The “internet of things” (IOT) has become a hot topic in the tech. press. While the term is relatively new, connecting devices to networks is certainly nothing new. IOT has become the catch-all name to describe smart-home sensors, embedded devices, and other smart-devices with internet connectivity, characterized by devices like the Nest thermostat but IOT includes everything from coffee makers to washing machines, lamps, wearable devices and about anything else you can think of. IOT has been envisioned for a long time. For example, consider the following from a 1999 paper related to Sun’s Jini project of the time:
The current trends in the hardware world point to a future in which more and more devices that we use every day will be enhanced with embedded processors and interconnected via networks.
That was 15 years ago… and we’re still waiting. If one looks through the various articles on the subject over the years, right up to today, we find very similar language, suggesting a glorious “internet of things” world right around the corner. Instead, we have a lot of fits and starts, balkanized islands, and no clear path to a viable ecosystem. A recent GigaOm article on the subject, rather than highlighting a bright future, to me summarizes the biggest problem: “Where’s the profit in the internet of things? Hint: not in the things.” It quotes a CEO of an IOT platform company as saying:
“The greatest opportunity in the internet of things is transforming a product business into a service business.”
And therein lies the problem. Companies may want to turn all of our “products” that we now buy and own into “services” that we pay for perpetually but consumers don’t necessarily want that – and neither do I.
I gave a talk at my former employer, EarthLink, circa 1999 describing this trend and explaining that an appliance company doesn’t really want to sell me a refrigerator anymore; they want to sell me a “food cooling service.” In the wild early days of the Internet, every business model was based on collecting $10 per month from users somehow. These days, it’s IOT startups who want their $10 per year, or some such subscription fee. The problem was then, and continues to be now, that there are only so many “$10 per years” to go around. The value isn’t there for most of these services so they never take off. And thus, IOT sits there, poised for stunning growth, but never realizing that growth in practice.
I want these devices to be self-contained. If they’re smart, they should be so smart that they are their own service. If I want to collect data from the device, I want to do that in my cloud, not in their proprietary silo cloud. The problem is, nobody can sell that business model to Silicon Valley VC investors. They want to collect our data, use our data, monetize our data. It’s the holy grail for IOT companies, to be the hub for everyone’s data. They all want to be that hub, that centralized place for all our data. But of course nobody wants to grant them that valued spot, so instead, in practice we end up with a bunch of proprietary silos.
My phone is smart enough and has the ability to find and connect directly with my connected garage door opener without a “service provider” in the cloud. That’s exactly how my connected garage door opener works – but that’s only because I built it myself to operate that way. None of the commercially available products work that way; instead, they provide a “service” to act as the intermediary. In some cases, those services are “free” in other cases they are provided for a fee – but in either case, our connected garage door opener “product” is now a “service,” reliant upon a service-provider in the cloud, and exposing our data to that service provider. Again, this is the only business model the industry seems able to conceive of. Nobody has been able to fund a business based on a self-contained, distributed, server-less model – a model, where the smart-devices themselves are the servers, as such.
We don’t need all these external services in the cloud. The hardware has gotten cheap enough and standards have evolved to the point where what one can embed into a device for under $100 (well under) represents as much computing and networking power as a typical cloud server just a few years ago, making that device plenty powerful enough to be totally self-contained, client and server, without the need for any “service provider” in the sky.
It’s a failure to communicate on many levels, but beyond the problems of lack of interoperability standards, platform battles, privacy concerns, and a slew of other issues, the fundamental problem is that what the IOT companies want people to want isn’t what people actually want. As long as IOT companies continue to ram “subscription” models down our throats, and as long as venture capitalists and the tech. media keep pushing for them as the only option, IOT will never expand beyond DIY hobbyists and niche markets.