With iiQKA.AI Vision, KUKA introduces new, integrated vision software for industrial robot applications. The solution supports both 2D and 3D vision and enables typical applications such as bin picking, assembly, quality control, and depalletizing. Since configuration is done intuitively via drag-and-drop, users can quickly implement productive vision applications without specialized knowledge.
Easy to get started; no additional hardware required
With iiQKA.AI Vision, image processing runs as edge computing directly in the controller on the KUKA AI Extension Platform, an integrated computing unit based on the NVIDIA Jetson Orin module. This ensures low-latency AI inference as well as secure, local data processing. In addition, sufficient computing power is available for demanding vision and AI tasks. No cloud connection, other external systems, or additional infrastructure is required.
The solution supports robust and reliable automation processes in industrial environments.
All data processing takes place locally within the system. Images are processed and stored on-board. As a result, users benefit from a high level of data security and independence from network connections. This not only reduces the integration effort but also lowers the total cost of ownership (TCO) in the long term and significantly simplifies the system architecture. Cameras and compatible components are available directly from KUKA and are tailored to typical industrial requirements.
The application is configured entirely within iiQWorks, using a standalone vision application with a graphical user interface. Thanks to drag-and-drop functionality, pre-built templates for typical applications, and guided workflows from setup to commissioning, vision applications can be implemented particularly efficiently and without in-depth specialized knowledge.
Robust image processing ensures process reliability with precise 2D and 3D object recognition, automatic position determination, and gripper point calculation. It’s reliable even with varying components and changing lighting conditions. Seamless integration with iiQWorks and the robot controller creates a unified system from engineering through operation, offering a consistent workflow from configuration to deployment, as well as centralized management and monitoring via the web interface.
AI-powered object recognition for robust processes
The AI algorithms used in iiQKA.AI Vision make it possible to precisely distinguish objects from the background and automatically calculate suitable gripping points. Precise position and gripping information is calculated based on image data or point clouds and transmitted directly to the robot. This results in seamless, stable processes from detection to execution and consistently high image quality.
With iiQKA.AI Vision, KUKA offers an efficient, robust, and ready-to-use image processing solution that significantly simplifies the transition to industrial vision applications and increases productivity.
Bringing Physical AI to Life: iiQKA.AI Vision at GTC Berlin
KUKA will present iiQKA.AI Vision at NVIDIA GTC Berlin from October 20 to 22, 2026. The demonstration illustrates how KUKA combines industrial automation expertise with AI and intelligent robotics to bring Physical AI into reliable, scalable, real-world applications.
