Today’s production plants would be unthinkable without robotic pick & place processes. Handling objects are picked by the robot and placed at the target position with the desired target orientation. An innovative possibility to significantly increase the workspace of a robot and to reduce process times is the extension of this process by the targeted throwing of handling objects. Possible areas of application are the sorting of products that are not damaged during the throwing process (e.g. screws or nuts) or whose damage is of secondary relevance (e.g. sorting out defective products, separating materials in recycling plants).
Our goal is to describe the robotic pick & toss process using a combination of classical analytical and machine learning models to make the innovative process industrially viable.
Contacts:
Nils Brückmann
Johannes Bolk
