Traffic and Enforcement
Multi-camera enforcement across intersections, congestion analytics, average speed, and two-wheeler and bike-lane monitoring.
See Multi-Camera Traffic Analytics
Most cities do not lack cameras. They lack one place where the cameras, the analytics, the alerts and the accountability meet, which is why so much municipal camera investment produces so little.
In Production
IREX is integrated into Kazakhstan’s national smart-city program, with the full catalog of AI modules deployed across participating cities. It serves as the central intelligence layer for the country’s urban modernization initiative, unifying cameras and sensors under a single management and analytics infrastructure with real-time alerting and inter-agency collaboration.
In Peru, the Lima and Miraflores multi-city command centers run on the same architecture, with a large-scale national rollout in partnership with Visium. Miraflores became the reference for the region, and IREX launched what it describes as Peru’s first real smart city.
The multi-city pattern is the important part. A municipality runs its own estate and its own command center, and decides what it shares upward to a national tier or sideways to a neighbor. Federation is a choice each tier makes, not a condition of joining. That is the model described in full on the Government as a Service page.
Municipal Use Cases
Multi-camera enforcement across intersections, congestion analytics, average speed, and two-wheeler and bike-lane monitoring.
See Multi-Camera Traffic Analytics →Weapon detection, watchlist alerts, missing-person lists, and crowd management at public events.
See the Video Analytics Overview →The long tail a city is liable for, described in a sentence rather than waiting for a vendor module: flooding, illegal dumping, graffiti, damaged fences, uncleared snow.
Explore StreamVLM™ →Cameras plus user-defined device layers for ATMs, legacy CCTV and intercoms, with indoor floor plans for civic buildings.
See Global Map & Floor Plans →Continuous export to Apache Superset: heatmaps, trend analysis, and compliance auditing of how the analytics are used.
See the Apache Superset Integration ↗Hierarchical roles so police, transport, emergency management and city services share one platform without sharing everything.
See Security & Compliance →On Camera
Four conditions no vendor ships a module for, each defined by an operator typing a sentence. This is the class of problem that decides whether a city program pays for itself, because it is what residents actually complain about.
A camera already pointed at the underpass answers “is it flooded yet” every few seconds. What the city buys is the hour of warning, not the picture.
A service-level agreement becomes a detector: uncleared snow, blocked curb ramps and fallen debris checked against a deadline, on cameras the city already runs.
Unlicensed street trading, obstruction and misuse of public space are things a city normally learns about only when somebody complains. Here each one is a prompt.
The stairwell was repainted recently: the patched squares are still visible on the wall. Catching the act instead of the aftermath is the entire economics of graffiti. Shipping in beta on selected instances.
StreamVLM™ →The Public Argument
No, and reusing them is the standard model. A citywide deployment is built around connecting existing ONVIF and RTSP cameras without replacing camera hardware. Some analytics, notably traffic enforcement, have stricter optics requirements and may need additions.
A single instance is proven at 100,000+ cameras, and the citywide proposal template is parameterized up to 10,000 cameras as a standard vehicle. The architectural ceiling is 1,000,000 cameras and 100,000 users without re-architecture.
Yes. Hierarchical role-based access grants each group specific cameras, watchlists, locations and monitors, with six roles from Viewer to System Manager, which is what lets police, transport and city services coexist on one platform.
With a pilot: one site, three to five use cases, six to twelve weeks, and two or three measurable success criteria agreed in writing before it begins. Accuracy is benchmarked on your own cameras, and the measured numbers go into the contract.
Keep Reading
An anchor deployment that proves the outcome in one district is how every national program we run actually began.