Getting everyday objects to think (Q&A)
Thomas Lee hopes to bring the Internet of Things into the mainstream. If the tech industry can play nice, he may succeed.
If the so-called Internet of Things ever becomes one of those proverbial next big things in technology, Thomas Lee, a card-carrying member of the cohort of big brains who populate Silicon Valley, figures to play a pivotal role.
Lee teaches electrical engineering at Stanford University and is the author of such page-turners as "The Design and Implementation of Low-Power CMOS Radio Receivers" and "Planar Microwave Engineering: A Practical Guide to Theory, Measurement, and Circuits." He also won the 2011 Ho-Am Prize in Engineering, informally known as "the Korean Nobel Prize."
But an ingrained entrepreneurial itch has taken Lee beyond the confines of the ivory tower. In 1998 he co-founded Matrix Semiconductor, a 3D-memory manufacturer that SanDisk bought in 2006. A few years later he founded ZeroG Wireless, which made low-power Wi-Fi chips. That company got acquired by Microchip Technology in 2010. The next year Lee took a leave from the university to put in a stint directing the Microsystems Technology Office of DARPA.
Now he's back in the game, this time as the co-founder of Ayla Networks, a startup offering a way for everyday products -- items like home controls or lights -- to interconnect so that consumers can check status updates or adjust their appliance settings over a wireless network.
This is a concept that's been around for at least a couple of decades. Former PARC chief scientist Mark Weiser wrote about "ubiquitous computing" in a 1991 article that appeared in Scientific American. Three years later, Kevin Ashton, then working for Procter & Gamble, expanded on the need for "an internet for things." Their ultimate goal was an ubernetwork of everyday objects that get turned into so-called smart objects with built-in networking capabilities and sensors -- and that, of course, connect to the Internet.
If the idea takes off, Ayla could rake in a windfall. The company's cloud takes care of the heavy lifting so manufacturers can quickly design devices that can communicate with each other; in return, Ayla would charge a small royalty on each device that gets connected to its Internet of Things platform. If you're looking for business analogies, recall that Microsoft grew rich charging royalties on each computer that went out the door with its DOS or Windows operating systems.
Is history likely to repeat? For now the only clear answer is that there's no shortage of hype surrounding the Internet of Things, aka IOT. Though Ayla presents its own cloud as the answer, other companies aren't blind to the opportunities. If they offer competing clouds, will they interoperate? If past is prologue and the tech industry repeats its history of squabbling over standards, the future of the Internet of Things might take a long detour. Then again, the industry might coalesce around common protocols for communications, similar to the way that Internet domains embraced the IP standard.
Ayla recently closed a $14.5 million Series B round of funding. It also gave an update on its earlier announced collaboration with Sina, which uses Ayla's cloud for a Wi-Fi weather station that offers updates and forecasts that Chinese consumers can access via smartphones and other devices. After the announcement, I chatted with Lee about how he envisions the what lies ahead for the Internet of Things. The following is an edited transcript of our conversation.
Q: Why is it so hard to define the Internet of Things?
Lee: If you were to poll 10 people, you would probably get 20 different definitions of what it is. The Internet of Things concept started to show up in the late 1990s. When I first heard it, it meant machine-to-machine communications. So back then you had industrial machines and petrochemical plants chugging away as islands unto themselves and then humans would ping them from time to time.
Since then, the trajectory's surprised people. Humans are now part of the loop in a very intimate way, and I think that will continue to be true as we inhabit a far richer, more interesting world where we're increasingly interconnected with things.
I see the rise of the smartphone as having become a kind of digital remote control for my life. And as IOT continues to evolve, it will be the thing that I use to turn things on and off.
As soon as I hop onto an airplane I start to fidget -- did I close the garage door or leave the sprinkler on? Those may be trivial apps, but I'm willing to bet they'll have wider appeal than some techno geeks are betting.
Why should we care? How is this going to affect peoples' lives?
Lee: It already has in lots of different ways. I was looking at just the smartphone average revenue per user and what fraction is due to data and how much is due to voice. There's been a nice convergence of these two curves. About a year ago, there was a crossover and we became a digital-consuming species. I think it will be recognized as a signal moment: when most of what we were doing was communicating with boxes to communicate what we wanted...like Netflix...It's clear in the data that we've become a digital species and there's no sign of a slowing of that trend.
Whenever I hear people talk about IOT, the most common items that come up are refrigerators or thermostats. What else will there be?
Lee: Aside from security and monitoring, there's a huge value-add for mashups of data across geographical zones that allow for apps that sit on top of that data. I spent two years at DARPA and the question was always, "Gee, we'd like early sign of epidemics that showed malintent by terrorists or if there was something wrong in the food supply." By the time we get to a level of conscious awareness that there's a problem in full swing, it's after the fact. But if you were able to mash up antidiarrheals that were being sold at one Walmart at a particular time, then it might indicate there's a bad restaurant in the area.
Another example: Municipalities are increasingly strapped for budgets and can't hire enough inspectors to guarantee that all bridges they're responsible for are getting inspected on a reasonable basis. We keep building new things and stuff keeps rotting. So our solution is to automate a lot of that with simple wireless sensors. Sometimes, all you need to know is whether we have a rust problem -- and yet the value to municipalities would be huge. Instead of driving around in search of a problem, now we'd only need a small skeleton crew and they can go directly to the problem.
And you think that we're talking about 1 trillion devices by 2030. How did you do the math on that one?
Lee: I looked at the rate of smartphone adoption and the rate at which we're consuming data and came up with a number not too different from what Cisco came up with for 2020. And then I extrapolated. There's a big range of possible numbers, but 1 trillion is a nice number. It gets people to say, "That's BS," and sparks conversation...at first, they'll think you're a lunatic. Afterwards, maybe they'll think you're only partly lunatic...Look, anyone who makes a prediction will immediately look foolish. When I was down at DARPA and mentioned the 1 trillion number, the director said, "That's it? Only a trillion?"
How much closer will we be within the next five years?
Lee: Sources differ as to where we are in the current deployment, but I think that it's in the neighborhood of 10 billion to 20 billion objects, counting anything not human that serves up things. My reading. From Cisco's point of view, that's their starting point, so let's split the difference at 15 billion. So five years from now, it will go up to 50 billion. I suspect that lots of those devices most consumers wouldn't care a rat's patooey about. But if any of this materializes, then we're talking about a brand-new world.
But who takes the lead? Will the development of these devices get done by the people who now make these things, like fridges?
Lee: I think that this will be a new category of folks. The ones with the need historically have not done a lot in the networking space. They've done some tentative forays, but this will require a new skill set. The folks who have networking and comms expertise are unaware of the specific needs in each of those categories. Neither of those groups alone have sufficient breadth and depth to get the job done alone.
What's to stop us from repeating the history of the tech industry, where standards wars interrupt progress and companies spend more time fighting with each other than making their products interoperable?
Lee: I wish that I could say that the industry has learned from that history, but history shows we make the same mistakes all over again. I hope that that foolishness will not repeat, but I suspect there will be walled gardens, religious wars, and shooting wars -- though I hope it doesn't lead to mutual annihilation.
Is there any doubt in your mind that the IOT era will come to pass?
Lee: I'm always full of self-doubt and I've been wrong about a lot of things. But I've also been right about a lot of them too...When the world moved to CMOS RF (a key enabling technology in wireless communication), I went from being an idiot to a visionary. I'm neither of those things.