The power grid, once a straightforward system, is undergoing a revolutionary transformation fueled by artificial intelligence. Recently, the US Department of Energy awarded $3 billion in grants for "smart grid" projects, marking a significant investment in AI-related initiatives.

One significant way AI is reshaping the grid is through expediting decision-making processes. The enormity of the power grid system, often hailed as the most complex machine ever built, necessitates a level of comprehension and prediction beyond human capability. Researchers, like Feng Qiu from Argonne National Laboratory, collaborate with grid operators to implement machine-learning models that optimize daily planning, reducing calculation time from nearly 10 minutes to a mere 60 seconds. These time savings, when applied to daily operations, can significantly enhance efficiency.

The application of AI extends beyond research labs to companies like Lunar Energy, where AI software is employed to optimize energy usage for individual customers. Gridshare software collects and analyzes data from tens of thousands of homes, creating personalized predictions of energy needs. This not only aids customers in saving energy and costs but also provides utility companies with behavioral patterns crucial for improving energy planning and overall grid responsiveness.

The surge in electric vehicle (EV) adoption poses a unique challenge for the grid, but AI is stepping in to address this energy demand. WeaveGrid collaborates with utility companies and automakers to collect and analyze EV charging data, identifying optimal charging times and turning EVs into valuable sources of energy storage for the grid. This innovative approach addresses the increased energy demand posed by EVs and enhances grid planning for a sustainable future.

AI's impact is also evident in critical operations, such as the inspection and management of physical infrastructure. PG&E in Northern and Central California utilizes machine learning to accelerate inspections, identifying areas requiring maintenance. Startups like Rhizome take it further, launching AI systems that predict the probability of grid failures resulting from extreme weather events. These advancements enable utility companies to make informed decisions on prioritizing projects for improved resiliency and disaster prevention.

As the energy sector embraces AI, questions arise about the possibility of fully automating the grid. Experts emphasize that significant hurdles, including security and data privacy concerns, need addressing. The stringent protocols and checks in place to prevent critical mistakes underscore the importance of maintaining human oversight. The potential for AI models to perpetuate biases also necessitates workforce training to ensure responsible and unbiased technology adoption. The journey toward a fully automated grid is ongoing, with the industry balancing innovation and the need for safety and reliability.

Source: technologyreview.com