<p>To facilitate efficient ship design, it is imperative to evaluate ship maneuverability with precision and expediency. Achieving this requires an accurate estimation of the flow field generated behind the vessel. Although fluid forces induced by steering can be derived from the pressure exerted by the rudder, such measurements typically require tank testing and computational fluid dynamics (CFD) analysis. These methods are tuned to the specific type of ship and, consequently, take a long time and much cost. In this study, we present a methodology for estimating the flow around a wing utilizing data acquired from a single observation point located posterior to the wing. We employ reservoir computing, a technique specifically customized for time series analysis, and rigorously validate our approach against CFD data. This methodology improves predictive accuracy by establishing a relationship between the two-dimensional movement of the wing and the information derived from the observation point, thereby facilitating a highly precise estimation of flow characteristics. Unlike a standard recurrent neural network methodology, which has been extensively explored, reservoir computing enables highly accurate predictions while significantly reducing computational costs.</p>