An AI‐Enabled Cargo Ship Monitoring System Based on Ball‐Joint Tandem Array Triboelectric Nanogenerators

Abstract Amidst the rapid development of AIoT, powering sensor nodes has become a major challenge for IoT. Moreover, research on self‐powered systems for complex scenarios like cargo ships is limited. This study proposes a cargo ship monitoring system based on ball‐joint tandem array triboelectric nanogenerators (BJTA‐TENGs), whose modular design is arranged in a similar way to the body segments of caterpillars. The BJTA‐TENGs consist of five TENGs connected in series using ball joints, featuring multi‐channel cross‐sensing capabilities. They can convert the vibrational energy generated as the ship plows through waves into electrical energy, while simultaneously outputting electrical signals that reflect the ship's navigation status. Upon testing, a single TENG generates an effective output power of 0.0057 µW (power density 23.351 µW m ^{− 3} ) and remains stable after 20,000 cycles. Furthermore, based on the 5‐channel BJTA‐TENGs and the GRU algorithm model, the monitoring system achieves recognition accuracies of 2 mm for vibration amplitude, 0.2 Hz for vibration frequency, 1° for tilt angle, and 0.1 Hz for tilt frequency, with an average accuracy rate of 98.76%. The trained model is deployed on a computer, with OpenBCI used for Wi‐Fi wireless signal transmission. The final recognition results can be visualized on a computer.