SENTRY Soft Target Security Series: Real-Time Threat Detection & Mitigation
On February 19, 2025, SENTRY hosted the latest installment of its Soft Target Security Series, focusing on the Real-Time Threat Detection & Mitigation research area. Led by David Castañón of Boston University, the webinar highlighted innovative research aimed at enhancing security measures for soft targets and crowded places.
SENTRY’s research in this area focuses on developing advanced decision support systems for command centers, leveraging video analytics, multisensor information fusion, behavioral modeling, and threat mitigation recommendation systems. This session featured presentations from three SENTRY researchers, showcasing varied approaches to improving emergency response strategies in active shooter scenarios.
Presentations
Humanizing Agents in Agent-Based Models for School Shooting Simulations
Richard John, University of Southern California
Richard John’s research explored agent-based modeling (ABM) for simulating school shooting scenarios, emphasizing the importance of modeling agents based on actual human behavior rather than theoretical assumptions. By running 81,000 simulations with probability-driven reaction times and movement speeds, the study demonstrated how different social influence conditions—such as whether people ran, hid, or exhibited mixed behavior—affected casualties. Notably, lower law enforcement dispatch times and more realistic human responses had the most significant impact on reducing casualties. This approach underscores the potential for empirical behavior modeling to refine emergency preparedness strategies.
Learning-Enabled Real-Time Strategies for Safely Avoiding an Active Shooter in Unknown Environments
Mario Sznaier & Milad Siami, Northeastern University
Sznaier and Siami presented a new predator-swarm-guided particle (PSG) model designed to guide groups of individuals to safety in active shooter scenarios. This model simulates real-time decision-making in complex environments, such as teachers leading students or ushers guiding event attendees. The approach uses state-of-the-art machine learning to enable adaptive evacuation strategies that continuously re-optimize routes as new threats or obstacles emerge. This research provides a framework for real-time, AI-assisted evacuation planning in high-risk environments.
Advanced Real-Time Evacuation Routing for Active Shooter Scenarios
Subhadeep Chakraborty, University of Tennessee
Chakraborty’s presentation focused on AI-driven decision-making tools for evacuation routing during active shooter incidents. The system can dynamically identify the safest evacuation routes by integrating real-world school layouts, AI-driven simulations, and equitable risk distribution models. Key developments include the strategic containment of threats, smart door-locking mechanisms, and real-time digital simulations to optimize response effectiveness. This research provides a valuable tool for first responders and security personnel in planning for and mitigating active shooter threats.
Presentations are available to view on our YouTube channel.
The webinar highlighted the power of AI, behavioral modeling, and advanced decision-making systems in strengthening security measures for soft targets. Integrating realistic human behavior data, adaptive AI-driven strategies, and real-world simulation models offers promising threat detection, mitigation, and emergency response planning advancements. By bringing together experts from academia, government, and industry, SENTRY remains at the forefront of developing cutting-edge security solutions to protect vulnerable spaces.
