A discussion on one of the lists I follow covers the question of “platform” support for the Internet of Things. Since consumer IoT is a kind of squooshy market right now with rapid changes in prices and no one obvious choice for what technology to adopt, it’s worthwhile taking some kind of review of the landscape for 2015 and sort out what to watch for next year.
This is organized by vendor, and loosely grouped so that similar systems are next to each other. It’s a highly dimensional space, but all elements have something in common: engineers are trying to change how we interact with simple devices like light bulbs through our hand-held phones, and no one has the full correct answer that’s obviously right. Indeed it’s not clear at all that there is a right answer for some of these controls; the Internet of Shit twitter account collects with some glee the number of devices that emphatically should not be on the Internet.
On to the platforms:
Amazon is hard to ignore. Many smaller cloud platforms are built on top of its AWS computing platforms, and Amazon has entered the fray with its AWS IoT service that offers a hosted MQTT message server. In addition Amazon has the Echo product that does voice recognition. Nearly impossible to ignore, Amazon is a clear central player in the IoT consumer space.
Google is a distant second in the IoT world, mostly because the Google Compute Engine isn’t the first choice for a compute platform, and the Android phone isn’t the first choice for a top end phone. Google does have the moxie to make it’s “Internet Thing” be a self-driving car. The Nest thermostat is the most obvious home point-of-entry for Google, and the platform and brand extension to watch is more stuff that works with Nest.
Apple has the phone to want, and the assistant (Siri) that people have learned how to talk to to get things done. If the Internet of Things is going to be controlled from an app, it’s an app that’s going to run on an Apple phone, and you’re going to be able to talk to it. Less clear is Apple’s in-home device hub strategy - the AppleTV never really has had that open platform feel to attract the first-mover developer.
Microsoft would dearly love to play in this world, but Windows 10 on ARM and the Raspberry Pi is a disappointment, and “big data” is mostly defined by problems too big for Excel. The Microsoft cloud platform, Azure, mostly seems to be used by organizations that treat it as a second or third data center, or who are locked in to Azure for business and partnership reasons. Microsoft’s Cortana is capable, but hardly anyone has Windows phones to talk with.
IBM is the wild card here. Their Bluemix hosted platform service is interesting if unusual and not really aimed at the consumer space. A very interesting Node-RED project has extended IBM influence into the Raspberry Pi and Bluetooth BLE sensor space, but it feels like an interesting technology play and not a product.
Next, to major players with hardware. “How many engineers does it take to change a lightbulb?”
Philips has the Hue line of home products, mostly lighting focused. There was a hue and cry about the Philips decision to disable support for some compatible products that did not have explicit partnership agreements in place, and the nature of the complaining made it clear that there’s an active hobbyist community around this product space.
Flex is the new player, with its recent acquisition of the assets of Wink from the bankruptcy of Quirky. Flex, formerly known as Flextronics, is a large contract manufacturer, and it knows how to build quite a few things; look for Wink’s technology to make its way into a wider range of products.
Samsung and their SmartThings brand has enough market share to make interesting efforts, but again to that point there’s no one breakout product in their portfolio that brings that brand much to mind unless you are using it.
The biggest winner in the IoT space for this year and probably for next year is an Internet tool with the unlikely name of IFTTT. “If this, then that” is the level of control you get with IFTTT, and surprisingly enough that’s enough to do all sorts of things - mostly because IFTTT speaks just enough of the protocols of over 100 systems that you can glue together a recipe that does something when something else happens even when the two vendors you are dealing with have never heard of each other.
Interoperability is the bugaboo of Internet of Things systems development. Many vendors want you to sign up to their big bear hug of a partnership agreement and work solely within their ecosystem. Others have narrow APIs and proprietary protocols that discourage casual reuse. Where IFTTT fits in is as a universal connector kit that lets you explore how parts fit together. The IFTTT recipes for the Internet of Things captures some of the essence of what can be done right now without much of a fuss if you have wide protocol support, even if it’s only very simple support. Many vendors in the “Internet of Things” space could benefit from thorough understanding of simple connectivity.