<p>Intravenous infusion is a common used method for administering medications and fluids to patients. Automated monitoring of the infusion process using sensors is crucial for ensuring patient’s safety and alleviating the workload of medical staff. Herein, we report a transistor-like triboelectric droplet sensor (T-TDS) for self-powered infusion monitoring. The T-TDS is innovatively designed by directly attaching a copper foil to the drip chamber of the primeval infusion tube and inserting a curved stainless steel probe to the drip chamber above the liquid level near the copper foil, thereby forming a transistor-like structure. This design generates a higher output voltage compared to the conventional single-electrode mode (e.g., only attaching a copper foil to the drip chamber), which contributes a higher resolution for droplet sensing. We optimized the tilt angle of the drip chamber to maximize output and verified the long-term stability of the signal over a 24-hour period. Integrated with the machine learning algorithms, we demonstrated the intelligent detection capability of T-TDS on drip rate, temperature, solution concentration and identification of the type of infusion medicine, the classification accuracy reaches 98%. The T-TDS presents the advantages of low-cost, high reliability, ease of implementation, and self-powering capabilities, making it a promising tool for large-scale infusion monitoring applications and improving medical management convenience.</p>