George Mason researchers enhancing emergency communications resiliency and effectiveness through AI

Paul Houser, professor, Geography and Geoinformation Science and executive director, Center for Resilient and Sustainable Communities (C-RASC); Lin Wells, executive advisor, C-RASC; Kathryn B. Laskey, professor emerita, College of Engineering and Computing (CEC); and Hemant Purohit, associate professor, Information Sciences and Technology, CEC; and Karina Korostelina, professor, Jimmy and Rosalynn Carter School for Peace and Conflict Resolution, received funding for the project: “Enhancing Emergency Communications Resiliency and Effectiveness Through AI.” 

The key objective of this project is to increase disaster resilience in communities while giving citizens a greater sense of confidence in their preparedness for emergencies by investigating the use of artificial intelligence (AI) to improve disaster resilience and support emergency preparedness of public safety agencies and citizens in a community. With an initial focus on Fairfax County and neighboring regions, this project aims to use AI wherever possible to reduce response times in emergencies by enhancing situational awareness for public safety officials, lightening the workload on 9-1-1 call center operators, improving enhance training processes of 9-1-1 call-takers by developing a customizable, personalized, and more realistic training capability, improving community stakeholder engagement, and extending support to some previously underserved communities. 

The project will leverage planned improvements in Fairfax County Department of Public Safety Communications (DPSC), reinforce their capabilities, and introduce new commercial tools to improve the training of call-takers.

One of the major challenges faced by DPSC, despite the comprehensive training program, is the limited training capability for diverse scenarios of emergency situations.

“This project will investigate the design of an AI-powered voice assistant to enable personalized conversation-based scenario training for trainees to go through various real-world scenarios at their own pace while accounting for complexities such as varying incident descriptions and multilingual callers,” Co-PI Purohit said.

Further, to improve the awareness of the situations for effective use of 9-1-1, the project aims to support the goals of DPSC for community engagement, enhancement of community education initiatives, and buy-in of the benefits of implementation of AI in public safety communications in the diverse communities of Fairfax County. “Community inclusion in the development and deployment of such capabilities is key -- ‘nothing about us without us’ is a basic principle,” Co-PI Wells said.  

Collectively the collaboration will: (1) improve the ability to identify and flag fake/fraudulent emergency calls; (2) provide real-time integrated detection of diverse languages; (3) seamlessly hand over calls to agents; (4) allow the triaging of calls and links to the Computer Assisted Dispatch (CAD) system; (5) automate handling of the workflow associated with alarm system testing and other system updating; and (6) provide general consulting services in AI. 

The broader impact will be the enhanced readiness of public safety communication and enhanced retention for new recruits through improved training processes as well as community support. The project team will employ a Design Thinking approach to collaborate with DPSC and conduct community engagement activities, which will proceed in five basic phases: (a) Empathize; (b) Define; (c) Ideate; (d) Prototype; and (e) Test.

“The Design Thinking process ensures that we are developing the solution the stakeholders want and need,” said Co-PI Laskey. 

Karina V. Korostelina, Professor and Director of the Sustainable Peace Lab at the Carter School for Peace and Conflict resolution said that the university will work with local officials on further outreach and communication efforts in support of the project.

Co-PI Korostelina said, “Our task is to create a systemic community engagement plan that involves Neighborhood and Community Services, Immigration Services, Fairfax County Public Schools and libraries in educating community members about the use of 9-1-1 and the role of AI. We also will conduct multiple focus groups and town halls to understand community concerns and facilitate participatory approaches to the improvement of Public Safety and Communications,” she said.

Her team, that includes a Ph.D. student and a post-doc, will implement cutting-edge practices of co-creation and community empowerment to enhance collaboration between communities and government agencies in Fairfax.  

These activities will be paralleled by extensive community engagement to gain the community support that is essential to success and sustainability, especially when potentially controversial features of AI services are involved. Community inclusion in the development and deployment of such capabilities is key (“nothing about us without us”). 

The project will include 15-20 interview sessions with first responders, while 5-7 focus groups with community representatives will explore community views and concerns regarding the use of AI in emergency management. These will be combined with education and training campaigns. 

“Based on the experience of this project in Fairfax County, we will explore the possibilities of offering the resulting technological capability to other 9-1-1 call centers,” co-PI Houser said.

The researchers received $961,000 from the National Institute of Standards and Technology for this project. Funding began in Sept. 2024 and will end in late Aug. 2025. The project was made possible through the efforts of Congressman Gerry Connolly (D-VA).