Ajjay Adhithya V

This paper proposes a conceptual framework for a LoRaWAN-based blood glucose monitoring system aimed at improving diabetes management in rural areas with limited healthcare access. Continuous glucose monitoring (CGM) sensors transmit patient data over a long-range, low-power network to a centralized server, enabling remote monitoring by healthcare professionals without relying on internet connectivity. The framework also envisions the use of machine learning algorithms to analyze historical glucose data, predict trends, and generate automated alerts for abnormal glucose levels. Together, these features offer a scalable and cost-effective approach to enhance timely diabetes care in low-resource settings.

This paper presents a comparative study of three lightweight transformer models: DistilBERT, ALBERT, and TinyBERT for the task of detecting AI-generated text. It evaluates the trade-offs between classification accuracy and computational efficiency, such as inference time and model size, against a traditional TF-IDF and Logistic Regression baseline. The study concludes that TinyBERT offers the most practical balance, achieving a high F1-Score of 0.9936 while being significantly faster and smaller than the other models. This makes it an ideal candidate for scalable, real-world deployment where resources may be limited.

This paper proposes an intelligent Road Surface Quality Rating (RQR) System that uses real-time data from multiple sensors to evaluate and rate road conditions. The system assigns quality scores to different road segments, enabling better monitoring, proactive maintenance planning, and optimized route selection. It also integrates hazard alert notifications, ensuring enhanced road user safety through GIS-enabled platforms. By shifting from reactive to proactive infrastructure management, the system improves transportation resilience, resource efficiency, and travel reliability.