Sensor Technologies for Arc Fault Detection in Naval Ship Power Systems: Applications, Challenges, and Future Prospects - A Review

Arc faults have historically posed a significant risk to naval ships because of dense electrical interconnections, confined operational spaces, and the complex behavior of arc propagation. This risk has become even more crucial with the adoption of next-generation medium-voltage DC naval power systems, which incorporate high-power components to supply energy to advanced propulsion, radar, and weapon systems. In such environments, arc faults can generate highly energized plasma discharges capable of damaging critical components, triggering cascading failures, and compromising mission reliability. Ensuring safe and reliable operation has thus become increasingly more consequential, driving the evolution of arc fault detection systems. These detection systems primarily rely on sensor-based monitoring to capture pre-arc and post-arc behavior. Although sensors form the foundation of effective arc fault detection, their evolution and overall contribution within naval detection frameworks remain insufficiently reviewed. Therefore, for the first time, this paper presents a comprehensive review of the development of sensor technologies used in different generations of naval vessels and analyzes their role in enabling intelligent detection strategies. Particular emphasis is placed on the electrical, environmental, and mechanical signatures produced by arc faults and their detection using different sensing technologies to enable predictive maintenance and condition monitoring of imperative components in medium-voltage DC shipboard power systems. The review further examines the adaptation of existing arc-detection techniques to harsh naval environments, and the sensing technologies are comparatively ranked based on their adaptability. The current challenges associated with these techniques, including electromagnetic interference, mechanical vibration, limited onboard computational capability, false alarms, cost, and detection latency, are also discussed. To address these limitations, emerging research directions, including hybrid learning-based methods, diverse sensing modalities, synthetic dataset generation, transfer learning, and quantum-enhanced diagnostic approaches, are explored. Overall, this review establishes a unified framework linking arc physics, sensing technologies, and intelligent diagnostics to guide the development of reliable and predictive arc fault detection systems for next-generation naval power platforms. The review is based on a comprehensive analysis of research articles, conference proceedings, government reports on historical shipboard fire incidents, and technical white papers.