CityIntegrator

Video Analytics

SmartCity AI Video Analytics

The analytics engine: detection models, rules and event delivery.

CityIntegratorVideo Analytics
SmartCity AI Video Analytics
Licensing
subscription
Latest version
3.4.0
SKU
CI-VA-CORE
Delivery
Certified local integrator

The analytics engine is the component that processes video. It runs detection models against each configured stream, evaluates the rules attached to that stream, and emits events to the CCTV platform, the Smart City dashboard or any subscribed system.

Rules, not just detections

A detection ("a person is in frame") is rarely useful on its own. Rules add the conditions that make it actionable: which zone, for how long, during which schedule, and how many.

Capabilities

Zone and line rules

Region-of-interest zones and crossing lines with direction and dwell conditions.

Scheduling

Rules active only during defined periods, so out-of-hours policies differ from working hours.

Event delivery

Events pushed to the CCTV platform, dashboard, webhooks or MQTT.

GPU inference

GPU acceleration for higher stream density per server.

Model management

Enable only the models a stream needs, which keeps hardware requirements proportionate.

Specifications

Specifications are maintained from the product database. Deployment-specific figures depend on hardware, camera placement and network conditions — your integrator will size the system.

Runtime

Operating systemWindows Server 2019+ / Ubuntu LTS
AccelerationCUDA-capable GPU recommended

Input

SourcesRTSP, ONVIF, VMS stream

Licensing

Licensed byStream, annual subscription

Verified integrations

  • ONVIF cameras
  • RTSP streams
  • Milestone XProtectMilestone
  • MQTT / IoT brokers

Release history

  • 3.4.0Latest01 Oct 2026
  • 3.0.0LTS01 Oct 2026

Frequently asked questions

Will it work on our existing cameras?

Often yes, provided the camera sees the scene at a usable resolution and angle for the rule you need. Your integrator validates this per camera during the survey.

How many streams per server?

It depends on the models enabled, the frame rate and the hardware. Your integrator sizes this from the stream list rather than a generic figure.