Leveraging Bayesian Network to Improve the Marine Insurance's Condition Survey Form for Passenger Vessel

M. Faishal*, R. O.S. Gurning, A. Santoso, D. H. Waskito

*Corresponding author for this work

Research output: Contribution to journalConference articlepeer-review

Abstract

To prevent the losses caused by accidents in passenger vessel, marine insurance is becoming one of the alternative solutions for the vessel's owner to retain their financial capability. One of the most essential phases in marine insurance is survey inspection, which utilises survey condition forms as guidance. However, the current form is not designed explicitly for particular vessel types, leading to inaccurate insurance assessment. The primary objective of this study is to identify and develop an adaptive and comprehensive condition survey form for passenger vessels in Indonesia by incorporating a qualitative and quantitative risk assessment, which will be the cornerstone of this study. The Bayesian Network (BN) approach was utilised to analyse several factors that lead to a sinking accident. Twenty-one accident data from Indonesian passenger vessels were utilised as base data for generating the conditional probability. The results indicated that vessel structural defect and unqualified crew are the two most significant factors contributing to the sinking accident's critical events, such as loss of stability and buoyancy. A practical improvement for the existing survey form has been proposed, which will benefit marine surveyor companies by enhancing the safety and accuracy of ship assessment.

Original languageEnglish
Article number012043
JournalIOP Conference Series: Earth and Environmental Science
Volume1423
Issue number1
DOIs
Publication statusPublished - 2024
Event4th Maritime Safety International Conference, MASTIC 2024 - Bali, Indonesia
Duration: 25 Aug 202428 Aug 2024

Fingerprint

Dive into the research topics of 'Leveraging Bayesian Network to Improve the Marine Insurance's Condition Survey Form for Passenger Vessel'. Together they form a unique fingerprint.

Cite this