AI Anomaly Detector

FactoryVision is a fault detection system with an optical camera for optical identification of manufacturing faults. Production line debris is caused both by the wear and tear of the production equipment and by accidental, primarily human errors. Current control systems typically specialize in the detection of the former, regularly occurring errors. With the advancement of applied informatics and optics, advanced control and control methods are gaining ground in manufacturing technologies. Complex systems that perform various sensory checks are currently operating worldwide, but they can test well-defined properties (color, size, component presence) under very limited conditions; if the product changes or the equipment enters another environment, the entire system needs to be re-tuned to fit the specific task, generating a small amount of time and cost. The visual quality control system eliminates these steps because it automatically defines the environment and the product in the unique production environment along with the problem: it creates a model of the product, does not need to predefine the element to be tested or creates a model from the environment. it can adapt so it does not require a fixed design, cover, shading, illumination. The models are refined continuously and can continue to operate on major changes without the need for operator or maintenance intervention. Able to manage multiple products at once, create multiple product models at the same time without specific (unique, advanced) settings. The system is able to identify scrap products early in the production process through self-learning modeling. The multimodel feature allows you to create multiple model profiles, analyzing products by category accordingly. With the help of advanced self-learning algorithms, the system was adapted to the changing circumstances. A special feature has been added to filter out noisy visual backgrounds. The device is capable of communicating with intelligent production control systems using standard industry 4.0 interfaces.

Industry 4.0 Communication Interfaces: RS485, USB, CANBUS, MODBUS, TCP / IP, RFID
Self-learning automatic modeling
Multimodel capability
Adaptation to changing circumstances
Noisy visual background filtering
Industry 4.0 connectivity (control, monitoring) on ​​standard interfaces: RS485, USB, CANBUS, MODBUS, TCP / IP, RFID
Product distance: 0.3 – 2 m
Product size: 0.2 – 1 m
Product Speed: 0 – 50 cm / s
Yield: 0 – 120 / min ”
Power supply: ~ 230V AC, 50Hz, 8A
Operating temperature: 10 to 40 ° C ”