‘Early Landslide Detection with AI’ Wins Grand Prize at Disaster Safety Competition

Pukyong National University’s Hayongsoo and Song Changho ... Propose 'Landslide Risk Detection Integrated System'

Hayongsu (second from the left) and Song Changho (third from the left), who won the grand prize at the BSI-Zone Disaster Safety Startup Idea Competition, are taking a commemorative photo after the award ceremony.

[Asia Economy Yeongnam Reporting Headquarters Reporter Dongguk Lee] Hayongsoo (PhD candidate), Song Changho (3rd year PhD student) from the Department of Ocean Engineering at Pukyong National University (Advisor: Professor Kim Yuntae) won the grand prize at the ‘This Year’s BSI-Zone Disaster Safety Startup Idea Competition’ hosted by Busan Technopark (BTP). They received a prize money of 3 million KRW.

The Pukyong National University students received the highest evaluation in this competition, which called for startup planning and commercialization ideas in the disaster safety industry fields such as safety accidents, flooding, fire, and landslides, with their project titled ‘Integrated Landslide Risk Detection System.’

The ‘Integrated Landslide Risk Detection System’ proposed by the students is an integrated system that creates landslide risk maps using deep neural networks and artificial intelligence (AI) techniques to extract landslide risk areas, and applies motion detection technology to detect landslide risks.

To overcome the limitations of existing landslide risk maps, they applied deep neural network technology and used 12 landslide-inducing factors?more than the conventional 9?to produce a highly sensitive landslide map for the Busan area, raising the landslide prediction rate to 60-70%.

Based on this improved prediction landslide map, the ‘Integrated Landslide Risk Detection System’ installs closed-circuit television (CCTV) incorporating AI technologies such as motion detection in landslide risk areas, which immediately recognizes cracks and alerts accordingly.

The Pukyong National University students said, “While conducting convergent research on landslide risk maps, retaining wall cracks, and ground cracks in the laboratory, focusing on landslide risk detection as a theme seems to have yielded good results,” adding, “Since we developed individual technologies, once the infrastructure such as facilities and communication is established, we expect to build a system beyond the idea stage and utilize it for actual landslide damage prevention.”

This content was produced with the assistance of AI translation services.