www.theverge.com
Matter launched with a mission to provide a reliable, secure way for smart home devices to connect to and communicate with each other across brands and ecosystems. But two years in, problems have plagued the connectivity protocol. Big players like Apple, Amazon, and Google have been slow to offer broad support for Matter device types, there are ongoing reliability and interoperability issues, and the Thread protocol Matter relies on has seen significant growing pains. Its been a bumpy start.The companies behind Matter hear the complaints and say this is the year those concerns will be put to rest. 2025 is going to be a really important year. Were looking to really hit reliability and make sure that things just work, says Tobin Richardson, president and CEO of the Connectivity Standards Alliance (CSA), the organization responsible for developing and maintaining the smart home standard.Its going to be a year of focusing on fixing these reliability and performance issues in Matter.To find out how Matter is going to improve, I sat down with the heads of the three organizations responsible for Matters success: Richardson of the CSA; Vividh Siddha, president of Thread Group and director of software engineering at Apple; and Kevin Robinson, president and CEO of the Wi-Fi Alliance. Thread and Wi-Fi are the two wireless protocols Matter works over, and the three organizations and their members have been collaborating on the standard since it began back in 2019. This is the first time the organizations have given a joint interview, and they had plenty to share about how they are working together to improve the smart home standard.According to Richardson, well start to see the Matter we were promised easy to use, reliable, and seamless potentially within the next 12 months. Its going to be a year of focusing on fixing these reliability and performance issues in Matter, he said. Siddha concurs, saying, 2025 is really pivotal for us to get all those issues resolved, to get all these product experiences to be more robust and reliable. Robinson feels confident that the collaboration between the three groups will help bring this about. I think where organizations like ours deliver value is by aligning [our industries], so everybody is singing from the same song sheet, he says.RelatedSolutions being explored include embedding software in home routers to ensure Matter and Thread devices can connect to networks seamlessly (solving the multicast issue with many home routers), adding Thread radios to home routers, potentially replacing Bluetooth with Thread for onboarding Thread devices, and pushing ecosystems onto Thread 1.4. Additionally, companies like Apple and Google are sharing data about their real-world deployment of Matter and Thread for the first time in an effort to help everyone implement the technologies successfully.These feel like positive steps for the standard especially the plans to fix how home routers deal with Matter devices but the proof will be in how quickly and effectively they can execute them. Apple, Google, and (hopefully) other ecosystems sharing what theyve learned in an effort to make Matter work for everyone, and not just the big four, also feels like a significant shift. Matter is supposed to promote interoperability, but today, the most reliable way to get it to work is to stick with one ecosystem. (This GitHub thread dives deep into the problems Home Assistant users are having with Thread devices that work fine with Apple Home.)After all, Matter is an industry collaboration, and the only way it will ever work as advertised is if all these companies continue to work together. Most of the big names here Apple, Google, Samsung, Amazon, LG, Comcast, along with silicon makers such as NXP and Qualcomm are members of all three of the organizations working to make Matter a viable solution for the smart home. If they fail, those companies will have failed, too.The following interview has been edited for clarity and length.Lets start with the basics. Explain why Matter chose Thread and Wi-Fi as its wireless connectivity protocols and not Zigbee or Z-Wave. And who made that choice?Tobin Richardson (CSA): The choice was 100 percent member-driven; those who founded Project CHIP [as it was called, founders include Apple, Google, Amazon, and Samsung SmartThings]. That was a two-year conversation. But Wi-Fi is already in every smart home device thats connected, so why go with a new technology? As Project CHIP came together, it was an evaluation. We knew we needed a common language, and we needed it to be IP-based. We looked at what is in the smart home already. Wi-Fi is in the smart home. But thats not all use cases. What captures 90-plus percent of use cases? Thread. So, with Wi-Fi and Thread, youre capturing a known. It comes down to Internet Protocol. That was a conscious choice to give a nod to Wi-Fi and Thread as they work over IP. Zigbee is great. It is lightning fast. Its been out for 20 years, but its not IP-based. Embracing Internet Protocol was really that key building block.Kevin Robinson (Wi-Fi Alliance): Its the Internet of Things, and the Internet of Things should be built like the internet, which means Internet Protocol. It was very simple: lets make the smart home like the internet and allow it to scale like the internet did. (The good parts!)Vividh Siddha (Thread Group): Part of it was also the ecosystem and silicon vendors having the support for Thread and Wi-Fi. Thats a big enabler that you require to make sure that this comes to fruition. Multi-protocol was a big factor because its not something that will take years to build. It was there already.KR: There were attempts within the Wi-Fi industry to do something similar to Matter, but it never really took off because it was Wi-Fi only. One of the genius things, and ultimately why were very optimistic about what Matter is going to deliver, is you have ThreadandWi-Fi. Now you have two major technologies that address pretty much anything youre going to want to do in the home, and you have an organization that is working on this upper layer (Matter) that makes it all work seamlessly. It was a very important strategic decision and the right decision to pick two technologies that cover the landscape and then say we are committed to both. Thats going to make Matter successful.So, you have two ways to use Matter: Matter-over-Thread or Matter-over-Wi-Fi? Which is better?KR: It depends! The Wi-Fi industry has been saying for years that the smart home is going to be a combination of technologies. No one technology can accomplish every use case. But all of us want, collectively, to be able to make the smart home opportunity reach its potential.The smart home is going to be a combination of technologies. No one technology can accomplish every use case.There are going to be cases for Wi-Fi if youre doing high-definition security camera streaming, I dont believe Thread can do that yet. Each technology is optimized differently, but theres going to be overlap. Ultimately, the vendors will decide whether it makes it into their product. Theres also the draw of Were already using one versus the other, so Ill just stick with that. But overall, its going to be driven by the specific application in the smart home, the capabilities that are required to address that application, and then the fit to one of our technologies.VS: We are getting into this new phase where people are looking at Wi-Fi and Thread as complementary not independent technologies. So, if you can take the example of a camera for streaming video, you definitely need to use Wi-Fi, but for some of the control functions in a camera stopping, starting, or doing other control functions you could potentially use Thread as well, which could be helpful if Wi-Fi is congested. There will be some interesting complementary ways that these technologies will be used. Were not there yet, but there are companies talking about how they could use that.I get the different use cases, but in Matter, we have product categories like smart plugs and smart switches where some manufacturers use Matter-over-Wi-Fi, and some use Matter-over-Thread. Thats where the consumer is confused about which one is better for them. I understand what Wi-Fi can do in terms of bandwidth and what Thread can do in terms of lower latency, but why do we have this distinction in devices where the use case isnt so obvious? Is that a manufacturer's choice based on cost? Simplicity? And which is better?KR: I dont think the consumer is going to choose that product based on Wi-Fi or Thread.Our readers do!KR: That may be true, but ultimately, if Matter accomplishes what it intends to, consumers are going to look for a Matter logo. Increasingly, were now seeing home access points that will include Thread as well. So, ultimately, your home network, if its a Matter network, is going to support either. Its a matter of, no pun intended, youre going to buy from a brand you trust that will drive decisions [more than protocols].VS: Where we are right now is people think about Matter-over-Wi-Fi and Matter-over-Thread, partly because Thread is in its initial phases of deployment. One of the key things you need is the infrastructure. As for Wi-Fi, its already a given that all people have a Wi-Fi router in their homes. Thread border routers, which is an essential function, are now pretty much in every ecosystems products, as well as many other companies providing border routers in the home.End users wont have to think about whether its Wi-Fi or Thread. It just works.With Matters HRAP initiative [Home Routers and Access Points], once access points start supporting border router functionality out of the box, then most people will have Thread functionality in the home. It will take some time, but I think once we get there, whether its Wi-Fi or Thread, from a Matter standpoint, it should just be agnostic. It will use the technology thats the best. Manufacturers will use the technology thats best for their product, and end users wont have to think about whether its Wi-Fi or Thread. It just works.KR: But then, outside this room, of course, Im going to be trying to convince manufacturers why Wi-Fi is appropriate for certain things, and [Threads] going to be convincing manufacturers why Thread is appropriate for other things. Itll all kind of shake out again by the vendors, those manufacturers making decisions of what really fits their use case the best.Those are obviously the manufacturers decisions. But for the consumer, its stability, low latency, and making everything work the way Matter has said it will. Thats what were here for. We want the smart home to work. Matter / Thread had a bumpy start. How soon is the dream of a stable, low-latency smart home where everything works going to be a reality? What is the timeline for when the consumer can buy a Matter product Wi-Fi or Thread and feel confident that it will work?VS: I would say 2025 is going to be a fairly important year from a Thread standpoint. We see this from our ecosystems telemetry as well. Most people do have a border router in their homes now, and Thread 1.4 will solve an important problem where [border routers from different manufacturers] dont necessarily talk to each other well. Thread 1.4 is going to definitely improve that. Once we get a deployment of 1.4, which we are accelerating by sunsetting earlier versions of Thread, it will allow us to get these border routers to work seamlessly between themselves. And once we get to that point, well see wider adoption of Thread just working.The deployment of Thread 1.4 is being accelerated by sunsetting earlier versions of ThreadYou mentioned sunsetting earlier versions of Thread. Does that mean that if you have a Thread border router, it should be on 1.4?VS: Yes, particularly for companies that are building border routers.Will there be scope for companies to be grandfathered in to earlier versions? They havent been super fast at updating new versions of Thread to date. How much power do you have here? [Companies manufacturing Thread border routers include Apple, Google, Samsung, Amazon, LG, and several smaller smart home companies.]VS: We definitely will listen to the ecosystems every company has its own limitations in terms of when they can update. But we have some policies in place where we expect them to implement 1.4, and we havent gotten much pushback. They understand this is an important thing to do. We expect that to happen fairly quickly, and its a small number of companies. Its not the entire set of companies that are building accessories. Its a small number of companies building border routers. So once the infrastructure is robust, the acceleration of products will happen.This year, we saw a number of different solutions to the issue of Thread border routers fromdifferent manufacturers not talking to each other, includingseveral ways to ensure they share their credentials and can form one Thread mesh network in your home. Thread has one solution, the CSA has one, and phone operating systems can also manage the exchange of credentials. Does the user need to worry about which solution theyre using? Is the end effect going to be the same, that theyre going to have a stable Thread network? Or is there a best solution?VS: I think the ecosystems will define what the user experience for that will be, but I expect there to be at least one minimum step where the user is involved in sharing [credentials]. For example, its similar to sharing your W-Fi network. Lets say I go to Kevins home, and he wants to share his Wi-Fi network with me. He has to share the password with explicit intent. So theres that step that will be there in a Thread network as well. How the ecosystems expose that experience to the end user will depend on their implementation. But the underpinnings of that, even if theyre different, dont matter.From the standpoint of the new smart home user, who Matter is targeting, currently the simplest way to use Thread in their home is the phone, thanks to Thread radios in mobile devices from Apple and Google. Is that a good solution for the smart home, or is that more of a temporary Band-Aid? Is it going to set them up for problems?VS: Both Apple and Google have made announcements of Thread radios in mobile devices. Its to enable an early adopter. If you take a very early on smart home user somebody who goes to Ikea, buys a bulb and a switch what happens with those users is they discover they cannot do automations. So, Thread on mobile can enable that evolution [to a border router]. But the other important factor is the ability to use your accessories, and some important ones, like your front door lock when theres a power outage. If you have no infrastructure your Wi-Fi router is down, your Thread border router is down and you still want the ability to get into your house or do other things that might be smart home related. Those are some use cases that its explicitly designed for, but its not limited to that.Where does Bluetooth fit into this conversation? Its part of Matter for onboarding, but is there potential for Thread to replace Bluetooth in certain applications? It is also a wireless, low-power, low-latency connectivity standard. Is it something that could do much of what Bluetooth does today in our smart homes?VS: Bluetooth still has its use cases and applications. For example, in high-end audio streaming, Bluetooth will still remain the dominant choice. But with Thread, you talked about provisioning or commissioning: Thread 1.4 now supports native Thread commissioning. At least for smart home products, we now have the ability to do native Thread commissioning, so we do expect to see that.If its a smart home product that requires Matter and Thread, it doesnt need Bluetooth. It allows [manufacturers] to do two things: one is to simplify their product, and the second is to reduce cost because they dont need to add Bluetooth, which takes memory. There is also a cost involved with adding an additional radio, so that can go away. If its a smart home product that requires Matter and Thread, it doesnt need Bluetooth.We were talking earlier about smart home use cases around health and wellness. Could Thread be used in devices in our homes, such as blood pressure monitors and other home healthcare devices that currently use Bluetooth?VS: Yes. In fact, we have some companies in Europe that are building exactly those types of products, where it gives you low power, low latency, and all that. There are those use cases where it makes sense because these are very small amounts of data that are being sent periodically, which is just an optimal case.HRAPs are now part of the Matter spec, which, due to incorporating Thread, Wi-Fi, and some Matter magic, should be a solid foundation for people wanting to use Matter in their smart home. Kevin, from your position in the Wi-Fi Alliance, what timeline can we expect in terms of seeing routers supporting Matter? It feels like an important step for adoption. But where Verge readers might change their routers once a year, people like my parents never do. Unless their ISP upgrades them, most people arent getting new routers regularly. What do you see among your members around adoption here and bringing that to the customer?KR: One key point to start off with is that theres close collaboration between our organizations and there are certain areas where we absolutely need to collaborate, and there are other areas where each organization moves at its own pace. And thats good. With HRAP, the industry is already delivering many of the things youre going to need in a home access point. For example, Wi-Fi 6 is a good baseline for todays access points. And as an industry, weve already, just this last year, shipped a little over a quarter of a billion Wi-Fi 7 access points. So that will be the predominant generation of Wi-Fi in most homes in a few years.What we need for a Matter HRAP device is A, B, C, and D, and weve already got A, B, and C. From our perspective, most of what Matter needs, particularly in an initial iteration, the industry already has. We can essentially flip a switch, and now its fully capable and fully Matter-enabled.So youre saying many current HRAPs in peoples homes could become Matter-enabled easily but not Thread-enabled? [Which is required for Matter certification.] What about Thread adoption in routers? Is that something your members are actively pursuing? Do you feel like thats something were going to see in the next year? Five? Ten?KR: In terms of the Wi-Fi capabilities that make up that HRAP device, yes. Most of that is already in the market today in the leading routers. [For Thread], it depends on each implementation and what radios are in there. In some of our service provider members, we see that they are already rolling out Thread on those devices. In some cases, the Thread radio is there, and its just waiting to be turned on. Its on a vendor-by-vendor basis, but we certainly expect, particularly with the service providers that are active in Matter, that they could enable that fairly quickly.TR: Were hearing the same thing. Its mainly from service providers. Companies like Verizon and others have indicated that direction. Verizon is now on the board of directors for the alliance.VS: A lot of the major service providers that we know about in North America are actively engaged, and they have plans and existing deployments that already have [an IEEE 802.15.4] radio, so it could just be turned on. For routers that dont have a radio, they might have a dongle they could stick in to enable this functionality.One of the challenges today is the way routers implement their multicast discovery can be broken, and that leads to a pretty bad experience for users.But even with routers that will support HRAP that dont have a Thread radio, theres an important aspect of some key functionality that routers have to support to enable not only the smart home but also Thread. These include features like IPV 6 and multicast discovery, which we want to make sure are consistent and reliable across routers, to make sure that they work.One of the challenges today is the way routers implement their multicast discovery can be broken, and that leads to a pretty bad experience for users. Matter particularly relies on that discovery mechanism to work all the time, and if it doesnt, then you have lots of issues in the ecosystems. There is a lot of value in the work that the HRAP group in Matter is doing.So this is something thats part of the Working Group for HRAP? If you have a Matter-enabled router, it should actually optimize your Wi-Fi for Matter in some way. Thats what youre working toward?KR: Yes, this is where the collaboration is... there are elements from Matter, and then there are capabilities and certifications were delivering around that to make sure its working properly. Its divide and conquer, right? Were able to deliver some value to the Matter ecosystem to allow it to be more successful.VS: That was the starting point for HRAP before we even added the Thread border router functionality. We need to make Wi-Fi work reliably for all of the IoT use cases. We drafted the initial set of things that we knew were broken or were not optimal. Specifically, the home routers. The issue is not with Wi-Fi as such; it is the layer above, the routing layer, that is broken in some of these vendor companies, and HRAP will allow us to fix this.Something a lot of Verge readers want is routers as Matter devices. Being able to control their router through their smart home turn it on, off, reboot it remotely, management of devices. Is the concept of having a Matter router be a Matter device on the roadmap, or is it just about the infrastructure?VS: Its more focused on functionality and manageability. What youre talking about is currently not in the scope.KR: You bring up some interesting cases. There are probably subsets of that that could make sense, where its more of the user-facing management of things. But theres a lot more that goes into managing a home network that Matter probably wouldnt want to get into.One possible area is the router itself as a sensor in the home. It has information about what devices came into the home. We now have Wi-Fi ranging, Wi-Fi location via multiple access points can pinpoint where somebody is. Wi-Fi sensing can tell when somebody is walking in the home. Did they fall? Leveraging that access point or that Wi-Fi network as a sensor of sorts, thats the type of application we do see members getting some interest around through Matter. But theres no work on it currently. There are other things that are more important to be focusing on.Can you give us a timeline of when Matter is all just going to work? What have been the major challenges in the first couple of years that you feel youve successfully surmounted, and which are you still struggling with?TR: 2025 is going to be a really important year for Matter. Were not looking to add 100 new device types this year. Were looking to really hit reliability and make sure that things just work. Thats been a strong focus, and thats why these relationships are so important. Because companies making cool experiences for the consumer will be enabled by the complexity underneath. We knew that coming in, and thats why I think its so important that Thread continues to do a lot of its work, engaged side by side with us and Wi-Fi.People ... who think they should be further along are not looking at past efforts what other initiatives like this accomplished in a given time horizon.When are we going to get to that promised land? Its months, not years. Is it two months? No. Is it less than 12? It depends. Your readers will see things faster than others because they understand whats going on. What was jointly addressed in [fixing] credential sharing with Thread Group is a great example of how we can move at a good speed. And what we did with the Interop Lab just recently, making it easier for developers and companies to get to market faster. Weve hit a lot of the unknown unknowns the ones you just dont know about on day one.What about for Thread? Its been around for 10 years, but this has been the first mass real-world deployment.VS: If you look at 2020, thats when we had our first border router products. Until then, we were just supporting specs. After Apple announced the first border router and all the other ecosystems now support it, I think were at a place where the border routers are now in peoples homes. With 1.4, the interoperability issues will be solved. Once we have done that, the infrastructure underpinnings will be pretty good for enabling Matter over Thread.There is some work we need to do with Matter on optimizing for the protocol, specifically for Thread because its a low bandwidth network, but thats something were collaborating on to help improve memory and release memory and stuff like that. A lot of that will get better, for sure. As Tobin mentioned, 2025 is really pivotal for us to get all those issues resolved, to get all these product experiences to be more robust and reliable.Something Ive heard from a number of manufacturers and companies is that Matter-over-Thread works really well in Apple and Google ecosystems, but when you take it outside of those platforms, thats when it starts to go wrong. Does that ring true to you?VS: Well, yes. Theres a lot of work the ecosystems do for their individual products. And Google has been working on Thread for a long time. Apple, particularly, has invested quite a bit in Thread, so they might be leading the efforts there. But all the ecosystems are putting a lot of effort in Thread.TR: If you have the ecosystems bringing that knowledge back in, not keeping it to themselves. Thats the benefit of why these large companies have invested in Thread, not just their own proprietary approach. But thats a process. To bring those learnings back and then ensure theyre picked up and understood by the other manufacturers. Whether or not those are performing well, the time for turnaround on that learning should be compressed, and thats one of the things that we are working on: making sure that it gets back out to the community so they can update and upgrade as appropriate. Thats the benefit of an open standard.VS: A key aspect for Matter is having a close eye on what end users are seeing. Member companies are sharing information with each other around key things that we need to prioritize. In Thread Group, we have started having ecosystems share data. In our previous member meeting, Apple shared data for the very first time about what they are learning about their deployments. At the prior meeting, Google shared some data. The ecosystems themselves are bringing a lot more internal data that will help us to understand how these things are working in the wild.On a recent Vergecast and on a Matter panel I hosted at CES with Amazon, Google, Samsung, and Bosch, I invited everyone to grade Matters efforts so far. What grade would you give Matter at the end of its second year, in its mid-terms?TR: I think Matter gets a B. When you look at a grade, what you obviously need to know is: are they getting their homework done? Are they passing the tests? Is there a good effort in there? Its a combination of those three, along with: are they ready to go to the next stage? And absolutely, Matter is ready to go to the next stage. Weve learned a bunch.These companies are trying to kill each other in the market ... But what theyre doing together is really impressive.The effort, thats an A. In terms of what people are doing in the Matter Working Groups and the Alliance. Its incredible, even heartwarming. Because these companies are trying to kill each other in the market, as they should, right? Thats their job, for their shareholders. But what theyre doing together is really impressive.VS: I think B. From a purely technology standpoint. The education aspect, I think theres more work to be done there. But I would agree that in terms of just the sheer enthusiasm, motivation, and dedication from the member companies to fix all the problems as quickly as we can and get them to market, its just unprecedented. Everybody comes to the table, and its all hands on deck to solve things.KR: I give Matter a B+, grading on a curve. People outside Matter who think they should be further along are not looking at past efforts what other initiatives like this accomplished in a given time horizon. From an outsiders perspective, the intensity and the amount of investment these companies are making to do the work is impressive to see. Things like this can get very territorial. Were different organizations. We all have responsibilities for the success of each of them. However, the CSA has been very easy to work with.Looking to the final grade, the thing that will be essential for me is ensuring that Matter does a good job of balancing the road maps of the respective technologies. As Im innovating, as Thread is innovating, the CSA needs to balance how we bring these innovations into the spec, into the products themselves, so that my industry can deliver the benefits of Wi-Fi innovation to people buying Matter products. Thats going to be a delicate balance that they will have to strike to get the A at the end of the semester.
/cdn.vox-cdn.com/uploads/chorus_asset/file/25843646/2158894635.jpg)
/cdn.vox-cdn.com/uploads/chorus_asset/file/25842090/257520_Matter_wifi_alliance_thread_CVirginia.jpg)
