KIss - Development of an AI-based outsourcing software for the automated generation of outsourcing strategies and dynamic strategy adaptation during unforeseen changes in order processing
The use of forklifts is part of everyday work in many German companies. These feature a high degree of flexibility combined with low investment costs. Accordingly, they are ideally suited to transport goods quickly and with little effort, which is why they will remain an important part of companies in the future. For this reason, safety and the operation forklifts are very important and the operators are regularly confronted with challenges. If the causes of severe and sometimes lethal accidents are considered, they are mostly due to driver misconduct and lack of operational organization. Although the number of accidents in general has been declining for years, this trend does not apply to accidents involving forklift trucks. If the particularly severe, lethal accidents are considered here, then these are by far attributable to the so-called "tipping accidents". In the context of the DGUV investigation, these types of accidents often result from cornering too fast and/or from improper loading of the forklift. Therefore, for a safe and efficient use of the forklift, the correct knowledge of the volume and weight of the forklift as well as the loading is required.
In addition to the safety aspects for the forklift, this information is also very important for the safety of adjacent processes, for example when loading a truck. This is essential data for the operator, as it allows him to estimate how many packages the forklift can be safely operated with on the one side and how a truck can be evenly loaded on the other. In the case of trucks, incorrect loading is also associated with avoidable disadvantages, e.g. higher one-sided wear and poorer driving characteristics, which in turn lead to higher risks in terms of driving accidents. For this reason, accurate determination of volume and weight is crucial for a risk-free and efficient work process.
However, determining these parameters is not only useful for the safety aspects, but also when considering the other processes in which the forklift is involved. Throughout the supply chain, the volume weight of the load is also of great importance. Optimal utilization of cargo space, correct invoicing of forwarding services, identification of the cargo, detection of damage and even compliance with safety standards when loading trucks or sea containers all go hand in hand with this.
The aim of the research project is to develop an automated solution for generating outsourcing strategies and dynamic strategy adaptation. AI-based software is to be developed for this purpose. The development includes both a static and an adaptive AI, which can be used to react to unforeseen changes during the automated generation of outsourcing strategies. The AI structure is designed to enable dynamic strategy adaptation. The dynamic adaptation is intended to increase the efficiency of stock removal. To do this, the system requires both a static neural network and a reactive neural network.
A simulation will be used to train and develop various AI models that are capable of developing optimized stock removal strategies. Once the simulation-based model has been developed, it is transferred to a real warehouse. Real data will be used to train the AI. The model created will then be used and tested in a real environment.
Contact Person
Korbinian Zöls, M. Sc