This paper demonstrates the application of a fully automated machine learning (ML) pipeline on time series data from the domain of production engineering. The workflow aims to streamline the machine learning (ML) process and reduce manual effort by incorporating automated machine learning (AutoML) and automated featurization techniques. A comparative study is created to evaluate the proposed workflow against existing state-of-the-art methods using six datasets from production engineering, five widely used ones and a newly presented one. The study compares the performance of the automated workflow with the manual data mining process described in the literature. The results lead to two main conclusions: first, an automated featurization process cannot match the performance of manual feature generation, but still delivers good results without requiring manual work. Second, if the manually generated features are used, AutoML can outperform the manual process in solving the combined problem of algorithm selection and hyperparameter optimization, leading to better prediction results. The presented approach serves as a useful reference for other methods, showing the practicality of baseline workflow without human bias. Overall, the proposed data processing pipeline demonstrates the potential of AutoML in production engineering and provides valuable insights for practitioners in this domain.