“If you begin to look at current crisis management infrastructures, they’re messy. It’s helter-skelter,” says James Llinas, director of the university’s Center for Multisource Information Fusion. The group is in the first year of a five-year project to make software systems that collect and interpret bits of disparate data-news broadcasts, 911 calls, satellite imagery, reports from fire and police departments, even readings from remote sensors attached to roadways and buildings-in a process known as information fusion. Currently, an official trying to ascertain road damage in the aftermath of an earthquake might have to keep one eye on the TV news while listening to both radio traffic reports and the police scanner. Since most of these data are available in digital form, the software could take them all in, process them and present a report outlining the best evacuation routes.
The center is using the 1994 Northridge, CA, earthquake as its first case study, but the software could be used for a variety of scenarios and tailored to any organization from the Federal Emergency Management Agency to a local fire department. At present, few of the agencies that respond to disasters use any decision support software. And, says Llinas, the center’s effort marks the first significant attempt to apply information fusion-long used by the U.S. military to streamline intelligence and surveillance operations-in civilian settings.
Steve Charvat, disaster recovery manager for the Washington, DC, Emergency Management Agency, is hopeful yet cautious about the new system. “The end product sounds great,” he says, “but until I actually see something better, the human brain still makes the best filter in the heat of battle.” But when the Buffalo programs become operational in a couple of years, Charvat and other emergency responders might just get some help in making the tough decisions.