Research & publications

Internet of Things (IoT)-based healthcare monitoring systems commonly depend on cloud connectivity for data storage, visualization, and remote access. This dependence limits their reliability in rural clinics, emergency camps, disaster-affected regions, and other low-connectivity environments. This paper presents the design, implementation, and preliminary evaluation of an offline-first IoT health monitoring prototype built on a Raspberry Pi platform. The system acquires four physiological and environmental parameters: heart rate, peripheral oxygen saturation (SpO2), body temperature, and ambient temperature. It integrates an MLX90614 infrared temperature sensor, a MAX30102 pulse oximeter module, a Python-based acquisition layer, a Node.js/Express local backend, local JSON storage, and a React dashboard served through a Raspberry Pi-hosted Wi-Fi hotspot. Users can access the dashboard from Wi-Fi-enabled smartphones, tablets, or laptops by entering the local IP address in a browser without requiring internet access. Quantitative validation against commercial reference devices produced mean absolute errors of 2 BPM for heart rate, 0.6% for SpO2, and 0.2C for body temperature under preliminary resting measurements. Benchmarking against recent IoT and edge-health monitoring studies shows that the proposed system contributes a compact, low-cost, multi-parameter, offline-first alternative for resource-constrained healthcare deployment. The results indicate that local edge processing and dashboard visualization can improve continuity, privacy, and usability where cloud-dependent health monitoring is impractical.