Situation: there are too many smart home standards Before I get into what the Connected Home over IP industry working group (CHOIP? Probably CHIP but CHOIP is funnier.) is trying to make, I think we need to take a minute thinking over what it is. It's rare to see all these companies in the same headline, and it's even rarer to see that headline be something they agreed on instead of something they're being investigated over. If nothing else, this working group is worth paying attention to because it has a better than average chance (for proposed open source standards, anyway) of succeeding. That's not often the case in the smart home space. In fact, it's usually not. For example, earlier this year Amazon put together a similar industry group designed to ensure voice assistants could interoperate with each other — so that voice assistants could coexist on the same speaker at the same time. Over 30 companies signed up. Google, Apple, and Samsung did not. And why would they? Right now it seems like the whole game in the smart home is getting as many people as possible locked into your ecosystem as quickly and firmly as possible. Generally speaking, if you see a company asking everybody to work together, that company is either winning so hard it can afford to be magnanimous (Amazon and Alexa) or losing so badly it needs to make sure it isn't completely pushed out of the market by the dominant players (Microsoft and Cortana). But now, we have all these companies actually agreeing to work with each other. Charlie Kindel, chief product & technology officer at SnapAV (and formerly on the Alexa team at Amazon) says "the fact that it includes Amazon, Apple, and Google is a pretty significant statement ... To get them together to talk at this level — which we've never had success in doing before — is a great start." A great start, but certainly not a complete solution — nor does it purport to be. The bare-bones website that was rushed up for the working group says that right now the goal is to "release a draft specification and a preliminary reference open source implementation in late 2020." Which means that actual implementation of the standard in consumer products is surely even further out. Think 18 months for real consumer products in the most optimistic scenario. Anybody who's watched gigantic companies try to work together to agree on standards for web browsers can attest to the fact that these things take time. So what exactly is this working group planning to propose? What is the thing they're actually making and why does it matter? That, as with all things in the smart home, is complicated. The group wants to create a better low-level networking system for devices in the smart home. Right now, there's a confusing mix of different kinds of radios (Wi-Fi, 802.15.4, Bluetooth, Bluetooth Low Energy, Z-Wave, the list goes on) and each tends to handle its networking in a different way. The internet and nearly all other networks you use, by comparison, are IP-based: each device has its own address and communicates via common, agreed-upon methods. So at its base, this is an agreement in principle that the smart home needs that same basic level of common plumbing. In many ways, it sounds quite a lot like the Thread and Weave systems that Google has completely failed to get any traction for — and in fact Google's blog post says explicitly that it will be contributing those standards to the project. An IP system is promising because everybody basically knows how it works and there's also just more knowledge about how to keep that kind of system secure. It also has the benefit of being agnostic to the kind of wireless radio it communicates over — it could be traditional Wi-Fi, 802.15.4, Bluetooth, or something else in the future. Kindel jokes that "the good thing about standards is there's so many to choose from" and that this this proposed standard "really is the poster child" for that idea. He notes that other smart home technologies like Z-Wave or even Amazon's proposed Sidewalk system won't be obviated by this, in part because they support longer ranges. This has all been a surface skim of a very technical area — if there's a takeaway for you it's that these companies recognize that it's still too difficult to know if a device will work with your setup and way too difficult to actually do that initial setup. In theory, this new proposal could help with that. If you've tracked the various competing technologies that gadgets have used to communicate over the years, you'd know it's always been a mess. Here's a story my colleague Jake Kastrenakes wrote over five years ago decrying the dumb state of the smart home. But a funny thing happened since then: Alexa. Well, Alexa and then the Google Assistant and then HomeKit, plus a side of Samsung SmartThings. All of those systems work by adding a layer of software abstraction on top of that wireless standards mess. So in some ways, the problem the group is trying to solve already feels, well, solved for many consumers. Most people don't need to bother knowing about Zigbee or smart home bridges to set up their light bulbs — their smart assistants handle that for them. Developers, however, are forced to make sure their product supports three or more different smart home software ecosystems. That's a lot of work. And it inevitably means some will get left behind, or busted, or worst of all: hacked. To be clear, even in a perfect scenario this new system won't mean a beautiful, interoperable world. The very layers of software abstraction that helped simplify the smart home are going to continue to exist and compete with each other. It just seems like Apple, Amazon, and Google would rather compete head-to-head than try to dragoon third-party smart home devices makers to their various sides in some kind of Risk-esque war of attrition. Plus, not for nothing, an agreed-upon solution for this level of smart home plumbing would mean that these companies could use the resources they have dedicated to that towards building other stuff. Kindel believes that lots of consumers are still going to need professionals and an OS that helps make different ecosystems work together — like those at his company Control4 (which recently merged with SnapAV). "We do believe that for it to really be refined, normal human beings aren't going to want to do it themselves. They're going to hire local professionals, just like they don't do their own plumbing or even paint their own rooms." The key to this whole endeavor might be Zigbee: the Zigbee Alliance is managing this working group and can act as a neutral arbiter between these companies. The working group is explicitly not adopting Zigbee 3.0 as its new spec, but what it's proposing seems not very far off from what Zigbee already does. Throw in a little of Amazon's penchant for openness, Google's Thread networking magic, and Apple's rigor around privacy and you might just have something good. Maybe. One thing's for sure, there are too many standards. Surely one more can't make it worse? | 
Post a Comment