Connected Assets: How Machine Learning Will Transform the Utilities Industry

James McClelland

Artificial intelligence (AI) and machine learning are poised to revolutionize the way utilities produce, transmit, and consume energy by powering the modern smart grid. Machine learning is an application of AI in which machines are given access to data and, based on this data, “learn” without being explicitly programmed. Machine learning may still be in the early stages of utility industry development and implementation, but its potential is enormous, according to Harvard University. Key machine learning benefits include more reliable energy, greater consumer choice and engagement, asset optimization, service restoration, outage management, and increased cybersecurity. Utilities that take steps now to modernize their infrastructure and adopt machine learning will gain a competitive advantage.

The need for the smart grid

In 2003, an overloaded transmission line in Northern Ohio sagged and hit a tree, causing the line to fail and shut down. Normally, this shutdown would have tripped an alarm at FirstEnergy Corporation, the Ohio-based utility company responsible for the line. The alarm system malfunctioned, however, and over subsequent hours, additional lines began to sag and fail. This seemingly mundane failure had a cascading and catastrophic effect, leaving 50 million people in the northeastern United States and southeastern Canada without power for two days, according to Scientific American. The massive blackout cost an estimated $6 billion and led to an official inquiry. The finding: the blackout was the result of human error and faulty equipment. Could machine learning have identified the elevated line fault risk before it ever happened and prevented the blackout?

This is one question that utility companies and machine learning experts are now trying to answer. Nearly 15 years after that 2003 blackout, the American power grid remains a vast network of more than 5,800 power plants and 2.7 million miles of power lines. The average power plant is over 30 years old and the average transformer is more than 40 years old.

In an effort to update and modernize America’s utility grid, the U.S. Department of Energy has invested $4.5 billion in smart grid infrastructure. This includes installing over 15 million smart meters to monitor energy usage and alert utilities to local blackouts. Artificial intelligence will be the “brain” for this smart grid.

Machine learning and the utility industry: Key benefits

The sheer volume of data collected by smart sensors can be overwhelming, making real-time analysis and action difficult. AI is solving this problem by becoming the “brain” of the future smart grid. Advances in deep learning algorithms are now making it possible for AI to instantly analyze real-time data. AI is able to spot patterns and anomalies in datasets, allowing utilities to make on-the-spot decisions about how to best allocate energy resources. These deep learning algorithms are revolutionizing both the demand and supply side for the energy economy in the following ways.

  1. Improve distributed generation management. The current system is not constructed to accommodate energy source diversification. For example, the rise of distributed generation is complicating supply and demand forces. When private users generate their own electricity from renewable sources such as wind and solar, utility companies can absorb the excess energy into the grid. This complicates supply and demand, however. AI can help utilities realize the next-generation grid through enhanced distributed resource management that automatically flows power through the grid to deliver more reliable energy and greater customer choice.
  1. Asset optimization. Utilities are developing algorithms based on industry intelligence that will predict the probability of failure. These algorithms take into account industry-wide early failure rates for equipment, creating a richer understanding of premature failure risks for enhanced asset maintenance, workflow, and portfolio management.
  1. Outage management. Utilities are using analytics-validating models to predict and identify outages. Machine learning and device automation allow for better resource management, reducing downtime and improving reliability. Self-healing grids can automatically detect and address vulnerabilities, reducing the likelihood of outages.
  1. Customer engagement. Utilities are mining data with the aid of machine learning to understand customer behavior and service needs. Using this data, utilities can provide faster and more intuitive interactive customer service via voice response, personalization, and service matching.

Learn how to innovate at scale by incorporating individual innovations back to the core business to drive tangible business value by reading Accelerating Digital Transformation in Utilities.

About James McClelland

James McClelland is the senior global director of SAP Utilities & Energy Industry Marketing, James has over 25 years of experience creating business strategy for the utilities industry. He is a graduate of the University of Toronto, Canada, holds a degree in Business & Commerce, and currently resides in Dallas, Texas.