Sandia fine-tunes simulation technology
- By Brian Robinson
- Sep 09, 2002
Sandia National Laboratories is developing a simulation and modeling tool that could greatly expand the ability of government agencies and emergency organizations to respond to terrorist attacks and other disasters.
Originally developed by Sandia's Weapons of Mass Destruction Decision Analysis Center in California to aid public health officials, the tool is now being adapted for other threats and applications.
The simulation and modeling tool is also an example of the kind of advanced technology-based developments that officials at the Office of Homeland Security have been pressing for as future aids for federal, state and local disaster response managers.
Homeland security officials proposed incorporating such tools into a nationwide Web-based network of resources that managers can use for training purposes or during actual emergencies.
An advanced version of the tool could be available in the next year or two, according to John Vitko, director of Sandia's Center for Exploratory Systems and Development department.
"People have [simulated these scenarios] through paper-based, tabletop exercises up to now, but they have to be prescripted and if you deviate from the script, then little extra information is available to the participants," he said. "Also, although people can learn valuable lessons from those tabletop exercises, they then have to go out and actively share those lessons with others."
Sandia's computer-based tool is far more flexible in that it allows users to add data at any time during the simulation. Plus, because it's designed as a distributed application, someone sitting at a terminal in Washington, D.C., can be as active a participant in exercises as people physically located in Sandia's Visualization Design Center "war room."
With the tool, an individual can model "what-if" scenarios just as easily as a group can, Vitko said.
One major advantage of the tool, according to participants in trial exercises, is that it can use both real and projected data. A recent exercise involving the release by terrorists of a biological agent in the San Francisco Bay Area, for example, used real data on diseases and deaths gathered from local hospitals and coroners' reports.
Using this and other data, such as air toxicity measurements or more detailed physicians' reports, Sandia officials feel disaster response strategies can be more effectively examined and tested by decision-makers.
"I'd been involved in several tabletop exercises before, but nothing as sophisticated as this," said Jeanie Perkins, earthquake project manager for the Association of Bay Area Governments, who took part in the exercise. "What made it so interesting was the way it was set up as a real-time simulation, with data flowing in the way it would in a real-life situation."
Gary Simpson, medical director for infectious diseases at the New Mexico Department of Health, came to the exercise as one of the few government executives who has experience dealing with a deadly disease caused by an airborne pathogen. The hantavirus, spread from the dust of dried mice droppings, was eventually identified as the cause of 11 deaths on Navajo reservations in New Mexico and Arizona in 1993.
"The challenge for people who have not been through that kind of situation is that it's really difficult to experience it in a way that enables you to make responses based on real-time data," Simpson said. "With the hantavirus, we saw otherwise healthy young people suffering from things such as respiratory distress and other symptoms, and it was a very complex situation to deal with."
Simpson said there is a lot of "wiggle room" with paper-based tabletop exercises, because you never seem to get enough good information to make realistic decisions. The Sandia tool sometimes provided too much data, he said, which was also good because it forced people to identify the most relevant information.
Another advantage of the Sandia tool, Vitko pointed out, is that it "learns" as people use it and over time will become more sophisticated in its handling of the assumptions and projections used for given scenarios.
Robinson is a freelance journalist based in Portland, Ore. He can be reached at firstname.lastname@example.org.
Brian Robinson is a freelance writer based in Portland, Ore.