Citing an anticipated surge in battery-powered vehicle charging on the grid in coming years, information technology giant Oracle Corp. is pitching artificial intelligence and big data analytics to help electric utilities manage that charging in a way that avoids unneeded investments in infrastructure and generation.
Utilities can also tap the results of Oracle's deep machine learning to engage customers who drive electric vehicles to better understand the impacts their refueling habits have on grid operations and electric bills and to make appropriate adjustments, according to the Silicon Valley-headquartered company.
"Better to get out in front of this now rather than get run over by it later," Dan Byrnes, senior vice president of product development at Oracle, said in an interview. "It's a really exciting and interesting challenge," he said, noting that Oracle uses artificial intelligence to collect, process and store "petabytes and petabytes of data" all sourced from smart meters.
As part of its $532 million acquisition of Opower Inc. in 2016, a cloud-based software provider to electric utilities, Oracle gained access to billions of data points on the household energy use of 60 million customers across 100 utilities. Pulling from that data haul and its multibillion-dollar annual research and development budget, the company has expanded its capabilities to detect when households charge EVs. The information, once confirmed by utility customers, can inform power companies about how to better plan their generation and network needs as they integrate increasing numbers of electric vehicles and can help utility customers save money by charging at off-peak times, according to Byrnes.
Oracle's utility-focused data analytics team has pilot projects underway with "a number of utilities" to test the EV detection technology, the company said in a May 29 news release. Byrnes declined to name the utilities but said they are "a handful" of companies based in North America. The pilots will provide additional data that Oracle plans to feed into its machine-learning platform.
"This is a huge opportunity for utilities," Byrnes said.
BloombergNEF expects global annual electricity consumption from millions of new EVs to jump nearly ninefold to 634 TWh by 2030, rising to 2,200 TWh by 2040, led by demand for EVs in China, the U.S. and Europe.
The Smart Electric Power Alliance, a nonprofit research and advisory group of which Oracle and a host of utilities are members, in a recent report called for "swift action" to eliminate political, technological and business barriers to widespread managed charging. Without smart charging, "we many look back in a decade and wonder what went wrong," the report's author said.