Einsatz von maschinellem Lernen zur Optimierung von Kalkulations- und Produktionsplanungszeiten von Faltschachtelklebemaschinen

Abstract

In this master's thesis, applications from the fields of artificial intelligence and machine
learning are used to predict production times in the industrial packaging sector. The
data for this comes directly from the production operations as well as from CAD
packaging development. Using this data, an attempt is made to predict the production
run and setup times of folding carton gluing machines for future production orders. The
predictions are made with various models such as neural networks or decision trees.
Furthermore, it will be investigated whether these algorithms can provide equal or even
better results than the currently used calculation program. The results of the individual
models are compared and evaluated. It turns out that the performance of the models
depends on the perspective taken. However, it can be emphasized that the XG Boost
algorithm and the neural network deliver the best overall results.

Date of Award	2024
Original language	German (Austria)
Supervisor	Gabriel Kronberger (Supervisor)