Answers are grounded in this paper and cite their sources.
The maritime industry is responsible for the transportation of over 80% of worldwide commodities. Currently, the industry heavily relies on traditional maintenance strategies, often leading to excessive and preventable downtime events, failing to ensure operational integrity and safety of the vessels. Machine learning (ML)-based predictive maintenance (PdM) emerges as a promising solution, utilizing vessel related data-driven insights to proactively predict machinery failures and optimize maintenance schedules. Towards this direction, this survey presents state-of-the-art PdM solutions, providing insights into different implementation approaches and their target components, while also mapping available datasets. Furthermore, considering the significant limitations of centralized ML (CML) in data privacy, security and scalability, the paper also explores decentralized architectures based on federated learning (FL). Finally, the paper presents a comparative analysis of the advantages and disadvantages of both CML and FL, with a particular focus on the potential benefits of FL for maritime PdM adoption, while also addressing open issues and future research directions in the field.
No takes yet. Share an insight, caveat, or question.
Kalafatelis et al. (2024) studied this question.