Increased transparency of material and product flows in warehouses will necessitate real-time determination of the properties of logistical units (e.g. small bins, parcels, palletized goods, intermodal containers). The key properties of logistical units are their unique identifiers and locations at a given time. Other properties such as dimensions, weight or temperature may also be relevant, depending on the use case. This paper presents a pallet monitoring system that determines the characteristics of pallets, namely storage location, storage time, dimensions and appearance. Technically, this is done by combining the MarLO vehicle positioning system that employs passive planar markers, an RFID identification system, a dimensioning system that employs depth sensors and a load change detection system mounted on vehicles. The proposed approach was developed and evaluated in a real world test bed. This enabled us to transfer the subsystems' accuracy to our new pallet monitoring system, i.e. we achieved a pallet positioning accuracy of up to 10 cm, a pallet dimensioning accuracy of up to 5 cm in each dimension and highly accurate pallet identification. By fusing the data from these subsystems, we were able to generate the aforementioned pallet information for subsequent monitoring and control of warehouse operations in real-time

No abstract

No abstract

In this paper we present an approach for optical vehicle positioning (e.g. fork lifters). Our approach is motivated by planar marker detection systems like ARTag or ARToolkit, in which poses of planarmarkers relative to the camera can be determined. In contrast to existing optical positioning systems(e.g. SkyTrax), we mount cameras on the ceiling and passive (non-electronic) planar markers on the top of the vehicles. The absence of complexelectronic components on the high stressed vehicles enables rapid process integration, which is particularly important for rental vehicles. We have evaluated our method keeping the most important user requirements coverage, costs and accuracy under consideration for three intra-logistic scenarios a) zone monitoring with zone precise positioning, b) storage aisle monitoring with storage place precise positioning and, c) complete driving range monitoring with maximum precision

The recognition of logistics objects is an essential prerequisite for the optimization of operational logistics processes and can be performed among others via image-based methods. However, the lack of available data for training domain-specific recognition models remains a practical problem. For this reason, we present an approach to solving this problem. The core principle of our approach is the automated generation of image data from 3D models, in which the appearance of the objects varies through variations of different parameters. The first results are promising: Without any real image data, we have created a neural network for recognition of real objects with a recall quality of 86%.