Why dumb devices rule the smart home: a cautionary tale from Apple's HomeKit

So if you hate Apple, you’ve probably been rejoicing in the trickle of news stories in the last week about various slippages regarding Apple’s HomeKit initiative.

Not one of the HomeKit-compatible devices announced for “Spring” has come to market and the promised featured role for Home Automation at the developers conference in June appears to be shifting to a slide show backup as Beats takes center stage instead.

Retail-ready HomeKit is coming certainly–but August 2015 is starting to look like a risky bet, and my own wild guess of late Spring 2016 looks more plausible every day.

But setting aside brand beating for a moment, if you look at this from a brand agnostic perspective and you just really like cool devices, there’s an important lesson in what may have gone wrong for this first stage of device development.

My candle burns at both ends, it will not last the night;
But, ah, my foes, and oh, my friends, it makes a lovely sight.

It often turns out to be simple: when battery-operated tiny devices get too smart, they die. I mean literally. Battery life starts to be counted in hours. Or a couple hundred minutes. Instead of the “6 months to a year minimum” most home automation trials have shown real people demand in their real homes.

HomeKit smarts reportedly required too much memory to be supported in the desired form factor. From door locks to contact sensors, it was an Edna St Vincent Millay spectacle of tiny candles burning brightly and dying quickly. Way too quickly for retail release.

Multiple device makers using multiple HomeKit chip sources all, according to rumor, with the same issue. Too much memory required for available battery size. So nothing came to market.

Meanwhile, zigbee and zwave devices chug along. They sleep a lot. They’re pretty stupid. Communication is asynchronous, and they know very little and usually say less. But they survive. For months, even for years. They guard the gates and light the lights and lock the doors and measure both heat and humidity.

So the next time you make a wish list of all the features you think a battery operated home automation device should provide, including more frequent reporting, and you plaintively ask the question, “how hard would it be to just…” think of all those tiny candles snuffing out, and know the engineering answer: “very hard.”

*Apple’s official position is that no announced dates were missed, which is true. Most dates were implied or rumored. But if we do get HA device releases in time for the June conference, look closely at the power sources. Are they battery-operated? Is there a second plug in piece handling security? Time will tell.


So far I have found exactly one HomeKit-compatible third party device you can actually own today (as oppose to just pre order): the Withings Home baby monitor camera. It’s a plug in, not battery operated, even though none of the marketing photos show the cord.

I can’t speak to the quality of the camera itself, but it is enough to allow Apple to say some HomeKit accessories are available now. Of course it doesn’t actually work with Siri yet, that has to wait for Apple’s release of the Home app.

I have yet to see Apple offer something that impresses me in the IoT… That being said, I am not a huge fan of Apple, but I will always recognize someone who is doing something right…

HomeKit has been all smoking mirrors to this point. I love Apple products as I own many of them including an IPhone, iPad, Apple Watch, MacBook, MacMini and AppleTV. I am taking a wait and see approach to see how HomeKit plays out but I am not hopeful. Insteon’s new hub claims to be compatible with it:

Insteon’s new HomeKit-enabled hub has been available for pre order for several months, but is still not shipping.


And Insteon says they haven’t released it yet because Apple has not yet released the controller app. from official responses to reviews on the pre-order page:

Customer 1: Homekit is available so I’m confused as to why you’re waiting. Trudy Muller at Apple stated the other day -

“HomeKit [hardware certification] has been available for just a few months and we already have dozens of partners who have committed to bringing HomeKit accessories to market and we’re looking forward to the first ones coming next month,”

Just release it already!

Smarthome:: We appreciate everyone’s excitement surrounding the release of Insteon’s HomeKit-enabled Hub. We also can’t wait to start controlling Insteon using Siri. The quote referenced above means that the hardware certifications of HomeKit-enabled hardware has been available but to date nothing has been officially released. She does say “first ones coming next month”. Stay tuned!


Customer 2: Enough with the delay!

Smarthome: We totally understand your frustration. We are also anxious to play with this HomeKit-enabled Insteon hub. The release of this product is tied with the public availability of the HomeKit app which is not yet released by Apple.

Yeah I have a co worker who has been taking up HomeKit for the past year and even pre-ordered the Insteon Hub. I’m just sick of hearing about it and want to see some tangible products and information. It’s all been very vague. As a shareholder, I hope they pull it off because the integration between Siri, Apple Watch and HomeKit can be special. Makes you also appreciate what Smartthings is doing as they are moving at a faster pace with much less people and resources.

IMHO, the ideal device is powered, with battery backup for at least a day or two (ideally a couple of weeks). While there is appeal for cheap, battery op’d devices, I am not sure I want to see the system designed around their limitations.

I would much prefer instant communication and programmable behavior that smarter, powered devices would bring. When you design for limited cloud traffic and limited power use by devices, you are depriving a home automation system of what it needs most to work well… immediate, and up to the second feedback.

Got no love for sitting in the driveway waiting for my presence sensor to log me home before opening the garage door. Before this was automated, I pressed a single button when I hit my street that had the door opened by the time I reached the driveway. This is hardly the improvement necessary to get the average Joe to install HA.


I agree that while energy is cheap, instant feedback and forced sequencing are well worth the cost for most people.

The biggest challenge, I think, is door locks. Unless we’re going to buy all new doors, or come up with a whole new design which moves the actuator to the wall instead of on the door, it makes sense to use batteries in a door lock. But only if those batteries last at least a year.

Sensors are a challenge from the other side, cost. Contact and motion sensors can certainly be easily used as plug ins. It just raises the cost of the device significantly. And for some homes there’s a convenience issue. The receptacles aren’t always near where the sensors go, and if you hardwire it adds a lot of cost.

But door locks are the ones where we’d need to rethink the whole approach.

Only if pin codes are needed. If not, we could go back to the '70’s technology of electric, latchable door strikes. And if pins are needed, a keypad could be installed.

While the all in one solutions are more convenient (except for the batteries of course), I have yet to see one that wasn’t an ugly, overpriced, piece of junk.

1 Like

If batteries are inevitable, perhaps devices with possibly short battery life need easy quick swap and recharge modules like on power tools. Off the self batteries are convenient to buy, but not the easiest to swap. Or we need to run “power over wifi”, right? If it works for Ethernet then why not wireless? :wink:

It does… http://ubeam.com/

Rather, “it may…someday.”

UBeam has some interesting patents, but the device is still just in the prototype stage. And then there’s this:

The uBeam charging capabilities do have some serious limitations, including the power transmitters’ inability to beam through walls. This means that unlike Wi-Fi hotspots, where a single device can transmit Internet to an entire house or small office, uBeam users would have to buy transmitters for each room.

Well it already works, so I would go with “it will…someday” (c;

Lots of prototypes work that either never make it to retail production, or make it and then fail because real world field conditions prove insurmountable.

For a recent example, see the original Lockitron.

The focus on HomeKit devices feels wrong. Yes, there were such devices shown off and expected, but the bulk of HomeKit is an app framework for iOS. Apple’s definition, copied and pasted from the HomeKit developer site:

HomeKit is a framework in iOS 8 for communicating with and controlling connected accessories in a user’s home. You can enable users to discover HomeKit accessories in their home and configure them, or you can create actions to control those devices. Users can group actions together and trigger them using Siri.

Yes, there is a hardware angle to HomeKit as well. But I think that if folks like us are going to find value in HomeKit, it’s going to be on the app side of things where those apps are controlling devices we already own. Again from the HomeKit developer site:

If your iOS app is primarily designed to provide home configuration or home automation services such as turning on a light or opening a garage door, learn more about the HomeKit APIs used for communicating with HomeKit accessories.

Check it out for yourself, and draw your own conclusions. I don’t claim to be a HomeKit guru, or anything close.

I think the real cautionary tale is in how many click-bait articles are written about Apple simply in order to drive ad impressions, not because they have any real insight.

1 Like

I’m confused by this comment.

:There’s a two-way authentication process required for HomeKit, and if you follow the links in the same passages you quote, a “HomeKit accessory” is primarily something certified through the MFi program that requires a special hardware chip to do that authentication.

The main exceptions that have been discussed are devices that “don’t control the home,” which means sensors that report data and a possible exception for zigbee lightbulbs that connect through an MFi certified bridge.

They’ve specifically said that, for example, Nest thermostats will not work with HomeKit.

There’s a required app that’s needed and has not yet been released that will discover devices and allow Siri to communicate with them. But the devices that can be discovered are limited to those that can do the authentication.

I haven’t seen anything coming out of Apple saying that HomeKit will work with a lot of devices we already own, with the exception of lightbulbs communicating through a HomeKit-updated bridge device. And, of course, anything running iOS.

It would be great if Apple opened up Siri control to a lot of the devices we already own, but I’m not seeing that in any of the information released so far.

Those HomeKit APIs are for “communicating with HomeKit accessories.” And so far that means devices that have the special chip. It’s not a firmware update you could apply to a zwave doorlock, for example.

There was one early article last fall that jumped on the first brief mention of a “bridge” and assumed that meant that a hub like SmartThings would be able to control all its devices via HomeKit. But as more details were released through the developer program in late December and January it became clear that the “bridge” idea was much more limited than people had first supposed. Instead, it looks more and more like the HomeKit “bridge” concept was intended for the Phillips Hue Bridge, assuming that device comes out in a HomeKit version.

But who knows? We’ll see what happens when everything is finally available to buy, including the Insteon Homekit-certified hub.

Yeah, HomeKit really does seem much more basic and wary of existing devices. And I’m being very hopeful of how it can add to HA - but that’s not an Apple strategy.

I’ve now read comments in articles saying that HomeKit is not intended for whole home control. Maybe that’s true of the initial offering, but I’ll bet that it is the intention for the long term.

The reason I’ve seen repeatedly for HomeKit not supporting the Nest thermostat is that HomeKit won’t bridge wifi devices because of security concerns. What exactly they’re referring to, and whether that’s the only reason, I don’t know.

There are a few interpretations of bridged devices not being allowed to control the home. You mention the “sensors, not actuators” definition (and I apologize, I’m greatly simplifying it).

One article claimed that controlling light bulbs falls controlling the home, and thus would not be allowed as a bridged accessory. I call BS on that because of the Hue example (Philips one of the early ones on the HomeKit bandwagon, right?). But I could be totally off and that refers to a complete redesigned Hue product for HomeKit.

My interpretation is the bridge/hub distinction where a bridge is dumb, and only translates HomeKit to z-wave/zigbee/insteon but doesn’t add any advanced logic or inter-device logic, instead deferring to HomeKit. But again, that’s me being optimistic. Oh, and devices that provide access to the home must not be bridged (locks/garage door opener).

Apple is all about the user experience. And they’re going to want to control and create that experience from the ground up. Existing devices don’t fit neatly into that strategy, and as much as I’d like otherwise, Apple doesn’t support things that don’t fit their strategy.

But I agree, we truly won’t know what it will and won’t do until V1 is fully released. Even then, we won’t know the roadmap for V2, V3, etc.


I first have to say that starting a thread with professed hatred of a company (a company that continuously releases incredibly useful and popular products) is a decent indication that every statement that follows is, at best, hugely biased speculation.

As to the speculation and conclusions, I think this thread will one day be given new life by someone pointing out how wrong it was. Apple is not a company that rushes products to the market. It doesn’t need to. It just has to offer something better than anyone else has offered. My guess is that people working on HomeKit have considered everything you have and a lot more. I also guess that they are capable of providing solutions and bringing a solid product (service if you prefer) to the market. That’s what happens when you have top talent and bottomless pockets.

I suspect you misread my first post.

I don’t hate Apple products–in fact, I’m almost totally dependent on them, and like them very much.

I am quadriparetic, with limited hand control. As I have posted often on this forum, Apple’s commitment to accessibility at the operating system level means it tends to work better for most people with similar issues than Android does. (I’ve posted links to other analysts explaining this issue in further detail for those who are interested.)

My point in the first paragraph was that while there are undeniably some community members with brand prejudices, the “candle at both ends” issue arises for many battery powered devices, including zigbee and zwave versions.

As for myself: I have an iPad mounted on my wheelchair, two iPad minis, an iPhone, and an Apple watch. Search the forums for any of the existing HomeKit discussions. I think my “nonhater” credentials are pretty well established in this community. :wink:

1 Like