What the 'Internet of things' really means

What do we talk about when we talk about the Internet of things?

Answer a call or go to a conference these days, and someone is likely trying to sell you on the concept of the Internet of things. However, the Internet of things doesn't necessarily involve the Internet, and sometimes things aren't actually on it, either.

In some cases, the Internet of things is simply a buzz phrase that companies use to sell whatever they've long had -- just as the cloud, green, Internet, e-, and mobile labels have long been abused. But there is a there there: The Internet of things has a real meaning that's useful to understand, as it will affect nearly every corner of both IT and consumer technology.

At its core, the Internet of things means just an environment that gathers information from multiple devices (computers, vehicles, smartphones, traffic lights, and almost anything with a sensor) and applications (anything from a social media app like Twitter to an e-commerce platform, from a manufacturing system to a traffic control system).

Basically, you need data and a means to access it -- that's where the "Internet" label comes from, though of course you don't need the Internet itself, or even an always-on network connection. The Internet may be the backbone of an Internet of things, but it's not the only bone in that body. Then you need something that works with that information to analyze it, act on it, or otherwise process it. That something is typically software, whether automated, semi-automated, or human-controlled.

The intrigue of the Internet of things

Where the Internet of things gets interesting is when you combine information from devices and other systems in novel ways, tapping into the huge processing capabilities available today to do the kinds of expansive analysis usually associated with the concept of big data -- meaning analysis of data not necessarily designed to be analyzed together.

Otherwise, you're talking about sensor networks and machine-to-machine (M2M) networks common in factories, hospitals, warehouses, and even streets (think the streetlights and "next bus" electronic signs) or network-connected product systems (like an Apple TV-based entertainment system, the Bluetooth stereo in your car, or iPod Touch-based cash registers in retailers) -- useful but not profoundly new.

To achieve the notion of the Internet of things, you need most of the following pieces in place:

  • Network connectivity, which is typically wireless
  • Sensors and/or user input that capture or generate data
  • Computational capabilities, at the device and/or back end

I say "most" because you could have a store-and-forward connectivity approach such as plugging a device into a USB port on a computer. Store-and-forward is essential in any case, because connectivity is not ubiquitous, so you need a way to send data captured when offline. That's a hallmark of the Internet, which was initially designed to allow communications even after a nuclear war through store-and-forward and auto-rerouting.

Putting the things in the Internet of things

You need things, but they need not be independent items like printers or earbuds or sneakers or golf clubs -- yes, there are golf clubs that monitor your swings and upload their data to apps that help teach you to golf better. A thing in an Internet of things could be simply status information, such as where you are or where the temperature is at a certain location or the engine temperature -- that may be collected through a general-purpose device such as a computer or smartphone. In other words, the thing itself need not be in an Internet of things, though data about it must.

And you need a purpose for having all these connected devices. There are thousands of possible purposes -- perhaps millions. That is why the Internet of things is not a thing but a concept that can be applied to all sorts of things. In most cases, those purposes are expressed through applications or services -- whether local, cloud-based, data center-based, or a combination of any or all of those.

In some cases, the services sift through huge amounts of data, which Hadoop and other big data technologies in combination with cloud services now makes possible. But an Internet of things doesn't have to involved big data -- there are small-data uses too, such as the Web of sensors on highways to detect chemical and nuclear weapons that are always monitoring but transmit only when an anomaly is detected. Combine that sensor network with traffic management systems, electronic highway signage and perhaps emergency broadcast notices, first-responder deployments, and so on, and you get a public-safety Internet of things.

Its versatility is what opens up so many possibilities for the Internet of things. For example, running an app like Foursquare or Google Now that monitors user locations takes an existing set of devices (smartphones), their sensors (location data), and their network connectivity to aggregate information to a data center somewhere in the cloud that uses that information for, in this case, ad delivery and market research. It's an example of how the Internet of things can simply be an application taking advantage of today's connected environment.

But an Internet of things can be more purpose-built, such as the devices that plug into your car's computer to transmit engine, speed, and other readings to your insurer (a bad idea!) or your smartphone (a better idea). At its most basic, this is just a sensor network in your car tying into a central transmitter. But the Internet of things twist is that some of that data would go to the government and private agencies that monitor traffic, feeding in real-time travel data to augment what they collect via in-road sensors and highway cameras.

Two (or more) is better than one

An Internet of things can enable hybrid uses. For the car example, multiple services might get pieces of that automobile and travel data for everything from traffic management to insurance rate-setting, from mechanics' diagnostics to road-repair prioritization.

As another hybrid example, think of all those health sensors available, like the Fitbit and Nike+ for personal health management, or like the Worthings blood pressure monitor or Agamatrix glucose monitor for medical monitoring. The personal ones expand the capabilities of the connected mobile world with a new sensor that sends data to an app in the cloud. But the medical ones may expand that same connected mobile world but send it to a medical provider's electronic health records (EHR) system. It's even possible that the two types of health sensors could cross-deliver, with your Fitbit data also going to the EHR and your physician-prescribed medical sensor also going to your personal health vault -- with each subset serving multiple purposes.

That notion of multiple purposes is probably the best reason for using the term "Internet of things," as the Internet is more than a resilient network but a conduit for any combination and collection of digital activities. The Internet started as a way for the government to communicate after nuclear war but has evolved to be much more than a network.

In many ways, the Internet has become a digital world that has gateways into our physical world. The Internet of things takes that concept to the next level, allowing multiple worlds -- some connected to others, some not -- that mash up physical and digital in all sorts of ways.

This article, "What the 'Internet of things' really means," was originally published at InfoWorld.com. Read more of Galen Gruman's Smart User blog. For the latest business technology news, follow InfoWorld.com on Twitter.

Copyright © 2013 IDG Communications, Inc.