Bio-Inspired Calorimetric Flow Sensor With Hierarchical Microfluidic Design For Flow-Acoustic Fusion Sensing.

High-sensitive flow and acoustic sensors with wide detection range are essential for the intelligence of underwater robotics. To tackle the challenge of limited detection range of flow sensors, here we present a bio-inspired calorimetric flow sensor (BioCFS) with hierarchical microfluidic design, mimicking the graded neuromast location of fish lateral line. The BioCFS achieves a high velocity resolution of 20 µm·s-1 and a wide dynamic range of 0-3 m·s-1. The high-sensitive and wide-range BioCFS enables the functions of yaw angle and velocity estimation, obstacle recognition, and vortex source localization. Moreover, the BioCFS can perceive low-frequency underwater acoustic signals, demonstrating a benchmark acoustic sensitivity of 1.31 V·Pa-1. By integrating flow-acoustic fusion perception with deep-learning models, the sensor achieves an accuracy of 99.6% in classifying 12 types of underwater scenarios. The excellent sensor performance and flow-acoustic fusion sensing function of the BioCFS offer significant potential for environmental perception and intelligent navigation in underwater robotics.