony. I doubt I will ever touch another smart device beyond these.


I consider any kind of DIY automation at home to be pretty abysmal.

But the few legitimate uses(Door sensors, cameras, checking if you left a heating appliance on, etc) are worth having the system for.


I got a Google Home as a raffle prize and didn’t see the point.

But my house was built in the 70s and has almost 0 ceiling lights in the room. So I started buying smart outlets for the lamps so I could just turn the room off by yelling at it.

My wife switched us to Alexa because she had an app that would let her do her billing through voice (start working on John Doe’s case, stop).

I’ve expanded the lights through the house, I use routines to do stuff like turn off my bedroom light, turn on the fan, and turn on a “rain on a tent” sleep noise thing.

We use it as an intercom for the kids and to make announcements like “dinner time.”

I got an Amazon TV on prime day because it was cheap. It’s kinda handy to be able to yell at the TV when the kids lose the remote. Not super useful.

We all use the things to listen to music. I have mine paired with nice speakers in my office.

I gave up on being spied on. Everything I own is feeding back to Google, Amazon, and Apple. Probably China too with my cheap outlets.

I stick smart devices on a separate 2.4GHz Wi-Fi network that also can’t route to my main one, but that’s just to keep inevitable compromises contained and not fill my network with a bunch of chattering nonsense.

I wish Alexa would let me name it whatever I want vs the choose 1 of 3 names. There are more satisfying names to yell.


I have finally automated (a) determining that the washer has finished washing clothes (smart plug monitoring amperage highs and lows) and then (b) reminding the teenagers in my house that the washer is done and they should put their dang stuff in the dryer.

(They have not yet attained the self-awareness to set a timer on their phone, and I tire of tracking down the culprit every time I find wet clothes in the washer.)

Setup is Home Assistant Docker image on a Raspberry Pi, MQTT broker, and Tasmota firmware on the smart devices. 95% cloud free except for the push notifications.

Still determining how to solder together a safe dryer temperature sensor module to detect when the dryer is done...


That's genius. For the dryer maybe try a vibration sensor on it? Should be possible to discern between laundry or not.


> What is the coolest thing you can do with a smarthome? The single biggest selling point?

There's no single selling point.

> Seems like you can basically just control appliances from your smartphone.

That might be the case for most consumers, but it's nowhere near the end of what you can do.

One day my 3d printer was working and my robo vacuum bumped into it. I realized that it wasn't a good idea to start a vacuum run if the printer was busy. Both vacuum and printer were already connected to Home Assistant, so the fix was trivial and took a couple of minutes.

I've setup my air filter to turn on whenever the air quality drops - currently watching outside particulate levels, but I'm assembling a device that will allow me to measure inside levels.

We have two portable air conditioners in the same circuit. They can often run together - but not always. They are wired to smart outlets that measure power draw. Automation will turn one of them off if they exceed the recommended power draw - before the breaker gets unhappy. They will also be turned on automatically based on internal temp sensors. So I don't have to worry about my dog.

The important thing is to connect everything that can be connected and start streaming the data. You'll start to find plenty of situations where doing <x> when <y> will be useful. Want to dim the lights when the TV comes on? Easy. Want to turn the coffee maker on and open the blinds when you wake up? Simple. Want to postpone your dryer or washing machine when there's people around? Not a problem. Want a reminder that you forgot the garage door open, or a door open? You can do it.

But what's important to you will depend on your use-case.


I set up water sensors under the kitchen sink and in bathroom vanities. Each of them cost about $20 and integrated into my Home Assistant setup without any issues. Three months later, I was on vacation away from home and my water leak alert triggered. I was able to alert my parents, and they came back to shut the water and fix the problem before it translated into thousands of damages.


At Home Assistant we're very excited about Matter.

Here are some highlights for us:

Open source reference implementation: Google, Apple, Home Assistant, we're all going to be running the same code to be a Matter controller. Chip manufacturers like Espressif and Nordic maintain implementations for their boards in there too, so anyone can now have the software to produce Matter compatible light bulbs.

It will be cheap: The software is freely available and works with a big audience. It's the same reason Android TVs from some manufacturers are cheap, the same will be the case for Matter lights and switches.

Multi-fabric: each Matter device is required to support 5 fabrics. A fabric is a Matter network. This means that you will be able to run multiple home automation controllers at the same time. So when run into the limitations of Google Home or Apple Home, you can try out Home Assistant without taking down your old system.

Easy sharing of devices: Because of multi-fabric, it will be as easy as hitting a share button to get a device added to another fabric. See this example of Android

Local: all communication for Matter is happening local between a device, a thread border router (if thread-based Matter), and the controller. Note that your controller can still decide to store your data in the cloud (ie Amazon, Google).

Supported by major systems: Amazon, Apple, Google and Home Assistant are all building the open source Matter code into their systems. It means that for manufacturers it will be easy to pick Matter as the protocol they want to support to reach most users.

Works over IP: Matter works over IP and doens't care how that IP-based device communicates. It means that you can have Wi-Fi based and Thread-based devices co-exist on your network. Thread is not required if you don't care about such devices.

Bridges are part of the standard: Devices like Philips Hue hubs are going to get an upgrade to expose all the Hue lights over Matter via the hub. This makes integrating a whole ecosystem at once into Matter very easy.

Thread: Thread is a mesh networking standard that connects to your Wi-Fi network via border routers. Where Zigbee and Z-Wave need to mesh communicate all the way to your controller, Thread messages will be delivered via Wi-Fi/ethernet as soon as possible. This means it is a lot more reliable and less traffic is going over the mesh. Expect cheap border routers (open source reference implementations) but also expect them in your future Wi-Fi routers, voice assistants, Wi-Fi connected TVs etc.

Paulus /Founder Home Assistant


This workshop was incredibly educational in how matter works


> Open source reference implementation: Google, Apple, Home Assistant, we're all going to be running the same code to be a Matter controller. Chip manufacturers like Espressif and Nordic maintain implementations for their boards in there too, so anyone can now have the software to produce Matter compatible light bulbs.

Will have to go through that documentation later, but are there any chips like ESP32/ESP8266 that hobbyists can use? Specifically for Thread.


ESP32-H2 supports Thread:

That's also the chip that Espressif has build it's reference implementation Thread Border Router around:

But you can also use ESP32s for Thread. If you have one that we support laying around, we have an installer website here that we use for testing




Thanks, this is the first explanation I've seen of how the ecosystem actually fits together, and what Thread and Matter would mean for end users.


Brought to you by our $20k/year annual membership fee that Home Assistant, an open source project, pays to be a member of the CSA alliance and be able to contribute to the open source Matter reference implementation.


Thanks for showing up and advocating for the open side of the world.

This made me curious about how Home Assistant is funded. Looks like it's mostly through Nabu Casa, which is a SaaS backend for Home Assistant that costs $6.50/mo or $65/yr in the US


I feel like it violates some core principle of FOSS if you have to pay to be allowed to contribute.

I mean, the purists will argue you can still fork it.

But then you'll eventually have the same problem all the Chromium forks have now with Manifest v3.

At least there's some form of "committee" in the case of Matter, but I'm slightly less excited about it now having learned this.


Ouch, is this from donations or is Nabu Casa funding that?


Do you have any fear that the connection of all these iot devices could be used in the facilitation of a widespread attack?

Also, all the devices in our homes running code which we have no way of scrutinizing. I'm a bit alarmed about that.

And the uprise in unchecked images people are flashing onto raspi's which could potentially be owned by a bad actor.

I could be paranoid, but these are things that have been on my mind.


There are enough products and projects available to build a smart home based on just open source code: Home Assistant, ESPHome and WLED for example. And soon you can add Matter to that mix. Product manufacturers could publish their source if they wanted to.

For the Raspi images: the build pipelines are often hosted on GitHub and all the build logs and source is available. If anything, open source projects that are big enough to offer their own OS often receive more scrutiny than most closed source products do.


Is there a way to force devices to use the mesh instead of going over the network? Lots of folks in this thread have concerns about allowing every lightbulb direct access to the internet and for good reason.


Wow, this post really highlights things I did not know about Matter, but which sounds great!

In all honesty, as a Home Assistant user I sort of feel that Matter/Thread is not important to me, but multi-fabric and bridges sound super nice.


Thanks for this comment - answered a lot of my questions


Something on this page prevents clicking the Security Webinar and Whitepaper download links on Firefox and Chrome, so here are those links:


Doesn't Matter require device attestation? To me, this says, "Matter will never be compatible with cheap DIY devices like ESP32 devices". And also makes me suspicious of being able to create a Chromecast competitor that removes ads, or whatever.


Matter certification requires that controllers do attestation (validate that the device says they are who they say they are) but you are allowed to offer an option to users to bypass it (like Apple does today for HomeKit today).


The official Matter SDK has examples targeting the ESP32, along with other low-cost chipsets:


Thank you! I'm happy to stop repeating/worrying about this and dive in more, and you've given me a great starting point.