Distributed hydrokinetic energy harvesting is essential for enabling sustainable self-powered sensing in aquatic environments. However, natural water flows are typically low-frequency, weak, and highly stochastic, which limits the efficiency of conventional energy harvesters. Here, a gray shark–inspired parallel flapping triboelectric nanogenerator (GS-TENG) is proposed for efficient hydrokinetic energy harvesting and environmental monitoring. A biomimetic flapping hydrofoil inspired by the streamlined morphology of the shark is developed and coupled with an energy storage–release mechanism to enhance energy capture under low-flow conditions. In addition, a deep-learning-assisted framework is employed to optimize key structural parameters of the flapping foil. The optimized biomimetic hydrofoil improves average power by 92.4% compared with the conventional NACA0015 airfoil. The energy storage–release mechanism converts irregular low-frequency flow excitation into stable rotational motion, resulting in a 71.7% enhancement in electrical output. The GS-TENG achieves a peak power of 46.65 mW and a power density of 29.32 W m^-3. Laboratory and open-channel experiments demonstrate that the GS-TENG can continuously power water temperature and water level sensors, highlighting its potential for self-powered water-environment monitoring.

As mine excavation deepens, ventilation systems often face the challenge of insufficient airflow, while the complex environment poses significant obstacles to powering monitoring and alarm sensors. Here, an integrated and efficient self-powered mine wind speed monitoring and alarm system (SLW-MAS) is proposed based on triboelectric nanogenerator (TENG). The SLW-MAS, featuring a centrifugal structure design, facilitates hierarchical control of the TENG module, thereby enabling differential responses to wind speeds. When the wind speed is lower than 1.5 m/s, the TENG module is maintained in a horizontal working state under the action of the centrifugal mechanism and produces a high voltage output; the switch circuit is selected through experiments, which makes it meet the alarm delay of 2 s and avoids the problem of inaccurate alarm caused by unstable airflow. This work provides the feasibility for the construction of an underground distributed Internet of Things monitoring and alarm system.