CIGRE National Committee of Slovenia organized its first workshop entitled “Artificial Intelligence: Are you ready for challenges in Power Systems?”. The one-day event took place in Ljubljana. Its purpose was to bring data analytics, machine learning and artificial intelligence (AI) closer to the power systems experts through examples of practical implementations in Slovenian utilities and power industry.
The workshop was divided into three sections with a round table discussion at the end of the day. The event was opened by CIGRE NC Slovenian President Mr. Marko Hrast; the session moderator was Tadej Šinkovec (Elektro Ljubljana).
The first section had three presentations from academia and research. Professor Zoran Bosnić (University of Ljubljana) presented the broad concept of today’s AI and the importance of having access to the right people in the industry. Leon Maruša (Milan Vidmar Power Research Institute) continued the section presenting the importance of proper data visualizations for finding patterns in energy data. Lastly, Miha Grabner (Milan Vidmar Power Research Institute) talked about usage of machine learning and profound studies in the power energy sector.
The second section hosted six presenters from power utilities. Firstly three DSOs presented their case studies with advanced solutions. Miran Rošer (Elektro Celje) presented their new data driven operations and showed obvious comparison between power systems of the past, present, and the new digitalized power systems of the future. Maja Savinek (Elektro Ljubljana) demonstrated how measurements from smart meters can be used for prescriptive grid management using customized data visualization platforms and methods of machine learning. Gregor Skrt (Elektro Primorska) spoke about the evolution of data models for implementing AI methods in everyday DSO operations. The section continued with presentation from a hydro generation company where Dalibor Kranjčič (Dravske elektrarne Maribor) presented usage of deep neural networks for turbine operation simulations. There was a presentation from energy retailer with Lev Prislan (GEN-I), who demonstrated two very interesting case studies using machine learning for classifying customer’s e-mails, as well as using satellite images of hydro dams to determine water levels which drives the prices on energy markets. This section concluded with a presentation by TSO – Uroš Kerin (ELES), who spoke about the new Diagnostics-Analytics Center which uses collected data for optimizing their day-to-day technical and business procedures.
The day’s third section had four presenters from the manufacturing and software industries for the power companies. Primož Sevčnikar (Troia) presented their solution for AI usage in advanced analytics for asset management. Maj Osolin (Landis+Gyr) demonstrated how a DSO can use smart meter data and optimize their low voltage networks using machine learning. There was an interesting talk by Peter Ceferin (SmartCom) on the topic of detecting cyber-threats in utility’s IT/OT systems using AI algorithms. The last presentation was from Marko Škufca and Boris Vezenšek (both ADD) who talked about proper hardware and software environment for supporting big data operations.
The event concluded with round table discussions with key panelists: Zoran Bosnić, Miha Grabner, Tadej Šinkovec, Rok Dolinšek (Troia), Gorazd Ažman (ELES) and the moderator of the round table Andrej Souvent (Milan Vidmar Power Research Institute). In conclusion, we heard that AI is already making a positive impact on the utility business and the fast technological advancement will continue to evolve rapidly. There are still some issues that we need to address such as technology transparency, proper data management, adapting business processes and finding the right people who can handle both worlds: AI methods and strong domain knowledge about power systems. Nonetheless AI is here, it’s changing the power industry and it’s up to us to decide how to make the best use of it.