CuidaconTIC: Data Capture and Use from Wearables for Quality of Life in Caregiving

Abstract

This paper presents the development of a system designed to collect, process, and visualize wearable health data using the Google Fit API. The system enables users to access their personal health metrics, such as step count, heart rate, and sleep patterns, while also offering personalized content and recommendations based on these metrics. Additionally, the system supports healthcare researchers by providing an AI-driven interface that translates natural language queries into SQL, allowing for efficient data retrieval and analysis. While the system leverages the broad compatibility of Google Fit, it also faces limitations due to its dependence on a third-party service, which could be affected by changes or discontinuation of the API. To mitigate the risk of errors in AI-generated queries, an exhaustive training process is required. However, the relatively simple database schema and limited query complexity help reduce the potential for significant errors. This system offers a practical tool for both personal health monitoring and research applications, with future work aimed at improving automation, resilience to external changes, and data query accuracy.