HP's grand vision: measure everything

NEW YORK (Fortune) -- Imagine walking down the supermarket aisle with a cheap device you could hold up to a tomato. If the sensor detects a pesticide residue, you'd know the "organic" label is a lie. Similar tools could track the chemical content of water in a stream, telling you if there was lead contamination and when it got there, or keep constant watch on a bridge and tell if a structural steel beam was at risk of collapse.

Such products are almost certain to become common in coming decades, according to Stan Williams, who heads Hewlett-Packard's Information and Quantum Systems Laboratory.

He aims to develop a panoply of microscopic-scale nanotech devices that will be able to measure essentially anything - and at low cost to boot. Viruses, bacteria, the chemical composition of molecules, vibration, moisture levels, particular sounds - these are just some of the things that the super-cheap devices he envisions will be able to detect.

In an exclusive conversation with Fortune, Williams described in detail a project HP (HPQ, Fortune 500) began way back in 1994. He will speak about it in public for the first time next week at Fortune's Brainstorm Tech conference. No other company he knows of is developing similar products.

"The theme for the lab is CeNSE - for Central Nervous System for the Earth," he says. Williams feels such a grandiose name is justified. Because these sensors will be built with standard semiconductor technology, they will ultimately be cheap enough to build "in the trillions." That will make it possible to deploy arrays of measuring devices anywhere at a reasonable cost.

In our era of rapidly decaying global environment, such tools could help us know with certainty how our world is changing, and help us make better-informed choices about how to respond.

Applying these devices to monitor the entire planet will take time. None of these nano-sensors will even come into existence for at least two years, and will not be deployed in quantity for 5-10 years. Before we monitor the planet we will likely use those sensors to monitor the strength of a bridge's beams. Others might maintain vigilance over a high-rise building's structure, or the steel in a ship's hull or a train track.

The first versions, emerging in the next several years, will be expensive and are likely to be used to monitor systems in oil and gas refineries and chemical plants, where investments in monitoring vibration and chemical composition will pay off the most. "Think of our sensors as stethoscopes," says Williams. "As soon as something started vibrating a little bit differently we'd know it."

HP has two fundamental types of nano-sensors under development in its lab. The first is a type of measurement device made from a relatively small number of atoms. Because of its small size, even the tiniest changes in the environment can perturb it. And that perturbance can be measured. "When you are able to craft matter at the nanometer scale," says Williams, "you have essentially achieved the ultimate level of control over directing matter, electrons or photons."

Such devices would be able to measure minute amounts of biological material, or ultra-tiny vibrations, with tremendous sensitivity. "We're working on being able to detect individual molecules of whatever you may be concerned about - or individual viruses or bacteria," Williams says.

Because these silicon-fabricated nano-measurers can be put by the millions onto one tiny chip, some of the products HP envisions resemble a nose - with multiple receptors for various "smells."

But another nano-device that Williams has high hopes for would be optical: "A very tiny laser would light up and we could look at the optical spectra of chemicals. Each one is like a fingerprint, with a unique spectral identity. That would be a single universal detector." Though a laser capable of such a task would today cost around $100, Williams thinks they can eventually be produced for about 10 cents.

HP already has working prototypes of various sorts of nano-detectors working in its 90-person lab.

The company would likely license the sensor designs to others and buy back the devices to integrate into its own measurement and control information systems. Only once such sensors are combined with sophisticated database and analysis technologies can their promise be delivered. HP sees its business opportunity in helping customers manage the vastly greater amount of information such monitoring will generate.

While Williams is confident HP has a huge head start (and he's not afraid to talk about the project -"We welcome competition," he says), this work can only proceed so fast. There simply aren't enough capable engineers in this highly rarified field, he says. Though the pathway is starting to seem clear, it remains long.