Spare parts forecast
Forecasting spare parts so that delivery times are shorter and customers are more satisfied? The Technology Transfer Center for Data Analytics and Artificial Intelligence Günzburg (TTZ Günzburg) has dedicated itself to this task together with practice partner AL-KO. In the following, the participants explain how AI comes into play and what solutions they have developed.
Nominated for the Bavarian-Swabian
Innovation Award Transfer 2024 of the Neu-Ulm University of Applied Sciences
What is this cooperation about?
The project is dedicated to forecasting the demand for spare parts for AL-KO Geräte GmbH, a company that manufactures various gardening tools - from hedge trimmers to lawn tractors. The aim was to use AI-supported forecasting to achieve short delivery times and thus greater customer satisfaction.
TTZ Günzburg and AL-KO are pursuing this goal in two phases: In the feasibility phase, various forecasting models from time series analysis are created and tested for a number of selected spare parts. The forecast results of the various models are compared with each other and with the results of simple (non-intelligent) calculations. The model with the best results will then be extended to the entire procurement process in the implementation phase.
The project team at TTZ Günzburg
The cooperation partner
The initiative for the contract research came from the company AL-KO Geräte GmbH, which wanted to solve its problem using methods of data analytics and artificial intelligence - so they were in exactly the right place at the TTZ Günzburg.
The collaboration between AL-KO and us was very close and on an equal footing. At times, there were even weekly project meetings. AL-KO contributed the data and the company perspective. As a result, we were able to build up a deep understanding of the data and were able to set up, test and optimize various AI models on this basis.
Interview with Sarah Dreher and Raphaela Erbel
[1] How would you describe your project in three key words?
Exchange-intensive, application-oriented, exciting
[2] What is special about your project?
Many people only talk about AI. We create concrete and customized AI solutions: As part of this project, we went through the complete development cycle for an AI application and realized an executable AI model at the end.
[3] What does transfer mean to you?
Transfer in this project means that no one can solve the problem alone. Implementation would not have been possible without the continuous exchange and expertise of AL-KO. This form of cooperation requires a high level of trust. There was a constant exchange of information and discussions with those involved in the project at AL-KO. New findings were always discussed in order to make a clearly defined, joint decision on how to proceed.
[4] What do you take away from the project - professionally, but also personally?
The most important professional insight was that the most complex model does not necessarily deliver the best results. It simply depends on the specific application. It was a great experience to have an interested and enthusiastic partner in AL-KO, who was actively involved in the project.
Personally, we can say that after successfully carrying out this project at the TTZ Günzburg, we are now certain that this is the ideal place to advance our research and thus our doctoral projects.
