The concept of a structural-acoustic (SA) sonar was first inspired by basic research conducted at the Naval Research Laboratory (NRL) which noted discriminatory features present in the broadband scattering signatures of elastic structures that could be exploited for target classification. The scientific underpinnings for this technology in the application of unexploded ordnance (UXO) remediation were developed under SERDP Projects MR-1513 and MR-2103. The project team estimated that more than 2 million acres (30%) of the over 400 underwater sites potentially containing munitions have water depths applicable to the autonomous undersea vehicle (AUV) base SA technology whose impact would be considerable given its ability to prosecute buried UXO. The objectives of this project were to: (1) Demonstrate the ability of AUV based SA sonars to detect and classify proud and buried UXO in a shallow water area which can act as a good surrogate for UXO-impacted areas in which the Department of Defense holds environmental responsibility; and (2) Access the performance and cost of this technology. The environmental versatility of the technology is highlighted by the relatively difficult environment regarding sediment type, indigenous clutter densities, and interfering recreational activities, fishing, and marine life at the demonstration site.

A primary innovation in the technology is acoustic color i.e. the echo’s spectral content versus angle swept out by the AUV. The SA sonar’s low frequencies penetrate the sediment to the buried targets whose echoes contain features effective in separating UXO and non-UXO. This project demonstrates such an AUV-based SA down-looking sonar against buried targets and a complementary side-looking SA sonar looking out sideways to greater distances providing higher coverage rate detection and classification against proud and partially buried UXO.

In blind testing, all 30 buried and proud UXO were successfully detected. The receiver operating characteristic (ROC) curves for both the buried and proud blind tests were each able to achieve a classification performance of Pc=0.90 well within a false alarm rate of Pfa=0.25. For this demonstration, coverage rates were 0.23 acres/hr for the down-look sonar and 2.3 acres/hr for the side-look sonar. The processing time was 600 hrs/acre and 90 hrs/acre for buried and proud target fields, respectively. Significant improvements to the buried target ROC, dig-list, coverage rate, and processing time performance parameters were realized in post-demonstration data processing studies. This demonstration utilized the NRL sonar payload on the available 21” diameter Bluefin-21 medium class AUV. With easily realized changes, the sonar technology could operate on smaller AUVs, such as the IVER 3. The larger Bluefin cost is ~$1.4M and the IVER 3-class cost ~$0.6M. Site preparation and data collection costs are $5.5K/Hectare with data processing costs about $13K/Hectare and $1.3K/Hectare for the down-look and side-look, respectively.

At-sea operations with the medium class AUV-based sonar are robust and well established. Implementation issues for the scaled-down sonar are related to source/receiver modifications resulting from movement to a smaller, less expensive AUV such as one of the IVERs. The design of the sonar may have to be modified to be compatible with the size, structure, and hydrodynamics of the smaller AUV. Based on the detailed studies which have been carried out in execution of this demonstration program - with respect to sonar design, fabrication, and classification performance dependence on source/receiver parameters - there is sufficient technical information to allow the successful procurement of a smaller, equivalent performance AUV-based structural acoustic sonar.