Volume 1, Issue 3 November 2001

October

July

Smart Buildings Admit Their Faults David Pescovitz photo When Steve Glaser isn't experimenting with Smart Dust Motes like those he holds in his hand here, he's most likely playing with his six-and-a-half-year-old daughter Nina or the 37-year-old Alfa Romeo he's painstakingly restoring. (Click for larger image.) Behind a building's manicured façade, disaster may be lurking. Minor earth tremors, for example, may not cause visible damage but they can create hidden cracks in support columns -- cracks that could eventually fail during a higher-magnitude quake. Even after a large earthquake, when beam buckling and structural bruises are likely, a building's true condition can only be determined by tearing down tons of sheetrock. Frequently, buildings are closed for months to undergo a detailed inspection. And even that is an imprecise science. But what if the walls could write their own bill of health? As part of the Center for Information Technology Research in the Interest of Society (CITRIS) program, UC Berkeley civil and environmental engineers and computer scientists are collaborating on technology to make buildings, bridges, and other structures aware of their own health. The result, says civil and environmental engineering associate professor Steven D. Glaser, would be smart structures that could tell us how they're holding up and automatically self-prioritize necessary repairs. Someday, he says, large buildings might be outfitted with stoplight-style signs that automatically announce structural soundness. "Of course we'd like to know a building's structural integrity at any moment, but we really want to be able to intelligently prognosticate how it will behave in the future," Glaser says. Smart Dust Motes, tiny and inexpensive sensors developed by UC Berkeley electrical engineers, are enabling Glaser and his colleagues to attain both of these goals. Outfitted with wireless radio transceivers and their own TinyOS operating system, the battery-powered matchbox-sized Motes can be built to sense numerous factors, from light and temperature for energy saving applications to location to dynamic response, the key characteristic for civil engineering. Even the tiniest movement of a supporting column in a building can reveal the structural soundness and, for instance, suggest that the column is handling more load than it should due to a problem elsewhere in the structure. Currently, wired seismic accelerometers - which measure movement - cost upwards of $8,000 each and are tricky to install. As a result, their deployment in buildings is kept to a minimum. The trouble with the current minimalist paradigm of structural monitoring, Glaser says, is that a handful of accelerometers in a large building can only provide a big picture view of a building's structural integrity. As a result, a problem only becomes visible once the entire building is affected and safety has already been compromised. Courtesy Steven Glaser A wireless Smart Dust Mote, marked with the red arrow, is surrounded by today's wired, bulky, and expensive commercially-available accelerometers in a seismic experiment at the Richmond Field Station. "You're monitoring globally and damage is local," Glaser says. But if sensors cost less than $1,000 and can be installed in minutes, "dense packs" of them can surround all critical beams and columns, providing extremely detailed structural data. In a recent test at UC Berkeley's Richmond Field Station seismic research laboratory, Glaser's team installed 15 Motes in the wood framing of a three-story model apartment building constructed on a "shake table" that simulates earthquakes. During the controlled quake, the Motes gathered seismic data from multiple locations in the building. That information was then compared to discern the way the tremors spread through the building and how the structure reacted. This kind of information, he says, will increase our understanding of earthquakes and how to prepare for them. In November, Glaser's students will conduct an experiment in Japan - ths time placing the sensors underground in liquefiable soil that will be subjected to earthquake-simulating dynamite blasts. A network of structural Motes can also act as an oracle of sorts, enabling researchers to use computer simulations of earthquakes, fires, or other structural threats to forecast the potential for damage. Civil engineering professor Gregory L. Fenves, a key participating investigator in the CITRIS disaster risk reduction research, expects the flood of data from the Smart Dust Motes to greatly increase the accuracy of his finite element analyses, a method of computer modeling where mathematical equations represent a structure's behavior under certain conditions. With a sensor network behind a building's walls constantly streaming data in real-time, the models can be updated throughout the life of a building and the prognostications will become much more than educated guesses. "As you learn the building, your models improve," Glaser says. The next step in the research is programming the sensor networks to deal with their own data. The Motes' TinyOS already enables them to automatically establish their own network and share information as soon as they're switched on. Eventually, the engineers hope the Smart Dust Motes will gain enough brainpower to process the raw data they collect before it even leaves the building. "In the end, we don't want data," Glaser says. "We want to know what the damage is. Let the sensors discuss the data among themselves and tell us where the problems are." Steven D. Glaser's home page Gregory L. Fenves's home page CITRIS Smart Dust TinyOS: An Operating System for Networked Sensors Lab Notes is published online by the Public Affairs Office of the UC Berkeley College of Engineering. The Lab Notes mission is to illuminate groundbreaking research underway today at the College of Engineering that will dramatically change our lives tomorrow. Lab Notes is written by David Pescovitz. Send comments to the Engineering Public Affairs Office: lab-notes@coe.berkeley.edu. © 2001 UC Regents. Updated 11/15/01.