Corrosion represents a prominent and multifaceted concern within the marine industry, exerting a substantial impact due to its contribution to economic detriment, structural degradation, and environmental compensations (Aa et al., 2022; Komary et al., 2023; Paredes, 2023; Pudar and Ivosevic, 2023; Refani and Nagao, 2023; Saleh et al., 2023). Given its multidisciplinary nature, corrosion imposes significant and enduring consequences, resulting in financial repercussions amounting to billions of dollars (Imran et al., 2023; Schindelholz et al., 2021; P. D. Sun et al., 2020; Xiang et al., 2023). The financial implications of corrosion are profound, manifesting in substantial expenses incurred for the maintenance, repair, and replacement of equipment (Antunes, 2023; BinSabt et al., 2023; Koko and Kusakana, 2023). These expenses encompass direct costs such as materials, labor, maintenance, and equipment replacement. Additionally, indirect costs come into play, which encompass production interruptions, environmental harm, transit delays, injuries, and even fatalities (Alwi Mohd Yunus et al., 2023; Davidson et al., 2021; Du et al., 2018; Hussein Farh et al., 2023; Lang and Han, 2022).

The global economy suffered an estimated loss of approximately 2.5 trillion US dollars, equivalent to 3.4% of the world's GDP, due to corrosion (Zhang et al., 2022). Fig. 1 (a) and (b) provide insight into the direct cost of corrosion in different sectors of China (Shang and Zhu, 2021). Notably, a report identifies the Arab world as the region most severely affected, by corrosion, with corrosion accounting for 16% of their total GDP, as shown in Fig. 1 (c) (Tom White, 2019). Importantly, a significant proportion of these expenditures, approximately 15–35% of the GDP share, can be mitigated, with inspection costs playing a pivotal role in this endeavour (Al-Moubaraki and Obot, 2021; Berdimurodov et al., 2022; Hossain and Ulčickas, 2021; Nayak et al., 2023; Zakeri et al., 2022).

Corrosion poses a serious hazard to marine structures, resulting in diminished mechanical effectiveness and compromised structural integrity, which can lead to challenges like hull failure, docks or offshore structural breakdown, pipeline leaks, and most critically potential threats to human lives (Lazakis et al., 2022; Negi et al., 2023; Prabowo et al., 2021; Soares and Santos, 2021). In fact, corrosion is responsible for approximately 90% of the expenses associated with maritime structural failures (Khayatazad et al., 2020; Woldesellasse and Tesfamariam, 2023; Zehra et al., 2021). An alarming statistic reveals that, roughly 25–30% of the steel produced annually succumbs to corrosion, incurring substantial direct costs amounting to a staggering $276 billion, equivalent to approximately 3.1% of the nation's Gross Domestic Product (GDP) in the United States. These costs encompass not only the repair and inspection of corroded surfaces and structures but also the disposal of hazardous corrosion waste materials. Furthermore, they include expenses related to the application of protective coatings like paintings and surface treatments (Kuo and Chang, 2011). According to a report by the British Hoar committee, corrosion costs account for 3% of the British Gross National Product (GNP), with the potential to reduce these costs by 23%. In industrialized nations, an estimated 3.5–5% of their income or GNP is allocated to corrosion-related expenses, covering losses, replacements, maintenance, and prevention measures. Corrosion also incurs various other costs, such as production losses due to shutdowns and leakages, product contamination, and maintenance expenditures (Ansari and Sharma, 2023; Bourzi et al., 2020; Dhakal et al., 2022; Durrani et al., 2020; Lin et al., 2021; Obi and Nwajana, 2022; Singh, 2020a, 2020b; Singh and Prasad, 2023; Thakur et al., 2022).

In the context of ships, the primary expenses associated with corrosion are primarily indirect in nature. These costs encompass increased mass, heightened workload during the design and construction, decreased performance, and the expenditures associated with repairs (Del Giudice et al., 2021; C. C. Gong et al., 2020). An analysis of ten different industries in the United Kingdom reveals that the shipbuilding sector incurs corrosion and corrosion protection costs amounting to 21% of its expenses. Nevertheless, by implementing improved design practises and superior protective coatings, it is estimated that up to one-fifth of this cost could be saved. Table 1 provides an overview of corrosion-related costs, encompassing both corrosion prevention and actual corrosion. For new cruise vessels, this cumulative cost amounts to $350 million, in addition to the annual expense for repair and maintenance for cruise vessels, which totals $337 million (Liisa Ojaniittu, 2019). However, it is noteworthy that comprehensive reports detailing the financial impact of corrosion in marine remain relatively limited.

Furthermore, the environmental repercussions of corrosion cannot be ignored (Abo Nassar, 2022). The release of corroded substances into the environment can contaminate water, contribute to solid pollution, and necessitate substantial economic compensation. This can exert a significant influence on the marine ecosystem. Corrosion results in the release of various pollutants, including heavy metals and chemicals, into the surrounding marine environment, which leads to degrading water quality, potentially causing harm to aquatic life, and disrupting the intricate web of marine diversity. Furthermore, corroded structures, whether they were intentionally submerged or accidently abandoned, can function as habitats for marine organisms. Continuous degradation of these structures can release pollutant into water, potentially alerting the ecosystem dynamics (Albeldawi, 2023; Al-Taai, 2022; Anthony et al., 2023; Fontes et al., 2023; Ogolo et al., 2022). These pollutants frequently invade the food chain, accumulating in the marine organism tissues through bioaccumulation, affecting both aquatic life and humans who rely on seafood consumption. Additionally, corrosion-related pollutants have the potential to contribute to eutrophication, disrupting imbalances in aquatic environments and impacting coral reefs and further fragile marine habitats (Chia et al., 2020; Issac and Kandasubramanian, 2021; Maser et al., 2023; Sonone et al., 2021). In order to overcome these limitations, long-term corrosion monitoring strategies are significantly essential.

Long-term corrosion detection and prevention are absolutely crucial for avoiding catastrophic marine structural incidents (Abbas and Shafiee, 2020; X. Li et al., 2022b; Shafiee and Soares, 2020). Beyond the financial advantages, early identification of structural deterioration significantly reduces risks to humans' safety and environmental harm while also preventing potential structural breakdowns (Farh et al., 2023; Jegadeesan, 2021; Lucy Li et al., 2021Lee et al., 2021; Verstrynge et al., 2022). Traditional corrosion control technologies, while somewhat effective, frequently face challenges related to adaptability, performance optimizations, and cost-effectiveness. According to estimations, employing the appropriate prevention measures can reduce corrosion costs by roughly 18–35% (Honarvar Nazari et al., 2022; Pourhashem et al., 2020). Moreover, current artificial technologies have the potential to decrease the yearly direct cost by 20–25% (Sheikh et al., 2021). Corrosion on ships alone costs the industry up to $80 billion annually, a staggering figure that could be mitigated through the application of artificial intelligence and advanced corrosion prevention technologies.

In recent years, the incorporation of machine learning approaches has revolutionized the field of maritime corrosion monitoring because of its ability to examine enormous datasets and identify complex patterns. These algorithms leverage historical data, diverse environmental parameters, and material characteristics to develop predictive models that can predict the initial and progression of corrosion (Lv et al., 2020; Rocabruno-Valdés et al., 2019; Roxas and Lejano, 2019a; Salami et al., 2020; Chengtao Wang et al., 2019; Xiao et al., 2022). Through this, they make it possible for preventative maintenance and intervention, which can significantly increase the lifespan of marine structures while simultaneously reducing associated costs. Additionally, as each marine location exhibits a distinct corrosion profile due to variations in salinity temperature and other environmental factors, ML algorithms can adapt to these particular environments. They can monitor corrosion-prone regions, provide real-time alerts, and ensure the implementation of appropriate corrosion control strategies that not only enhance cost effectiveness but also bolster safety measures by reducing the risk of structural breakdown (Alamri, 2022; Coelho et al., 2022; Jia et al., 2022; Kaidarova et al., 2023; Liu and Li, 2023; Ma et al., 2023; Momber et al., 2023; Parjane et al., 2023; Pezeshki et al., 2023; Roy et al., 2022).

Furthermore, the process of early detection using machine learning entails the capability to identify corrosion abnormalities well in advance of them developing into substantial structural problems. Through the examination of historical data and real-time sensor inputs, machine learning models have the capability to detect minor indications of corrosion, facilitating prompt intervention (Ahuja et al., 2021; Parjane et al., 2023; Rocabruno-Valdés et al., 2019). Precise estimation of corrosion rate is a crucial element of corrosion control that is enhanced by the use of machine learning. These models can accurately measure the rate of corrosion by considering several environmental and operational conditions (Alwi Mohd Yunus et al., 2023). The utilization of data-driven estimation enables efficient maintenance planning, assuring optimal allocation of resources and minimizing downtime. As a result, operating costs are minimized, and the lifespan of assets is extended (Coelho et al., 2022).

Moreover, ML algorithms particularly effective in improving asset maintenance planning by detecting issues early and accurately estimating failure rates. Efficiently prioritizing maintenance chores for vital assets guarantees that they are promptly addressed (Coelho et al., 2022; Jia et al., 2022). Consequently, this enhances overall operational effectiveness, diminishes the enduring financial load of maintenance costs, and guarantees a more sustainable strategy for managing assets (Ma et al., 2023; Pezeshki et al., 2023). Material selection and coating optimization are essential aspects of corrosion management, and machine learning can offer vital aid in this field. Through the analysis of large datasets on material performance and coating efficacy, these models aid organizations in making well-informed decisions about the utilization of corrosion-resistant materials and the application of protective coatings. Making well-informed decisions not only decreases the expenses associated with corrosion over time but also improves the overall strength and endurance of assets (Ansari and Sharma, 2023; Bahrami et al.).

The aims of this review paper encompass delivering a comprehensive overview of machine learning approaches for corrosion studies in marine, offshore, and oil and gas infrastructure. Our aim is to explore the potential of ML techniques associated with corrosion detection and prediction, with a specific focus on the analysis of the two most widely used algorithms and architectures in this field. Moreover, we intend to identify future research opportunities and advancements by gaining insights into the strengths and limitations of the existing methods.

The chapter outline of this paper, Chapter 1, provides a framework by establishing the study context, including an evaluation of corrosion effects and associated costs. In chapter two, we explore the methodology utilised in this research, including the research question, search strategies, and search results. Findings are illustrated in Chapter 3, including the application of artificial neural networks and random forests within the context of marine, offshore, and oil and gas corrosion research. Moving on to the next chapters, we discuss the challenges and research gaps in existing research; furthermore, future research potential is provided before the paper's conclusion.

Materials and research design

The methodology employed for this review paper involves a structured approach consisting of four essential steps. First and foremost, we meticulously established the research questions, refined the search strategy, defined selection criteria, and outlined the procedures for data collection, extraction, and analysis. The research questions were meticulously formulated to serve as the guiding principles of our investigation. The research questions are as follows: RQ1: What are the specific