The origin of this article is

Berkeley 911

How a Berkeley professor and his team helped
locate and rescue missing persons during Hurricane Katrina.
By Patrick Dillon

SUSAN LINDELL RADKE ANSWERED A CALL for help at her Orinda home the morning of September 1, more than 72 hours after Hurricane Katrina struck the Gulf Coast. On the other end in Jackson, Mississippi, was Ed Hanebuth, a business colleague, phoning from the state's emergency management center, where he'd volunteered his expertise in computer mapping. "I could hear the commotion in the background when Ed called," she said.

Emergency service workers were overwhelmed, Hanebuth told her. "Volunteers were taking names, addresses, pertinent information such as the last time someone saw somebody, notating special circumstances such as medical conditions," he said. "They were documenting all this on paper forms. There was tons of paper. They were about 1,500 forms behind when I walked in."

"It was just crazy. Cell phones failed, the 911 system wasn't working, phone lines were going nuts," recalled U.S. Coast Guard Lieutenant Commander Byron Thompson, who had been trying to coordinate search and rescue operations with the Mississippi Emergency Management Agency (MEMA-the state's version of FEMA). "We were getting frantic calls: 'My aunt Matilda's on the roof and the water's rising.'Our aircraft had been flying randomly doing a lot of 'search and jerk, 'just seeing someone in trouble and pulling them up."

Hanebuth sized up the chaotic MEMA operation and saw an immediate need for an electronically distributed information application to manage records of missing persons. That's why he called Susan Radke. The two had worked together for eight years developing and distributing educational software to school districts throughout the nation. The software overlays local maps with local information-street grids, parks, hospitals, census information, even what she calls "cultural mosaic," information about ancestry, race, spoken languages. "Kids respond visually and spatially. This software helps them do what any eighth-grader in America should be able to do-analyze geographic information," she said. "Students are able to map their neighborhood roads, even the mailboxes and basketball courts, with this software. They locate their own relationship to the world beyond their block."

Susan Radke hung up the phone and turned to her husband, John Radke, an associate professor of environmental planning and geographic information science at Berkeley. Within 24 hours, they would gather a team and, using computer systems programmed, managed, and based in Berkeley, organized and directed data that helped locate thousands of missing persons and pinpointed stranded survivors nearly 2,000 miles away.

Radke's team integrated Internet database tools to build an electronic vehicle allowing anyone with access to the Web to click a "missing persons" button and enter their own information or information about someone they were concerned about. This was then collated electronically with lists of people turning up at shelters, census data, latitude and longitudes of personal addresses, pharmacies, and hospitals. It was the output of this database that helped rescue and relief workers to map the last known locations of people reported missing.

"In an emergency, you have to be able to size up all the pieces of the puzzle," he said. "We tried to devise a system whereby all the pieces showed up. For instance, when people arrived at the emergency centers, they would have to be identified, counted, and referenced. They would be asked or they would volunteer whether someone they knew or were related to was missing."

"We were getting four or five hits a minute," Hanebuth added. "And we were getting matches of people missing to people found or showing up in emergency shelters."

This spontaneous, remote relief effort demonstrates the efficiency of a custom designed application on an off-the-shelf operating system, matched with human wherewithal, which government emergency agencies found themselves initially lacking during the chaotic ordeal of Hurricane Katrina.

"We could call up the information system, make the recovery, and even determine where to take people," recalled Thompson, the Coast Guard search and rescue liaison officer. "We'd even be able to find a hospital or a soccer or football field for a landing zone. The data would tell us where to go."

During the next four days more than 9,500 names, personal records, and geographic locations would be entered into an electronic database. More than 2,000 air and ground-level rescues would be initiated and recorded-with helicopters swooping low over mostly submerged houses that revealed no signs of life, until rescuers lowered themselves and punched holes in roofs, retrieving people who might otherwise have been Katrina's certain victims.

The program worked so well that by Wednesday, September 7, MEMA downloaded the data tables and gave them to the Red Cross. Susan Radke has only one lament. "We designed, developed, and delivered a system in 24 hours that should have been in place immediately after 9/11. The federal government has had four years to develop something like this."

John Radke, her husband, estimated that the Berkeley Web-based program cost less than $4,000 to develop. Iterations are simple to achieve. Their cost, a mere $100 or less. "The bottom line is that we were able to step in and help from anywhere and do it automatically," he noted. "You can't avoid a hurricane but you can avoid chaos. We learned that you can't wait around for bureaucracy."

Patrick Dillon is the executive editor of California Monthly.

For a more comprehensive view of the technology powering the Katrina rescue and relief operation from Berkeley, see Berkeley-based rescue and relief computer program.