Applied Machine Learning in Hydropower

One of Europe’s largest generators of renewable energy and leading company in hydropower set out to explore the potential of digitalization for its hydropower plants. Our team of consultants developed a machine learning platform, the surrounding backend infrastructure as well as visualization tools to explore and analyse sensor data for both specific use cases as well as across the entire fleet.

Hydropower

We explore the data in close collaboration with the maintenance team on site and apply state-of-the-art machine learning algorithms to monitor various kinds of failure modes from slowly developing failures over time to repeating failures in the auxiliary equipment. This kind of condition monintoring supports the daily maintenance by improving the predictability of wear and asset failures, thus reducing operating costs and the need for costly and time-consuming manual inspections.

Consultants:

Marty Cochrane

CTO of Expert Analytics. Marty is well known in Norway and much further afield for his knowledge when it comes to Edge Analytics, predictive maintenance and energy trading. He’s spent his last ten years building automated control systems, leading Nordic software development at Statkraft, solving complex Edge analytics problems for maritime, manufacturing, construction & energy and has been an executive technical lead for a range of organisations. At weekends you’ll find Marty in different countries either competing racing superbikes or using his low level software skills to modify motorcycle engine control systems where he’s built his own analytics platform and predictive traction control systems.

Pia Zacharias

Pia is a physicist with more than ten years of experience in scientific programming and data analysis. She holds a PhD in physics, which she obtained from the University of Freiburg, Germany in 2010. She has been working as a researcher, mentor and lecturer in astrophysics at different research institutions across Europe before she joined Expert Analytics in 2018. Her areas of expertise include numerical simulations of complex physical systems, statistical data analysis and signal processing, as well machine learning applications and visualization and handling of large datasets.

Vinzenz Gregor Eck

Vinzenz is a curious, pragmatic, creative problem solver. He holds a PhD in biomechanical engineering, focusing on stochastic simulations of the blood flow in the human arterial system. This interdisciplinary work included working within engineering, medicine, biology, software development and statistics.