In a cavernous machine shop at Fort Belvoir, in northern Virginia, stands an eleven-foot-high bulldozer. Originally manufactured by the Swiss construction-equipment company Liebherr, it has been customized with thick steel plating on the undercarriage, a protective cockpit, and a special rotating tiller that is guided by ultrasonic depth-control sensors. The tiller is designed to "scarify the ground and bring large objects to the surface," Mike Collins, the Army engineer who adapted the machine, told me. The revamped bulldozer is for clearing land mines. Its final cost is likely to be half a million dollars.
Fort Belvoir is home to the Army's Night Vision and Electronic Sensors Directorate, which in turn houses the Humanitarian Demining R&D Program. Although the United States has so far refused to sign the 1997 Ottawa Convention banning the use of anti-personnel mines (it has cited, among other rationales, the need to continue mining the border between North and South Korea), in October of 1997 the Clinton Administration pledged to eliminate the global threat to civilians from land mines by 2010. Since then the United States has spent $255 million on "humanitarian mine action," which includes civilian mine-awareness programs, training for foreign de-miners, and development of new de-mining technology. Some $54 million of that money has gone to the Pentagon's R&D efforts, chief among them the program at Fort Belvoir.
Research efforts are under way outside the Pentagon as well, in both the private and the public sector. Government-supported efforts include the Multidisciplinary University Research Initiative and the Army Research Lab program, which award basic-research grants to universities around the country, and the Defense Advanced Research Projects Agency (DARPA, the organization that developed the early version of the Internet), which is pursuing highly experimental solutions that could take years to produce results. Among the programs DARPA has funded are a $25 million quest to develop an artificial dog's nose (at present dogs are the only mine detectors that can sense explosives rather than metal) and a $2.5 million effort to investigate whether honeybees could be trained to pinpoint land mines and other military targets.
The engineers at Fort Belvoir, in contrast, focus on developing tools that can be used by de-miners currently in the field. Those tools include the Air-Spade, which rapidly clears the earth covering suspected mines; the Rhino, a remote-controlled tiller for mechanically unearthing mines; and the mini-flail, a remote-controlled vehicle that thrashes the ground with spinning chains to detonate mines.
Ingenious as these devices are, many of them will never find their way to a minefield. And some of those that do—for example, a twelve-foot-long weed trimmer for clearing brush-have proved to be unreliable, useless, or even dangerous in real-world conditions.
L and mines are easily detected, but exceedingly difficult to distinguish from the tin cans, spent cartridges, and other metallic debris that litter the world's war zones. Fewer than one percent of all signals from de-miners' metal detectors indicate land mines. Moreover, the technology used in the field has barely advanced since World War II. By far the most prevalent system remains the so-called creep-and-probe method: a de-miner uses a metal detector to locate an "anomaly" underground and then probes the ground with a stick to determine whether the anomaly is a mine. Once discovered, mines are carefully dug up and detonated in place or removed. The process is laborious and inefficient as well as dangerous.