Abstract

“Advanced Sonar-Based Deep Learning for Underwater UXO Remediation” by Dr. David Williams (MR21-3543)

An unfortunate legacy of former military activities is the contamination of aquatic environments with military munitions, which pose a serious threat to both humans and the environment. In the United States, more than 400 underwater sites, spanning an area in excess of 10 million acres, potentially contain such munitions. The objective of this project is to develop novel detection and classification algorithms for unexploded ordnance (UXO) to support remediation efforts at these sites. The new algorithms are based on deep-learning techniques and leverage multiple nonconventional sonar data representations within a convolutional neural network framework. These methods should enable higher probabilities of detection and classification, at much lower false alarm rates, than is possible with existing approaches, thereby reducing costs during remediation. Results of the new algorithms will be shown for three-dimensional volumetric sonar data collected by two systems spawned from SERDP/ESTCP funding, namely the Sediment Volume Search Sonar (SVSS) and the Multi-Sensor Towbody (MuST).

Speaker Biography

Dr. David Williams is an associate research professor in the sensor analysis and data modeling department at the Applied Research Laboratory (ARL) at The Pennsylvania State University. His current research interests include the intersection of machine learning and underwater acoustics. Prior to ARL, Dr. Williams was a scientist in the mine countermeasures division at the NATO Science and Technology Organization, Centre for Maritime Research and Experimentation (formerly NATO Undersea Research Centre), in La Spezia, Italy, where his work focused on sonar-based pattern recognition and algorithm development for autonomous underwater vehicles. Dr. Williams was the recipient of a James B. Duke Graduate Fellowship and a National Defense Science and Engineering Graduate Fellowship, and has been recognized with the Excellence in Review by the IEEE Journal of Oceanic Engineering twice. He received a bachelor’s degree, master’s degree, and doctoral degree in electrical and computer engineering from Duke University in Durham, North Carolina.