A day trip to Fourmile Canyon in early October to study a forest fire with a series of wireless sensors represented a technological breakthrough for University of Colorado professor Rick Han and his graduate students.

In the canyon west of Boulder, Han and his students learned that the wireless-sensor software they had developed in a laboratory worked in the real world. The sensors successfully monitored a controlled fire, tracking the burn's temperature and movement.

"There is so much power we can leverage with this technology," said Han, a computer science professor who last summer co-founded wireless sensor company Tendril Networks.

The rapidly developing industry of wireless-sensor networks is gaining traction nationwide as startups and technology giants bet on big returns in the future.

But before the nascent technology makes a large-scale splash, engineers and company executives face a handful of hurdles, experts said.

Advertisement

The price of the components required for each point or node on the network is still too high, with each node selling for between $25 to $125. Generating enough commercial demand is another concern, as is identifying a killer application that would spur mass adoption, said Tim Enwall, Tendril's chief executive.

The company has not made any sales. It has, however, attracted money from a handful of investors, although Enwall wouldn't say how much or from whom.

"You can't run a company on fumes," said Enwall, who anticipates the company's first sales next year.

The future uses for the technology are tantalizing.

British Petroleum recently deployed sensors in the engine room of an 850,000-ton oil tanker. The wireless network monitors vibrations and when rumblings reach a certain threshold, maintenance crews are alerted, fixing mechanical or electrical problems before they create havoc.

Scientists at Tendril are mainly focusing on developing software. Startups Dust Networks Inc., Millennial Net and Ember are some competitors.

"It's still in the early phases," said Rob Conant, vice president of business development for Berkeley, Calif.-based Dust Networks, arguably the industry leader. But by the end of next year, "we expect (the technology) to be all over the place."

Dust Networks launched its software about a month ago, following pilot programs primarily focused on energy efficiency in grocery stores and other commercial buildings. Test clients lowered their energy consumption by 10 percent to 25 percent per building, Conant said.

Wireless networks are made up of a series of nodes that monitor things like temperature, light, pressure and vibrations. Each node requires a variety of components, including a radio chip, an antenna, sensors and a small battery that can supply power for as long as two years. The nodes gather environmental information and broadcast the data back to a central computer.

Han said the cost per node most likely will plunge to less than $10 within four years, and more sensors or even cameras could be added to each node.

By decade's end, wireless sensor networks will generate $7 billion to $10 billion in sales, up from less than $500 million now, said Chris Mathias, an analyst with the Farpoint Group, a Massachusetts-based advisory firm that specializes in wireless communications.

Still, sensor technology itself is nothing new. It's been used for years in a variety of arenas, including home thermostats.

But today's technology is significantly different, Mathias said. For starters, traditional sensor networks require extensive wiring, manpower and time to install.

Wireless sensors, on the other hand, can be deployed in minutes, not months, using Velcro or tape. Also the wireless networks are smart-wired, meaning that when new nodes are added, the whole system adjusts. The networks can be reprogrammed remotely, allowing operators to change the setup on the fly.

"There are homeland-security applications all over the place," Mathias said.

To capitalize on that growth, many technology giants, including Intel, Siemens and Samsung, have invested heavily, banking on the new industry feeding demand for more and more powerful computers.

"It's moved out of the lab," said Kevin Teixeira, a spokesman for Intel's research and development division. "It's reached a point where a lot of companies are experimenting with them."

For Han and Tendril Networks, future experiments include avalanche detection applications this winter.

"Now we want to take steps into the outside world under more robust situations," Han said. "The ultimate goal is to outfit an entire home or building."

Staff writer Will Shanley can be reached at 303-820-1473 or wshanley@denverpost.com .