Next-generation energy powered by artificial intelligence
Artificial intelligence (AI) is revolutionising the energy market, with impacts being felt all along the value chain – from generation, to distribution and ultimately consumption. This encompasses a wide field of applications, including smart grid optimisation, load and flexibility forecasting, demand response management and predictive maintenance, alongside various storage and trading functions. However, according to Spiros Mouzakitis from the National Technical University of Athens and coordinator of the EU-funded I-NERGY project: “This explosive demand is not being met with enough robust, mature, industry-driven AI products. Those on the market are usually expensive to integrate and introduce significant cybersecurity and ethical risks.” I-NERGY’s ambition was to contribute to European efforts to develop an AI-on-Demand platform (AIoD), by creating novel assets and services that electrical power and energy systems (EPES) stakeholders could cost-efficiently adopt and extend, improving processes and competitiveness.
Suite of AI-enabled analytical energy services
The I-NERGY team designed the system’s architecture based on user requirements and analysis of available state-of-the-art technical solutions. Components such as data services, machine learning algorithms and models were then developed, alongside training/evaluation tools and finally AIoD functionalities. The resultant prototype was tested across 9 pilots with 15 use cases, covering the entire energy value chain with its varying objectives and requirements, from utility systems operators, to investors and policymakers. The cases straddled three domains of AI use: optimised electricity and district heating networks; renewable energy in buildings, districts and communities; synergies with other energy and non-energy domains. Specifically, the use cases included AI for: predictive maintenance of network assets; operational planning through network load forecasting; consumption and flexibility prediction for a local community and electric vehicle (EV) charging stations; and de-risking of energy efficiency investments and prediction of climate change impacts regionally. According to Mouzakitis all pilots returned promising results. For example, the AI fault analysis and predictive maintenance for circuit breakers, accurately classified incidents about 96 % of the time, above the target of 80 %. Meanwhile, the AI-driven management of network loads and demand forecasting proved accurate even for fluctuating load profiles. “Sharing knowledge and tools with stakeholders, especially transmission system operators, has generated very positive feedback on its usefulness and impact,” explains senior researcher Nuno Pinho da Silva. In parallel, I-NERGY organised two open calls targeting SMEs set up to develop AI-driven energy services. Winners went through technology transfer programmes, with guidance from I-NERGY’s technical partners leading to 25 services and over 76 AI assets uploaded to the AIoD platform, ranging from predictive maintenance of PV panels, to optimised public lighting.
Practical benefits build trust in AI
As I-NERGY’s services can improve energy efficiency and reduce waste, they help tackle both the climate crisis and energy poverty, supporting a range of global initiatives, such as the Sustainable Development Goals, especially Goal 7 for affordable and clean energy. Closer to home, they also support the EU’s AI Act which includes provision for services to be evaluated against ethical guidelines for trustworthy AI. “While AI has recently caused concerns, our practical use cases showcase its benefits, offering more flexible energy pricing for consumers which reflects individual usage. Predicting demand for EV charging stations means that incentives can be introduced for charging at other times, when electricity is generated by renewables for example,” says Mouzakitis. Currently, I-NERGY’s services are being fine-tuned through the EU-funded DeployAI project, set up to create an AI platform for industry and the public sector, guided by the principles of trust, ethics and transparency.
Keywords
I-NERGY, energy, AI, data services, machine learning, algorithms, power, forecasting, electric vehicle, EV, charging stations, cybersecurity, technology transfer
