ABCNEWS.com October 18, 2002
Sounding Out Snipers
By Paul Eng
The military has long understood the dangers of enemy snipers - especially those lurking in urban settings. In places like Sarajevo, sharp-shooting gunmen hidden among the city's buildings and rubble claimed many lives during the Bosnian conflict. And starting around 1993, the Defense Advances Research Projects Agency (DARPA) began work on a way to find these lethal predators.
By far, such systems rely on sensors that track the sniper's trademark calling card: the snap and flash of a high-speed rifle bullet in flight.
One prototype system developed by BNN Technologies, a division of telecommunications company Verizon, uses a network of inexpensive microphones, portable computers and compasses connected to a central processing station.
Each audio sensor can be mounted on a soldier's helmet, a vehicle, or lamppost and is tuned and is tuned to detect the "crack" - actually a tiny sonic boom - of a rifle shot. When two or more sensors pick up those acoustic vibrations, location and time data from each sensor is sent back to the central processing station.
A computer there coordinates the data and mathematically determines the trajectory, distance, speed and elevation of the bullet. Following the bullet's flight path backwards thus reveal the location of the shooter.
Another system developed by the Maryland Advanced Development Laboratory in Greenbelt, Md., tracks bullets in a different manner.
Instead of audio sensors, the Viper Counter Sniper System uses an infra-red video camera that can be mounted on aircraft - including unmanned robot planes. As the aircraft circles over a given area, the camera detects the invisible infra-red light and heat signals that are generated from the muzzle of a fired rifle. Once such a muzzle blast is picked up, the system notes the shooter's location using Global Positioning Satellite signals and beams the information down to nearby troopers.
Measurable Results?
DARPA hasn't released any information on the accuracy of such systems to nail down the exact location of a sniper. But similar gunshot detection systems have been proving useful in some municipalities for other purposes.
Since 1995, the police department of Redwood City, Calif., has been using a system called ShotSpotter to detect troublesome gunfire in one area of the municipality. The acoustic system is similar to BNN's setup and uses just eight microphones to cover an area of 1 square mile.
Ward Hayter, an administrative manager for the city's police department, says that the $200,000 ShotSpotter system has helped cut down on the amount of random gunfire incidents. "Our total number of gunshots incidents was around 300 to 334 in the 1995, 1996, 1997 timeframe," says Hayter. "Now it's about 180."
Unlikely to Nail Serial Sniper
But experts are quick to point out that the technology is highly unlikely to help police in the Washington, D.C., area trying to find a sniper who has struck 11 times since Oct. 2, killing nine.
John Pike, founder of GlobalSecurity.org, a defense think tank in Alexandria, Va., says that such counter-sniper systems are useful for military operations where battle commanders might have a rough guess of where enemy snipers might be hiding. But it would be extremely difficult to adapt such systems for the serial sniper which has struck locations in Maryland, Virginia as well as the District of Columbia.
"All of this stuff is basically useful over ranges of thousands of feet," says Pike. But, "Look at the area he's operating in, it's just too big."
He notes that if investigators could predict where the serial sniper most like might strike next - say, another strip mall parking lot or schoolyard - such gunfire detection systems might help. But even then it might be an impossible proposition.
"The criteria would appear to be a lighted location with a clear line of sight of a few hundred feet [and] multiple avenues of egress," says Pike. "Well, that's basically any shopping center or service station and in an urban area with a population of about five million people, how many such locations would exist?"
Copyright © 2002 ABC News Internet Ventures.
