GenAI for energy transition: paving a promising pathway for tomorrow's revolution?

Clément Le Roy Partner

Thanks to Artificial Intelligence (AI), we can produce, manage, and consume energy more efficiently. However, this promise can only be realized if energy players focus on decisions that prioritize security, robustness, and long-term sustainability.

Consider it a mini revolution, foreshadowing many others. In summer 2023, Enedis (French Distribution System Operator), garnered recognition for its AI solution that predicts outages across its grid. Its CartoLine BT tool received an award at the 2023 ceremony organized by the ISGAN, an organization emanating from the international energy agency and working to promote innovative solutions for the benefit of smart grids.

Enedis’ AI-powered solution analyzes data collected by Linky smart meters to anticipate incidents and provide intervention recommendations on the low-voltage network before they affect customers. Over 95% of the detected anomalies were confirmed on-site.

Tackling pathbreaking complexities

This success story represents one of AI’s areas of expertise where it can help energy companies gain perspective on their infrastructure and assets. Admittedly, energy operators did not wait for AI to appear before conducting a granular analysis of production and consumption. They have spent years identifying customer needs to improve their service. Nevertheless, today’s market faces.

The ramp-up in renewables is reshaping the energy mix, with the latter now harder to control since it no longer exclusively includes large-scale production units. Nowadays, multiple factors are at stake, with everything interconnected. For example, a small cloud in the Netherlands or a voltage drop in the Balkans can impact the Old Continent’s entire energy system.

Staying ahead of the trends

The energy consumption market is also increasingly fragmented. Though the competition is growing, it has been somewhat curbed by energy prices of late. Consumer preferences are evolving, driven by increasing environmental awareness, the rise of electric mobility, and heightened vigilance in response to rising prices.

To navigate this new landscape, energy players must adapt their production and consumption methods. AI has a key role to play in facilitating their grasp of such trends and staying ahead of them to deliver more effective management.

For instance, AI is vital to energy production. Already, the industry leverages supercomputers to assess weather data, long-established demand statistics, and energy prices. However, AI’s real-time capabilities in analysis and prediction will pave the way for new supply-demand balance models, resulting in visionary and informed trade-offs as well as more effective resource allocation.

Learning from other regions

The same scenario applies to demand. Granted, current prediction models are robust. But they also lack flexibility with a need to consider new consumer trends and uncertain climate conditions. With its agile methodology and vast streams of data, AI exposes the shortcomings of conventional algorithms.

A case in point is heat waves which are now abnormally hot and locally concentrated, with unpredictable impacts. Energy companies are currently collecting data from regions as far away as California and Australia, which paints a picture of France’s potential future climate. This data is then given to the AI which generates more accurate prediction models.

Making the right trade-offs

Another important segment centers on creating innovative energy ecosystems. Today’s energy universe combines traditional production facilities and consumers-turned-producers (“prosumers”), with their solar panel roofs, personal storage units, delayed charging systems and the list goes on! As the ecosystem becomes more complex, its linear structure is replaced by a free-form web.

This means new networks are an absolute must, which involve extensive calculations and modeling to consider technical specifications along with social and economic impacts. For example, if a is developed in a rural area, it will require the redesign of an energy network which was previously built for low energy consumption and minimal production. AI serves to make the right trade-offs, including in times of emergency, with recommendations such as how to restore a network wiped out by a storm.

Personalizing eco-responsibility

Despite being energy-intensive, AI remains pivotal to adopting eco-friendly practices. Not only does it monitor energy consumption in buildings and manufacturing facilities, but it also finds ways to make savings and adopt best practices for maximum energy efficiency.

AI does the same for private individuals, personalizing their eco-responsibility based on real-time monitoring of their consumption habits. It’s also a tool used for “smart charging.” The latter refers to the process of charging electric vehicles when it is greenest and cheapest, factoring in driver needs, energy prices and grid charging capacity while protecting battery life.

Lastly, R&D labs are deploying AI to assist them in overcoming challenges involving insulation, batteries, and solar panels which continue to overwhelm mainstream technologies.

Forging ahead cautiously, in an essential industry

For energy companies, AI is now a top priority. In this respect, France should take pride in its innovation ecosystem. Across the board, stakeholders focus on tests before implementation to scale, as exemplified by Enedis whose AI technology now operates the High Tension “A” (HTA) network of medium-voltage overhead lines.

Forming one of the cornerstones of the economy, the energy industry will forge ahead, albeit cautiously. Progress depends entirely on the fulfillment of specific conditions, namely foolproof cybersecurity, full transparency on personal data regulations, and trustworthy partners who seek to safeguard sovereignty.