Predict Failures

With the assistance of an onboard DataHub, our remote analytics cloud platform, MarineInsight, creates a unique digital twin of your vessel. Using this digital twin, MarineInsight immediately starts monitoring threshold-based alerts, such as high fuel pressure and bearing temperature and records real-time analytics.

After less than a month of data collection, MarineInsight’s automated anomaly detection (AAD) system uses artificial intelligence to automatically generate machine learning models that monitor assets for malfunction. These behavior models are deployed to the onboard DataHub to identify anomalous behavior in each asset. With these models deployed, the AAD system will detect anomalies and send alerts before conventional alerting systems have even detected a problem!

Though the AAD system is already detecting anomalies, protecting assets, and predicting failures, it doesn’t stop learning. As MarineInsight collects more data, the AAD system’s artificial intelligence continues to learn and progresses in sophistication. At all times, the AAD system monitors its own performance and when necessary, generates new machine learning models to ensure continued asset protection and failure prediction. Eventually, the AAD system builds several types of advanced multivariable machine learning models and competes these models against each other, enabling the highest performing model.

Throughout the entire machine learning process, MarineInsight is doing the hard work of detecting anomalies, protecting your assets, and predicting failures. In the process of doing that hard work, MarineInsight’s artificial intelligence continues to find more sophisticated ways to predict failures, protect your vessels, and optimize your fleet.