부산대학교 도서관

The airflow sensor enabled by triboelectric nanogenerator (TENG) is significant for intelligent lab safety and human–computer interaction applications. However, the reported airflow/wind sensor focuses on enhancing the sensing materials and structures, lack of high resolution, and smart signal analysis. Herein, we present a self-powered micro-airflow sensor and its artificial intelligence (AI) system, applied for lab safety and human–computer interaction. The as-fabricated sensor has a high sensitivity of $0.6258~\mu \text{A}$ /(m/s) and a linearity of 0.9968. Attributing to the Venturi effect, the minimum detection velocity of the sensor is 0.13 m/s. Given the sensor performance, we develop a real-time pipeline gas leak location system with an AI user interface, which achieves a potential low detect error $\le 2.9$ cm. In addition, we successfully explore other applications, including human exit–entry counting, ventilation alarm, and breath-based smart aid communication. Above all, the airflow sensor exhibits tremendous potential in the AI and Internet of Things.