Control of Marine Robots in the Era of Data-Driven Intelligence

The control of marine robots has long relied on model-based methods grounded in classical and modern control theory. However, the nonlinearities and uncertainties in robot dynamics, coupled with the dynamic nature of marine environments, increasingly reveal the limitations of conventional control methods. The rapid evolution of machine learning has opened new avenues for incorporating data-driven intelligence into control frameworks, prompting a paradigm shift in marine robot control. This article provides a comprehensive review of recent progress in the control of marine robots through the lens of this emerging paradigm. It covers both individual and cooperative marine robotic systems, highlights notable achievements in data-driven control methods, and summarizes open-source resources that facilitate the development and validation of advanced control methods. Finally, several future perspectives are outlined to guide research toward achieving high-level autonomy for marine robots in real-world applications. This article aims to serve as a roadmap toward the next-generation control framework for marine robots in the era of data-driven intelligence.