On the Development of a Digital Twin for Fire-Incident Onboard Guidance

A shipboard fire can arise from many potential sources, including but not limited to human error, equipment malfunction, damage or grounding, or even external factors. The crews carry out fire management using a physical approach, which inherently has limitations regarding danger and slowness. To address those concerns, streamlined methodologies that could be used to forecast the evolution of room availability are proposed in this work. This approach, implemented within an onboard installed digital twin, could be used to provide the crew with vital information to analyze the operational status of the vessel. Different alternatives have been studied to accomplish this task, from the most complex ones (FDS) to simpler ones (CFAST), that together with artificial neural networks should allow the system to work with onboard sensors in real-time.