The use of land mines to impede the movement of enemy troops has a long history, dating back to 1277 when the Chinese used explosives to repel the invading Mongols. Regrettably, the use of land mines has gained in popularity ever since, and it is estimated that there are about 110 million active land mines in the world today, killing or injuring an average of 70 people every day. Most of these are civilians; about 30% are children. The removal of these mines is a slow and exceedingly dangerous task; if no further mines are placed, removing the existing mines is deemed to take more than 500 years, costing more than US$33 billion. Tragically, about 20 new mines are currently placed for each mine that is removed.

Significant work has gone into finding efficient ways to detect land mines, for instance using ground penetrating radar (GPR), chemical sensors, metal detectors, and even trained bees. Most of these methods have serious disadvantages; metal detectors, for example, have difficulties in magnetic soils and with mines of low metal content, and GPR in clay or wet and conducting soils and with mines very close to the ground surface. As a result, to be able to detect the mines, the detector's sensitivity often needs to be significantly increased, thereby also increasing the likelihood of a false detection; as much as 99.96% of all alarms during the clearing of a mine field may be due to false alarms. A promising alternative would be to construct a detector based on nuclear quadrupole resonance (NQR). Such a detector would allow for both reliable detection and the clearing of false alarms. The idea of using NQR to detect explosives, and in particular land mines, goes back to the early 1950s, when the British Army suggested the possibility to researchers working with Nuclear Magnetic Resonance.


NQR is a radiofrequency (RF) technique in which the observed frequencies depend on the interaction between the electric quadrupole moment of the nucleus and the electric field gradient generated at the nuclear site by external charges. All common high explosives contain 14 N, a quadrupolar nucleus generating three sets of resonance frequencies, providing an unequivocal method of detecting and identifying an explosive, as well as estimating its quantity and depth. Because of its high specificity there is little or no interference from other nitrogen-containing material that may be present - such as the mine casing or fertilizer in the soil.

However, the NQR signals can be very weak, particularly from the common explosive TNT, and there is an urgent need to find techniques for improving the probability of detection, as well as combining the technique with another, faster detector to produce a more reliable and rapid method of clearing minefields. NQR is also of particular interest due the possibility of using the technique to detect explosives at airports and other public places, as well as for detecting narcotics, or for pharmaceutical quality control; for instance, different crystalline structures can give rise to different pharmaceutical properties in medical drugs.

In close collaboration with the NQR group at King's College London, we work on developing robust and reliable detection algorithms for NQR signals. Recently, our work has focused on multi-sensor systems and active interference cancellation techniques, and on exploiting the possible presence of more than one crystalline structure. The group has published extensively and holds several patents in the area. The project is, or has recently been, funded by the Swedish Research Council, Carl Trygger's Foundation, Vinnova, STINT, Innovationsbron, and the US NSF.

Links related to the detection of land mines and to NQR:

• Article on NQR in Scientific American.
• The international campaign to ban landmines.