Surveillance Platform Domain

Overview#

A city transport authority runs 2,400 cameras across its rail network. When a weapon detection alert fires on a platform at 23:47, the Surveillance Platform has already done the heavy lifting: the AI pipeline verified the detection against a confidence threshold, created an incident record with a video clip, and fired a dispatch alert to the nearest security team, all within 15 seconds of the event appearing on camera. A supervisor watching the live event stream from the control room sees the alert, reviews the clip, and confirms the response. Every step, the camera feed, the AI verdict, the incident record, and the dispatch notification, flows through the Surveillance Platform as a single integrated system.

Key Features#

Camera Fleet Management: Onboard, configure, and manage cameras across facilities with support for multiple protocols, batch operations, and lifecycle tracking from installation through decommissioning.
AI-Powered Event Detection: Process video frames through multiple AI models simultaneously to detect objects, weapons, fire and smoke, faces, licence plates, and anomalous actions in real time.
Automated Incident Creation: Automatically create incident records when verified threats are detected, including video clip extraction with contextual pre and post buffers.
Alert Rule Configuration: Define configurable rules that determine how AI-detected events are handled, including confidence thresholds, event categories, notification targets, and dispatch triggers.
Real-Time Monitoring: Subscribe to live event streams for real-time situational awareness with filtering by zone, camera, event type, and severity level.
Dispatch Integration: Route critical events directly to dispatch systems for emergency response, enabling rapid reaction to detected threats such as weapons, fires, or violent activity.
Health Monitoring: Track the operational status of camera fleets, AI processing pipelines, and storage systems to ensure continuous surveillance coverage.
Feature-Gated Deployment: Control surveillance feature availability per organisation through feature flags, enabling phased rollout and customised capability sets.

Use Cases#

An integrated surveillance platform with AI detection and dispatch integration is valuable wherever continuous monitoring of large physical spaces is operationally necessary. Key industries include public safety and law enforcement, transport and infrastructure, and defence and government facilities.

Facility Security: Monitor building perimeters, entrances, and sensitive areas with AI-powered threat detection and automated alerting for security teams.
Public Safety Monitoring: Deploy AI-enhanced video surveillance across public spaces to detect threats, track incidents, and support law enforcement operations.
Critical Infrastructure Protection: Provide continuous monitoring of critical facilities with automated detection of unauthorised access, hazards, and suspicious activity.
Incident Investigation: Search recorded video with AI-assisted analysis to locate specific events, objects, or individuals relevant to active investigations.

Integration#

The Surveillance Platform connects with safety and security operations across the platform:

Camera Management: Camera fleet onboarding and lifecycle operations
Surveillance AI: Multi-model video analytics processing pipeline
Zone Management: Geographic zone-based camera and detection configuration
Incident Management: Automated incident creation from verified detections
Dispatch: Critical event routing for emergency response

Open Standards#

ONVIF (Open Network Video Interface Forum): Camera onboarding, device discovery, media profile enumeration, and RTSP stream URI retrieval all use ONVIF SOAP services at the device and media service endpoints.
RTSP (RFC 2326 / RFC 7826): The primary transport protocol for ingesting live video from IP cameras; the platform probes cameras via RTSP OPTIONS, constructs authenticated RTSP URLs, and normalises multi-profile paths (primary, preview, analytics).
WS-Discovery / WS-Security (OASIS): Multicast WS-Discovery is used to locate ONVIF cameras on the local network, and OASIS WS-Security UsernameToken is applied as a fallback authentication mechanism when HTTP Basic probe fails.
SUNAPI (Hanwha/Wisenet): The Hanwha camera HTTP management API is supported as a first-class protocol alongside RTSP and ONVIF, enabling device-info and RTSP URI extraction from Wisenet camera ranges.
RTMPS (Real-Time Messaging Protocol Secure): Live camera streams are re-published to Cloudflare Stream via RTMPS ingest for low-latency browser playback and clip extraction.
MP4 / ISO 14496-14: Verified incident video clips are extracted and stored as MPEG-4 Part 14 (mp4) containers with H.265 encoding, serving as tamper-evident evidence records.
WebSocket (RFC 6455): Operators and control-room dashboards subscribe to real-time event, alert, and health-status topics over persistent WebSocket connections for live situational awareness.
ISO 8601: All incident timestamps, frame timestamps, clip boundaries, PSAP dispatch timings, and heartbeat records are serialised in ISO 8601 UTC format throughout the domain.

Last Reviewed: 2026-02-24 Last Updated: 2026-04-14