A Comparative Study of Forecasting Methods in the Context of Digital Twins

Abstract

This paper describes and compares different forecasting techniques usedto build a real-world Industry 4.0 application using concepts of DigitalTwins. For this experiment, real data collected from a temperature sensorduring the initial stages of a manufacturing process is used. This raw datafrom the sensors is preprocessed using state-of-the-art time seriestechniques for gap removal, normalization, and interpolation. Theprocessed data are then used as input for the selected forecastingtechniques for training, forecasting, and tests. Finally, the rates of thedifferent techniques are compared using accuracy measures to determinethe most accurate technique to be used in the application to support itsforecasting use cases. This paper also explores different areas that canbe used as topics for future work.