Context-Aware Multisensory Monitoring for Aging-in-Place Applications: An MQTT-Based Indoor Positioning System

Abstract

Background: The aging population in Canada and globally faces numerous chronic and neurodegenerative conditions, with limited mobility being a significant predictor. Collecting longitudinal mobility data presents a potential solution, aiding in the proactive diagnosis and management of related health issues. However, current wearable and IoT-based solutions encounter adoption barriers due to usability, data accuracy, and network reliability. Additionally, the digital divide in healthcare impacts older adults, limiting their access to health-monitoring technologies.

Objective: This thesis aims to design a real-time, sensor-based indoor positioning and health monitoring system that is efficient, scalable, and user-friendly, especially for older adults. The design emphasizes reliable data acquisition and transmission, as well as key considerations for stakeholders, including older adults, their caregivers, and healthcare teams.

Methods: The system comprises a smartwatch emitting a BLE signal and equipped with sensors to collect physiological data, and ESP32-based beacons that gather ambient sensing data and the BLE signal from the smartwatch. MQTT is used as the data transmission protocol from beacon to Raspberry Pi-based hub. Key system optimizations include data transmission frequency tuning, epoch-based timestamp synchronization, and load-balancing to reduce network congestion. The system has been evaluated against performance metrics including data accuracy, latency, and scalability tested under different sensor loads.

Results: The system proves feasible for real-time health monitoring, demonstrating performance within requirements in a short time frame for minimal data loss, time synchronization, and network stability. Challenges include long-term performance pertaining to memory and load management on the ESP32 as well as BLE emission configurations on the smartwatch for RSSI consistency.

Conclusion: This system offers a real-time, low-latency health monitoring framework that addresses usability and efficiency. Contributions include increased accessibility for stakeholders and reliability in data acquisition and storage. Future work will focus on refining the system’s long-term performance.