Vision AI platform
Bina is Niadad's vision AI platform that analyses images and video at the edge, from plate recognition to parking occupancy.
- Intelligence
- Niadad
- In development
Problem
Cameras in car parks, streets and compounds produce hours of video that nobody watches. Centralised analysis depends on bandwidth and adds latency, and the results never reach operational systems such as payment and access control.
Platform overview
Bina is Niadad's computer-vision platform: it runs vision models on edge devices and in the core, extracts meaningful events from video (plates, entry and exit, bay occupancy, anomalies) and delivers them as events to other systems. Models are trained and versioned on Rayon.
Key capabilities
Plate recognition
Reads vehicle plates across lighting and angle conditions and matches them against permitted lists.
Occupancy and counting
Reports bay status and vehicle or people counts per zone continuously.
Edge inference
Runs models on edge hardware so analysis does not depend on central bandwidth.
Event and anomaly detection
Turns unexpected movement, long stops and entry into restricted zones into events.
Camera and model management
Manages cameras, analysis zones and the model version of every device from one console.
Architecture
Edge agent with on-device model execution
Video pipeline and frame extraction
Detection, tracking and plate-reading models
Event layer and delivery to target systems
Camera and model management console
Integrations
Use cases
Related platforms
Let's build together.
If your organisation, bank or industry is ready to turn data into decisions, start the conversation here.