Home / Blog / Computer Vision and AI can make a difference – I

Computer Vision and AI can make a difference – I

A fly on the blade of a wind turbine tower. The photo was taken by an autonomous drone inspecting the tower
using AI and computer vision to detect cracks as small as 3mm. Image credit: Clobotics

Some years ago I was involved in evaluating the possibility to use drones to inspect telecommunications infrastructures, namely radio antennas. This is a time consuming job but it turned out the the regulation at the time was enforcing the tethering of the drone (in Italy) and the resolution of the camera was not sufficient for the inspection.

Now I stumble on a Chinese/US start up focussing on the use of drones for inspecting wind turbines, Clobotics. They are addressing the problem by leveraging two technologies: advanced computer vision and artificial intelligence.

The problem is a big one, with wind turbine tower as tall as 100m, taking a team of 5 technicians to spend at least six hours in checking the integrity of the tower and blades. It is necessary to inspect them on a sub cm scale to detect possible cracks. Think about even a medium size wind farm with tens of wind turbine, not to mention larger ones with hundreds of them, and you get a feeling of the problem and resource needed for the periodical inspection (add to this the intrinsic danger of inspecting such a tall structure…).

Using drones would seem a no brainer. Yet it is a tough challenge. Wind turbine have the nasty habit of being installed in … windy areas, and that is not a good place for flying a drone. Manual control of the drone proved to be tricky and unfeasible in most situations. Clobotics developed an autonomous flying system based on AI and computer vision that allows the drone to get close to the structure in a safe way. The computer vision enhances the camera resolution looking specifically for small details that would signal a potential crack. You can see in the photo the detection of a fly, a speckle of just 5mm on the big blade.

The AI and computer vision software plus the data analytics software is running on board the computer drone. This software is for the autonomous control of the drone, to get it close to the structure and to make sure that the whole structure is inspected. The captured images are sent to the Microsoft Azure cloud for processing.

Notable the increase in speed of inspection. A single drone completes the whole structure inspection in just 25 minutes, a factor of 10 gain versus human inspection.

About Roberto Saracco

Roberto Saracco fell in love with technology and its implications long time ago. His background is in math and computer science. Until April 2017 he led the EIT Digital Italian Node and then was head of the Industrial Doctoral School of EIT Digital up to September 2018. Previously, up to December 2011 he was the Director of the Telecom Italia Future Centre in Venice, looking at the interplay of technology evolution, economics and society. At the turn of the century he led a World Bank-Infodev project to stimulate entrepreneurship in Latin America. He is a senior member of IEEE where he leads the Industry Advisory Board within the Future Directions Committee and co-chairs the Digital Reality Initiative. He teaches a Master course on Technology Forecasting and Market impact at the University of Trento. He has published over 100 papers in journals and magazines and 14 books.