WEND: An Efficient WSN-Assisted Emergency Navigation Algorithm for Dynamic Hazardous Ship Indoor Environments

Abstract

A reliable and efficient emergency evacuation of a damaged cruise ship is essential for passenger safety. Although many advanced evacuation approaches based on wireless sensor networks (WSNs) are capable of exploring dynamic environmental hazards and provide real-time navigation service, such approaches are mainly used in buildings on land, taking no account of the complex ship structure, deadline for ship survival, and dynamic ship inclination, and therefore may fail to evacuate passengers before ship capsizing. This article proposes WEND, an efficient WSN-assisted emergency navigation algorithm for dynamic hazardous ship indoor environment, which informs each passenger about a hazard-avoid route that minimizes the total dynamic typical delay with error bound guarantee while meeting the deadline for ship capsizing in a real-time manner. To achieve this aim, WEND investigates ship interior layout to construct a 3-D topological model and analyzes the time-history of the roll motion of a damaged passenger ship to evaluate ship survival time and the dynamics of pedestrian movement. Then, an efficient adaptive emergency navigation algorithm building on the ideas of Hassin’s algorithm is presented to provide a hazard-avoid path from each passenger’s current location to boarding stations such that the worst-case delay along this path is no greater than the specified deadline, and the total dynamic typical delay is small. We evaluate our algorithm by conducting extensive simulations. The results demonstrate that WEND improves the navigation success ratio by approximately 30% and 10% compared with the state-of-the-art navigation methods, namely the expected number of oscillations (ENO)-based oscillation-free method (OPEN) and the deadline-aware adaptive emergency navigation strategy (ANT), respectively.