SEA-Scan: Structural and Environmental Algorithms for Sensing of Offshore Systems

This project will develop a novel, multi-modal system for detecting entanglements in floating offshore wind infrastructure. To accomplish this it will: (A) integrate sensors into the mature Spotter sensing system; and (B) develop a low-power artificial intelligence/machine learning node for on-board processing and machine learning training, which will limit data transmission expenses; and (C) develop machine learning algorithms, to enhance anomaly detection and characterization of fishing gear entanglements and marine animal detection in noisy, difficult-to-detect settings; and (D) integrate the data and machine learning with a physics-based numerical digital twin to alleviate data-limited constraints, enhancing the efficacy of the entanglement detection. The digital twin will also be used to output information on estimated mooring damage and remaining fatigue life.