AI-enhanced fire detection and suppression system for autonomous ships

Abstract

The rapid propagation of fire after a flashover can result in significant economic losses and human casualties. Meanwhile, the shipping industry faces crew shortages, making unmanned fire detection crucial for reducing manpower while ensuring seamless operations. Although autonomous fire detection and suppression systems (AFDSS) have been applied in tunnels and building infrastructure, no specialized AFDSS exists for unmanned autonomous ships, where minimizing false alarms and ensuring precise fire suppression under irregular wave conditions are critical. In this paper, we introduce an innovative AFDSS that integrates RGB, infrared (IR), and ultraviolet (UV) sensors to reduce false alarms and employ a reinforcement learning algorithm to optimize water spray under sea conditions. We present the design, system integration, and fire-extinguishing experiments, demonstrating the effectiveness of the system in enhancing the fire safety of autonomous ships under simulated sea-state conditions.

Keywords

• Autonomous fire detection and suppression system
• Artificial intelligence
• Sensor fusion
• Stereo vision
• Ship motion