ThingMagic’s Rollercoaster Journey—From the Internet of Things to the Calculus of Reality
ThingMagic is a 10-year-old technology company whose core idea seems as fresh today as when it first started. The bad news: that means it may have been too far ahead of its time. The good news: times are changing.
The Cambridge, MA-based firm was founded by five MIT Media Lab alums, who had the goal of “adding magic to everyday objects”—hence the company’s name. This “magic” came in the form of radio frequency identification (RFID) tags, which enable wireless communication by way of tiny electronic chips that can be embedded in things, accompanied by readers and software to make sense of what each tagged item is, and track its whereabouts. The big vision was to create an “Internet of things,” so people could retrieve information about the objects around them—everything from product inventory on shelves to stuff in your home, office, or car.
Sounds a bit far out even today, right? Well, it was far more so in the early 2000s, and the evolution of RFID ever since has been quite a rollercoaster ride. To make a long story short, the technology was strong but its business use was overhyped, so it got stuck on the adoption curve. Tech companies based around RFID have come and gone, but some have endured, such as Alien Technology, Impinj (which I wrote about here), and ThingMagic, which all have raised a fair amount of venture funding.
In July 2008, my colleague Wade wrote an in-depth piece on ThingMagic, focusing on the company’s progress in shrinking its RFID readers (its main product) down to a size where they could be put into places like offices and hospitals—a big step toward realizing the Internet of things. Around the same time, Mark Roberti, the founder and editor of RFID Journal told me that it would be another two to three years before the RFID market would take off, because end users were still figuring out the physics and economics of tags and readers.
Well, it has been two years, and I’m wondering what has changed in the RFID world. To get some answers—and an update to the company’s story—I recently sat down with ThingMagic co-founders Yael Maguire and Ravi Pappu, and director of marketing Ken Lynch, to talk about what lessons they’ve learned over the past decade. We met at the company’s new digs at One Cambridge Center in Kendall Square.
“After 10 years, we’ve seen literally hundreds and hundreds of ideas for using the technology,” says Maguire, the company’s chief technology officer. “It may not be ubiquitous, but in most cases it’s caught up with people’s imagination. People are focusing on how to deploy it.” Pappu, who runs product development and implementation, puts it this way: “The story is changing from RFID on everything to RFID in everything. That’s always been the vision of the Internet of things.”
It’s certainly in ThingMagic’s interest to promote all the new ways RFID is being used. But beyond any PR spin, there’s something real going on. Yes, there is the recent news that Wal-Mart plans to put RFID tags on apparel like jeans and underwear starting this month, to help track inventory more efficiently. Some observers are calling this a “comeback” for RFID, because these kinds of retail applications were hyped in 2004-2005 but failed to materialize in a big way. (ThingMagic had worked with Wal-Mart and other retailers on using RFID to increase supply-chain efficiency, but ultimately these deals fizzled in part because of economics and consumer privacy issues.)
But meanwhile, other uses of the technology have emerged. The new Disney Family Cancer Center in Burbank, CA, is using RFID badges and readers to help keep track of patients and customize its treatment rooms in real-time by tapping into room temperature, lighting, and nurse call systems. Greenville Hospital in South Carolina is using tags and readers to keep an eye on surgical tools in operating rooms. The Florida State Attorney’s office uses RFID to track case files. Ford and DeWalt use tags and readers to help customers track tools in the back of pickup trucks. Car rental companies use RFID readers to check in returned vehicles. And the MIT Media Lab and other places are experimenting with RFID to create smart, personalized computer interfaces for visitors logging in. All of these are ThingMagic customers.
So although the latest Wal-Mart news is good for RFID companies—as far as I know, Wal-Mart hasn’t said whose technology it’s using—ThingMagic has learned not to put all its eggs in one retail basket. “We got burned pretty early by the supply chain stuff,” Maguire says. Pappu adds, “In some sense we created it. We were out there beating the drum. It took much longer, and it happened in different ways. We placed a very big bet. Before that, we were profitable and required no venture capital. We went in and raised VC to support that. That didn’t pan out. But fortunately we were able to learn from it.”
Back in the days when ThingMagic was focused on supply chains, 80 percent of its revenue came from its top two customers, Pappu says. Now the company is much more diversified across other areas, such as building RFID reader modules for a wide range of customers, and consulting. What’s more, he says, 35 to 45 percent of its sales come from overseas. In terms of revenues, Pappu says, they “grew last year compared to the prior year, by a tiny amount. Things are going much better this year than that. People have turned the corner. In our quarterly revenue projections, you can see when the collapse began, and when it started to turn around.”
And more broadly, looking back over its 10 years, ThingMagic has reinvented itself a couple of times. It started as a consulting company, became focused on retail supply chains, and now has settled in to what seems like a good business delivering reader modules. The company currently has 32 employees, down from about 60 at its peak in 2006. Its venture investors have included Tudor Ventures, The Exxel Group, Cisco Systems, Morningside Technology Ventures, and In-Q-Tel.
A big reason why ThingMagic and other RFID companies are still around is that the hardware has finally become cheap enough (tags cost less than 5 cents each) to see widespread use—not to be put on everything, but on a lot of things. “They’re cheaper and also work better,” Pappu says. “At this point, you can read tags at 50 feet in free space, they come in a number of shapes and sizes, and can withstand any operating environment…The readers can be anywhere too now. They’re not tethered to a desk or portal, but in handhelds, in label printers. The third thing is, it’s not just about the RFID. It’s about reaching out to the tag, reading data, and connecting to other systems, like Bluetooth, GPS, Wi-Fi…By combining those, you now have the Internet of things.”
OK, with their business issues seemingly under control, I could ask these MIT Ph.D. guys a little more about their grand vision and thinking. For instance, are Google, Bing, and other Web search companies getting interested in building a “reality search engine” that incorporates RFID data?
“Google is really interested in mapping, but they haven’t crossed the boundary from the street to the building,” says Maguire. “We’re starting from inside. There should be overlap quite soon.”
Pappu took it a bit further. “Think about all the things you could look for. Think about the calculus of reality,” he says. “There’s a certain scale you can put Wi-Fi on—a laptop, phone. But the next level down does not admit batteries.” He’s talking about the idea that tiny sensors could be placed in all of the items you interact with every day, and information from those sensors could feed into a centralized database that keeps track of the physical state of everything in the world. “The interesting premise is, don’t make any changes to the interface, especially the interface to the human. Let them be how they are, and see how you can do this calculus without affecting them.”
It’s not really clear yet what the best applications of all this would be. But tying RFID into the exploding sector of location-based services and mobile applications certainly seems intriguing. You can see this technology as a form of artificial intelligence, Maguire says. “The computer is there observing the world in a very distributed way. Every object is like a little Twitter feed, and something has to make sense of it.” (Depending on whom you ask, that last bit is either the most useful, or most annoying concept ever.)
But back to the founders’ own calculus of reality—specifically, the future of ThingMagic. Pappu and Maguire are squarely focused on their next big deals. Pappu talks about the big opportunities for sensors in healthcare, and posits that within five years, there will be at least one hospital that is completely RFID-enabled. “Somebody has to place that bet and create that infrastructure, and prove to themselves it’s going to happen,” he says.
Maguire sees at least one short-term business trend that also bodes well. “Scale is one of the things we’re seeing,” he says. “The demand from customers is unprecedented this summer. For the first time, we couldn’t meet demand.”