RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2
Project

Drones that detect power line defects began to be sold in Russia

Customers: Rosseti Center (IDGC of Center)

Moscow; Power



Project date: 2025/07

The Russian company VizorLabs has developed and commercialized a software and hardware complex for automated detection of defects in power lines using unmanned aerial vehicles and neural network technologies. The system was implemented in PJSC Rosseti Center and made it possible to reduce the cost of network inspection by 64% compared to manual monitoring while improving the accuracy of damage detection. This became known in mid-July 2025.

The total length of 35-110 kV power lines requiring annual inspection is more than 63 thousand kilometers. About 43% of all lines run through hard-to-reach areas, including river crossings, marshy and wooded areas.

Drones that detect power line defects began to be sold in Russia

Traditional methods for monitoring the state of power lines, based on visual surveys, required significant time and human resources. VizorLabs' automated system solves this problem through the use of unmanned technologies and artificial intelligence.

VizorLabs specialists have created a specialized software "BVS Flight Organization," which automates the creation and approval of applications for drones. The system also manages the activities of flight teams and coordinates the infrastructure monitoring process.

The key element of the complex was neural network models that provide high-precision detection of damage in automatic mode. A unique set of detectors has been developed that allows you to record various damage to power lines with high accuracy.

Svyatoslav Sterlikov, head of the VizorLabs unmanned systems department, explained that the monitoring uses cameras that allow detecting all types of violations. The system identifies trees growing in the protected zone, fallen insulators and other infrastructure defects.

Sterlikov noted that the neural network model, trained on thousands of images, finds defects much faster than humans. This is especially important when analyzing thousands of high-resolution photographs obtained per drone flight.[1]

Notes