AI Sweden is releasing a further developed version of Behovskartan. The purpose is both to improve the discussion surrounding Sweden’s future power and energy needs and to create a framework for modeling and visualizing the energy system.
"Through visualizations based on scenarios from the Swedish Energy Agency, we are giving Swedish organisations a new tool and framework," says Filip Kjellgren, Head of Adoption - Emerging Sectors at AI Sweden.
Behovskartan 2, an interactive tool for exploring scenarios for Sweden’s future electricity needs.
Access to energy and power is a matter of competitiveness and resilience for Sweden. Uncertainties about what future needs will look like are a challenge for those planning expansions and new connections at electricity grid companies. With funding from the Swedish Energy Agency, AI Sweden has therefore continued to develop Behovskartan.
The first version was launched in 2024 and was primarily a prototype showing how data could be visualized, from a national level down to individual areas in the country's municipalities.
![]()
The new version is based on data from the Swedish Energy Agency, including their long-term forecasts that extend until 2050. Furthermore, we are now giving users more options to filter the data and vary different assumptions, allowing them to create views more relevant to them.
![]()
Filip Kjellgren
Head of Adoption - Emerging Sectors at AI Sweden
He continues:
"One changed assumption that we are making possible is to investigate the impact that the expansion of data centers is expected to have in different parts of the country."
In parallel with the new version being made available at www.behovskartan.se, the underlying source code is also being released, including a framework for how data can be modeled and visualized.
![]()
The energy transition relies on many stakeholders making decisions based on the same reality. We hope this framework sets the tone for researchers, analysts, and decision-makers to choose the path of open analysis – where assumptions can be scrutinized and built upon, not just accepted as a number.
![]()
Viktor Bengtsson
Developer of Behovskartan at AI Sweden
Filip Kjellgren hopes that Behovskartan will also contribute to ultimately creating better conditions for using artificial intelligence in forecasting.
"At AI Sweden, we see poor data quality as a recurring challenge when our partners want to start using AI to perform various types of analysis. Through this tool, we make it possible to create immediate value when users start working with their data—a value that can increase over time if layers of AI analysis are also applied," says Filip Kjellgren.
Related articles