Deep learning-based estimation of truck Turn Around Time at container port

  • Shin, Byongchan; 
  • Min, Yusung; 
  • Lee, Gunwoo; 
  • Yang, Hyounseok; 
  • Cho, Bumchul
Citations

WEB OF SCIENCE

7
Citations

SCOPUS

7

초록

In South Korea, most import and export cargo are mainly transported by sea. With so much cargo being transported through the ports, congestion from trucks is particularly severe at city ports, which are located close to cities. Accordingly, port terminals require systematic and efficient terminal operation, since congestion leads to problems such as air pollution. This study aims to estimate the Turn Around Time (TAT) of trucks at container terminals. Accurate truck TAT estimations have the potential to mitigate congestion, since truck drivers can avoid congested times. The dataset was constructed by combining truck DTG (Digital Tacho Graph) data collected in Korea in 2021, shipping in port data collected from major ports in Korea such as Busan New Port and Incheon Port, and weather data. Several artificial intelligence models were used to estimate the TAT, and their performance evaluation indicators were compared. From the comparison results, the artificial intelligence model with the best estimation performance was identified. As a result of estimating TAT through artificial intelligence models, Busan New Port had the highest TAT estimation accuracy. Through this study, we found that it is possible to estimate TAT using DTG data, which is installed in most Korean trucks. © 2025 Informa UK Limited, trading as Taylor & Francis Group.

키워드

container port; Deep learning; DTG (Digital Tacho Graph); Turn Around Time (TAT)
제목
Deep learning-based estimation of truck Turn Around Time at container port
저자
Shin, Byongchan; Min, Yusung; Lee, Gunwoo; Yang, Hyounseok; Cho, Bumchul
DOI
10.1080/03088839.2025.2470437
발행일
2025-02
유형
Article in press
저널명
Maritime Policy and Management
권
53
호
2
페이지
183 ~ 199