Globally active companies have to assert themselves in a changing environment and at the same time overcome problems in supply chains. Shifting production back to the domestic market is increasingly an option. This requires not only a certain degree of resilience, but also compliance with strict environmental regulations and cost-efficient strategies to make domestic production competitive. If you want to ensure the competitiveness of domestic production, you also need to overcome staff shortages.

Automation through robotics has long since become the driving force here, and artificial intelligence is increasingly playing a key role. This technology is developing just as rapidly as the pressure to automate is increasing. In order to map production processes in your own company with AI, the simplest possible integration and the shortening of training phases are already decisive factors. This is where the British start-up Cambrian Robotics comes in with a fully AI-based solution for various robotics applications in production. It handles fast bin picking or pick and place, the precise feeding of parts for machines and various work steps in material handling.

Analyze and recognize

The easily integrated system consists of a module for robot arms, a computer unit with pre-installed, intelligent software and a camera module, each equipped with two uEye+ XCP cameras from IDS. "The task of the cameras is to capture an image of the area with the objects to be handled. Based on the images, the software can analyze the scene and identify exactly where the objects are located," explains Miika Satori, founder and CEO of Cambrian Robotics. The images are processed using the heart of Cambrian Vision, a self-learning software for predicting the position of parts and their pick points. This ensures image matching on an AI basis so that no classic 3D point cloud is required. Based on simulated data, the AI learns independently and localizes the pick points and parts extremely precisely. The AI models for part recognition and communication with the robot are controlled by a powerful GPU. "With the Cambrian software package, pick points for new parts can be defined and the application configured in just two to five minutes," emphasizes Satori.

The associated camera module is equipped with two space-saving XCP industrial cameras. "The two IDS cameras use the stereovision principle to provide images of the object scene from different viewing angles. The challenge is to determine the position of the part to be gripped as accurately as possible from these images. This, in turn, is the task of the AI," says Satori. The combination of image acquisition, AI models and special image processing makes it possible to determine pick-up points and positions with particular precision. "Standard CAD applications for 3D bin picking often use structured light or sensors that project something onto the environment, generate a point cloud and then try to find the part within it. Cambrian uses just two standard industrial cameras instead of a 3D camera.

High accuracy and speed

With an accuracy of less than 1 mm, Cambrian-Vision is also more precise than other systems. "The system reliably detects a wide range of parts, including shiny, reflective or transparent components, where conventional vision systems often reach their limits. At the same time, it remains robust against external light conditions," says Satori, describing the special requirements placed on the cameras, which are a fundamental part of the solution. "The system is also super-fast, as the inference speed is less than 170 ms, whereas comparable solutions often take more than 1,000 ms." The fast calculation time enables cycle times of 2-3 s in a bin-picking setting. "This ensures efficient, precise and accurate execution in a single pass," emphasizes Miika Satori. The One-Shot system is therefore currently one of the fastest AI image recognition systems on the market.