Objective To monitor sudden changes in the physical status (normal to fall) of the elderly in real time and provide intelligent alerts, enabling guardians to intervene promptly. Methods The system employs an MPU6050 three-axis accelerometer and a digital motion processor (DMP) to form the body state sensing unit, which captures the acceleration along the X, Y, and Z axes and calculates the resultant vector. Data are transmitted via the inter- integrated circuit (I 2 C) bus to an nRF51822 microcontroller and then sent to the data processing module using Bluetooth 4.0. The module uploads the data to a cloud monitoring platform via GSM/4G wireless communication. The platform uses a threshold-based detection algorithm to analyze the data and sends alerts to the guardian’s mobile terminal when anomalies are detected. Results When a fall or other abnormal event occurs, the system immediately triggers an alert and sends the location of the monitored individual, allowing timely intervention. Conclusion The system establishes a “sensor detection–Bluetooth transmission–cloud analysis–intelligent alert” closed-loop monitoring process, offering strong real-time performance, low false alarm rates, and flexible deployment. It effectively improves the response speed and treatment success rate for the elderly sudden change in the physical status.