Stony Brook and Local Firm Win DTRA Grant to Build Radiation AI

On January 7, 2026, Stony Brook University and Redshred announced they had been awarded a Phase I Small Business Technology Transfer contract by the Defense Threat Reduction Agency to develop RADIANT, the Radiation AI Decision and Information Assistant for Nuclear Tasks. Led by principal investigator Manoj Mahajan, the project will create an AI system that extracts and interprets complex nuclear and radiation data and delivers field-ready, tactical recommendations for military, emergency-response and civilian nuclear-safety operations.

RADIANT is intended to translate dense radiological documentation into actionable guidance such as exposure time limits tied to dose thresholds and other immediate operational decisions. The platform is being designed to run on small local devices for on-site use, enabling responders to receive timely information without depending solely on remote connectivity. Stony Brook’s Center for Excellence in Wireless and Information Technology and Stony Brook Medicine will contribute health-physics expertise, AI development and operational integration. Redshred, a company based in CEWIT on the Stony Brook campus, brings prior Department of Defense contract experience to the collaboration.

For residents of Suffolk County, the project has several practical implications. If RADIANT progresses beyond Phase I feasibility and into deployment, local emergency responders could gain new tools to make faster, more precise decisions during radiological incidents, potentially reducing unnecessary exposures and improving triage and sheltering choices. The technology also has civilian applications in nuclear power plant operations, facility inspections and international treaty-monitoring, areas that intersect with regional infrastructure and workforce needs.

The Phase I designation signals an initial focus on feasibility and prototyping rather than immediate wide-scale rollout. That means local agencies should anticipate a period of testing, validation and training before any operational adoption. Integrating an AI health-physics companion into emergency protocols will require coordinated planning among county emergency management, public health agencies, hospitals and municipal first responders to ensure guidance is accurate, understandable and actionable in high-stress situations.

Beyond technical performance, the project raises policy and equity questions about access, oversight and accountability. Ensuring disadvantaged communities receive equal protection during radiological events will require deliberate distribution strategies, training resources and transparency about how recommendations are generated. Local economic effects are likely positive: the partnership reinforces Stony Brook’s role as a regional research hub and supports job growth in tech and health-physics work on Long Island.

As the Phase I effort moves forward, Suffolk County officials and health systems will need to monitor progress, engage in validation exercises and plan for how an AI decision companion could be incorporated into existing emergency-response frameworks to enhance community safety.