‘Smart Dust’ is a term you’d expect to hear in a Mission Impossible movie or Michael Crichton’s 2003 novel Prey. It conjures up images of the unthinkable, technology so small and so intricate that it could be anywhere and everywhere and turn absolutely anything into a digitally recognised substance. Surprisingly, it’s been around for a long time.
“It was kind of a joke - everything in the US and LA at that time seemed to be “smart”, smart bombs, smart houses, smart roads,” says Kris Pister, who coined the term and co-authored the paper ‘Smart Dust: Autonomous sensing and communication in a cubic millimeter’ in 1997.
Pister is a Professor of Electrical Engineering and Computer Science at University of California, Berkeley and the founder and CTO of Dust Networks, a company set up in 2002 and acquired by Linear Technology in 2011. He has been working on his smart dust idea since the mid-90s, which, given the rapid evolution of so many technologies, protocols and standards we take for granted today, shows a dogged determination.
“As a grad student at Berkeley in 1992 it was clear to me that Moore's Law, the communication revolution, and MEMS technology were all driving in the same direction: the size, power, and cost of a wireless sensor node were riding exponential curves down to zero,” says Pister.
The project ultimately set out to develop a cubic millimetre device with a sensor, power supply, analog circuitry, bidirectional optical communication, and a programmable microprocessor. That was in 2001. A list of possible applications, which were ahead of their time, included battlefield surveillance sensor networks, virtual keyboard, inventory control and environmental monitoring.
Intelligence on a grain of rice
Fifteen years is a long time in technology so how has Smart Dust progressed?
“When I got to Berkeley and we wrote the Smart Dust proposal in 1997, we were making small optical communication systems,” says Pister. “We showed that you could make a complete autonomous wireless sensor in a five cubic millimetre package, about the size of a grain of rice.”
At the same time, Pister says, they also started making “the smallest off-the-shelf wireless devices that we could, that would use radios instead of optical communication. They were about the size of a pair of AA batteries, because that's what powered them. Since then we've done a lot of integration, so all of the silicon is on a single small chip now, and the power requirements of the hardware have dropped one or two orders of magnitude, and the networking algorithms have reduced the power required (radio duty cycle) by another two or three orders of magnitude. So now, even something running on a coin cell battery can last for years, and the smallest standards-compliant wireless sensor nodes are around a cubic centimetre.”
The advancements made at Dust Networks have been enough to propel Smart Dust onto the Gartner Hype Cycle for emerging technologies right at the start of the line, in the ‘innovation trigger’ phase. You probably can’t get more ‘hype’ than that and Gartner has whispered the idea that Smart Dust is the future of IoT.
Wireless evangelist Nick Hunn agrees the Smart Dust concept is a “very beguiling one” but adds that it has “a whole set of practical issues” of which the greatest is power.
“We’re probably at the point where you can drop sensors which could be autonomously solar-powered, but there’s a trade-off in solar cell size and processing,” says Hunn. “If they want to monitor constantly, they need a larger cell; alternatively they would just be waking up and sampling occasionally, which is a subset of what dust offers.”
Power isn’t everything
Pister agrees that power is important but says “reliability is paramount. If you had a system that burned zero power but only got the data through half the time, no-one would buy it. Or worse, they would buy it, try it, hate it, and then think that wireless sensors don't work.”
At Dust Networks, Pister says the business started out claiming 99.9% reliability in its “best effort” (UDP-like) networks but claims that today the company now talks about five nines (99.999%) reliability. However, he contends that most of its networks operate at significantly higher reliability than that.
Pister also claims to have overcome the issue of requiring a self-forming, self-healing network to manage the motes. SmartMesh he says, does all that, monitoring and managing network performance and security while exchanging data with a host application.
So what are those applications? It’s not actually too far away from the vision Pister had in 1997. He says that while he doesn’t believe the business is currently selling anything to the US Department of Defense, the company is actively selling into industrial process automation, parking monitoring, EV battery monitoring and the building sector.
There are, of course, limitations and the current problem most companies in the sensor space are facing is return on investment. Government and industry are natural opportunities but for the moment at least it’s difficult to imagine this spreading much further.
“I’m happy to be proven wrong, but 15 years on, I don’t see it [Smart Dust] taking over the world.”
That’s perhaps not great news for the science-fiction writers but then we have only just started the story. If Pister has proven one thing, it is that he has staying power and a taste for the long game.
PREVIOUS ARTICLE«Konica Minolta wants to get in on the ‘workforce collaboration’ act
NEXT ARTICLEFree money: The answer to a post-automation world?»
Adrian Schofield sheds light on tech in South Africa
Mark Chillingworth on IT leadership