Artificial intelligence for a successful energy transition
Data-driven risk assessment offered by variante.energy aims at paving the way to a hundred-percent self-sufficiency with green electricity
The Adlershof-based founding couple behind variate.energy provide decision-making support for companies, communities, and other organisations that seek to decrease their carbon emissions through wind and solar projects. To do so, they use algorithms to create historical and forecasted weather data for the planned location and combine them with suitable technology based on the customer’s energy needs. This data-driven risk assessment aims at paving the way to a hundred-percent self-sufficiency with green electricity.
Charlotte Huang and Joachim Reinhardt are a finely attuned team. When explaining the plans and goals of their start-up project, they complement each other mid-sentence, their thoughts flowing seamlessly. Huang will finish Reinhardt’s sentences, and vice versa. It feels completely intuitive, without any pushing and shoving. And so, it is fitting that their officially registered company name is ReinhardtHuang GmbH.
However, the name has a different reason altogether: “We had already jointly started a data science consultancy before this project,” Reinhardt explains. Huang looks back even further. Their paths had already crossed as prospective economists at Free University Berlin, where they first encountered statistical methods and data science applications in finance and insurance. Fascinated by the benefits of these methods, they set out to apply them to other areas. Their focus became the energy sector. Huang headed to Fraunhofer Institute for Systems and Innovation Research ISI in Karlsruhe to write her master’s thesis. Reinhardt did the same at the Fraunhofer Institute for Solar Energy Systems ISE. The focus of their: optimising energy projects using modern statistical methods.
Pushed forward at the Adlershof Founder’s Lab, this is where the variate.energy project comes full circle. “We noticed that when planning wind and solar projects, comparatively little consideration is given to the variability of the underlying data,” explains Reinhardt. This is different in finance and insurance. “When they do risk assessments, they include historical data as well as possible future events if they are within the realm of statistical possibility,” Huang adds. The founding couple now wants to establish such model-based scenarios in the energy sector in order to base investment decisions on a more real-world foundation. Until now, high-resolution weather data have always been the most important basis. On the one hand, however, they don’t reach back very far. On the other hand, they don’t take the probable future climate and thus weather changes into account. The energy transition’s statistical base falls short. “Even the Federal Audit Office criticised this recently,” says Reinhardt.
The duo plans two things to change this: On the one hand, they use algorithms to derive more specific, albeit synthetic, data on solar radiation and wind conditions from general historical weather data. Using these synthetic time series, they can expand the data base for planned project locations. Secondly, the founders are committed to identifying the most promising solar and wind technology to match every location and to finely attune the timeframes for energy production with the customer’s energy needs to the highest degree possible. If this proves difficult, they instead identify the storage requirements needed to achieve full self-sufficiency using self-produced, climate-neutral electricity.
“We want to provide decision-making support to companies, communities, and other organisations planning wind and solar projects to reduce their carbon footprint and to help provide them with more realistic risk assessments,” says Huang. Using AI algorithms to optimise matching historical and synthetic weather data with their customers’ energy needs as well as adequate solar, wind, and storage technology, their long-term aim is to get more and more companies to get into self-produced renewable energies and have them gradually make the switch. The closer the preliminary analyses of variate.energy match the actual yields, and the more directly energy production matches energy needs, and the more precise their AI-based cost-benefit analysis becomes, the faster companies and communities will go forward with their very own energy transition.
As a start-up on the first mile, variate.energy is interested in pilot projects—be it with players from the energy sector or companies mulling over wind and solar projects. “We consciously decided for Adlershof because of its entrepreneurial spirit and local networks,” says Huang, “and because the energy transition needs to become a successful model, especially in Berlin,” says Reinhardt, finishing her sentence.
Peter Trechow for POTENZIAL – The WISTA Magazine
- Adlershof Founder’s Lab: Teams 2021