Learn more
Artificial intelligence and machine vision
We train custom models using data collected from your site and optimise them for the edge device with TensorRT. Camera, lighting, processor, network and PLC integration are designed as a whole; the system operates under field conditions without dependence on the cloud.
Counting and sorting systems
Camera systems that count sacks, cases and steel bars on the conveyor. SmartBag Counter records every loading session on video together with the plate, customer and count, and exports the report to Excel; on steel lines bars from 8 to 50 mm are counted and separated. The recording settles shipment disputes.
Machine vision
Deep-learning models integrated with industrial cameras. Millisecond-level object detection, measurement and counting; results are transferred directly to the PLC via OPC UA, Modbus and SLMP.
Quality control
Detection of scratches, cracks, colour deviations, missing parts and assembly errors. Accuracy improves as the system sees more samples; operator feedback reduces false positives.
Licence plate recognition and vehicle access
OCR engines trained specifically for the Turkish licence plate format operate under variable lighting and weather conditions. The recognised plate is sent to the PLC, and the barrier opens automatically for authorised vehicles; records are stored in a local database.
Facial recognition and access control
Facial data is not stored on a central server; only irreversible biometric vectors are retained at each point. Turnstile, door, lift and barrier control is integrated with the PLC.
Edge AI
Models run on an edge device inside the factory panel. They are unaffected by internet outages, data does not leave the site, and latency is kept below 30 milliseconds.
Artificial intelligence and machine vision
We bring artificial intelligence to the factory floor: real-time defect detection, counting, measurement and recognition.