Power companies use AI for cleaner, reliable electric grid
Major electric power companies are teaming up with technology firms to use artificial intelligence (AI) to improve the electric grid. This new collaboration aims to meet rising energy demand while reducing greenhouse gas emissions. The announcement was made during NVIDIA's Global Technology Conference in San Jose, California. The partnership, involving key U.S. power suppliers and tech giants like Microsoft and Oracle, seeks to develop AI models specific to the challenges faced by the power sector. Jeremy Renshaw from the Electric Power Research Institute (EPRI) mentioned that AI can enhance the safety, reliability, and affordability of energy systems. EPRI, a nonprofit energy research group, is bringing together experts from both the energy and tech fields. AI models, previously created with EPRI's data, will be improved using the vast datasets collected by power companies over several decades. As energy demand grows rapidly, especially after a stable period, power companies face significant challenges. Severe weather and wildfires, aggravated by climate change, put further strain on the electric grid. Companies are also working on integrating renewable energy sources like wind and solar power, which can be unpredictable. AI could help optimize energy generation based on demand forecasts, potentially replacing dirtier energy sources. It could also streamline the permitting process for new power sources and transmission lines. With the rise of data centers for AI operations, electricity demand is expected to soar. Interestingly, using energy-intensive AI to solve power challenges poses a paradox. NVIDIA is focused on improving the efficiency of the chips used in AI computing to help reduce energy consumption. EPRI and its partners aim to drive the transformation of the power sector through innovative AI applications. Meanwhile, the upcoming AI Impact Awards will spotlight notable AI solutions across various sectors, recognizing efforts to tackle critical issues. Entries are open until April 25, with winners to be announced in late May.