ORNL and the Smart Sensor Revolution

Through its developments of sophisticated instruments, electronic chips, and computer algorithms, ORNL is advancing the smart sensor revolution.

Dick Anderson studies a typical silicon microsensor developed at ORNL. Shown in the background is another example, a “critters on a chip” sensor (described in the Incredible Shrinking Labs article) that processes signals from bacteria attached to the chip that are genetically engineered to light up in the presence of certain pollutants.

By Dan McDonald
Jewels of light from a city at night and smoke from factory stacks are enduring images of the industrial revolution. This prevailing phenomenon started in the late 18th century and spawned the industrial plants that provide us today with power, products, and unprecedented, widely distributed wealth. The industrial revolution replaced our muscle power with mechanized tools of production driven by steam and electrical power. At ORNL, in spite of our late start in the 1940s, we helped this revolution roll along by developing designs for nuclear power plants and other energy sources. We developed the first control systems to operate these facilities efficiently and safely. And we invented new instruments needed to monitor and control these facilities.

In his 1980 book The Third Wave, Alvin Toffler argued that the basis for wealth has entered a new era, or “wave,” dominated by information technology. (In the first wave, wealth was based on the ownership of land and agriculture. In the second wave, the industrial revolution, wealth was created through machines.) The information revolution described by Toffler took off in the mid-1970s with the advent of desktop microcomputers to enhance our brain power by doing clerical and intellectual tasks much more quickly, reliably, and accurately.

Within two decades a large majority of the nation’s professionals had a personal computer (PC) in the office and many owned one at home. ORNL researchers developed ways to use digital computers to control and analyze reactors and other processes. The availability and increasing power of PCs revolutionized the development and use of instruments at ORNL and elsewhere. The reason: PCs can control the measuring instruments and sort through and compare streams of data and make calculations much more quickly and more accurately than a human can. As a result, computerized instruments are becoming faster, more sensitive, and more accurate than previous instruments and are extending measurement capabilities into altogether new areas.

As suggested by Paul Saffo, director of the Institute of the Future in Menlo Park, California, this marriage of sensing and computing in the 1990s will be the basis for a new revolution in which computers, provided with direct sensing capabilities, can interact directly with their environment. These smart sensor technologies, some of which have been developed at ORNL and are described in this special issue of the ORNL Review on measurement technologies, extend our senses, allowing us to “see” the stars and manipulate atoms and even single electrons. These sensors receive and respond intelligently to a signal or stimulus, like ultrasonic sensors that enable grocery store doors to “see” you coming and open automatically to let you in.

Some ORNL-developed sensors “hear” sounds that tell us that a submarine is operating too loudly, that a machine is malfunctioning or is about to malfunction, or that a person is ill with a specific respiratory disease. Other sensors “sniff out” chemicals to warn of the presence of toxic gases at hazardous levels. Still others use lasers and optical fibers to “see” cancerous tumors in parts of the body. Special cameras combined with software provide “machine vision” (pattern recognition) that can be used to detect flaws in fabrics while they are being woven or to detect and classify defects in semiconductor wafers during manufacture.

Many advances in sensor technology are driven by the automobile industry’s need for inexpensive, reliable measurement and control technologies. In modern automobiles, sensors feed data to microprocessors controlling actuators that activate mechanical devices, ensuring smooth operation of anti-lock brakes, cruise and traction controls, and fuel injection systems. Even more advanced sensors, actuators, and computer controllers will be incorporated into future cars for satellite-based navigation systems to help drivers reach their destinations faster and for radar collision avoidance systems to reduce the number of automobile accidents. In future airbags, chips that detect an abrupt change in acceleration will also sense a person’s weight and size and adjust the airbag inflation force accordingly.

The goals of ORNL research in sensors and measurement include miniaturization and intelligence—designing smaller, smarter sensors on silicon chips. Such chips will combine sensing with computing, or signal processing, as well as signal transmission to an external receiver. These smart sensors are making possible smart cars, smart buildings, and smart machines.

Measures of Our Success

ORNL has traditionally been deeply involved in the measurement sciences. Our multidisciplinary organization is a particular strength in developing advanced sensors. Almost half of the R&D 100 awards given by R&D magazine to Oak Ridge researchers (who lead the other Department of Energy laboratories in the number of such awards received) have been for developments of measurement techniques and devices.

We have years of experience developing instruments, beginning with radiation detectors needed to monitor ORNL’s reactors and our employees. Many of the fundamental principles of reactor control and protection systems developed at ORNL for the Laboratory’s early reactors are used widely in commercial nuclear power plants. In 1975 our noise analysis measurements, with which we learned to detect nuclear plant anomalies using the reactor’s “noise” signature, explained why the General Electric (GE) Company’s boiling water reactors were experiencing some internal damage. Our results led to a governmental decision to keep them operating, but at lower power, until GE solved the problem. Later ORNL noise analyses in the mid-1980s showed it was feasible to measure the stability of boiling water reactors, and this measurement concept was later marketed by GE.

In another energy area, we developed measurement technologies to determine the temperature, energy transfer, and other parameters in fusion energy plasmas. ORNL measurements of the electron density profile at the edge of the plasma have advanced the science and technology of heating fusion plasmas with radiofrequency power.

We measured the radioactivity emitted after the 1979 accident at the Three Mile Island nuclear power plant. We measured uranium, mercury, and many other pollutants in soil, water, and air. We found that the health risks of the public’s average exposure to environmental tobacco smoke may not be as high as other studies suggest.

We invented the routinely used commercial device for identifying and measuring concentrations of body fluid constituents that indicate disease states. We developed instruments to detect carcinogens in the environment and DNA indicators of the presence of diseases. We used radioactive isotopes and various instruments to follow the movements of nutrients and pollutants and the carbon dioxide and water vapor exchange rates in forest ecosystems. Our monitoring technologies are helping us predict the effects on forest productivity of air pollution, increased concentrations of atmospheric carbon dioxide, and the expected effects of global warming—changes in daily temperature and rainfall patterns.

Panos Datskos (left) and Slo Rajic show the prototype of the calorimetric microspectrometer they developed. Called CalSpec, this device can detect chemicals indicating the presence of natural gas as well as vapors from explosives, toxic materials, and chemical warfare agents. For this development, the two researchers and their colleague Chuck Egert received an R&D 100 award in 1998.

ORNL’s Capabilities

We have smart people from many different fields who work together to devise smart sensors to solve tough problems. For example, we have applied our multidisciplinary strengths to develop a variety of approaches to solve one of the world’s worst pollution problems—buried land mines that kill and maim people and prevent the use of large tracts of land. Three of our sensor technologies highlighted in this issue of the Review are being developed to detect the chemical signature of plastic explosives that have leaked out of mines into the soil, from which they may vaporize into the air. (1) Our direct-sampling ion trap mass spectrometer, which is being reduced from the size of a desk to that of a briefcase, can sniff out explosive molecules in the soil or air. (2) Microcantilevers (tiny springboards attached to electronic chips) coated with platinum bend or vibrate in the presence of the explosive TNT. When TNT is attracted to the platinum coating, it reacts, causing a detectable mini-explosion and resulting deflection of the microcantilever. (3) ORNL researchers have genetically engineered bacteria to light up in the presence of trace amounts of TNT from land mines.

Helping Our Customers

Originally, ORNL developed measurement and control systems to meet the Laboratory’s needs. Today we have many outside customers because of our expertise in mixed analog and digital chip design. Mixed-analog-digital chips are the keys to the integration of analog transducers and electronics with digital computers and signal processors on the same chip. Several ORNL divisions are designing and building entire “instruments on a chip,” and much progress has been made toward the next level of integration and complexity—the “lab on a chip.” In these microscopic systems, analog and digital electronics are being combined with the ability to handle micro quantities of liquids or vapors. DNA and proteins can be analyzed on a single chip, providing information that could identify crime suspects and lead to the development of new drugs.

We have a growing expertise in advanced signal processing—directed at extracting very weak but meaningful signals from a noisy environment—and in developing algorithms (step-by-step procedures to enable a computer to recognize patterns). As a result, we are inventing devices capable of machine vision. For the U.S. government, we have developed an automated inspection of stamps and currency, to weed out defective items, and a heartbeat detector that can spot intruders or escaping prisoners concealed in closed vehicles. For the medical community, we developed a lung diagnostic system for detecting and classifying respiratory disorders. We are developing an algorithm-guided laser system to locate and destroy burned tissue to reduce pain and hasten healing for burn patients. Our machine vision expertise has enabled us to support industry through the development of software and hardware systems for detecting and classifying defects in semiconductor chips and textiles. In this way, we are helping the semiconductor and textile industries reduce production of defective items, increase quality, and cut costs.

We are working with industry in other ways, too. For example, Perkin-Elmer is collaborating with us in commercializing our “critters-on-a-chip” technology for environmental applications. This chip hosts an electronic sensor that responds to bacteria on the chip that are designed to emit light in the presence of specific chemicals, such as soil pollutants.

For the police community, through videotape enhancement techniques, we have extracted valuable information from videotapes of people present during store robberies. This information has been used as evidence in court to convict one suspect.

We are developing new measurement techniques in support of the Department of Energy’s Industries of the Future program in which DOE helps the energy-intensive industries that produce aluminum, chemicals, forest products, glass, metal-casting, mining, steel, and agricultural products rethink how they manage technology. For the steel industry, for example, our engineers have devised a technique for determining surface temperature during critical stages of the process of producing galvanneal steel for making rust-free automobiles. Also for DOE, we have developed instruments for ensuring that nuclear material is being safely stored and accounted for and for showing that weapons-grade material is converted into reactor fuel. ORNL research on glow discharge ionization, electrospray and quadrupole ion trap mass spectrometry has led to tools that could help detect hidden explosives and tiny airborne particles that could cause lung disease, monitor hydrocarbon levels in low-emission vehicles, and determine structures of biomolecules such as DNA and proteins. We are developing snapshot laser radar as a remote sensing technique for possible use in detecting chemical compounds in smokestack plumes and in characterizing air vortices to guide the safe spacing of aircraft during landings and takeoffs.

Alan Wintenberg (left) and Glenn Young check a fixture for testing ORNL-designed integrated-circuit electronics for a silicon detector being developed for the quark-gluon plasma experiment at Brookhaven National Laboratory’s Relativistic Heavy Ion Collider. The experiment could help scientists better understand the actions of subatomic particles in the universe at the start of the Big Bang.

For the nuclear physics community at DOE’s Brookhaven National Laboratory, we are developing detectors to help physicists better understand the universe a few seconds after its birth. A Spallation Neutron Source is expected to be built by DOE at ORNL by 2006 to measure changes in atomic-level structure and interactions among molecules in materials. Such measurements could lead to stronger materials, faster electronic devices, safer and faster transportation, longer-lasting body implants, and more effective drugs.

For the functional genomics community, we are developing instruments to help relate the structure of a gene to its function. We are designing a wireless sensor that can be implanted under the skin of a mouse to measure its temperature, pulse rate, and activity in a cage to help determine the physiological and behavioral effects of its defective genes. We have developed a miniaturized CAT-scan device to image mouse mutations such as skeletal defects, fat deposits, enlarged kidneys, and other abnormally shaped organs, as well as growing tumors and other manifestations of disease.

For the U.S. military, we are developing a more rugged and sensitive mass spectrometer system for detecting biological and chemical warfare agents. We are designing telesensors to help medics locate wounded soldiers and determine their condition (e.g., detect rises in body temperature and drops in blood pressure). We are developing an ultrasound sensor to monitor brain injuries. Instruments, detectors, and sensors are getting smarter and smaller. ORNL is playing a big role in these improvements as we help the newest technical revolution to roll along.

Dan McDonald is director of ORNL’s Instrumentation and Controls Division. The work he describes reflects the contributions of a number of ORNL divisions.

Where to?

[ Next article | Search | Mail | Contents | Review Home Page | ORNL Home Page ]