An official website of the United States government
Here's how you know
A .mil website belongs to an official U.S. Department of Defense organization in the United States.
A lock (lock ) or https:// means you’ve safely connected to the .mil website. Share sensitive information only on official, secure websites.

Home : Media : News : Saved News Module
NEWS | Dec. 4, 2019

NSWC Crane, Old Dominion University machine learning research could significantly improve warfighter capability to detect drones

By NSWC Crane Corporate Communications

CRANE, Ind. – Naval Surface Warfare Center, Crane Division (NSWC Crane) and Old Dominion University (ODU) are researching machine learning (ML) based drone detection methods to enhance warfighter capability in the field. Through a Cooperative Research and Development Agreement (CRADA), NSWC Crane and ODU’s research uses ML techniques to detect and identify drones based on their classification of Radio Frequency (RF) signals.

This collaborative research method aims to work in remote and constrained environments and is portable – which is useful for warfighters. This method takes the latest technology and applies it to RF signals.

Dr. Sachin Shetty is an Associate Professor of the Department of Modeling, Simulation, and Visualization Engineering and the Associate Director for the Virginia Modeling, Analysis, and Simulation Center at ODU. Dr. Shetty is also an NSWC Crane temporary faculty employee while working on this research. He along with his undergraduate student research assistant, Mr. Michael Nilsen, a senior in the Department of Electrical and Computer Engineering at ODU developed machine learning techniques for adaptive detection of drones. The technique has also been successfully evaluated at a couple of outdoor test ranges.

“The benefits of using this machine learning technique is that no matter what types or models of drones made in the future, the technique can detect them,” says Dr. Shetty. “We didn’t want to create a technique that would be tied to a certain drone model and have to constantly change it.”

Dr. Shetty says ML examples in industry include facial recognition, like what you find on social media.

“Typically, you see industry having the computing power necessary for ML in supercomputers,” says Dr. Shetty. “Supercomputers are powerful but also take up a lot of space. Where the warfighters needs to use rapid drone detection technology, there is often no internet. They need a mobile, lightweight, and packable solution that works in a resource-constrained environment.”

He explains the way they are able to implement this technology in resource-constrained environments is due to the rigorous algorithms tested and used.

“This project is at the intersection of cyber, autonomous, and artificial intelligence technology,” says Dr. Shetty. “We created a library of algorithms which have been rigorously tested in a variety of urban and rural environments and different conditions. Essentially, we are able to give the warfighter an RF classification toolbox on a device the size of a phone.”

Dr. Shetty says the algorithms are trained to detect the key features of RF drone signals, which is more efficient, economical, does not require many resources to make, while being effective.

“The goal of this research is to help the warfighter improve their capability of detecting every drone. Improving detection is essential to mitigate any threats from adversarial drones. If we can detect it, we can mitigate those threats ahead of time.”

About NSWC Crane

NSWC Crane is a naval laboratory and a field activity of Naval Sea Systems Command (NAVSEA) with mission areas in Expeditionary Warfare, Strategic Missions and Electronic Warfare. The warfare center is responsible for multi-domain, multi- spectral, full life cycle support of technologies and systems enhancing capability to today's Warfighter.