Internet of Things and Edge AI for Smart Healthcare: Enhancing Real-Time Monitoring and Predictive Decision-Making

Abstract

The study used a simulation-based research methodology and prototype development to assess the deployment of IoT and Edge AI in the smart healthcare monitoring system. A wearable or IoT device-based sensor system was connected with the Edge AI architecture to process local data and make predictions to aid in healthcare monitoring. The sensor data we collected were related to health care and were retrieved from either public datasets or simulated IoT devices with edge-level machine learning algorithms. The proposed system was tested for predictive accuracy, response time, latency, bandwidth consumption and computational efficiency and its performance was compared with the conventional cloud-based processing approach. The results showed that the Edge AI configuration had a similar or higher predictive accuracy, but with significantly reduced response time, end-to-end latency, and bandwidth use, and better computational efficiency per energy used, compared to the cloud-based approach. The greatest improvements were seen in latency sensitive applications like fall detection and cardiac anomaly detection, where timely response is critical. The study concludes that Edge AI can be a good and sometimes better solution than traditional cloud-based processing for real-time healthcare monitoring in situations where cloud-based architectures are not feasible due to network reliability, bandwidth or response time requirements.