This research aims to develop a prototype motion detection device and apply the Blynk IoT platform to transmit real-time emergency alerts to smartphones. The ESP32 microcontroller processes signals from a passive infrared (PIR) sensor to transmit data. Accuracy and response time were evaluated in 50 experiments (25 people per group). Confusion matrix analysis showed that condition B had the fastest alert speed, with an average response time of 1.05 seconds. Furthermore, Condition B achieved an overall accuracy of 92.00%, precision, recall, and F1-score of 93.33%, significantly outperforming Condition A (80.00% accuracy). However, this heightened sensitivity yielded an average of 4.60 false positives, demonstrating stable emergency alert capability, effectively alleviating caregiver concerns and enhancing security.