Year: 2025
Type: data-paper
Abstract: Data availability represents a critical bottleneck in the development of data-driven analysis tools, particularly for domain-specific applications in manufacturing. This paper introduces a comprehensive dataset of crimp force curves, captured during the production of crimp connections and commonly used for in-line quality control in industrial settings. The dataset comprises 2,439 crimp force curves, obtained from a semi-automatic crimping machine. Each curve has been annotated by both a state-of-the-art crimp force monitoring system, capable of performing binary anomaly detection, and by the authors who provided a more detailed classification into multiple quality categories. The paper introduces this novel dataset with the objective to enhance data-driven quality control systems in manufacturing. Specifically, the dataset serves two specific purposes: it provides a robust foundation for developing domain-specific machine learning models in the context of crimping processes, and it offers a benchmark resource for univariate time series analysis in data-driven applications.
Source: Scientific Data
Authors: Bernd Hofmann, Patrick Bründl, Jörg Franke
Institution: Friedrich-Alexander-Universität Erlangen-Nürnberg
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,490
Award: G968210948
OpenAlex ID: W4413638425