Concept of automated load detection for de-palletizing using depth images and RFID data

In this paper, we present a novel concept for the detection of loading positions of parcels or bins on a pallet to enable automated order picking using knowledge about the packing pattern model. The approach comprises (1) a new combination of pattern model data and PMD-camera-generated point clouds and (2) a novel concept of RFID data management using a Binary data on Tag/Schema on Net and semantic coding approach. The latter enables the use of additional services like storing of loading positions on auto-id devices (RFID-tags) in a wider concept of the Internet of Things, while the former presents an alternative approach in the context of contour check and position detection of unit loads for automated de-palletizing.