MIDS Capstone Project Spring 2022

PeraML: Anomaly Detection for Preventative Maintenance

PeraML comes from the Indonesian word peramal, which means clairvoyant or fortune teller. Here at PeraML, we endeavor to endow all of our customers with the ability to tell their own futures by providing predictive power into health and safety of their operating machinery. For decades, maintenance and repair have been primarily reactive with organizations having little insight into the real-time health of their equipment. Modern demands for just-in-time delivery and disruptions in global supply chains make it more important than ever for organizations to be proactive in performing maintenance and eliminating unplanned downtime.

Our project uses autoencoders to identify anomalies in sensor data for rotating machinery and classifies these anomalies according to whether or not they are anticipated to lead to failure. This can allow businesses to schedule maintenance before failures occur and thus fulfill our mission of eliminating unplanned downtime.

More Information

Last updated:

April 9, 2022