Design and Implementation of a Vision Intelligence-Based Safety Management System for Ship Block Transportation

In the shipbuilding industry, the safe transportation of heavy blocks is a crucial factor directly linked to productivity and has been highlighted for the last few decades. However, current operations heavily count on module Transporter(TP) operators' expertise and experiences, posing persistent safety risks due to limited visibility and operator fatigue. This study proposes a smart Health, Safety, and Environment(HSE) system based on visual intelligence to address these challenges, implements this system and demonstrates it. The proposed system integrates FHD cameras attached on all four sides of the TP, a You Only Look Once(YOLO)-based real-time object detection AI model, an edge computing-based image processing system, self-luminous Personal Protective Equipment(PPE), and LED guide beams for vehicle width monitoring. The system includes real-time processing for detecting five types of obstacles - people, vehicles, traffic lights, street lamps, and trees. The system incorporates optimized image processing algorithms for reliable performance while implementing QR code-based automatic location tracking, enhancing transport history management. During a field demonstration, the system significantly improved workers' visibility and consistently detected potential hazards during transport operations, substantially reducing accident rates. These results illustrate that smart technology can effectively enhance worker safety and operational efficiency, establishing a foundation for future smart shipyard implementations.