Abstract
This study explores the potential of synthetic data generated through clustering and Generative AI (GenAI) techniques to enhance Predictive Maintenance (PdM) services. Traditional maintenance practices incur high costs due to unplanned downtime, emphasizing the need for PdM, which proactively uses sensory data to predict failures. However, PdM's effectiveness relies on sufficient anomaly data, which is often lacking in specific sectors. This study addresses data scarcity by generating synthetic anomalies and creating realistic, privacy-preserving datasets for PdM. The findings highlight synthetic data’s value in optimizing PdM models and fostering collaborative service ecosystems, where shared synthetic datasets drive innovation, cost-efficiency, and data security across industries.
Recommended Citation
Mihale-Wilson, Cristina; Cordes, Anjana; and Lowin, Maximilian, "Synthetic Data Generation for Predictive Maintenance Services" (2025). SIG SVC Pre-ICIS Workshop 2024. 5.
https://aisel.aisnet.org/sprouts_proceedings_sigsvc_2024/5