Conventional electric fans lack the ability to perceive environmental and user conditions, making it difficult to provide personalized thermal comfort. To address this issue, this study aims to design and implement a caring-type smart fan system equipped with intelligent sensing and dynamic control capabilities, integrated with an Internet of Things (IoT) platform to enhance caregiving interactivity. The system adopts a dual-sensing architecture that simultaneously captures visual and thermal data. In visual recognition, the YOLOv8n-Pose model is employed for human pose estimation, and real-time feedback is transmitted through the n8n platform to LINE, allowing caregivers to remotely monitor user posture and activity. For temperature assessment, body key points are mapped onto the thermal image array, and a Mean Skin Temperature (MST) algorithm with a distance compensation model is applied to correct infrared attenuation, thereby improving measurement accuracy. The core control logic employs a proportional control strategy that dynamically adjusts fan speed based on temperature data, with device integration via Home Assistant. Experimental results show that the system effectively detects user presence, posture, and skin temperature, providing continuous and adaptive airflow regulation. Compared with conventional fans, the proposed system demonstrates high potential in enhancing thermal comfort and intelligent caregiving applications.