AI and Drones Improve Power Line Inspections

Power line inspections are necessary but dangerous and expensive. Using drones, machine learning, and humans to help train the AI, image-based inspection solutions are safer and more effective than sending techs in the field.


Article Key

Image credit: Hitachi Vantara

Energy is essential to our way of life. Power enables us to work, communicate, travel, eat, and entertain ourselves. While our reliance on power is increasing, our infrastructure is aging. Energy companies strive to ensure their power lines are in proper working order, a task that requires constant vigilance and a substantial budget. Routine inspections cost thousands of dollars and put employees in dangerous working conditions.

In some cases, inspectors work from cherry pickers, climb towers hundreds of feet tall, or hang from helicopters to examine power lines. Wind and inclement weather further increase the risk, making power linemen one of the most dangerous jobs in the US. The use of drones is helping.

Eye in the Sky

Drones are lightweight and agile, allowing them to travel along power lines easily. With 3D mapping capabilities, drones can find the most efficient routes and avoid obstacles. They collect images along the path and, if outfitted with the appropriate sensors, can even detect heat signals. Drones reduce, if not eliminate the risk to linemen, who are free to analyze the images for defects or potential hazards.

That analysis takes time when facilitated by trained personnel and can be inconsistent. Linemen have to sift through hundreds of images and assess the stability of the equipment therein. Assessments can be subjective, as each employee might have a different threshold for equipment replacement or repair.

Artificial intelligence can further improve the benefit of drones. Image-based software systems bring efficiency and consistency to an otherwise time-consuming and subjective process.

Drone flying near utility pole

Image credit: Hitachi Vantara

Knowledge is Power

One company, Hitachi Vantara, an Intel® IoT Solutions Marketplace partner, is tackling the inspection problem for energy companies. The Hitachi Image-Based Inspections software uses artificial intelligence and machine learning to enable linemen to monitor power lines from the ground. Designed for transmission and distribution assets, Hitachi Image-Based Inspections can identify power lines, cracked insulators, bent dampers, and other potential problems that can have a devastating impact on the power supply and the environment.

The Hitachi Image-Based Inspections software automatically recognizes assets, identifies defects, and assesses the severity of each risk, and prioritizes maintenance. The solution collects and analyzes thousands of images per second, far faster than the human eye. It uses machine learning to identify anomalies or failures with a 360-degree view of the equipment.

Experienced employees and subject matter experts can train and retrain the algorithms to ensure that repair and replacement thresholds are appropriate and consistent across the company.

Power Play

Built on Lumada, Hitachi’s Industrial Internet of Things software-as-a-service (SaaS) platform, the Image-Based Inspection solution operates on the cloud or on-premises, depending on the client’s needs. It integrates with existing performance management platforms. Customizable dashboards display asset risk scores, potential points of failure, and possible mitigation tactics.

Using map-based visualization, power personnel can locate compromised equipment and be dispatched to fix it quickly before it fails. Data can be fed through machine learning algorithms to forecast potential failures, leading to more accurate predictive maintenance schedules, and a more reliable power supply overall.

Coupling drones with AI technologies, power companies can increase inspections and reduce the likelihood of failure, while limiting the risk for their employees.


Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries.