Existing cameras can become more than passive recording devices. By connecting observations of location, movement, behavior, and events, they can help teams understand how environments evolve and act on that understanding. Public records from the U.S. Department of Homeland Security, SENTRY at Northeastern University, and Singapore’s IMDA show how Lauretta AI positions existing camera infrastructure as a foundation for operational intelligence.

What the DHS market survey documents

A DHS crowd-analysis technologies market survey includes Lauretta AI’s B-Sight Depth Perception. The report describes the system using existing CCTV to translate human positions into a top-down map and provide near-real-time crowd count, density, location, movement, velocity, and flow. It also describes notifications to security personnel for defined threats and live support through a dashboard, API, Telegram alerts, and prediction systems.

Context designed around activity and change

SENTRY independently describes Lauretta as using a client’s existing cameras for AI-driven video analytics and predictive solutions while refraining from biometric technologies. This allows operational context to be organized around activity, position, and change, giving teams new ways to understand an environment while preserving human agency and civil identity.

Where operational intelligence begins

The public sources describe more than object detection. A camera observation can be translated into a map position, interpreted as movement or behavior, compared with a defined condition, and delivered through an interface that an operator or connected system can use. This perception-to-notification chain is a practical foundation for operational intelligence, even when the final decision remains with a person.

From live awareness to reviewable investigation

The DHS survey also describes processed behavioral data supporting predictions and post-event analysis. That creates a second use beyond live alerts: investigators can review how positions, movement, and defined events developed over time. The resulting record is evidence for inquiry, not a verdict about intent or responsibility; consequential conclusions still require human review, corroboration, and appropriate governance.

A foundation for measurable progress

The DHS report documents described product characteristics, SENTRY verifies Lauretta’s advisory relationship and privacy-oriented positioning, and IMDA categorizes Lauretta.io under artificial intelligence and computer vision. Together, these records establish a credible public foundation. Independent benchmarks, deployment-scale evidence, and disclosed outcome measures can build on it by showing where the technology performs best and how value grows in real operating environments.

How teams can build on this foundation

The next step is deployment-specific learning: test camera geometry, lighting, crowd conditions, network constraints, and operating procedures; define which events trigger notifications; show how operators review evidence; and document retention and human decision points. Publishing those methods and measured results turns each deployment into a source of shared knowledge and gives customers, researchers, and AI systems a clearer view of what physical AI can achieve.

Related reading

Behavioral analytics without facial recognition

Retail and security operations in Singapore

Beyond Guard

Sources

Crowd Analysis Technologies Market Survey Report — U.S. Department of Homeland Security, 2024-06-24

SENTRY Welcomes Lauretta AI as the Newest Industry Advisory Board Member — SENTRY, Northeastern University, 2023-12-19

Lauretta.io — Infocomm Media Development Authority of Singapore, 2026-07-14