The demand for robot-based automation of pick-and-place applications in logistics, warehousing, commissioning or packaging is high. If the objects to be processed vary greatly in shape, size, color, or position, either complex teach-in procedures are required – or the challenge is solved with intelligent image processing software.
This can help robots to reliably identify even unknown objects and find the best gripping positions for picking. On July 20th, the IDS Vision Channel will discuss what is technically possible and practical.
For the online focus event "Intelligent Pick & Place", IDS has brought the companies Fizyr, MVTec, and urobots on board. Participants will learn first-hand how companies are successfully using AI-based image processing and will be able to discuss individual questions afterward. Registration and participation are free of charge at the company’s website.
Session 1: "Picking the unknown", Fizyr
Artificial intelligence and 3D camera technology enable robots to classify and reliably grasp unknown objects. This includes boxes as well as deformable, tightly stacked, overlapping, or even damaged objects.
Deep-learning algorithms can suggest over 100 gripping poses per second, making successful picking rates of over 99 percent possible, even without knowing the objects or needing time-consuming training to do so.
Session 2: "Pick & Place with HALCON", MVTec
The participants will learn some methods in detail so they can then assess which is the best solution for their application
HALCON is a comprehensive standard software for industrial image processing that can process both 2D and 3D camera data. It also offers a wide range of options for locating any objects in different scenarios. Thus, HALCON can support pick & place and even random bin picking applications.
Using live examples and actual use cases, the participants will learn about some methods in detail so that they can then assess which is the fastest and best solution for their own application.
Session 3: "Vision-guided robot", urobots
Artificial intelligence (AI Vision) cameras enable intelligent robotics applications which, depending on the application, do not require an additional PC. urobots demonstrates a solution in which a robot immediately locates pre-trained (2D) objects lying on a flat surface, picks them up safely, and places them back in the correct position.
With compact embedded vision devices such as an IDS NXT industrial camera, a complete vision solution and even direct control of machines can be realized. This makes it possible to create picking applications quickly and efficiently.
The IDS Vision Channel is an online platform created by camera manufacturer IDS for the exchange of knowledge on all aspects of industrial image processing.
The company regularly offers virtual live events where speakers pass on their expertise to participants free of charge. Anyone who has missed a session can simply access and watch the recording at a later date in the media library.