General Anomaly Detection of Underwater Gliders Validated by Large-scale Deployment Datasets

Underwater gliders have been widely used in oceanography for a range of applications. However, unpredictable events like shark strike or remora attachment can lead to abnormal glider behavior or even loss of the glider. This paper employs an anomaly detection algorithm to assess operational conditions of underwater gliders in the ocean environment. Prompt alerts are provided to glider pilots upon detecting any anomaly, so that they can take control of the glider to prevent further harm. The detection algorithm is applied to abundant datasets collected in real glider deployments led by the Skidaway Institute of Oceanography (SkIO) in the University of Georgia and the University of South Florida (USF). In order to demonstrate generality, the experimental evaluation is applied to four glider deployment datasets. Specifically, we utilize post-recovery DBD datasets carrying high-resolution information to perform detailed analysis of the anomaly and compare it with pilot logs. Additionally, we implement the online detection based on the real-time subsets of data transmitted from the glider at the surfacing events. While the real-time glider data may not contain as much rich information as the post-recovery one, the online detection is of great importance as it allows glider pilots to monitor potential abnormal conditions in real time.