AI makes rooftop solar panels extra efficient

• Rooftop solar panels come to be a lot more energy efficient if placed precisely

Mohammad Aslani, scientist in geographical information science at University of Gävle, has actually developed AI methods that in numerous steps improve the placement of rooftop photovoltaic panels.

Figuring out just how much electricity rooftop solar panels can create is difficult. For specific house owners, photovoltaic panels can be a great financial investment for their house. The obstacle is to decide on the number of panels and also exactly how to position them to maximize the amount of electricity that can be created on the roofing.

" Way too many photovoltaic panels on the roof generates a surplus of electrical energy, however it does not offer the owner a lot of cash when it is sold on the marked. Too few photovoltaic panels, nevertheless, does not generate sufficient electrical power in connection with the total cost of financial investment. Solar panels are costly, so we need to find a balance to make the financial investment as profitable as possible for the consumer," Mohammad Aslani says

Within the framework of Mohammad Aslani's doctoral thesis, Aslani and Stefan Seipel, teacher in computer science, have actually developed AI methods which can be made use of in a process consisting of several steps to assess the potential of rooftop solar panels. These methods can be beneficial on a micro degree for specific house owners, along with on macro degrees for urban planners as well as realty business.

"This research is useful as a basis for decision making. If an urban planner wishes to assess the potential in solar energy for a whole neighborhood, it is vital that the assessment is as accurate as possible, both for financial and also environmental factors," Mohammad Aslani says.

By utilizing publicly offered cadastral maps, aerial photos as well as point target data, the AI engine can, in a primary step, identify all the roof coverings in an image of a neighborhood. In the next step, it assesses the slope and orientation of the roof coverings. In a third step, it thinks about the structure of each roof, consisting of chimneys and also angles. Finally, it recommends the number of photovoltaic panels should be installed on each roof covering, and also just how they need to be positioned to be as efficient as possible. In total, 6 various AI engines are used: 2 for every of the three steps of the process laid out above.

In two study, one in Gothenburg and one in Uppsala, the new AI methods showed to be a lot more accurate than previous ones.

"Existing methods have a tendency to be too confident; they overstate the energy efficiency since they are not as described as ours. They do not take spatial details that affect or obstruct the positioning of the photovoltaic panels right into account as high as our methods do," Stefan Seipel says.

The following step in Mohammad Aslanis' study is to develop the AI to make it weigh in the energy demand of a building when calculating the variety of photovoltaic panels.

"I would also like to turn our thinking around and instead study on exactly how to construct buildings as well as roofings to obtain the most out of photovoltaic panels," Mohammad Aslani concludes.

The research is released as part of the Proceedings of the 8th International Conference on Geographical Information Systems Theory, Applications and also Management.