Small bridge sensors will give early warnings of anomalies, weaknesses

roughly 100 million times less than a typical light bulb, Kalantari told Dresser.

For the little power it does use, the silicone-coated sensors would not rely on batteries but on tiny supercapacitors with no chemical fuel to run out. Kalantari said that for now his company would guarantee the sensors for at least ten years but suspects they could last much longer with the help of advanced adhesives that would hold them in place through all types of weather.

Individually, the sensors would cost only about $25 to $30 to manufacture, Kalantari said, but it could take thousands to cover a span the size of the Bay Bridge. The most lucrative opportunity, he said, would be in selling the software and the services of monitoring and interpreting the data the sensors transmit.

Neither Kalantari nor his colleagues envision a product that would replace human inspections. Rather, they say, it would let inspectors track changes in the condition of bridges between the inspections they routinely undergo every year or two.

“The sensor will never 100 percent replace the human intelligence,” Kalantari said. “We don’t want to give the impression we want to replace the bridge inspectors with our sensors.” He said, though, that particularly with older bridges, a device that could monitor them constantly for signs of weakness could be useful. “If a bridge is 35-40 years old and has a history of structural deficiency, it may be time to put on something to give you real-time information,” Kalantari said.

Dresser notes that based on that description alone, the market for such a technology could be vast. With an interstate highway system that was largely built out in the 1950s through the 1970s, there are tens of thousands of bridges that fit that description in the United States alone. Takshi said such sensors could be particularly useful in areas of high earthquake activity, providing signals of problems before inspectors could be dispatched.

Kalantari, who lives in Bethesda, said he was on the UM faculty in 2006 — about a year before the Minnesota bridge collapse — when he began to explore some of the potential uses of sensor networking. He said he found there was significant theoretical work being done on the development of sensors but little on the creation of wireless networks of the devices.

“I like to see what can be done in a practical sense,” he said. Discussions with friends pointed the way toward monitoring