Self-learning AI camera TIV: Turck brings AI directly to the production line
Industrial vision for quality assurance, object recognition, and code reading—without AI expertise
With the new AI camera TIV (Turck Intelligent Vision), Turck makes it much easier to get started with industrial image processing – and brings artificial intelligence directly to the production line.
Instead of complex programming, just a few sample images are enough for the camera to independently learn patterns and differences. The TIV reliably detects good and bad parts, different variants or classes, and thus supports stable, automated quality assurance in the ongoing process.
At the heart of the solution is a powerful 12-megapixel global shutter sensor (Sony Pregius S, 4th generation) combined with an NVIDIA Jetson Nano GPU with 4 GB RAM.
Training and execution of the neural networks take place entirely on the camera – in real time and without additional edge hardware or cloud connectivity. Companies can thus use AI vision directly on the machine without having to extensively rebuild their existing IT and automation architecture.
Four AI apps for typical tasks in production and logistics
The TIV12MG-Q110N comes with four pre-installed AI apps that cover key industrial image processing tasks:
Difference Check – detects deviations and missing or incorrect components
Classifier – assigns components, products, or variants to different classes
Detector – locates objects in the image and outputs their position
Code Scan – reads 1D and 2D codes for tracking, traceability, and logistics
Inspection areas can be defined and evaluated separately. The camera delivers coordinates, confidence scores, and IO/NIO signals directly to PLCs or IT systems. This allows inspection, sorting, or picking processes to be efficiently automated without the need for a complex vision system in the background.
Robust, decentralized, and close to the process
The TIV is designed for direct use in industrial environments. The C-mount connection allows common industrial lenses to be used flexibly and adapted to different working distances. An optional protective tube brings the system up to IP67 protection – ideal for harsh environments, splashing water, or dusty production areas.
Control, lighting technology, sensor technology, and power supply are decentralized and robust. M12 connectors for power supply, network, trigger, and I/O allow the camera to be seamlessly integrated into existing systems—directly on the production line, without additional control cabinet space or external vision controllers.
Easy operation via browser and integration into the Turck Automation Suite
The TIV is operated intuitively via a web browser. Users
set up image sections and inspection areas,
select the desired AI app,
train the system with a few sample images
and check the results in real time.
No specialized AI knowledge is required. Integration into the Turck Automation Suite (TAS) also simplifies:
device management and firmware updates,
integration into digital maintenance and monitoring processes,
management and distribution of data sets and trained neural networks.
On-premise operation, transferable models, and data sets enable scalable rollouts to additional cameras—without additional licensing costs for external AI platforms.
AI vision as a building block for industrial IoT
With the self-learning AI camera TIV, Turck is expanding its portfolio for industrial automation with a powerful edge vision solution. The combination of robust hardware, integrated AI, and simple operation enables users to deploy industrial vision where the data is generated: directly on the production line, on the machine, on the conveyor belt, or in the packaging and logistics line.
Whether for quality control, object recognition, completeness checks, or code recognition, the TIV (Turck Intelligent Vision) AI camera lowers the barrier to entry for AI-supported image processing and helps companies make their processes more transparent, stable, and efficient.