Soon, artificial intelligence and big data analytics will help Japan’s largest utility provider carry out maintenance of their infrastructure, reveals Innovation General Manager at Tokyo Electric Power Company Holdings Inc. (TEPCO)
“We can use big data to come up with a more effective maintenance, like predictive maintenance,” says Hirokazu Yamaguchi, Executive General Manager of Global Innovation and Investments of TEPCO. It plans to start a proof-of-concept for predictive maintenance next month.
To put things into context, the events of the past few years have spurred this focus on innovation. TEPCO nearly went bankrupt following the meltdown of its Fukushima Daiichi Nuclear Power Station in 2011, Yamaguchi says on the sidelines of the Asian Utility Week conference. The utility must now regain customers’ trust, while keeping up with new technologies.
Predictive maintenance
In 2012, TEPCO received a government bailout of ¥1 trillion (~US$8.95 billion) and is now semi-nationalised. In the years since, as TEPCO works to decommission the Fukushima nuclear reactors and clean up debris, “we have been undergoing a lot of reform,” Yamaguchi says.
Yamaguchi’s team is looking for ways to “disrupt the utility business model”, which includes focusing on distributed energy resources, and software and IT for managing these resources. “We need to disrupt ourselves first before someone disrupts us,” he says.
The predictive maintenance pilot will help the company preempt any failure to its infrastructure and avoid downtime for customers. “We can use all sorts of data, including weather data, sensor data, and temperature to analyse risk of breakdowns,” Yamaguchi explains. TEPCO will provide the company carrying out the pilot with data regarding transmission, distribution and maintenance. “It’s going to be in certain areas. We’re trying to find out the risk of vulnerability of the transmission, substation data,” he says.
