A Smart Fuzzy Risk Assessment System for Epidemic Prevention on Cruise Ships

Cruise ships are a popular vacation choice, but their usually crowded conditions and enclosed spaces make them vulnerable to outbreaks of airborne infectious diseases. Risk assessment (RA) of epidemic outbreaks on cruise ships has traditionally relied on knowledge-based and probabilistic models, focusing on long-term predictions, with only a few studies considering the effect of short-term disease transmission. However, these methods face the following limitations: a) they do not utilize sensors for real-time monitoring, b) they require information that is unavailable in real-time conditions, c) they consider a limited number of risk factors in the RA process, and d) they lack comprehensive control measures. However, these methods face the following limitations: a) they do not utilize sensors for real-time monitoring, b) they require information that is unavailable in real-time conditions, c) they consider a limited number of risk factors in the RA process, and d) they lack comprehensive control measures. To address these challenges, a novel sensor-based RA system for real-time monitoring and disease control is proposed. The proposed RA system, named Healthy Ship Risk Assessment (HSRA), includes the following components: 1) a network of smart sensors for continuous monitoring of indoor ship areas and detection of symptomatic passengers; 2) a data analysis module intended to utilize deep learning methods for symptom detection; 3) a fuzzy rule-based decision module, developed using domain knowledge to compensate for data limitations; 4) a risk index derived from the inference process that corresponds to the spread of the disease onboard; and 5) adaptive safety guidelines for the ship personnel based on the risk estimated by the system. The system is validated using an agent-based model (ABM) and achieves an accuracy of 87% for experiments conducted across areas of a representative cruise ship with different characteristics, demonstrating its effectiveness in real-world conditions.