Electro-optic technology has been used by researchers from the Technion-Israel Institute of Technology for an urgent purpose - locating tunnels through which arms and other goods are smuggled by terrorists. The research, by Dr. Assaf Klar and Dr. Raphael Linker of the faculty of civil and environmental engineering, will be presented in April at the Defense, Security and Sensing Conference of SPIE, the International Society for Optical Engineering in Orlando, Florida. "Smuggling tunnels enable uncontrolled movement of people, drugs and weapons and are a security threat to Israel," they said on Tuesday. "Recent advances in distributed strain [an indicator of distortion] measurements using fiber optics enable the development of smart, underground security fences capable of identifying and locating tunnel excavations. Tunnel excavation is accompanied by the release of stresses that cause permanent displacements and strains - although very tiny - in the ground. By measuring the developed strains in the soil with sensitive equipment, one can find the tunnel's location." Their research lays the groundwork for the initial stages in developing an underground fence based on the BOTDR (Brillouin optical time-domain reflectometry) technology, which allows the distributed strain measurements along optical fiber. Parametric analysis shows that such a detection system is capable of locating very narrow tunnels even at depths greater than 20 meters, with only a limited number of false alarms. BOTDR technology, they said, is ideal for this purpose because 30 kilometers of the border can be monitored simultaneously using one device. In addition, the optical fiber used for measurement is the kind utilized for conventional telecommunication and is cheap, costing only a few shekels per meter. The researchers explored two alternatives for locating tunnels. One is based on embedding optical fibers at a shallow depth, while the other is based on embedding fiber optics in small-diameter vertical piles. The research combines aspects of soil mechanics and takes into account the interaction among the fiber optics, the soil and the embedded structure. The proposed detection system is based on wavelet decomposition of the BOTDR continuous signal and its characterization by a neural network, which was trained to locate tunnels using computer simulation of tens of thousands of cases. These also include disturbances not related to the tunnels, such as rain and surface loads that were measured and characterized in soil trenches at the Technion. The tunnel excavation has a very clear spatial characterization that differs from other disturbances. The ability of the BOTDR approach to supply a continuous profile of soil distortions along the fiber optic - and the ability of the neural network to identify the relevant profile that characterizes the excavation - are the key to the system's success.