Machine Learning

ML Heatmap

Servo indication

ML-distribution

Learning curves

Machine learning is a field of computer science where models are trained on data to predict an output given a set of input values. We use these methods to train models on historical production data in hydro power plants to predict suboptimal operation and abnormalities. If a component breaks down, we are able to detect similar pattern in the current plant, but also if the same component is used in different plants.

We have decades with in-depth experience building turbine governors and excitation systems and hydraulic systems, as well as control systems, valve control systems  and simulators. Take a look at our “Products” section. We are combining this experience with PhD-level experience in mathematics, statistics, physics and data science working with sensor data. This gives us a non-matched experience matrix.

Our condition monitoring system is designed from ground up specifically for running our machine learning pipeline for predictive maintenance.

Contact us to learn how our system can utilize machine learning to help you improve your operation.