FURI | Fall 2025

WhiskerSense: Decoding Hydrodynamic Trails for Covert Underwater Tracking

This research focuses on the dynamic classification of underwater objects by using a seal whisker-inspired sensor to interpret hydrodynamic trails for marine monitoring. To improve classification accuracy, a novel spiral-perforated base amplifies the sensor’s vibrations, boosting the signal strength from weak wakes by up to 51 times. The resulting high-fidelity data is used to train a convolutional neural network (CNN) to classify the shape and motion of marine animals. Future work involves validating this system in real-world conditions to provide a new, intelligent tool for non-invasive conservation efforts.