 |
BERKELEY, Calif., Feb. 26, 2003 — Most of us think of motes, if we think of them at all, because one was in someone’s eye in the Bible. We’ll soon think of them differently if Kris Pister has his way.
 |
Pister, a University of California, Berkeley, professor is a pioneering developer of wireless sensor networks built around tiny wireless radio and semiconductor devices known colloquially as motes. Pister finally jumped from academia in January to create Dust Inc., where he is chief executive, and is dedicating himself full time to commercializing motes.
“I never thought I’d do a startup. I just got to the point where all I could think about” was figuring out how to implement motes for potential customers, he said.
 |
He plans to go back to teaching electrical engineering at Berkeley at the end of this year, though he’ll stay at Dust longer if he needs to. “I have no illusions about my ability to run a company. I’m the visionary, and we need that right now,” he said.
Networks of Dust motes run on TinyOS, an operating system deployed in mesh networks that requires very little power. While Pister’s research often focused on microelectromechanical systems (MEMS), used to help the motes communicate with each other, Dust isn’t using MEMS. Pister said they simply aren’t ready for commercial applications. Instead, Dust aims to be an enabler for other types of small technology, such as nanoscale industrial sensors, or RFID tags such as those made by Alien Technology Corp.
“MEMS is about making sensors smaller and cheaper. What we do is make installation of sensors cheaper and easier,” said Robert Conant, Dust’s head of business development. Conant said new types of sensors may drive down the cost of the actual sensing devices, but it still takes wires to connect those sensors to systems that can read the data, and that means significant installation costs. Dust aims to sharply reduce those costs.
What Pister sees is an opportunity to save costs. For instance, the current generation of sensors under development at companies like Nanomix Inc. promise to cut the costs of sensors dramatically. But to get information still requires wires to be run to them, and labor to install it. Pister believes that using tiny wireless sensing networks from Dust will cut installation costs by as much as four times.
Similarly, RFID tags are wireless themselves, but they don’t have much range. A company could have a warehouse of pallets and goods full of RFID tags but not be able to read them without either running wires through the facility or sending out a worker with a handheld reader. A network of Dust sensors could act as the equivalent of cell phone towers, tracking where the RFID-tagged pallets or goods move within a warehouse.
Motes are being deployed now, and Pister said the company is pulling in checks from customers “for thousands of dollars.” It may net $1.5 million in sales this year.
Longer term, Dust is in a market that 10 years from now could generate $50 billion in sales, according to Laurel Donoho, an analyst at Frost & Sullivan. “This technology will change the way we live,” Donoho said. The potential for cheap networks of tiny sensors constantly generating information is “really exciting.”
Before such profound changes occur, obstacles exist for Dust and related wireless sensor network companies such as Ember Corp. and Millennial Net Inc. For starters, Donoho noted, the technology is brand new. A great deal of education and testing needs to take place, particularly in industrial sensors, where users tend to adopt new technologies slowly.
Boston-based Ember, founded in 2001, has raised $28 million, but has found it hard to get customers to move beyond pilots. “Initial interest is great, but the next step, commercialization, has been slowed by the economy,” said Robert Poor, Ember’s chief technology officer.
For now, Dust is pursuing three main application areas: security, tracking and industrial monitoring. Honeywell International Inc. is Dust’s first commercial customer. It’s begun using BLUE Motes in supermarkets to gauge electricity usage. If a supermarket uses 50 refrigerators, it’s a major cost issue if one of those runs out of coolant. Dust’s motes, which sell for between $50 and $100 apiece, are small enough to attach to all the refrigerators and report on electricity usage so the store can keep tabs on costs.
Besides figuring out ways to apply the technology today, Pister is focused on raising funds. Currently, Dust has less than a million dollars in funding (one of its angel investors is tech legend Gordon Bell), although the company said it is about to close a Series A round. Dust won’t disclose how much funding it wants, but noted that it will ultimately need about as much capital as, say, Ember.