Skip to main content Skip to Footer

BLOG


April 12, 2017
Meet your new utility colleague: The drone
By: Jeff Lewis

Electric utilities have always faced a cost versus coverage challenge with respect to inspection of linear assets in the field. With an aging infrastructure, the risks associated with this challenge have been rising significantly. Inspection of transmission lines is a good example of where the costs and safety risks of operating helicopters to get the job done has resulted in only a fraction of lines being inspected each year at some utilities. 

Fortunately, advances in technology have ushered in a better, much more cost-effective option. In this new paradigm, helicopters are replaced with drones, which are not only highly economical, but also eliminate risks to crew safety. When coupled with artificial intelligence (AI), this drone-based approach to visual surveillance and data gathering revolutionizes the inspection process.

Some forward-thinking utilities are eager to take advantage of the drone-AI combination. Accenture has worked with one utility that previously could only inspect 5 percent of its transmission lines every year. We replaced their helicopters with drones and equipped them with cameras and ultrasound devices that would not only visually inspect the company’s lines for defects, but also identify specific problems, such as the corona effect that indicates energy leakage.

Drones and cameras are certainly an integral part of this new approach to inspection. But equally critical are the AI algorithms and technologies that convert imagery into actionable insights. For the utility mentioned earlier, we taught the AI algorithm to recognize patterns in the video feed that streamed from the drone cameras to the server. We also taught it to get smarter over time. With machine learning, the algorithm became more precise in its interpretation of data each time it was used. Pretty soon, the algorithm became so intelligent that it was able to capture data that an engineer manually inspecting lines might have missed. Finally, we tied AI into the utility’s work management system. That meant any AI-detected problem automatically issued a defect alert, GPS location and work order—all of which was sent to field crews’ mobile devices in real time.

This client and the handful of others now using drones and AI to manage their line inspections have been able to significantly reduce their inspection costs, while increasing the number of assets and miles of line they were able to inspect. Further, AI has enabled them to improve the precision and usability of their visual data evaluations.

Using drones and AI to transform the process of inspecting transmission lines represents an exciting opportunity for utilities looking to contain costs and improve system reliability. But line inspections are just the tip of the iceberg in terms of how utilities can apply drone-AI technologies. There is, for example, no reason that drones cannot be used to:

  • Capture Light Detection and Ranging (LiDAR) data to measure the elevation of T&D lines and identify where visual markers are needed to comply with FAA regulations.
  • Capture LiDAR data to identify trees that are in danger of encroaching on power lines.
  • Perform detailed visual inspection of boilers at generating stations.
  • Identify meter-to-transformer connections. This is a serious issue for many utilities, since the inability to understand the connections diminishes the effectiveness of transformer load management analyses and other grid analytics.
  • Assess remote storm damage. Many utilities already use predictive analytics to estimate where a storm is likely to impact the T&D network. In the future, they will use drones to gather actual damage-assessment data after the storm has passed to quickly confirm or adjust this storm recovery plan based on actual data.

Of course, there are still challenges associated with using drones and AI in the utility industry. Chief among them is the uncertainty about how government regulations might affect the viability of certain drone use cases. But the potential value of doing something far outweigh the possible risks. And our research reflects the desire for utilities to make greater use of AI, specifically machine learning. In Accenture’s 2017 Technology Vision, 38.1 percent of utilities executives surveyed believe their organizations will be making extensive investments in machine learning for AI in the next three years.

It’s time for utilities to begin piloting use cases to better manage their operations and workforce productivity. With drone and AI technologies now so readily available, utilities no longer have a reason to sit on the sidelines. Those that get in the game sooner rather than later will be the first to reap the rewards—and set the standards that others will follow. 

Popular Tags

    More blogs on this topic

      Archive

        SUGGESTED CONTENT