We all know that utilities and power grid operators face massive challenges. Aging power grids, regulatory changes, outages and power demanding loads result in grid investments, expensive operations, and increasing maintenance costs. Use of artificial intelligence (AI) can contribute to reduced costs and failure rates, and extended asset life, among others.In the US, the electric grid delivers energy to more than 144 million end-use customers. It is a vital component of the critical infrastructure and serves as an essential foundation for the American way of life.
The number one cause of power outages in the United States is severe weather. It costs billions of dollars a year in lost output, delayed production and damage to grid infrastructure. Most of the grid was constructed over a period of more than one hundred years. The aging nature of the grid makes it even more vulnerable in severe weather.
Governing laws require that 10 percent of the line be inspected each year, which, today, is a time-consuming manual job. In addition is the unforeseen inspections after storms and extreme weather. The traditional inspection methods include inspectors walking the lines and climbing towers, manually looking for faults or potential problems.
Streamlining line inspections with AI
An increasing number of utilities and power grid operators are using low flying helicopters equipped with cameras or unmanned aerial vehicles (UAVs) for power line inspections to improve the efficiency and reduce the risk of accidents. It is undoubtedly the way to streamline line inspections, and the data collected from UAVs or helicopters are essential parts of the digital information utilities need to get a full overview of their business.
But how can utilities and power grid operators better utilize the data?
The enormous volume of data makes it a challenge to release the full potential and value. Capturing the data is relatively easy – analyzing and processing of the images and data is hard and time-consuming. By taking advantage of AI and machine learning it’s possible to drastically improve the capacity to analyze powerline data, reduce costs and improve inspections and storm assessments.
EPRI (Electric Power Research Institute) is already pursuing more than 20 initiatives and projects to explore AI’s potential and limitations. One of them is evaluating and enabling AI algorithms that are trained to recognize potential faults and pinpoint problems in the power lines. In this article, you can read more about EPRI's AI initiatives.
Unique insight and the possibility to predict future problems
With software trained to analyze images for rot in wooden poles, broken insulators, strand damage, rust and more, utilities gain a unique insight into the current status of their infrastructure. This insight improves the process of problem finding and helps them build an updated Intelligent Asset Library. It also empowers field workers to make better decisions in their daily work.
The combination of many different data sources including the line of business systems, images, statistical and historical information, and weather forecasts enable utilities not only to identify faults but also to predict the future. Field data from the air and existing asset data combined with AI algorithms enable insight not only in existing but also potential future power line problems.
The massive challenges in the energy industry open for new opportunities and possibilities. With AI and technologies like image recognition, we can find exactly which areas of the power grid the utilities need to invest or do maintenance.