February/March 2004
http://www.coe.berkeley.edu/labnotes/0204
Commuter-friendly version

According to graduate student John-Michael Wong, "although a lot of the fundamentals of this project are based in theory, this system could someday help people in making decisions about the structures that they work and live in." (David Pescovitz photo)

When the "check engine" light on your automobile's dashboard flashes on, you know immediately that the car may not be safe to drive. Now imagine that your office building was outfitted with similar technology. A quick look at the building's Web site could notify the property manager to call in an engineer because a support column is in need of a check-up. Or a display screen mounted on the front door might warn occupants about to re-enter the building after an earthquake that the entire structure is on the verge of collapsing.

John-Michael Wong, a graduate student in UC Berkeley's Department of Civil and Environmental Engineering, is designing precisely this kind of "dashboard for buildings." The project is sponsored by the CUREE-Kajima Research Program, a joint venture between the Consortium of Universities for Research in Earthquake Engineering (CUREE) and Japanese construction firm Kajima Corporation.

Formally called a "Framework for Integration and Visualization of Structural State Data," the system analyzes the raw data from wireless sensors installed in a structure and translates those calculations into easy-to-interpret graphical displays.

"With our system, you can display data on very different levels," says Wong, a student of CEE professor Bozidar Stojadinovic. "A normal user might only be interested in the overall condition of the building, while an engineer needs access to more detailed data and analysis output."

This image links to a Web-based demonstration of a display that a property manager or structural engineer would refer to for information about a building's structural health. (courtesy the researchers)

The data for the system comes from a network of tiny wireless sensors installed in key structural locations around a building. The devices can be outfitted with accelerometers to detect vibration or strain gauges that measure the bending or twisting of a beam, for example. Those numbers are then wirelessly transmitted to a central computer for processing. Wong's collaborator on the project, graduate student Jan Goethals, has spent several years developing the structural sensing system with CEE professor Steven Glaser and others.

The problem though, Wong says, is that the sensors' flood of data is too raw for efficient analysis, even by an experienced structural engineer.

"A stream of numbers isn't useful to anyone," Wong says.

To classify and analyze the data so it can be intuitively displayed, Wong developed a novel database storage system and metadata schema. Metadata is literally data about the data. For example, the metadata attached to a measurement from a particular sensor might describe the type of sensor and where in the building it's located. Wong's system can also perform calculations on independent pieces of data to provide information that's useful for evaluating the building's health.

For example, story drift, the rocking motion that occurs between stories in a building, is determined by calculating the difference in position between the two stories. Wong's software automatically performs that calculation on the raw data from two sensors so that the story drift can be taken into account when the system diagnoses the building's overall structural health. A dangerously high amount of drift might then trigger a "check building" alert on the property manager's display. He would then call in an engineer who might drill deeper into the data in attempt to identify the cause.

Wong's system will be put to the shake table test within the next six months when a twenty-foot-square, single-story structure is instrumented with sensors and then subjected to a simulated earthquake on UC Berkeley's giant shake table. The researchers will then assess their system's accuracy in diagnosing the specimen's structural integrity.

Currently, most models that predict how a particular building will perform in an earthquake are based on data collected from shake table experiments and other simulations. Wong's system will help civil engineers study the effects of real earthquakes, windstorms, and other phenomena though, hopefully aiding in the design of better buildings.

"Our system is geared so that the real world can be used as an experimental test-bed," Wong says. "That way we can refine our modeling of structures to match up a building's expected performance with its actual performance during a quake."

Professor Tarek I. Zohdi's swarm mechanics research also has long-term applications in medicine, specifically aiding in the design of new drugs that employ swarm behavior to collectively attack cancer cells. (David Pescovitz photo)

To become intimately acquainted with the research of UC Berkeley professor Tarek I. Zohdi all one needs to do is yell "fire" in a crowded theater. More than likely, the audience will stampede toward the emergency exits, bumping and bouncing off each other as they push to safety. If you had a bird's eye view, the chaos might resemble a laboratory physics experiment with hundreds of magnetic particles attracting and repelling one another. Ironically, Zohdi explains, there isn't much difference between the two scenarios. In both, the interacting objects--people or particles--exhibit what's known as swarm behavior. And understanding its mechanisms could impact everything from building design to robotics.

"We would like to control groups of individuals simultaneously so that their collective behavior achieves the particular goal that we have in mind," Zohdi says.

The individuals Zohdi refers to are not necessarily biological entities, although they could be. Essentially, he says, the aggregate motion of a large group of individuals--flocks of birds or crowds of human beings--does not come from a top-down plan directing the emergent behavior but from the simple one-to-one interactions between the individuals.

Zohdi's aim is to computationally tease out the simple laws of one-to-one interactions that affect the motion of an entire group so that the swarm behavior can be modified for the better. For example, he explains, accurate models of panic behavior could help architects design stairwells and exits that reduce dangerous bottlenecks during emergency situations.

The key to modeling swarm behavior on a computer, Zohdi says, is mathematically treating each individual in a swarm as a simple particle governed by very specific physical laws. For instance, he explains, a particle representing a human may be restricted by a law saying that "people in our culture don't like their personal space invaded within two feet."

This video* depicts an optimized swarm of particles that after twenty generations learned to navigate a three-dimensional obstacle course in a minimum amount of time. The swarm's ultimate objective is to land within a pre-specified distance of a target (the white dot on the far left) while following certain rules: the swarm members could not make physical contact with one another or get too close to an obstacle. "Initial swarm members would impale themselves on the obstacles," Zohdi says. "However, later generations have 'learned' by adjusting their interaction laws to bunch up and squeeze through the obstacle course." (courtesy the researcher) *video requires QuickTime player

"The whole ball of wax boils down to what is the correct description of forces between individuals?" Zohdi says. "If you can map the characteristics of an individual into some kind of force, you can start to design the behavior you want. For example, you might say, I want a team of robots to swarm around one object and avoid another."

Indeed, instilling a bit of swarm intelligence into robots is one of Zohdi's ongoing efforts. In a project for the Office of Naval Research's UC Berkeley-based Center for Collaborative Control of Unmanned Vehicles, Zohdi is collaborating with professors Karl Hedrick and Raja Sengupta on the design of unmanned robotic vehicles that can avoid obstacles and fly in swarm-like formations. Similar technology could also be employed by self-driving vehicle systems like those pioneered by researchers in Berkeley's Partners for Advanced Transit and Highways (PATH) program.

When Zohdi begins a computer simulation, he seeds the virtual swarm with a series of laws based on his own educated guesses. The beauty though is that the moment he first unleashes the swarm to tackle whatever task he's simulating--an obstacle course, for instance--the particles begin to learn from experience.

"I use genetic algorithms," Zohdi says. "Just like with living creatures, the swarms store and rate previous experience."

After each simulation, the swarm undergoes the software equivalent of "biological mutation." For example, a law describing how closely particles should remain together when maneuvering an obstacle course may be randomly tweaked between a simulation. If the slightly changed law leads to better swarm performance, it gets passed down to the next strain of particles. The evolution continues with each simulation until a set of laws emerge that is best suited for the task at hand.

"Approximately, 20 separate laws are all I've ever needed to predict any kind of complex movement," Zohdi says. "My focus is on designing the laws so that they match physical reality."

Graduate student Josei Lee holds the first prototype of her loom to weave electronic textiles. (David Pescovitz photo)

Someday soon, dressing smartly may take on a whole new meaning. Electronic textiles--fabric containing microprocessors, sensors, and actuators--could lead to shirts with pores that automatically open and close depending on the temperature, army fatigues with chameleon-like color-changing properties, or tents that sniff out environmental contaminants. Josei Lee, a UC Berkeley graduate student in Electrical Engineering and Computer Sciences, and Professor Vivek Subramanian recently built the world's first flexible transistors directly on fibers. Their success is a leap toward the future of computer couture.

The aim of the project, part of the Center for Information Technology Research in the Interest of Society (CITRIS), is to weave large bolts of fabric that incorporate a smart grid-like network of interconnecting wires. Various devices such as chemical sensors or blood pressure monitors could then talk to each other through the network of fibers, much like desktop computers, servers, and printers communicate on an office network.

"If one device is on the fabric's grid, it's connected to everything else on the grid," Lee says.

While other researchers have embedded entire circuits into fabric, that kind of "one-off" approach is not cost effective or efficient enough for widespread adoption of the technology. That's where Lee and Subramanian's fiber transistors come into play.

"When an electric textile is first switched on, it most likely won't know where the arms or legs or peripheral components are located," Subramanian adds. "But if you have dynamic switching, you can identify the signal paths. That way, the fabric can say, 'I'm the CPU, and in location X is a sensing unit, and this is the best way to connect us together.' "

An array of transistors fabricated using the weave-based masking process. Strips of the semiconducting coating are visible between the gold contacts. (courtesy the researchers) A close-up view of an e-textile sample during the weaving process. The thin wires containing the transistors and the perpendicular interconnect wires are woven between thicker Teflon threads that make up the bulk of the fabric. (courtesy the researchers)

Lee's fabric transistors act as the switches, enabling a grid of fibers to establish the path a signal takes from one device to the other. For example, the switches could determine that the most efficient route for a signal traveling from a temperature sensor on the wrist to a microprocessor in the breast pocket is across the shoulders. The dynamic switching even enables the e-textile to be fault tolerant.

"If I tear the sleeve of my electronic shirt, does that mean I lose all functionality?" Subramanian says. "No, because it adaptively routes around the rip."

The first step in fabricating the transistors is to coat a group of parallel hair-thin aluminum wires with an organic material called poly-4-vinylphenol (PVP). The PVP acts as a flexible insulator so that the fabric can be bent without breaking the circuit. Then, a semiconducting coating is added. After that, another layer of fibers, placed perpendicularly over the first group, acts as a mask so that gold contacts can be patterned onto specific portions of the wire. The result is an array of functional transistors. Interconnecting wires are then woven across the contacts so each transistor in the grid can be individually controlled. Traditional fibers like cotton can also be threaded through to fill out the fabric.

"The entire transistor is made without conventional lithography," Subramanian says. "The weaving is what sets the location of the transistor."

Until recently, Lee individually hand-stitched the wiring for each transistor. Now though, she's also experimenting with a tiny loom built in UC Berkeley's machine shop that will enable her to more quickly produce small samples of the e-textile.

"The loom is similar to what kids use in kindergarten," Lee says.

Lee is also working to improve the reliability of the transistors while ensuring that they're durable and flexible enough for real-world fabric applications. Meanwhile, she's developing additional solution-based processes that will enable the researchers to coat the wires with the organic materials by pulling them through a liquid, just as fibers are colored in the garment industry.

"Our key goal is to do everything just as it's done in textile manufacturing," Subramanian says.

Billy Kluver with a battery-powered neon letter he created for Jasper Johns's Field Painting (1964).

Born Johan Wilhelm Kluver in Monaco on November 13, 1927, he grew up in Salen, Sweden, where his father built the first ski hotel in the country. After graduating in electrical engineering from the Royal Institute of Technology, Stockholm, he worked on projects ranging from the television antenna atop to the Eiffel Tower to an underwater video camera for Jacques Cousteau. Kluver received his PhD in electrical engineering from UC Berkeley in 1957 and taught in the department the following year. From 1958 to 1968, he was on the technical staff of Bell Telephone Laboratories in Murray Hill, New Jersey, and it was during his years at Bell Labs that he became immersed in New York's thriving avant-garde art scene.

Kluver’s informal collaborations with artists reached their pinnacle in 1966 when he and Rauschenberg organized "9 Evenings: Theatre and Engineering." The performance series incorporated new technology developed by 10 artists working with more than 30 Bell Labs engineers.

Billy Kluver working on Oracle (1965), a collaboration with Robert Rauschenberg.

Encouraged by the success of "9 Evenings," Kluver and Rauschenberg then founded Experiments in Art and Technology (EAT), a non-profit organization designed to match artists with engineers and scientists. In 1970, Kluver led the EAT team to design the Pepsi Pavilion at Expo '70 in Osaka, Japan, a tour de force of high-tech installation art.

Kluver went on to write and edit several books, most recently A Day With Picasso, published by MIT Press in 1997. Kluver discussed this work during his last visit to UC Berkeley in the fall of 1997, as a speaker in professor Ken Goldberg's Art, Technology, and Culture Colloquium. Coincidentally, the presentation was scheduled on Kluver's 70th birthday. After his energized presentation, Kluver was surprised with a cake and a special gift--a bound copy of his dissertation, presented by his PhD advisor, EECS professor Emeritus John Whinnery.

"Billy Kluver was an inspiration to everyone in the art and technology community," says Goldberg, who is also an internationally recognized artist. "He showed that a serious engineer could make serious art."

Billy Kluver (center) in 2003 at the New York City opening of Mori, an art installation directed by Ken Goldberg (second from left).

When Kluver died, he and his wife Julie Martin were working on a social history of international art communities from 1945 to 1965 in the United States, Western Europe, and Japan. Kluver is survived by Martin; daughter Maja Kluver of Brooklyn, NY; son Kristian Patrik Klüver of Boulder, Colorado; half brothers Björn Tarras-Wahlberg and Lorentz Lyttkens; and half sister Ase Lyttkens all of Stockholm, Sweden.