Uncertainty-aware decision support for real-time damage stability assessment in maritime emergencies

Tools

Louvros, Panagiotis and Boulougouris, Evangelos and Vassalos, Dracos and Htein, Nay Min and Stefanou, Evangelos (2025) Uncertainty-aware decision support for real-time damage stability assessment in maritime emergencies. In: 20th International Ship Stability Workshop, 2025-10-01 - 2025-10-03, Porto Platanias Beach Resort & Spa, Platanias Village.

[thumbnail of ISSW-2025-Uncertainty-Aware-Decision-Support-for-Real-Time-Damage-Stability-Assessment-in-Maritime-Emergencies]
Preview
 Text. Filename: ISSW-2025-Uncertainty-Aware-Decision-Support-for-Real-Time-Damage-Stability-Assessment-in-Maritime-Emergencies.pdf
Accepted Author Manuscript
License: Strathprints license 1.0
Download (714kB)| Preview

Abstract

Effective decision-making in maritime emergencies relies on the accurate assessment of flooding progression and damage stability, yet significant uncertainties often impede timely and reliable predictions. This research presents a novel probabilistic decision support framework for real-time damage stability assessment, integrating advanced uncertainty quantification methods to enhance situational awareness and risk-informed decision-making. Building upon existing machine learning (ML) and case-based reasoning (CBR) methodologies, this study introduces a Bayesian inference and Dempster-Shafer theory-based approach to quantify and propagate uncertainty in flooding scenarios. By incorporating probabilistic reasoning, the proposed framework provides confidence intervals for key stability indicators, such as time-to-capsize (TTC) and progressive flooding rates, ensuring that decision-makers are informed of the reliability of each prediction. A Monte Carlo-based sensitivity analysis further assesses the impact of input variability, improving the robustness of predictions under diverse damage conditions. The decision support system is validated through extensive numerical simulations, demonstrating its ability to provide actionable insights with quantified uncertainty. Results indicate that integrating probabilistic methods into damage stability assessment significantly enhances prediction reliability compared to deterministic approaches, reducing the likelihood of unnecessary evacuations or delayed emergency responses. By advancing uncertainty-aware decision support in maritime safety, this research contributes to improving emergency preparedness, aligning with evolving International Maritime Organization (IMO) safety frameworks.

ORCID iDs

Louvros, Panagiotis ORCID logoORCID: https://orcid.org/0000-0001-8623-0680, Boulougouris, Evangelos ORCID logoORCID: https://orcid.org/0000-0001-5730-007X, Vassalos, Dracos ORCID logoORCID: https://orcid.org/0000-0002-0929-6173, Htein, Nay Min ORCID logoORCID: https://orcid.org/0009-0007-9534-4509 and Stefanou, Evangelos ORCID logoORCID: https://orcid.org/0009-0005-7586-4676;
CORE (COnnecting REpositories)