It's the new gold rush: the Internet of things, particularly the home automation segment. You have to watch out for fool's gold, bandits, price gougers, and bad stakes -- whether you're a user, provider, or developer.
Perhaps the biggest danger is that of the common platform. At this week's GigaOm Structure conference, a lot of those dangers are on display as vendors jockey for attention. Dozens of vendors are proposing "smart" home-automation hubs that promise to be your one-stop shop for controlling your thermostat, lights, door locks, windows shades, intercoms, monitoring cameras, security systems, and more.
Let's put aside the question of whether you need to remotely control or program all those items -- that control can be very handy for some home devices, though perhaps not for as many as vendors would like.
The fundamental question is whether you need a central controller for all those gadgets or whether they should be independent devices. If you step back from the hype, it should become clear there are only a few good reasons for a central controller.
One is to program related devices, such as lights you can set to turn on and off in natural patterns while you're on vacation to fool would-be burglars. If you ever used old-fashioned plug timers, you know how iffy the results can be. You might also want to manage security sensors, cameras, and door locks as an ensemble -- for example, to allow access to a neighbor or contractor.
Another is to provide a firewall to remote-controlled devices to reduce the risk that your home -- and its Internet-connected system -- gets hacked.
But there are better ways to address these few real needs. And there are very good reasons why a central hub is a very bad idea.
The dangers of the home-automation hub
Every vendor salivates over the prospect of creating a locked-in ecosystem. Unfortunately, many abuse their imprisoned customers. Think Comcast, Time Warner Cable, Charter, AT&T, Verizon, many insurers, some utilities, some government agencies, and many airlines.
Every potential home-automation buyer should be especially scared that the phone and cable providers are offering home-automation systems that cost real money every month and hold you hostage to their devices and apps. What are they providing that justifies a monthly fee?
Not every hub provider is a carrier, of course. But they offer the same core trap: A limited selection of devices and a single app. When I look at the limited device compatibility hub vendors provide, it's clear that buyers will need to replace most existing gadgets they want to control or end up with a hub that controls very little. You might swallow $250 for a thermostat, $50 for each switch, or $150 for a garage-door controller, but not $6,000 for a heating system or whatever an IoT washer/dryer might cost. Even those "little" items' costs add up to serious money if you have to replace more than a few of them.
Forget about interoperability. A new device may not work with your hub, and a new hub may not work with your devices (or apps, if you are a developer who counts on specific hubs' APIs). In fact, you can bet they won't be compatible, putting thousands of dollars of investment down the drain.
Also, you can forget about rapid app evolution in the hub approach.
People complain about the lock-in of Apple's $100 Apple TV or Google's forthcoming $99 Nexus Player. That's peanuts compared to what the home automation hubs' lock-in will cost you.
The better way to manage home automation
Plus, chances are you already have a hub: your Android or iOS device. It can run all the apps that connect whatever devices you have, not only those supported by a specific hub. A smartphone or tablet is not quite as universal as the Web, but it's darned close -- and much more suited as a remote control. Just as in the Web you go to the website you want, with mobile devices you run the individual apps for the iOT devices you want. There's no hub to get in the way or limit you.
With your mobile device as your "hub," adding new iOT devices are no problem. And each developer maintains its own apps, so the risk of poorly designed apps ruining your whole experience decreases. (If you've used a provider's DVR or website, you know how bad a provider's apps are likely to be.)
The missing link in the "your mobile device is your hub" approach is that different home devices won't be able to be managed collectively, such as in the scenarios I described earlier. But Apple's HomeKit APIs and new extensions capability should allow developers to provide such collectivity, and Google's interapplication communication support in Android, along with its hinted APIs, should do the same.
At some point, there needs to be a vendor with a de facto choke hold on the ecosystem by virtue of being that nexus. But I believe Apple and Google are better gatekeepers than the telcos, cable providers, and so on, whether you are a user, device maker, or developer. Unfortunately, Apple has yet to capitalize on its Apple TV as a HomeKit-based hub, as sort of a shared controller for iOS devices and Macs -- maybe next year.
As for the security need, that's both tougher and easier. It's easier because iOS is quite secure, and Android 5 Lollipop looks like it may be secure. But it's tougher because the frontline defense -- your broadband modem -- is as likely to come from your broadband provider, which could use that device to put competitors at a disadvantage, or simply be poorly designed in terms of security, as has been the history for such "consumer-grade" devices.
If Google and Apple used their TV devices as the firewall, that could sidestep the broadband modem risk. If not, I'd hope that Apple updates its very nice AirPort routers to be home-automation-savvy -- and Google hire a firm like Netgear or D-link to create a home-automation-savvy Nexus router.
If there must be a hub, it should be something designed as a flexible, broad, open-to-all platform. That rules out the hubs being hawked today.