The Open Logistics Foundation‘s Open Logistics Repository is a library of open source software developed by the Foundation‘s community. It also contains code that has been donated to the Foundation by research organisations, for example. “This software usually has a specific focus and fulfils a clear task. Such small but effective tools have not yet led to communities in logistics – but they are regularly downloaded from the Repository and used by companies and their developers,” says Jens Leveling, Technology Advisor at the Open Logistics Foundation. This is the case, for example, with the ML Toolbox software kit, developed as part of the large-scale “Silicon Economy” research project at the Fraunhofer Institute for Material Flow and Logistics IML. The ML Toolbox is a toolbox for AI-based image processing. With its various algorithms for recognising and locating vehicles or load carriers, for example, companies can train AI models independently. This usually requires highly qualified specialists.
However, the ML Toolbox includes a “Guided Training Service” that guides companies safely through the entire machine learning process. The trained AI model is then immediately ready for use. Mannheim-based Contargo, one of the world‘s leading container hinterland logistics networks
and a member of the Open Logistics Foundation, has used the ML Toolbox in a research project to automate remote crane control centres using artificial intelligence. The automation and remote control of cranes is a promising approach to optimising handling, avoiding bottlenecks and alleviating the shortage of skilled labour. The long-term goal is to develop fully autonomous cranes that can also operate at night. The safety requirements for crane operation are high, as the cranes are operated in open areas without fencing. The ML Toolbox was successfully used in the project to train an AI model to support gantry crane control systems at inland ports from different manufacturers so that obstacles – be they people, trucks or load carriers – can be reliably detected at the terminals.
Flexible fleet management
The open source software libVDA5050++ of the Fraunhofer IML is also a quick win from the Open Logistics Repository. The software fully maps the VDA5050 standard, which aims to standardise communication between driverless transport vehicles and the control system. “Providers can build on this instead of blocking already scarce resources for further in-house implementation,“ explains Nathalie Böhning, Innovation and Project Manager at the Open Logistics Foundation.
