The Challenge:

Electric energy companies need to perform tasks related to the inspection of thousands of kms of their transmission networks as well as the monitoring and maintenance of transmission lines that can run through multiple regions and ecosystems.


The challenges that energy and utility companies face with managing infrastructure at this scale means they must automate inspection and gather information faster.

The Solution:

Addressing the challenges of first, monitoring large scale electric infrastructure and second, automating the inspection and detection of anomalies requires two crucial components:

UAVs equipped with various sensors/cameras to inspect thousands of Kms of transmission lines and towers as well as substations.

Insight-AI computer vision neural networks to automatically detect faults on the various assets throughout the network.


The Technology:

Insight-AI is used as the deep learning backbone for automated image analysis and inspection. Computer vision algorithms are first trained with thousands of images categorized according to fault types for specific infrastructure components: power lines, electric towers, sub-stations. Images captured by UAVs or drones and labelled by human experts are used to train the neural network.

Once deployed, UAVs flying inspection missions can transmit the images in real-time to Insight-AI, for analysis and inference, or they can be later uploaded into Insight-AI to be analysed by its trained neural network to detect faults or anomalies on any of the electric network infrastructure.

The same optical images can be used to re-train the neural nets and thus improve their accuracy or add new fault categories. Insight-AI for electric network inspection is trained to identify the following types of faults:

Transmission Towers: corrosion, missing pieces, damage equipment, lightning damage, vegetation overgrowth. Components inspected include pylons, switches, capacitors, fixtures, foundations, communications panel.

High Voltage Transmission Lines: missing pieces, damaged equipment, vegetation overgrowth, clearance issues, over-heating. Components inspected include lines, clamps, spacers, vibration dampers, aerial markers.

Substations: corrosion, missing pieces, damaged equipment, overheating. Components include transformers, bus, circuit breakers, disconnect switches, lightning arrestors.


Explore Datumcon Products

Neural Networks for Video, Language, and Geospatial Analytics.

Schedule a demo

Discover more about how Datumcon uses drones for tower inspections.