Why the 'Internet of Things' may never happen - Mike Elgan

Why the 'Internet of Things' may never happen - Mike Elgan 2014Jan18

It's also a lousy name for a great idea that is doomed from the start. Here's why.

Computerworld - Research firm Gartner says the "Internet of Things" will have 26 billion connected devices by 2020.

Maybe. But connected to what? And how? Here's what you need to know about the "Internet of Things" phenomenon.

There will be no 'Internet of Things'

The label "Internet of Things" is used to describe Internet-connected devices that communicate without human involvement.

For example, as you read this article, you're using the regular Internet. You're a human being who is communicating with another human being (Yours Truly), and this communication is facilitated by many other human beings (editors, web designers, engineers, etc.). Like Soylent Green, the Internet is made out of people -- and computers whose main purpose is to help people use the Internet.

The "Internet of Things" is different mainly in that it's not made out of people.

Let's imagine a scenario 10 years into the future when the "Internet of Things" is supposed to be established. You come home with a hypothetical "smart toaster," which connects to the Internet. You plug it into a kitchen outlet. The toaster boots up, finds the home Wi-Fi network and sends out a query to all the other smart devices registered to you. Your alarm clock, smart toothbrush, TV, smartphones, tablets, PCs, smart glasses, smart smoke detector, home automation base station, smart clothes, smart fridge, smart washer and dryer and smart kitty litter box each in turn introduces itself to the toaster, telling its unique identifiers and what they're capable of doing. The toaster responds in kind. In the future, the toaster can send and receive instructions from other devices.

For example, you have friends over for breakfast and make several slices of toast. There's a lot of heat and a little smoke, and your smart smoke detector suspects a fire. So it sends out a message to the other devices saying, in effect, "is anyone creating heat and smoke?" The toaster can respond the equivalent of: "Yeah, it's me. No fire here and nothing to be alarmed about." So the smoke alarm doesn't sound.

"Things" are connecting to each other and interoperating without human involvement. That's one consumery example of the "Internet of Things." (There will be industrial and other applications on a massive scale.)

The "Internet of Things" is a bad name because "things" don't have their own Internet. They use the regular Internet. There is no separate "Internet of Things."

"Things of the Internet" would be closer. And "things that interact with other things without human involvement" would be even more accurate.

Another reason why the "Internet of Things" is a bad name is that the devices can make these connections without using the Internet. Some can connect peer-to-peer, or over a local network, without going online. The ability to connect to the Internet is not a necessary criterion for inclusion in the "Internet of Things" category.

Oh, and one more (fatal) problem

There's one more problem with the label "Internet of Things" -- it implies Internet-like compatibility and universality of communication standards that may never happen.

The basic standards for the Internet were developed before there were powerful companies with a vested interest in excluding competitors from markets. By the time the big Internet companies were rich enough to throw billions of dollars around to get their way, the standards, such as TCP/IP and others that make the Internet universal, were already well established.

This is not the case for the Internet of Things. The phenomenon is arising in an industrial environment of powerful companies that each want an unlevel playing field in their favor, or that have strong and mutually exclusive ideas about how the industry should work.

Former Apple executive Jean-Louis Gassée calls it the "basket of remotes" problem.

His "basket case," if you will, is the ugly fact that most TV owners have a "basket of remotes" that remain forever confusing, incompatible, unprogrammed and mostly unused. Even technical users who can, don't, Gassée pointed out in a recent blog post.

The industry came up with multiple schemes, products and standards to usher in "universal remotes." But this dream has gone mostly unrealized.

If we can't figure out a simple problem like self-programming TV remotes, how are we going to build a self-programming, universal and compatible "Internet of Things"? If we can't make one TV remote communicate with three home entertainment devices, how are we going to get 26 billion devices to all talk to, and work with, each other?

Gassée points out what is obviously true: The dream of the "Internet of Things" isn't going to happen without universal standards for discovery and communication between devices.

His vision is to add every "Internet of Things" device to a detailed database that tells what it is, what it does and how it does it -- a "definitive, curated, source of systematic knowledge about connected devices."

He goes further. The data output from all connected devices would be applied to the Wolfram Data Framework, uploaded to the Wolfram Cloud and would be visualized and analyzed by the Wolfram Language.

Wolfram, Wolfram, Wolfram.

Stephen Wolfram is brilliant and capable of doing all this. His intentions are good. But this is a first salvo of a powerful person or organization saying: "Hey, we're doing to do the "Internet of Things" OUR way.

Other geniuses and other organizations will step forward and say: "No, we're going to do it OUR way." And so on.

Another organization launched this week with its own vision of how to create a universal system for connecting, a kind of DNS system for the Internet of Things.

In the same way a website has an assigned and universally agreed-upon name -- say, computerworld.com, for example -- "Internet of Things" devices can each register and receive a unique identifier for addressing and communication by other devices.

Called The Wireless Registry, the service is already selling participation in the scheme. The first registration is free, and after that it's $4.99 per year per device.

Hmm, let's see. Multiply 26 billion devices by $4.99 and The Wireless Registry is looking at a $130 billion business. Per year!

Does anyone think the rest of the industry is going to cede this market to a faceless startup? There will almost certainly be others stepping forward to establish their own incompatible systems of device registration.

It seems to me that the so-called "Internet of Things" will be littered with multiple, warring, incompatible standards, protocols and systems for connectivity, making it very unlike the actual Internet and looking more like Gassée's basket of remotes.

These divisions will split industries and alliances within industries. Nations and regions will disagree. The U.S. will try to control the global system, and the American models will be rejected out of fear of NSA spying. China will create its own set of standards in order to maintain control. The EU will regulate the standards to death, stifling innovation. Google will spend billions establishing its own vision, which Apple will reject in its totality. All that will happen if tomorrow's industry acts anything like today's.

Unless something changes, there will be no "Internet of Things." Just a lot of things that connect over the Internet, but not necessarily to each other.

It's too bad, because the "Internet of Things" would have been really awesome.

This article, Why the 'Internet of Things' may never happen, was originally published at Computerworld.com.

Mike Elgan writes about technology and tech culture.

.END