A large portion of the cost associated with unexploded ordnance (UXO) cleanup projects is attributed to carefully excavating innocuous items in response to false alarms produced by sensors. Extremely conservative survey standards must be employed to avoid missing buried targets of concern, but focusing on discrimination, or the interpretation of signals, can help determine the object or type of object that has produced them. Broadband electromagnetic induction (EMI) sensors have shown great promise in this realm. Operating as low as some 10s of hertz and as high as 100s of kilohertz in the frequency domain, the cutting edge in instrumentation development can be exploited to obtain the data diversity necessary for fundamental discrimination progress. Data diversity implies an array of connected measurements over which patterns may be recognized. To complement the already wide and growing frequency/time ranges being implemented, different viewing angles or spatial variation of responses should enlarge the picture. Beyond that, new technology is required to add reception of all three vector components of the magnetic field signal. This would complete the definition of a target's EMI response. At present, high-scoring systems like the GEM-3 typically receive only a single (e.g., vertical) magnetic field component.

The objective of this SERDP Exploratory Development (SEED) project was to modify the GEM-3 sensor head and the necessary control and signal processing so the complete magnetic field vector can be defined for a target's EMI response.

EMI signal transmitters, whether they correspond to the sensor or to a responding target, send out a fan of curving magnetic field lines. An elongated magnetic object tends to orient this fan in the direction of its long axis. Conventional EMI sensors in their usual orientation have horizontal transmitter and receiver coils, which inevitably miss some of the signal of interest. In this project, a set of vertical coils were built into the center of the GEM-3 sensor head, orthogonal to each other and to the original horizontal receiver coil. The new receiver coils are oriented perpendicular to the transmitter coils, thereby largely escaping contamination by the transmitted primary field. The new coils also reside largely in the transmitted field minimum, near the center of the head, further helping to isolate them from the primary field.

The data produced by this instrument or its successors in the course of close interrogation should enable a new generation of processing, inversion, and classification. Improved discrimination will lead to a reduction in false alarms and more positive identification of unseen target types, the benefits of which include cost reduction and increased speed of surveys during UXO cleanup projects. (SEED Project Completed - 2004)