@inproceedings{Phi2021,Author = {Philipp Borchers, Dennis Lisiecki, Patrick Eschemann, Linda Feeken, Mehrnoush Hajnorouzi, Ingo Stierand},Title = {Comparison of Production Dynamics Prediction Methods to Increase Context Awareness for Industrial Transport Systems},Year = {2021},Pages = {49-55},Month = {10},Publisher = {Reproduct NV, Ghent, Belgium},Booktitle = {Modelling and Simulation 2021},Organization = {EUROSIS},type = {inproceedings},Abstract = {The performance of factory-internal logistic systems plays acentral role for the overall productivity of the factory of thefuture. A key element is the optimization of logistic systemsbased on predictive analytics of transport tasks in order toanticipate and to adapt to changes of the production flowsin the factory. Although this information might be derivablefrom production plans and machine configurations, the datais often not available for transport systems like automatedguided vehicles provided by third parties due to data protectionconcerns. Here, prediction needs to be based on observableinformation. In this paper, we investigate methods for theprediction of future transport tasks between machines based onrecorded task histories. For this, we modify several predictionmethods from the literature for making them applicable topredict timestamps and destinations of future tasks. The resultsare compared with evaluation data taken from a factorysimulation via multiple metrics, which are revealing suitablepredictors based on achieved prediction accuracy.}}@COMMENT{Bibtex file generated on }