Pallet detection and picking

Motivation and scope

Warehouse logistics is a common and growing application for mobile robots. A central task for self-driving warehouse trucks is to detect and pick pallets (which are then transported around the warehouse). This project consists of implementing and testing one or several pallet detection methods, which should work under varying conditions, independent of whether the pallet is empty or loaded with objects, or if it is on the floor or on a shelf.

Specific tasks

The first step will be to verify an existing method for detecting empty pallets on the floor, and its integration with physically picking up the pallet with a robot. The next step will be to use colour and depth camera data to detect pallets in other configurations.

Necessary skills

Good programming skills, preferrably in C++.

Benefits

Develop experience with 3D perception, mobile robots, and the Robot Operating System (ROS)

Contact

Martin Magnusson, martin.magnusson@oru.se