FROM THE collapsing of city walls during medieval sieges, to the laying of giant mines filled with explosives under enemy trenches during the first world war, to the supply networks of the Ho Chi Minh trail during the war in Vietnam, to the Tora Bora cave complex, used first by the Mujahideen to oppose the Soviet invasion of Afghanistan and then by the Taliban to oppose the American invasion, tunnel-digging in times of conflict has a long history. These days, secret tunnels are used to move weapons and people between Gaza and Egypt, and by Kurdish militia operating on the frontier between Syria and Turkey. But the same principle applies. What happens underground is hard for the enemy to observe. Digging for victory is therefore often a good idea.
That, though, may be about to change. Real-time Subsurface Event Assessment and Detection (RESEAD), a project being undertaken at Sandia National Laboratories in New Mexico, uses novel sensors to make accurate maps of what is happening underground. This will, no doubt, have many civilian applications. But Sandia is principally a weapons lab, and it is military matters that are uppermost in the mind of the project’s leader, Chet Weiss.
RESEAD grew out of work that was looking at ways of monitoring earth tremors by means of sensor networks. The researchers involved found that the analytical techniques they had developed to handle data so collected were fast enough to process those data more or less in real time. This would mean they could detect movement underground in a way that would be militarily useful.
The versions of these networked sensors used in RESEAD are embedded in the collars connecting sections of borehole casings, and communicate with the surface via wires. This makes them easy to deploy and use. The sensors themselves are a mixture of accelerometers, which pick up vibrations, current detectors, which measure the electrical-resistance of rocks and soil, and subsurface radar.
RESEAD’s real secret, though, is in the way it handles the incoming data. First, these are pre-crunched by powerful hardware built into the sensors themselves. This step, a form of parallel processing, greatly reduces the time required by the central receiving computer to finish the job. Second, the mathematical technique employed, a species of finite-element analysis, divides the volume represented by the incoming data not merely into blocks, as is normal in this approach, but into blocks the faces and edges of which can have different properties. This means a block can straddle, say, the interface between a metal pipe and the rock surrounding it. That considerably reduces the number of blocks needed, and further simplifies the calculation.
To test their system Dr Weiss’s team decided to look at an oilfield. Most oilfields have only one or two wells per square kilometre, which is hardly challenging. The team, however, chose in Bakersfield, California, which has a dense network of oil-related infrastructure. The square kilometre they picked contains more than a hundred wells, many underground storage tanks and tens of kilometres of steel pipeline. RESEAD made short work of this challenge. It produced an accurate map of the area in just ten minutes.
Exactly how RESEAD sensors would be put in place in a zone of active conflict remains to be seen. But the system could certainly be useful for other sorts of security. In particular, America has a problem with tunnels under its border with Mexico being used to smuggle drugs and migrants into the country. RESEAD would be able to detect existing tunnels and nip new ones in the bud.