Citation

Youn IH, Park DJ, Yim JB. Appl. Sci. (Basel) 2019; 9(1): e4.

Copyright

(Copyright © 2019, MDPI: Multidisciplinary Digital Publishing Institute)

DOI

10.3390/app9010004

PMID

unavailable

Abstract

Maritime accidents remain a significant concern for the shipping industry, despite recent technological developments. In the Republic of Korea, the leading cause of maritime accidents is navigator error, particularly in collisions and groundings; this cause has led to 79% of maritime accidents, according to a recent assessment. The reduction of navigator error is crucial for accident prevention; however, the lack of objective measures to monitor navigator error remains a challenge. The purpose of this study was to develop an objective classification of navigation behaviors in a simulated environment. The statistical model of classification of lookout activity was developed by collecting participants’ lookout behavior using a Kinect sensor within a given scenario. This classification model was validated in non-scenario experiments. The results showed that seven standard lookout activities during a lookout routine were accurately classified in both the model development and validation phases. The proposed model classification of lookout activity using an optical sensor is expected to provide a better understanding of how navigators behave to help prevent maritime accidents in practice.

Language: en

Keywords

lookout behavior classification; machine learning model; maritime accidents; optical sensor; simulation environment