Artificial intelligence: Axpo Italia and the University of Genoa enter into a partnership

Axpo's Italian subsidiary has entered into a research cooperation with the University of Genoa to closely investigate potential artificial intelligence applications for energy trading. Algorithms that analyse huge data volumes could help Axpo’s energy traders in Italy improve the forecasting of the gas and power demand. Axpo Italia is supporting this academic pioneer work and financing a doctoral and research stipend for the innovative project.

In recent years, the University of Genoa’s computer science department, the Dipartimento di Informatica e Tecnologia, Bioingegneria, Robotica e Ingegneria dei Sistemi, or DIBRIS, has made a name for itself in the area of research on the trend topic of artificial intelligence and machine learning. DIBRIS's outstanding expertise will now benefit Axpo Italia, which is one of the leading energy traders in Italy.

A laboratory for big data analysis

The parties have agreed to establish a state-of-the-art laboratory at the University of Genoa in order to develop new models, algorithms and digital tools that can be used to find solutions to questions in the area of energy trading. The focus is on research in the area of forecasting in dynamic systems, which includes the processing of large data volumes. The development of algorithms also requires sophisticated mathematical modelling and research work. The main areas of application will ultimately include models that support traders in making decisions for trading and portfolio management.

Simone Corbo, Head of Energy Management & Trading at Axpo Italia, explains: "The goal is to maintain and develop our competitive advantage on the Italian market by using the most advanced data analysis methods in artificial intelligence and machine learning to support our daily decisions."

From trader to techie

Together with Professor Lorenzo Rosasco's DIBRIS team, Axpo Italia will be analysing huge data volumes. The objective is to learn from past experience and to best forecast future power and gas consumption. This will support Axpo Italia's traders in their decision-making on when and how much energy to buy and pass on to their customers. To do so, a highly complex software based on logical and mathematical operations is used: "We want to be at the forefront in this area all over Italy," says Simone Corbo.

How important data analysis in energy trading has become is also indicated by the fact that the sector now puts greater value on technical know-how. Simone Corbo experienced this change himself: "When I started my professional career in energy trading, financial expertise was required. Today, we are mainly looking for computer scientists and mathematicians."

Win-win for private industry and research

Lorenzo Rosasco can confirm this development: Artificial intelligence has become established in more and more economic sectors and energy trading is taking on a pioneering role. "So I'm all the more pleased about the cooperation with Axpo Italia: The exchange between leading companies and researchers like us is essential."

Under the cooperation with DIBRIS, Axpo Italia is financing a three-year doctorate as well as a research stipend, and will also acquire project results as intellectual property. The team of Lorenzo Rosasco, who teaches at the University of Genoa and also conducts research at MIT in Boston, will act as the academic partner under the cooperation and will be responsible for the processing and modelling of the technical questions. The fundamental findings, analyses and models will remain with Axpo Italia.

That all this goes hand in hand is very important to Lorenzo Rosasco: "Here at the University of Genoa we have developed a great deal of know-how in the area of artificial intelligence in the past years and we now want to share it with the business world. At the same time, we can offer young researchers a concrete opportunity. Ultimately we all benefit."

Link