Anomaly detection for an international manufacturing Corporation

Portfolio Description


Client: Confidential

Industry: Manufacturing

Year: 2023

Artificial Intelligence can be used to identify anomalies and machinery downtime

Case study
Identify anomalies starting from machinery production data

In the fast-paced world of modern manufacturing, optimizing production processes and minimizing downtime are a must have. One of the game-changing technologies that's making this possible is anomaly detection.

By applying anomaly detection techniques to production machines and sensors, we developed a new tool to support the Company employees in their day-to-day operations, allowing the engineers to spend more time fixing the anomalies rather than finding them, generating value for the Company and its customers.

In order to properly support the team in identifying & fixing the anomaly, the users have access to the “outlier data” that the algorithms identified as an anomaly.

The support team can then give feedback as “correct” or “not correct” to improve the Anomaly Detection models, creating an unique collaboration between the Tram and the A.I. algorithms.