Cookie Consent
By clicking “Accept”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Privacy Policy for more information.
PS AI Labs and Devron created a predictive model that trained on the SCADA sensor data and achieved a 500% boost in predictive performance above the baseline. Thus enabling the client to identify issues more accurately and faster.
Download Case Study
The Sira-Kvina power company is one of Norway’s primary renewable energy power providers, producing approximately 5% of the country’s power from seven hydropower plants. Sira-Kvina struggled to tie their data together from disparate and heterogeneous operational systems and thus could not make informed decisions. On average, one of the hydro generators going down creates a loss of approximately $1 million per day, as well as negative customer sentiment and trust. Sira-Kvina needed a trusted data and analytics partner and engaged PS AI Labs & Devron to design and deploy predictive models to determine hydro generator component failure risk to enable preemptive maintenance operations and system monitoring.
PS AI Labs' dedicated data science and engineering team worked hand-in-hand with Sira-Kvina leaders, operators, and stakeholders to design and build machine learning models using various operational and maintenance data sources. Using the Devron federated machine learning platform our team developed an abstract model of the generator systems that could be linked to key operational and sensor data systems. Our focused data discovery and exploration identified that the generator had moved into a damaged state eight power cycles earlier than had previously been observed by the Sira-Kvina maintenance team.
PS AI Labs & Devron created a predictive model that was trained on the SCADA sensor data and achieved a 500% boost in predictive performance above the baseline. Thus, our models helped Sira-Kvina operators identify issues more accurately (fewer false positives and false negatives) and much earlier, allowing operators to strategically approach each maintenance issue in a cost-optimal fashion. Sensor metrics were reduced from hundreds to ten, which provided the best predictive performance, making the results easier to interpret and more actionable for the business.
