AI and UAVs to Monitor Lithuanian Road Infrastructure
AB Kelių priežiūra – a state-owned company that ensures maintenance of more than 21.000 km of national roads – joins forces with technology companies Thrust and Agmis to introduce an AI-powered drone-based system for automated road inspection. This joint project, called GreenBee, aims to improve the quality and scope of road network inspection and also to reduce CO2 emissions created while performing inspection tasks.
Currently, road inspections are performed by specialist vehicles driving slowly and visually inspecting the road infrastructure. Depending on the type of road, these inspections are performed monthly, weekly, or sometimes daily. Each year road inspection vehicles travel thousands of kilometers, emitting tons of CO2.
With GreenBee road inspection, vehicles will be replaced by UAVs equipped with high-definition cameras and multiple sensors. The drones will fly along the inspection routes and gather data for further analysis. Acquired video and sensor data will be analyzed by employing AI and Computer Vision technologies to identify defects and trouble spots.
“By rapidly covering large areas and pinpointing areas of potential defects, UAVs and AI analysis will act as a reconnaissance service. This will allow us to efficiently plan manned inspection routes and to dispatch teams to the areas requiring prompt action” – noted Head of Quality and Technology at AB Kelių priežiūra Jolita Mackienė.
According to Jolita Mackienė, the introduction of UAVs will help to reduce the CO2 emission footprint of inspections by 90%. It will also help to divert maintenance resources from initial surveying to more critical tasks.
The UAV platform for the project will be developed by technology company Thrust. Thrust designs multifunctional and customizable drone systems for client-specific aerial inspection tasks.
“The length of the EU road network stretches for approximately 5 million kilometers. We estimate that the current road inspection framework requires driving 93,7 million kilometers and committing 3,1 million man-hours every year, with up to 37,5 thousand tons of CO2 emitted into the atmosphere. Employing UAVs for the task cut these emissions by 90%” – said the Head of Business Development at Thrust Gintautas Nedzveckas.
“AI video analysis is a game-changer in infrastructure inspection. From power lines to gas pipes, solar and wind farms, and other industrial objects, automated inspection frameworks help tackle even the largest projects, which otherwise would not be feasible. In addition, it uniforms the defect spotting process, eliminating the human factors in the process. By reducing costs and improving efficiency, AI analysis is quickly becoming a standard staple in today’s infrastructure diagnostics” – added Head of Product Development at Agmis Žygintas Šitkauskas.