Berkeley Wireless Research Center researchers examine a printed circuit

E-WALLPAPER DESIGNERS: Ana Arias with colleagues Elad Alon (center) and John Wawrzynek of the Berkeley Wireless Research Center make rooms more connected and are developing other applications for flexible electronics. (Photo by Peg Skorpinski)

From smart dust to smart rooms

Coming one day soon to a big-box store near you: e-wallpaper that instantly recognizes and responds to you and your mobile devices. Thanks to steady advances in wireless technology, that idea doesn’t sound as far-fetched as it once did. Small and inexpensive wireless sensors placed throughout our physical world are capturing and transmitting streams of information about conditions in places, things and even our behavior.

If Berkeley engineers are successful, our walls may become alive with sensors, transmitters and other electronics that are communicating among themselves and with our smart phones and mobile devices.

“The room would know you’re coming,” says associate professor Ana Arias, a Brazilian-born physicist who joined the electrical engineering and computer sciences department last year to spearhead Berkeley Engineering’s efforts in the cutting-edge arena of flexible printed electronics. Working with a team from the Berkeley Wireless Research Center, she is fashioning a mesh-like material that would have remote sensing components printed directly onto its bendable surface.

“The idea is you would go to Home Depot and buy tiles of this wallpaper and design capabilities in your room,” says Arias, who envisions electronic wallpaper lining the walls of airports, factories, offices and other spaces.

Sensors gather data by measuring changes in voltage or other electrical properties that take place when the sensor is exposed to whatever it’s sensing. Linked into self-organizing networks, wireless sensors collect and relay their information as low-power digital signals that hop from one device to another. The e-wallpaper project, which includes Arias’s colleagues Elad Alon and John Wawrzynek, is one of the latest examples of Berkeley’s pioneering and far-reaching efforts in the field of sensor technology.

SmartMesh moteSPIN-OFF: Kris Pister formed the company Dust Networks to commercialize his “smart dust” invention. The company’s SmartMesh network is a self-forming mesh of nodes, or “motes.” The low-power network includes the hardware and firmware necessary to form a reliable and secure wireless network and is the basis for a number of networking standards. All motes in a SmartMesh network—even the routing nodes—are battery-powered, enabling energy harvesting and allowing flexibility in placing sensors exactly where they need to go, with low-cost “peel-and-stick” installations. (Photo courtesy Linear Technology/Dust Networks)Kris Pister never imagined a mote of dust could create such a stir. Fifteen years ago, professor Kris Pister and his EECS colleagues began packing sensors, microprocessors, radios and power into progressively shrinking chips. Eventually, they unveiled a 5-cubic-millimeter sensor—roughly the size of a grain of rice—that was wireless and fully operational. Pister aptly called the invention “smart dust.”

Today, mass-produced versions of such sensors are on the job in many places. At oil refineries and other industrial plants, they’re monitoring operations and watching for early signs of equipment failure. They track shipments of goods, turn off the lights in empty homes and offices, improve energy efficiency in power-hungry data centers and have begun steering drivers to vacant parking spots in California cities.

The prospect of ubiquitous sensing—commonly known as the Internet of Things or the Industrial Internet—has launched ambitious sensor-deployment initiatives like IBM’s Smarter Planet and HP Labs’ Central Nervous System for the Earth.

“A lot of universities and a lot of companies contributed, but Berkeley played an absolutely central role in that evolution,” says Pister (M.S.’89, Ph.D.’92 EECS).

For starters, Pister and David Culler, professor and associate chair of EECS, “open sourced” the hardware and software designs, so any interested developer could pick up the technology and run with it.

Many did. Berkeley’s sensor “motes” became “sort of the workhorse of the industry,” says Pister, who founded a company, Dust Networks, based on the technology in 2002. TinyOS, the operating system developed by Culler’s team to enable the motes to communicate, is utilized by thousands of researchers and developers worldwide, and provides the foundation for emerging standards.

Their contributions reflected an EECS tradition. In the early 1970s, the late professor Donald Pederson made Berkeley’s SPICE computer simulation software an open-source program, which became the universal standard for integrated circuit design. “It’s in our DNA,” Pister says. “You put your research results in the public domain.”

The Swarm Lab heralds a new era of ubiquitous sensing. Rob Gilmore, vice president of engineering at Qualcomm Research, notes that Berkeley researchers have also been leaders in developing low-power radios with innovative means of harvesting energy for fuel. “Berkeley just seems to have a staggering number of outstanding, talented professors who have contributed to the whole revolution,” Gilmore says.

Last fall, campus and industry leaders built on that legacy with the opening of a new Cory Hall research facility. Called the Swarm Lab, the collaborative venture will create and disseminate research on wireless sensing and the immense wave of connectedness it is expected to generate. The lab received major support from Qualcomm Inc.

Illustration of sensors monitoring parking spaces in HollywoodSPIN-OFF: Kris Pister formed the company Dust Networks to commercialize his “smart dust” invention. The company’s SmartMesh network is a self-forming mesh of nodes, or “motes.” The low-power network includes the hardware and firmware necessary to form a reliable and secure wireless network and is the basis for a number of networking standards. All motes in a SmartMesh network—even the routing nodes—are battery-powered, enabling energy harvesting and allowing flexibility in placing sensors exactly where they need to go, with low-cost “peel-and-stick” installations. (Photo courtesy Linear Technologies/Dust Networks)Professor and Swarm Lab director Jan Rabaey is an expert in next-generation wireless technology and an enthusiastic champion of its future. “Sensors have the ability to revolutionize how we do things,” he says.

Rabaey and others envision a time when trillions of intelligent sensing devices are scattered throughout our environment, transforming how we interact with the physical world, use energy, care for our health, run transportation and otherwise conduct our daily lives.

Some 124 sensor-related projects are taking shape at the Berkeley Sensor and Actuator Center (BSAC). Founded in 1986, the facility supports university and industry research with commercial application and relevance. Extending across two floors of Sutardja Dai Hall, the sleek Marvell Nanofabrication Lab serves as a state-of-the-art nursery for the latest in sensor designs emanating from BSAC and beyond.

Many campus scientists are harnessing sensors to tackle problems in such areas as energy, the environment, health care and infrastructure. At the Center for Information Technology Research in the Interest of Society (CITRIS), where research targets some of today’s most pressing issues, “every one of our projects has some form of sensing at its core,” says CITRIS director Paul Wright.

Steven Glaser can measure snowpack from the comfort of home. A network of wireless sensors is being tested in the Sierra Nevada’s American River basin, and another 60-sensor network is collecting information in the mountains around Shaver Lake, about an hour’s drive northeast of Fresno. Backed by a $2 million National Science Foundation grant, civil and environmental engineering professor Steven Glaser and other Berkeley and UC Merced researchers are deploying the instruments to collect data about California’s snow pack.

When completed, thousands of sensors will blanket a 2,000-square-mile forest, forming the world’s largest ecological sensor network. Each node will measure such characteristics as snow depth and soil moisture. Fresh data is transmitted wirelessly every 15 minutes “and it ends up in real-time on my desk,” says civil engineering doctoral student Branko Kerkez.

Combining the sensor information from the American River basin with satellite information gathered by regional utility districts will allow water managers to make better decisions about things like dam releases. Accurate and always on duty, the smart gadgets could replace old-school snow surveys by state hydrologists who continue to trudge to mountain locations with measuring poles. “Climate’s changing,” Glaser says. “We can make better use of each drop of water.”

David Culler is training a “macroscope” on energy management. Culler likens sensor technology to a new scientific instrument—a “macroscope.” Through its lens, he says, we can observe (and respond to) global-scale activity in fine detail.

One of the first places that Culler and others are focusing the macroscope is on energy issues. Smarter systems could improve how we manage the electric grid by heating and cooling indoor spaces more efficiently. Advanced sensor networks are really just another tool to solve engineering problems. In the case of energy, Culler says, sensors are “an information-age solution to an industrial-age problem.”


Reach the editors at berkeleyengineer@berkeley.edu