Deep learning assisted ternary electrification layered triboelectric membrane sensor for self-powered home security

Abstract

In the era of the Internet of Things (IoT), home security has become increasingly critical. Here, we have developed a ternary-electrification-layered triboelectric membrane sensor (TEL-TMS) as a cost-effective approach for self-powered home security. This triboelectric membrane sensor holds a collection of compelling features, including decent flexibility and transparency, enabling a tight attachment to the curved surfaces. By integrating it into household devices such as doors, windows, and safe cases, comprehensive monitoring coverage for the entire home can be achieved. The results showed that the TEL-TMS has a speed range detection of 5–165 mm/s, a response time of 0.32 s, an error rate of less than 1%, and exceptional stability (>10,000 cycles). With the further introduction of machine learning algorithms, the sensor can identify different motion states and activity patterns with a classification accuracy of up to 99.2%. Moreover, it is easily detachable and reusable, offering wide applicability. To facilitate practical applications, a custom mobile application (APP) based on built-in algorithms has been developed for one-click status monitoring and intelligent home environment recognition. The development of this TEL-TMS represents a solid step towards self-powered smart home security.