Artificial intelligence helps solar shine

December 2, 2022 Wolf Ruzicka

Far from dystopian science fiction, artificial intelligence (AI) now serves as a pervasive enabler across industries and markets. Although we may not think about it, AI and machine learning algorithms are already an integral part of common day-to-day activities, but smart technology is also very much at work on a larger scale.

The potential of AI in the green energy sector is rapidly growing. As another year of extreme weather hits the world, the urgency to stem the tide runs deep. While there is no one-size-fits-all solution to the problem, maximizing clean energy generation and storage is one of the things that can hopefully help us protect our planet.

RRenewable energy sources such as solar power are complex and unreliable due to ever-changing weather conditions, but AI can be a positive force in this particular industry, helping to break down barriers and liberate the true power of solar energy.

Solar forecast

In 2018, the Australian Energy Market Operator (AEMO), the central entity responsible for forecasting the country’s electricity production, has begun to allow national solar farms to “self-forecast”. A number of large-scale solar sites have chosen to partner with Solcast, an Australian solar forecasting provider that offers a model with an artificial neural network (ANN) – a complex, trained, multi-layered computing system with an architecture directly inspired by the biological brain – to predict the hourly solar energy production of a given photovoltaic (PV). The sophisticated Solcast satellite-based “Nowcasting” system detects and predicts weather and sky features through geocodes, quality checks, solar irradiance estimates, real-time solar energy data streams and sky-to-ground images of local cloud cover. The system produces over 600 million calculated forecasts per hour that are shared within seconds with customers. This forward-looking approach pays off for customers a 10-15% improvement over previously used AEMO predictive models.

Asset Maintenance

Another area where AI is proving very useful is asset maintenance, which includes inspection and cleaning. Traditionally, maintaining a solar farm would require a team of workers to manually inspect the panels. The process can easily become tedious and expensive and cannot be always accurate or performed daily for full deployment. Operators have turned to AI to speed up the process and improve accuracy, reducing quality control costs for entire facilities. Unmanned aerial vehicles (UAVs) – or drones – are the most effective and common tools used by AI in solar to collect data and provide aerial imagery. The captured images are then subjected to a deep learning algorithm trained to identify and classify visible signs of defects from a large amount of previously processed labeled image datasets.

 Soil Mitigation Solutions

Cloudy weather isn’t the only natural cause that can hamper solar power generation. “Soiling,” which occurs when dust, bird droppings, pollen, and other particles accumulate and harden on photovoltaic panels, hinders sunlight penetration and can even sometimes damage components . A recent study estimated that the global cost of fouling could reach over $6 billion by 2023. Typically, solar farm operators would turn to fouling stations to solve this problem using a method of comparative measurement of energy losses. However, these stations can incur unnecessary cleaning costs due to inaccurately measured soil levels. Today, AI-powered platforms can use the vast streams of data from solar panels to determine when and how panels should be cleaned, helping to increase efficiency and reduce costs.

Thanks to AI, the creation and use of sustainable energy has never been in a better position to proliferate. Importantly, these examples illustrate how AI can increase the reliability of solar power and enable more farms to operate successfully to help mitigate fossil fuel generation and climate change.

Wolf Ruzicka is President of EastBanc Technologies, a provider of software solutions that enable customers to win in today’s digital economy. Wolf is a technology industry veteran with over 25 years of experience leading business strategy and innovation. He joined EastBanc Technologies in 2007, initially as CEO. During his tenure, Wolf also served as president of APIphany, a division of EastBanc Technologies, through its acquisition by Microsoft. Follow Wolf on LinkedIn.