"ChatGPT moment for robots" has arrived with 80x faster path planning and 3D-guided picking solutions.
With "physical AI," artificial intelligence is moving out of its digital confines to power robotic applications with humanlike perception, creating solutions that can handle variation without prior teaching or programming.
"We are entering an era of intelligent automation where robots can learn from and adapt to their surroundings in real-time," says Ujjwal Kumar, Group President of Teradyne Robotics, parent company of Universal Robots (UR). "This opens new possibilities for complex, creative and collaborative tasks, transforming robots from tools into intelligent partners that enhance workplace productivity and creativity."
50-80x Faster Path Planning and New 3D-Guided Picking Solutions
UR's AI-powered autonomous inspection solution showcasing robotic path planning 50-80x faster than today's solutions has been made possible through NVIDIA's accelerated path planning tools. Teradyne Robotics has collaborated closely with NVIDIA incorporating the AI capabilities into UR's soon-to-be-released PolyScope X operating software as a foundation for developments.
"As we continue to develop in-house physical AI capabilities, developers have also been turning to us as the preferred robotics platform for AI solutions," says Kumar. Two of those address new approaches to bin picking, a task that in more than 90% of applications is still performed manually.
One of them is Photoneo, a UR Marketplace partner. The company has unveiled a unique new, AI-enabled multi-view localization solution, that leads to exceptional results of 3D bin picking of difficult objects. The Bin Picking Studio features the newest MotionCam-3D, incorporating a UR cobot and two statically independent PhoXi-3D Scanners to capture the scene from multiple viewpoints, enabling true automation in motion -- such as picking in dynamic applications.
The other partner, Siemens, has debuted SIMATIC Robot Pick AI, a pre-trained, deep learning based vision software for piece picking robot applications that is used for autonomous bin picking primarily in warehousing and e-commerce applications. The system does not need to be trained on the parts being picked -- any unknown part can be picked from a bin, which means the number of SKUs being picked are endless.
Want more information? Click below.