ABSTRACT – With the advent of Industry 4.0 and equipment connected to the Internet of Things (IoT), predictive maintenance—in conjunction with diagnostic systems and intelligent fault detection based on data and Artificial Intelligence (AI)—has been a subject of study and research for several years. Utilizing these tools with data originating directly from monitoring sensors aims to guarantee equipment availability and identify signs of failure even before an experienced technician can detect them. Currently, maintenance planning and control systems may not provide the precision expected for effective action. This occurs because information input is performed manually by the operator and may be done incorrectly, whether due to the non-performance of the proposed service or the language presented in the manual being misunderstood by the operator, hindering the evaluation by the maintenance planner and the diagnosis of an imminent failure. By opting for data-based predictive maintenance, the goal is to improve assertiveness, reduce costs from unnecessary parts replacements in preventive maintenance, and reduce the probability of breakdowns. In the maritime field, an increase in the search for these solutions is observed in vessels due to the increased data collection capacity of equipment with new onboard technologies. During vessel operations, the challenge of applying intelligent monitoring systems stands out primarily due to the complexity of the equipment and non-constant operating profiles, unlike industries with static assets. Consequently, maritime companies currently rely on preventive maintenance scheduled by the manufacturer and corrective maintenance, which can be operationally and financially costly. The application of fault and anomaly detection methods in marine diesel propulsion engines is essential to complement preventive methods, ensure safety, and maintain vessel operation. These engines are usually large-scale, making their maintenance complex and non-trivial, requiring hours or days of downtime to perform the work. This work presents a preliminary literature review concerning fault detection methodologies in marine engines, with the goal to apply in tugboat operations, serving as an initial component of a research project shared for academic discourse rather than as a formal peer-reviewed publication.