VLM output
A fixed CCTV camera overlooks a port container terminal's gate and access roadway in daylight, where container trucks and tractor-trailers move along the multi-lane road while others are parked in rows beside stacked containers. Several workers in high-visibility vests stand along the lane markings and cone lines directing traffic, and over the frames a container semi-trailer drives from the gate area across the frame toward the lower right.
- Port Access Road VisibleYes
- Container Trucks PresentYes
- Moving Truck Count2
- Traffic Congestion LevelLight
- Queue At GateNo
- Traffic Marshals PresentYes
- Pedestrians In RoadwayYes
- Lane Cones DeployedYes
- VisibilityClear
Customer challenges
Trucks arrive together when vessel schedules cluster, backing gate queues into access roads and the city
Port authority, terminal operators and city traffic bureau each manage one segment with no shared traffic data
Illegal parking, double-parked waiting and dangerous-goods vehicles mix on port access roads, raising accident risk
Gate processing times and queue lengths are reported manually, giving no live picture and no peak forecast
Incidents or roadworks leave no live detour guidance; a single event blocks all port access
No objective data supports decisions on road widening, additional gates or appointment systems
Customer benefits
24x7 AI-driven awareness of hinterland truck flow
Port and city share one live traffic picture for the first time
Gate queues and access-road backups become predictable and dispatchable
Violations and dangerous-goods vehicles tracked live, lowering accident and community impact
Data to support investment decisions on gates, roads and appointment systems
Runs on existing CCTV, sharing compute and the digital twin with the port safety platform
Key features
Truck Flow Perception
Recognizes container, tanker and dangerous-goods vehicles, tracks them across cameras, measures gate queue length, waiting and processing times automatically, and detects illegal parking, double-parked waiting and wrong-way driving.
Access Corridor Event Management
Detects incidents, breakdowns, debris and lane-occupying works, grades event severity automatically, reasons the cause of congestion, and pushes events in real time with CMS traveller-information sync.
Hinterland Flow Analysis and Dispatch
Projects live traffic with hotspot analysis, forecasts peaks with detour recommendations, and recommends which gates to open and which appointment windows to use.
Hinterland Digital Twin and Dispatch
Projects hinterland traffic in the digital twin and recommends dispatch actions.
Before & after
| Before | After — with Linker AI Nexa |
|---|---|
| Before Queue length reported manually | After — with Linker AI Nexa Gate queue and waiting times computed automatically in real time |
| Before Port and city manage separate segments and trade blame | After — with Linker AI Nexa A shared digital twin and traffic dashboard |
| Before Relief attempted only once the peak arrives | After — with Linker AI Nexa Peaks forecast from vessel schedules and historical flow, dispatched in advance |
| Before Illegal parking and dangerous-goods trucks found by patrol | After — with Linker AI Nexa VLM detection with cross-camera tracking |
| Before Detours improvised after an incident | After — with Linker AI Nexa Detours and CMS messages triggered by the event |