Using Machine Learning for Anomaly Detection on the Shop Floor

Presentation by Martin Liebig, Sr. Director Data Solutions at Altair as part of the 2025 ATCx AI for Engineers conference.

Modern manufacturing environments generate vast volumes of data from machines, sensors, and production systems. Amid this complexity, detecting anomalies—such as equipment malfunctions, process deviations, or quality issues—in real time is critical to maintaining efficiency, reducing downtime, and ensuring product integrity. In this session we explore how advances in machine learning are enhancing anomaly detection on the shop floor by identifying subtle patterns and deviations that traditional rule-based systems often miss. Through real-world examples, we’ll demonstrate how manufacturers are using machine learning to move from reactive maintenance to predictive operations, enabling smarter decisions and continuous improvement.