The picking of goods in shelving storage warehouses still involves a lot of manpower. The employee has to move to the storage location, pick the required goods and bring them to the collection point. Up to now, this type of order picking (“person to goods” principle) has hardly been automated as it is technically demanding to precisely recognize and grip different kinds of objects exactly.

The aim of the project KomRob is to develop a mobile robot which moves autonomously in a shelving storage warehouse, picks the ordered goods from the shelves and delivers them to a collection point. For this purpose, an object recognition as well as a gripping algorithm is to be developed with the aim of keeping the number of ungraspable objects to a minimum. The robot has to work as a plug and play system without requiring extensive conversions or adaptations of the warehouse. In addition, the robot has to be suitable for hybrid application scenarios, i.e. it can work cooperatively or parallel to human order pickers.