Fraunhofer Institute for Manufacturing Engineering and Automation IPA

The Deep Picking project develops modular and flexible software and hardware solutions for the picking of known and unknown objects in different delivery states in production processes, which can be quickly and robustly integrated and set up.

The operation of the objects becomes more reliable and robust due to the developed technology. The previous knowledge about objects and their properties is not important.

In today's production processes, a large number of objects must be operated. For product personalization, scalable technologies are necessary to maintain the automation potential.

The use of machine learning enables the handling of unseen and unknown objects. Difficult object geometries can be handled robustly and the efficiency of the process can be increased.