This paper presents a novel AI-enabled Internet of Things (IoT) framework to transform healthcare delivery and emergency response in developing countries. The proposed system combines wearable sensing based on TinyML, deep learning inference on-device, and triage powered by reinforcement learning in a hierarchical edge–cloud architecture optimized for data privacy and low latency. To enable adaptive edge–cloud orchestration for real-time analytics and predictive diagnostics without jeopardizing patient data, federated learning is used for decentralized, secure model training. By assessing vital signs, geolocation, geographical location, and resource availability, a reinforcement learning agent dynamically prioritizes emergencies, maximizing response times and triage effectiveness. In a simulated rural Sub-Saharan African environment, field validation produced diagnostic accuracy of over 93%, outperforming conventional cloud-only systems in speed and efficiency. The user-friendly interface facilitates multilingual, cross-modal communication through voice confirmation, app alerts, and SMS for populations with low literacy and internet access. The system's 2.3-second real-time warning latency makes it particularly useful for tracking infectious diseases, chronic conditions, and maternal care. This scalable, morally sound paradigm offers a guide for AI-driven health innovation in underprivileged environments, and it is in line with the UN Sustainable Development Goals (SDGs 3 and 9).
Bali et al. (Sat,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: