Researchers receive NSF grant to develop novel gunshot detection technology

Vanderbilt researchers have received a U.S. National Science Foundation CIVIC Innovation Challenge grant to develop innovative gunshot detection technology using machine learning that could become a model for communities nationwide.

Will Hedgecock, principal investigator at Vanderbilt’s Institute for Software Integrated Systems, is leading the development. Co-PIs of the $698,169 grant are Akos Ledeczi, professor of computer science and electrical and computer engineering at Vanderbilt University; Charles Loeffler, associate professor of criminology at the University of Pennsylvania; and postdoctoral researcher Kathleen Padilla in the School of Criminology and Criminal Justice at Arizona State University.

The acoustic gunshot locator (AGL) system consists of small, wall-powered acoustic sensors that residents can plug into outlets on their porches or exterior walls. Each device connects to the cloud via cellular technology, requiring no internet setup from homeowners. When gunshots are detected, machine learning algorithms running on Vanderbilt’s patented sensor fusion technology precisely locate the source of gunfire within about 15 yards.

“We need to solve the complex challenge of accurately detecting and localizing gunshots in irregular, community-based sensor deployments,” said Ledeczi, who has been conducting acoustic gunshot research at Vanderbilt for over 20 years. He is also a senior research scientist at the Institute for Software Integrated Systems.

The six-month pilot deployment will take place in the North Austin Civic Association (NACA) area, a diverse community of 27,000 residents that experiences about 300 violent crimes annually.

What sets the Vanderbilt-led research apart is its comprehensive approach combining technical innovation with deep community engagement. The Vanderbilt team conducted extensive surveys and workshops with residents, police, and city officials to ensure the technology meets real community needs. If successful, the system could provide a scalable model for communities nationwide struggling with gun violence.

“This isn’t just about better technology,” Hedgecock said. “It’s about creating a system where communities feel ownership and control, which we believe will lead to more effective reporting and better police-community relations.”

 

Contact: Lucas Johnson, lucas.l.johnson@vanderbilt.edu