Acoustic target recognition has always played a central role in marine sensing. Traditional signal processing techniques that have been used for target recognition have shown limitations in accuracy, particularly with commodity hardware. To address such limitations, we present the results of our experiments to assess the capabilities of AI-enabled acoustic buoys using OpenEar™, a commercial, off-the-shelf, software-defined hydrophone sensor, for detecting and tracking fast-moving vessels. We used a triangular sparse sensor network to investigate techniques necessary to estimate the detection, classification, localization, and tracking of boats transiting through the network. Emphasis was placed on evaluating the sensor’s operational detection range and feasibility of onboard AI for cloud-based data fusion. Results indicated effectiveness for enhancing maritime domain awareness and gaining insight into illegal, unreported, and unregulated activities. Additionally, this study provides a framework for scaling autonomous sensor networks to support persistent maritime surveillance.

While study of ocean dynamics usually involves modeling deep ocean variables, monitoring and accurate forecasting of nearshore environments is also critical. However, sensor observations often contain artifacts like long stretches of missing data and noise, typically after an extreme event occurrence or some accidental damage to the sensors. Such data artifacts, if not handled diligently prior to modeling, can significantly impact the reliability of any further predictive analysis. Therefore, we present a framework that integrates data reconstruction of key sea state variables and multi-step-ahead forecasting of current speed from the reconstructed time series for 19 depth levels simultaneously. Using multivariate chained regressions, the reconstruction algorithm rigorously tests from an ensemble of tree-based models (fed only with surface characteristics) to impute gaps in the vertical profiles of the sea state variables down to 20 m deep. Subsequently, a deep encoder–decoder model, comprising multi-head convolutional networks, extracts high-level features from each depth level’s multivariate (reconstructed) input and feeds them to a deep long short-term memory network for 24 h ahead forecasts of current speed profiles. In this work, we utilized Viking buoy data, and demonstrated that with limited training data, we could explain an overall 80% variation in the current speed profiles across the forecast period and the depth levels.