Touching Tomorrow atPukyong National University

NEW BEGINNING, NEW INSPIRATION

Pukyong Today List

NOTICE
PKNU student won grand prize with the idea of 'landslide detection'
WRITER 대외협력과 WRITE DAY 2022-05-25
COUNT 216
작성자,작성일,첨부파일,조회수로 작성된 표
PKNU student won grand prize with the idea of 'landslide detection'
대외협력과 2022-05-25 216

'Early detection of landslides with AI', the winner of the 'Disaster and safety contest, BSI-Zone 2022'

- Ha Yong-Soo and Song Chang-Ho from PKNU … suggested 'integrated detecting landslide system '

Ha Yong-Soo (second from left) and Song Chang-Ho (third from left) are taking a commemorative photo after receiving their awards at the prize-giving.

 

Ha Yong-Soo (doctoral graduate) and Song Chang-Ho (3rd year in Ph.D. course) (advisory prof. Kim Yoon-Tae) of the department of Ocean engineering, Pukyong National University, took the honor by winning the grand prize at the 'Startup idea contest on the disaster and safety, BSI-Zone 2022' hosted by Busan Technopark (BTP). The prize money was 300 million won.

 

Pukyong National University students received the most excellent evaluation as the 'integrated landslide detection system' in this contest, where ideas for startup planning and commercialization in disaster and safety industries such as safety accidents, floods, fires, and landslides were held.

 

The 'integrated landslide risk detection system' proposed by them is a system that creates a map of an area at risk of landslides with a deep neural network and AI technique, extracts landslide risk areas, and detects the risk of landslides by applying motion detection technology.

 

They applied deep neural network to overcome the limitations of the existing landslide risk map and produced a sensitive landslide map for the Busan area, which raised the landslide prediction rate to 60-70 percent by using 12 more landslide triggers than the existing 9 triggers.

 

'Integrated landslide detection system' is a system that is based on the landslide map that has increased the prediction rate in this way, after installing a closed circuit (CC)TV combining AI technology such as motion detection technology in a potential landslide risk area, a method of promptly recognizing a crack and notifying it as an alarm.

 

The students from Pukyong National University said, "I think that we got good results because we conducted a convergence study on the topic of landslide risk detection while several studies on cracks in retaining walls and cracks in the ground were being conducted simultaneously in our laboratory, including a landslide risk map study," and added, "As a result of this, unit technologies have been developed, and as long as the infrastructure such as facilities and communication is provided, we are very anticipating that it will be possible to build an actual system beyond an idea and use it to prevent damage from landslides." <Pukyong Today>