by 
I’m employing Ian Betteridge’s law to respond to Parker Ortolani’s opinion that “HomeKit is doomed to fail if Apple doesn’t make their own automation hardware“. This is a topic that keeps coming up. Bradley Chambers writes it in “Why 2023 is the time for Apple to release its own smart home products.” Should Apple make accessories? (No.)
Ortolani begins by quoting the Alan Kay quote that Jobs used when introducing the iPhone. You know the one, “People who are really serious about software should make their own hardware.” Which has some truth to it. If you’re trying to craft a special experience, where every detail matters, it may also require making your own hardware to support that experience. The point is that the experience you’re trying to deliver drives the software and hardware requirements, not the reverse.
Apple has made all sorts of accessories: Wi-Fi routers, wireless headphones, dongle connectors, and in years long gone now, laser and inkjet printers, and digital cameras. It’s worth noting that Apple has stopped making Wi-Fi routers, printers, and digital cameras. Apple stopped making these things because they couldn’t add anything to the experience above what other companies were offering. Printers made sense when Aldus PageMaker and LaserWriters were revolutionary. Routers made sense when Cisco and Netgear’s Wi-Fi offerings were so terrible.
To be fair, the flip side of the argument is, Apple makes iPhone cases, iPad keyboard cases, and headphones, crowding in on the accessory maker third party turf. The answer is, Apple thinks they can do something better in these spaces than the third parties do.
Ortolani wants Apple to make accessories because Google and Amazon do. This is a mistake that ignores how and why Google and Amazon have these things. Google acquired Nest and Dropcam, and effectively started the modern SmartHome era. Google has since spent a ton of focus on having third party accessories support Google Assistant.
This is the same point that Chambers makes, saying, “If someone were to ask me, what are the best products for a Google smart home, I’d say the products that Google makes. If someone were to ask what the best products for a smart home built on Amazon, I’d say products like Ring, Eero, and other products owned by Amazon.” The point both make is that it’s easier to recommend best-in-class accessories if they’re made by the 1st party (Amazon, Google, Apple.)
Amazon, for their part, acquired Ring and Blynk, but also make a small sample of devices compatible with Alexa. The reason is because Amazon’s Echo speakers work with Zigbee, and Amazon wants to show third party manufacturers examples of how to make these devices. The Alexa-enabled microwave is not an amazing microwave, but it shows how Alexa could be integrated with appliances.
Ortolani has two major complaints: One, that HomeKit products are of varying quality, so he has a hard time recommending them, and two, that devices don’t interoperate with other assistants, so each assistant provider, (Google, Amazon, Apple) should make their own.
One, everything that’s being sold should be better. Companies that made bad accessories, and know their stuff is bad, are bad. This usually happens because they over-optimize on cost – setting a cost target at the beginning of a project, and then looking for ways to cost-reduce at the end of it, until there’s nothing left.
Two, Ortolani mistakes “devices that don’t interoperate with other assistants” today for ‘this is the way things must be, will always be.’ Instead of presuming that and using it as a reason to have Apple make first-party accessories, the better path is one that is already beginning: what if accessories were compatible across all the major platforms?
Chambers proposes Apple buying a company that were started by ex-Apple employees, namely Level. Level makes a smart home door lock, and then more or less copying the Amazon/Google recipe of making one of each type of accessory: an outlet switch, a thermostat, a door lock, indoor/outdoor cameras with HomeKit Secure Video, and a mesh router system.
Chambers is partly right; There are a world of mesh router systems, from ones that hide all the complex options to ones that expose them all, and it’s hard to know which of these will just work with HomeKit out of the box. Reddit and forums are littered with advice like making sure mDNS is enabled. Apple got out of the router business because they weren’t sure what they could add. Now that’s more clear again.
Chambers is also correct about cameras. Following Eufy’s security failings, you’d have to look at Logitech, Eve, Aqara, or Ecobee. How do you know which one’s good? How would you have known that Eufy would handle their security so badly? For many Apple users, it would be easier to trust an Apple-branded camera than Eufy.
“If it wants to win in these new categories, it needs to follow its own philosophy of making the whole widget. HomePod mini isn’t much of a home controller if its accessories are mediocre at best, that is to say if anyone actually buys any.”
Ortolani
Ortolani is right to be concerned about the quality of third party accessories. Chamberlain’s MyQ garage opener has gotten poor reviews for ages. Insteon’s HomeKit devices were terrible in terms of reliability, with an excellent app early on. Belkin’s WeMO smart plugs were the first device to be HomeKit compatible without shipping with the 8 digit HomeKit code, and were unreliable for ages. Successive firmware updates have caused them to be stable, and reliable in use. Most accessories have been pretty reliable.
Ortolani is partly right about the second problem: Because many accessories are only made for use with Amazon Alexa and Google Assistant, there is a shortage of accessories for Apple HomeKit.
This is about to change. Apple, Google, and Amazon have been working together for the past couple of years on the Connected Home over IP project, which has been renamed Matter.
Matter is a way of unifying smart home interoperability by providing an agreed-on set of design criteria. It will work with WiFi and Thread devices. This means that any Matter certified device will work with any other Matter certified device, regardless of if you use HomeKit, Alexa, or Google. It will use Bluetooth for the initial provisioning of the device, and then use Thread to communicate the rest of the way. Thread is an evolution of Zigbee that was pioneered by Google back in 2015-2016.
Thread is the radio communication happens over (like Wi-Fi or Bluetooth, but Thread.) Matter is the way the communications are formed – a common lingo, so every accessory uses the same kind of packets and signals, and pairs to the local network in a common way so every major platform can host them.
The net result is that devices will be able to add the way Ortolani wishes (he envisions using Apple’s proprietary U1 and H1 chips to handle Bluetooth communications to get a device on the network) without using proprietary chips to do so. The experience should be good, should be cross-platform, and result in more accessories for HomeKit devotees, easier purchasing experiences, and really no need for Apple to get into making accessories.
So far, Matter does not support all the categories of devices that HomeKit supports, and it doesn’t support all the features: HomeKit has adaptive lighting for smart bulbs, that adjust color temperature throughout the day. It’s still early days, and we can expect an explosion of Matter and Thread devices over the next year.
Apple does not need to make HomeKit accessories. We just need more Matter / Thread devices. The goal should be a vibrant ecosystem with every device being first in class, so that you don’t have to know which devices are better than others.
