The Internet of Things Revisited

By Mark Roberti

RFID is not really about connecting objects to the Internet; it's about capturing data needed to manage things that aren't being managed today.


I had the privilege of joining an illustrious panel on May 19 to discuss the so-called “Internet of Things” at the 2010 MIT CIO Symposium. The session was great—it was one of those rare panels that was both entertaining and informative. But I think the idea of the Internet of Things might have outlived its usefulness.

The panelists were Robert LeFort, CEO of Ember, a company that manufactures RFID sensors used to form mesh networks; Sanjay Sarma, a professor and co-founder of the Auto-ID Center at MIT; Bob Metcalfe, Ethernet inventor and founder of 3Com, who is now a partner at Polaris Ventures; and me. The panel was moderated brilliantly by Michael Chui, a senior fellow at the McKinsey Global Institute.

Sarma, a great speaker and an all-around smart guy, opened by explaining how he and David Brock came up with the original concept of putting low-cost RFID tags on everything and linking them to data stored in Internet databases. He said Kevin Ashton, the Auto-ID Center’s director, coined the term the Internet of Things.

Metcalfe was funny, charismatic and intelligent. He pointed out that 10 billion to 15 billion microcontrollers are shipped every year (chips used in everything from cars to coffeemakers). “These are not tags,” he stated. “Most are not networked. The value of these devices would increase exponentially if they were networked. There’s a law about that,” he said, referring humorously to Metcalfe’s Law.

LeFort made the point that we shouldn’t underestimate “the unbelievable power of convenience.” He said his RFID sensors could be used to adjust the temperature in different areas of a room or facility automatically, to maximize energy efficiency. People wouldn’t have to do anything.

We were all in agreement that RFID is a powerful technology that will deliver significant benefits for both business and consumers. I made the point, however, that while the Internet of Things is a useful term in that it helps people grasp the concept of being able to connect objects to the network, it probably has outlived its usefulness.

My main gripe with the term is that it discourages people from examining how RFID can deliver value to their companies today. First, it sounds too futuristic. Currently, the only “things” connected to the Internet are computers and cell phones. The idea that everything will be connected to the Internet one day—every chair, shirt, pallet of Tide and so forth—sounds like it won’t be real for 30 or more years.

Second, the term puts the emphasis in the wrong place. The Internet of Things suggests that what is important is connectivity, but the real value to business is in the data RFID provides. RFID systems tell you what an object is, where it is and perhaps its condition (too hot or too cold, for example). It allows you to collect that information automatically and cost-effectively. And it enables you to manage all of your business’ mobile assets that you were previously unable to see and manage.

Finally, the term suggests that if businesses connect objects to the Internet, they will automatically achieve some benefits, just as linking a computer to the Web provides immediate access to tons of information and entertainment. Companies need to focus not on connecting things to the Internet, but rather on collecting the data they need to better manage their business’ mobile aspects. RFID will shoulder most of that duty, collecting information on many things automatically. (For items that are too small or too inexpensive to justify RFID tags, bar codes will be a better option. And for things that travel too far to be tracked with a network of readers, GPS will be required.) So in my mind, it’s more important to think about RFID as an automatic data-collection technology than as a networking technology (though it is that as well).

Unfortunately, I don’t have a better term than the Internet of Things. But as someone who believes passionately that RFID can deliver tremendous business value, I don’t think it’s helpful for businesspeople to think of RFID as an abstract concept of soup cans and hairbrushes linked to the Internet. I’d like them to think of the technology as systems that can extend the reach of their enterprise resource planning applications, so that they can identify, track and manage assets, tools, work in process, parts, raw materials, vehicles and other mobile objects that aren’t presently well-managed.

If anyone knows of a term that captures this concept, please e-mail me at I promise to give due credit, of course.

Mark Roberti is the founder and editor of RFID Journal. If you would like to comment on this article, click on the link below. To read more of Mark’s opinions, visit the RFID Journal Blog, RFID Connect or the Editor’s Note archive.