An IoT-Device Based on KNN for Heatstroke Illness Prevention

Abstract

Using Internet of Things (IoT) technologies and Private Cloud data processing, this study aims to (1) find out how heatstroke occurs in people who are outside, especially when they are dehydrated or have impaired thermoregulation; (2) create an automated real-time monitoring and heat warning system; and (3) assess how well the system works to prevent heat-related illnesses. The study involved creating a prototype that can independently use external environmental sensors to evaluate temperature, humidity, and air pressure. A mean fusion technique was utilized to combine sensor data with meteorological data sourced from the OpenWeatherMap API, thereby improving the accuracy of the study. The K-Nearest Neighbor (KNN) method was utilized to analyze the aggregated data and evaluate the likelihood of an individual experiencing heatstroke. The model achieved peak performance at k = 3, demonstrating an accuracy of 86.67% in recognizing high-risk heat scenarios. The system was equipped with automated notifications that delivered accurate real-time alerts. Groups participating in outdoor activities, such as students in outdoor classes, athletes undergoing training, and workshop attendees exposed to sunlight, took part in field trials. The evaluation findings indicated that consumers expressed a high level of satisfaction, achieving an average rating of 4.37 out of 5.00 (87.44%). The device control function achieved an impressive satisfaction rating, with an average score of 4.54 (90.80%). The findings demonstrate that the method we proposed serves as a dependable, cost-effective, and scalable solution for tracking local temperature levels and reducing the risk of heatstroke. This method can be utilized in educational settings and various outdoor locations where proactive management of heat-related health issues is essential.