As the use of robotics in industrial automation grows, so do the demands placed on reliable image processing. One of the key challenges here is to reliably detect objects in varying positions, locate them precisely and make them available for subsequent process steps. Conventional image processing solutions often reach their limits, especially when confronted with irregular shapes, reflective surfaces or unorganized bulk materials. At the same time, achieving stable results often entails considerable effort for integration and parameter configuration.
This is exactly where the new 3D Handling System comes in. SensoPart's camera hardware provides robust data for different distances and types of objects. The AI software from Cambrian processes the image data, using it to generate precision 3D information for the robot guidance. The AI models are trained in a cloud environment, enabling consistent results without the need for extensive parameter configuration.
Typical applications involve detecting and gripping complex components in boxes or other containers. The components are also recognized reliably when their shape, orientation or surface is challenging, even in changing illumination conditions. This includes, for instance, metal parts, different plastic components, transparent packaging or housing elements.
"Our joint approach demonstrates just how powerful modern AI-based image processing can be today," says Martin Stengel, Head of Global Key Account Management at SensoPart. "Users are provided with a tool that reliably handles even demanding tasks and makes it easier to get started with complex 3D robotics applications."
