Patrolling motorways in real time using deep learning
On January 29 (Wed), Daegu Infrastructure Corporation joined the Daegu Smart City Support Center* and RGBLAB** for a trial deep learning-based patrol using drones along the motorway between Seobyeon Overpass and Gochongyo Bridge (Seobyeonnam-ro in Buk-gu).
* A provisional organization created based on a project operated by the Institute of Advanced Convergence Technology at Kyungpook National University (smart city planning, policy establishment, etc.)
** A venture-backed company dedicated to the research and development of industrial drone platforms for logistics, security, and pest control services using drones
Trial Drone Patrol
On this day, using LTE communications network and an AI system, the corporation conduct a trial patrol of a motorway between Seobyeon Overpass and Gochongyo Bridge in real time with a drone mounted with remote control and automatic analysis technologies.
Deep learning, which was used for the trial patrol, is a learning method that allows computers to think and learn like humans. Even without being “taught” by humans, this technology allows computers to learn by itself and predict situations that may arise in the future
Through this trial patrol, the corporation has confirmed the applicability of deep-learning technology. In the future, it will work with local organizations that hold advanced technologies to collect and analyze information, such as confirmation of fallen objects on the road, crackdown of badly loaded trucks, and analysis of damaged roads, in an accurate and timely manner to provide safer roads and environment for the people.