The probability of unexploded ordnance (UXO) detection on documented U.S. Army test sites can exceed 90% in carefully executed geophysical surveys; however, despite recent advances in geophysical technologies and signal processing algorithms, false alarm rates remain unacceptably high. One of the challenges faced by geophysicists is development of improved electromagnetic induction (EMI) sensors and algorithms capable of recording target responses, isolating them from geological noise, and processing them to discriminate the nature of the target during cleanup operations.

 The objective of this project was to develop a prototype, time-domain EMI multi-receiver sensor along with its supporting interpretation software.

This project consisted of three main tasks—sensor prototype development, software development, and integrated system testing. The system hardware includes a transmitter, a multi-channel receiver, a power supply module, a transmitting coil, multiple receiving coils, and a signal processing unit. A 16-bit data acquisition (DAQ) card is used in the data acquisition and processing module to acquire signals associated with the EMI responses of buried targets. Labview processing code is used along with the DAQ card to control the transmitter and multi-receiver subsystems, acquire data from the DAQ card, and write the data into a disk file in a laptop computer. The system enables multi-channel acquisition of transient EMI responses from buried targets. The UXO/clutter/fragment discrimination and UXO classification rests on an interpretation of the EMI transients in terms of target parameters such as location, depth, orientation, shape, electrical conductivity, and magnetic permeability. The EMI interpretation is cast as a parametric nonlinear inverse problem. The discrimination and classification algorithms are treated as pattern recognition tasks. Integrated system testing of the multi-receiver EMI prototype system was conducted at the Texas A&M Riverside campus field site.

A multi-sensor EMI system prototype was developed. The transmitter, receiver, and power supply constitute the system’s main hardware and are completely designed on a single circuit board, thus making the system very compact and man-portable. EMI interpretation software also has been developed to (1) discriminate UXO from non-UXO such as clutter or exploded metal fragments and (2) classify UXO into various categories such as mortars, artillery shells, submunitions, etc. Field tests showed promising indications that the prototype system will be a useful instrument for buried UXO target discrimination and classification.

The multi-sensor EMI system represents an innovative tool for buried UXO target discrimination and classification, enabling a reduction in remediation costs through increased probability of detection with lowered false alarm rates. A rigorous system calibration is required before the system can be deployed in the field.