A Novel Bio‐Inspired Whisker Sensor System for Robotic Navigation With Enhanced Environmental Adaptability

ABSTRACT Vision and LiDAR sensors, though widely used in robotic navigation, perform poorly in low‐visibility or complex environments. Existing tactile sensors, despite some progress, lack the flexibility and adaptability of biological neural systems. This paper proposes a novel bionic whisker tactile perception system that demonstrates exceptional anti‐interference capability and environmental adaptability. The system integrates a spiking neural network (SNN) with a central pattern generator (CPG) to accurately emulate the active tactile sensing mechanism of rodent whiskers. Based on SNN and CPG, the sensory system achieves noise suppression capability through a dynamic neuron allocation mechanism. The robotic navigation system equipped with this module can operate stably in low‐visibility environments, even in complete darkness. Extensive experiments in simulated and real‐world environments, including scenes with low lighting and visual failures, demonstrate the system's superior performance by robot.