Solar potential determination using drones and AI

Initial problem

Due to rising energy prices and climate change the use of renewable energy sources, such as wind, water or solar power, becomes inevitable. Farmers in particular suffer from high electricity prices. The installation of a solar or photovoltaic (PV) system offers a welcome alternative. Solar or PV systems can save electricity and heating costs in the long term and can create added value (by feeding surplus electricity into the grid) under favourable conditions.

However, there are always problems in determining suitability and planning for the installation of these systems by companies. Since solar as well as PV systems usually have to be installed on roofs, it is difficult for workers to see the available roof space. In addition, a lot of effort has to be made to get a good overview. The necessary scaffolding and cherry pickers are expensive and inflexible to reach the roofs. It is difficult to install solar panels on large and highly angled roof areas and to plan such an installation with manually taken pictures or old analogue drawings from archives. Roofs have air outlets, lightning conductors as well as chimneys and it is often difficult to remember exactly where these were located and manual measurements have limited credibility.

Manual work on roofs is dangerous and health and safety measures need to be taken to avoid accidents such as falls. A thin grid of points is needed to create the terrain model, which usually has to be captured by time-consuming, inefficient surveying of fields by surveyors. Surveyors often lack a sufficient overview of solar issues. Thus, in addition to the risks workers must expose themselves to in order to gather the necessary information for system planning, there is also the risk of costly planning errors due to incorrect data or misjudgements.

Our solution

AIRXBIG Projektentwicklungs GmbH has set itself the task of offering its self-developed drone platform, consisting of the 6 components drone, hardware, software, sensor technology, analysis and personnel, as a "one-stop shop" in the form of a Drones-as-a-Service (DaaS) for customers and enabling the use of drones in all areas. This includes using the drone platform to analyse roof surfaces for their potential to install solar and photovoltaic systems. The captured drone images are first photogrammetrically modelled using an in-house developed analysis process and then subjected to a spatial analysis to determine the solar potential of the roof surfaces of the particular building.

The analysis process is illustrated in the following 4 images: The 1st image on the left shows the orthophoto of the centre of the municipality of Hornstein. In the process, building areas are subsequently filtered (2nd image from left) and roof areas are identified (3rd image from left). The current final result can be seen in the last image on the right, the solar potential of the roof areas.

The calculated potential of the roof areas and related data such as the available roof area, orientation, slope, number of possible panels, CO2 savings and financial-related data such as estimated system and installation costs, average monthly payments and payback time of the system will be available in the AXB dashboard in the future. Based on this information, customers or residents of a community can plan the installation of a solar or PV system on their roof in cooperation with local energy suppliers or an installation company active in this field.

request project