Machine Learning Enhanced Worker Safety Product: Potential Integration with Smart Jackets
G. GOPU, Adel Mariam A, ANU V ARSHINI R P, K Hemmasri · 2024
In today's era., the importance of health monitoring has reached unprecedented heights., necessitating innovative solutions to ensure timely intervention and proactive care. However., challenges include ensuring data security and privacy., managing the high initial costs and maintenance demands., addressing interoperability issues., training complexity., and navigating ethical considerations for responsible deployment. To overcome this the work introduces an innovative intelligent health monitoring system., leveraging IoT with advanced methods of machine learning to enhance worker care and preventive health management. Utilizing sensors for vital sign monitoring., including temperature., pulse., blood pressure., and oxygen levels., and a Raspberry Pi micro controller for data processing the system analyzes data and makes predictions about health status using machine learning methods like random forest., logistic regression., gradient boosting classifier., KNN and SVM. Demonstrating superior performance., the Random Forest algorithm achieves 95% accuracy., 87% precision., 100% recall., and a 92% F1 score while done in software. When it comes to hardware and software setup with real time data handling SVM provides the best result with accuracy 97%., precision 92%., 100% recall and 95% F1 score. By identifying abnormalities and alerting healthcare providers., this proactive approach ensures rapid responses to critical health situations., thereby enhancing worker care.