Interior Reconnaissance
and Inference System
Where a unit conducting a building breach is able to gain access to the structure's closed-circuit cameras, IRIS ingests the live camera feed and fuses every camera into a single, continuously updated interior picture: the number and location of every person in the area of operations, their movement, and how the situation evolves in real time.
IRIS is a single software core running across a family of compute form factors; primarily backpack- and vehicle-borne. Feeds enter; a perception pipeline performs detection and tracking; a fusion layer resolves the camera observations into a single interior picture; an output layer pushes that picture to the assault element's display.
What IRIS can report depends directly on the data it is given. From CCTV alone, IRIS performs two functions dependably: it detects people, and it tracks them across the camera network. Each additional data source adds a specific capability. Which of them apply is a function of what the mission can provide.
[ IRIS CIV ARCHITECTURE ]
Civilian Facility Security
Most facilities already have far more cameras than anyone can parse. A person at a security desk can follow only a handful of feeds at once, so at any moment most of the building goes unobserved, and the cameras end up recording far more than anyone actually sees in real time.
During an incident, responders arrive with no interior picture of the building, no count of who is inside, and no idea where people are, beyond what someone has managed to describe over radio. During an evacuation, headcounts are pieced together by hand at the moment in which speed and accuracy count. IRIS closes this gap by turning the cameras a facility already has into a live, continuously updated picture of where people are.
Security staff watch one clear picture of the whole facility, and responders can be handed that same picture the moment an incident begins.
Core Capabilities
Every IRIS deployment includes the following.