Evaluation of Helmet Wearing Compliance: A Bionic Spidersense System-Based Method for Helmet Chinstrap Detection
He Xu, Zhen Ma, Ziyu Wang, JieLong Dou, Yi Qin, Xueyu Zhang · Preprints.org · 2025
With the rapid advancement of industrial intelligence, ensuring worker safety has become increasingly important. Wearing safety helmets is one of the key protective measures. The demand for detecting the wearing status of helmet chinstraps during production processes is growing. Existing detection methods often suffer from discomfort to the wearer or privacy invasion issues. To address these challenges, this paper proposes a non-intrusive helmet chinstrap wearing detection approach based on a bionic system inspired by the mechanosensory hair arrays on spider legs. This method employs multiple MEMS inertial sensors to simulate the sensory function of spider leg hair, enabling efficient data acquisition of helmet wearing states. Unlike traditional vibration-based detection, posture signals can reflect spatial structural features, but their fusion from multiple sensors increases signal complexity and background noise. Therefore, an improved adaptive convolutional neural network (ICNN) combined with a long short-term memory network (LSTM) is utilized to classify and recognize the tightness of the helmet chinstrap based on single-sensor and multi-sensor data. Experimental validation was performed using data collected from 20 subjects engaged in wall-climbing robot operation tasks. Results demonstrate that the proposed method achieves a high recognition accuracy of 96%. This study provides a user-friendly, privacy-preserving, and efficient solution for helmet-wearing detection in industrial environments.