DESIGN OF AUTONOMOUS DISINFECTION SYSTEMS FOR HIGH-TRAFFIC AREAS ON SHIPS

Abstract

Regular cleaning of damaged ship hulls during dry dock repair ensures the efficient functioning of the shipping sector.Implementing an autonomous mechanism for corrosion removal by waterdisinfection is a viable strategy to alleviate manual labor demands while minimizing water, time, and energy expenditure.This research proposes a water-disinfectionmethod for a unique robotic platform termed Hornbill, which utilizes an ongoing magnetic attachment mechanism and sensor fusion for self-location to navigate effectively on vertical surfaces.Therefore, the research suggests a Comprehensive Way-point Path Planning (CWPP) to re-disinfectthe self-synthesizing Deep Convolutional Neural Networks (DCNN) on the corrosion heatmap through initial disinfection.The ideal CWPP issue, which encompasses minimizing trip length and duration to conserve water and power whilst ensuring all designated waypoints are visited, is formulated as the classical Traveling Salesman Problem (TSP).The Pareto-optimal path for the specified TSP has been derived using Reinforcement Learning (RL) techniques, incorporating a suggested reward system depending on the robot's performance during disinfection operations.The findings from experiments at the shipyard indicate that the suggested RL-based CWPP produces a Pareto-optimal path, allowing the water-disinfection robots to utilize around 12% less energy and 8% less water compared to the second-greatest evolutionary-based optimizing approach across diverse workplaces.