With more people needing electricity, AI helps stop power cuts.
Back in February 2021, Aseef Raihan woke up in San Antonio, Texas, feeling super cold. He had to use his military sleeping bag to stay warm because the power was out. It was freezing, around -19C, because of winter storm Uri. The storm messed up the electricity grid in Texas. Wind turbines froze, solar panels got covered in snow, and a nuclear reactor had to be shut down.
Over 4.5 million homes and businesses lost power for hours, even days. Aseef couldn’t heat his place or cook because everything ran on electricity. It took more than two weeks for things to get back to normal in Texas.
The storm showed how easily our electricity systems can break. We usually expect power all the time.
Many places don’t have winters as harsh as North America’s, but everywhere, we want more electricity. More electric cars and air conditioning mean more power.
Countries are switching to renewable energy, like wind and solar. But these sources aren’t consistent. No wind or sun means less power.
The UK’s Energy Secretary, Claire Coutinho, said we might have blackouts without new gas power plants as backup.
Another way to make power systems stronger is by adding big batteries to the grid.
When there’s extra electricity, batteries can charge up. Later, they release electricity when more power is needed. Texas is doing this.
“Since the storm, we built over five gigawatts of battery storage in Texas in three years,” says Dr. Michael Webber, an energy professor. It’s akin to having the capacity of four large nuclear power plants.
To be helpful, batteries must know when to charge and when to release electricity. This means predicting future electricity needs.
Gavin McCormick, who started WattTime, says the weather and electricity demand matter most. His company, based in Oakland, California, uses AI to predict electricity needs in an area. This helps batteries charge and discharge at the right times. It also helps people use electricity at cheaper rates in their homes.
For example, if you have an electric car that needs 8 hours to charge but only takes 2-3 hours, the software finds the best times to charge it. It looks for times with extra or clean energy, charging the car little by little overnight. This ensures it’s ready by morning.
AI can predict things like weather, holidays, work schedules, and even football matches. People often take breaks during halftime to make tea. A Danish company called Electricity Maps also uses AI to forecast weather patterns such as clouds, wind, temperature, and rain.
This helps them understand how much electricity wind turbines or solar panels can generate. Google uses their forecasts to know how clean the power grid will be in the next few hours. They adjust their data centers’ electricity usage accordingly.
Also Read: Gemini AI Licensing: Know Fantastic Apple, Google’s Sensational Discussion