Anomaly Detection
Predict when critical equipment parts will go bad to prevent failures and downtime.
The Challenge
Producers must be increasingly competitive to protect their market share. Strong competition means machinery must operate at peak performance for as long as possible - and without interruption. Manufacturing plants are a complex array of components and automated systems finely tuned to work together. Critical parts are monitored for proper functioning, with sensors providing data at regular intervals. In order for companies to ensure that maintenance occurs at exactly the right time, they must know about impending issues far enough in advance in order to take action.
Our Solution
Based on readings taken from sensors while parts are functioning correctly, a model is trained to detect anomalous data, thereby predicting impending breakdowns. This data is read into a KNIME workflow which is automatically executed daily on KNIME Server. In the case of an anomaly, the model determines whether a first or second level alert should be activated.
Why KNIME Software
A KNIME workflow deployed on KNIME Server as a Guided Analytics Application (hosted in the cloud), makes vast computational resources available to deploy predictive analytics on sensor data, predicting breakdowns up to ten weeks in advance and setting off appropriate alerts to Production Managers.
More Solutions
Find out how other companies are using KNIME to solve their data challenges.
