AI Traffic Management
Real-Time LPR, Vehicle Flow Analytics and Congestion Detection
Traffic management decisions require data that is real-time by definition — a congestion pattern forming now, a plate that needs to be matched now, a parking bay that filled two minutes ago. Okiff Reflex delivers vehicle counting, LPR/ANPR, and flow analytics with sub-10ms AI inference, processed on-site without a cloud round-trip. Traffic intelligence that responds at the speed of traffic.
< 10ms
Vehicle detection and plate recognition latency
100%
RTSP / ONVIF camera compatibility — no hardware swap
Real-time
Flow analytics and congestion alerting
Why traffic data arrives too late to act on.
Traffic management systems generate enormous volumes of data — vehicle counts, dwell times, queue lengths, plate reads — and make almost none of it available in real time. The data is typically aggregated in batch cycles, uploaded to central analytics platforms, and surfaced in reports that describe what happened 15 minutes ago.
Three failures define the current model:
→ LPR latency disqualifies real-time enforcement. A licence plate reader that routes its reads through a cloud API for matching against a blocklist has already allowed the vehicle to proceed by the time the match result returns.
→ Vehicle counting without flow context is a vanity metric. Knowing that 3,200 vehicles entered a car park today does not help the operator manage the queue forming at the entry barrier right now.
→Traffic incidents are documented, not detected. A vehicle stopping in a prohibited zone, a pedestrian entering a vehicle-only area — these events appear in footage review after the operational disruption has already occurred.
Okiff Reflex: traffic intelligence that operates at the speed of the vehicles it monitors.
LPR / ANPR — Instant Plate Recognition
The Okiff Reflex ANPR engine reads and classifies vehicle licence plates in real time at the point of camera capture. Plate reads are matched against configurable local databases — access allowlists, enforcement blocklists, time-restricted categories — with the classification result available before the vehicle has cleared the camera frame.
Real-Time Vehicle Flow & Occupancy Analytics
Object Counting monitors vehicle counts per zone, per lane, and per entry/exit point in real time. Occupancy thresholds trigger automatic Rules & Events alerts — configurable as progressive responses at 70%, 90%, and 100% capacity.
Smart Zone Traffic Enforcement
Prohibited stopping zones, vehicle exclusion areas, and emergency access clearance zones are defined as Smart Zones on the camera view. Any vehicle stationary in a prohibited zone beyond a configured dwell time triggers an immediate alert.
What are Okiff's key capabilities?
LPR / ANPR — real-time
Plate recognition and database matching completing before the vehicle clears the camera frame
Multi-lane support
Simultaneous plate processing across multiple entry/exit lanes from a single ANPR instance
Object counting — vehicles
Real-time vehicle count per zone, per lane, per entry point — not batch reports
Congestion zone alerting
Configurable occupancy threshold alerts with Rules & Events multi-channel notification
Prohibited stopping detection
Smart Zone dwell-time monitoring for enforcement of no-stopping zones
Evidence Manager
Tamper-evident export for traffic enforcement proceedings and insurance investigations
Vehicle classification
Car, truck, bus, motorcycle — classified per zone with configurable alert logic
Pedestrian-in-vehicle-zone
Smart Zone detection of pedestrians entering vehicle-only areas
Cross-zone vehicle tracking
MTMC maintains vehicle identity continuity across multi-camera traffic environments
Connected vehicle integration
OBD / BLE / WIFI integration for connected vehicle data correlation
Historical traffic analytics
Full footage archive indexed via AI Video Description — searchable by vehicle type, plate, time
Use Cases
Corporate Campus (800-Space Multi-Deck Car Park)
A corporate headquarters manages an 800-space multi-level car park with 6 entry/exit lanes. Okiff Reflex ANPR processes every plate at each lane simultaneously — matching against a staff allowlist and visitor pre-registration database. Object Counting tracks per-level occupancy in real time, with digital signage integration via REST API directing drivers to available levels before they enter the structure.
Urban Traffic Management Centre (Municipality)
A city deploys Okiff Reflex across 80 intersections and 12 major vehicle access points in the city centre. Object Counting generates real-time traffic density data per intersection, fed to the traffic management centre dashboard via REST API. Smart Zones cover bus lanes and emergency vehicle corridors — alerting the traffic management team when unauthorized vehicles stop in these zones.
Logistics Hub (High-Volume Delivery Management)
A national distribution centre processes 1,200 delivery vehicles daily across 24 loading docks. ANPR identifies each inbound vehicle at the perimeter gate and matches it against the day's scheduled delivery manifest. Unscheduled vehicles receive a Rules & Events rejection flag before the vehicle reaches the loading area.
FAQ
Does Okiff Reflex LPR/ANPR require specialized traffic cameras?
No. The Reflex Router Agent provides universal RTSP/ONVIF compatibility — ANPR operates on any existing IP camera with a clear line of sight, at angles of 30° or less to the vehicle approach.
How does Okiff Reflex handle multiple vehicles in different lanes simultaneously?
The ANPR engine supports multi-lane configurations, processing concurrent approaches from multiple cameras simultaneously on the same edge server — with no cloud API rate limit constraints.
Is vehicle flow data stored locally or transmitted to a cloud analytics platform?
All data is stored on-premise in AES-256 encrypted edge storage. Real-time data is accessible via REST API for external platform integration, but source data never leaves your infrastructure.
Can Object Counting data feed directly into third-party traffic management systems?
Yes. Real-time count data and events are available via REST API and Webhooks for direct integration with traffic dashboards, digital signage, and smart city platforms.
How does Okiff Reflex handle partial plate reads in traffic environments?
ANPR supports partial plate matching with configurable tolerance thresholds for lighting, angle, or debris conditions. Confidence scores are included in every plate read event.
Traffic intelligence that acts at the speed of traffic.
Join us and make your company a better place.