Patient Indoor Positioning: Leveraging Wearable Technology for Location Awareness

The accurate determination of the position of individuals and objects in indoor environments represents a significant challenge, given the ineffectiveness of traditional Global Navigation Satellite Systems (GNSS) in such contexts. This thesis addresses this problem by designing, implementing, and experimentally evaluating an indoor positioning system (IPS) based on Bluetooth Low Energy (BLE) technology, ESP32 microcontrollers, and wearable tags. The main objective is to provide a reliable, precise, and energy-efficient solution for real-time tracking, with particular attention to applications in healthcare. The developed system adopts a hierarchical architecture that integrates wearable BLE tags (specifically Global Tag Disk Beacons) as mobile transmitters and a network of ESP32 microcontrollers as intelligent scanning/receiving stations. Communication is managed via MQTT protocols for data exchange between ESP32s and the backend, WebSocket for real-time updates to the frontend, and RESTful APIs for resource management. Positioning algorithms are based on RSSI-weighted multilateration techniques, with the integration of advanced filtering techniques such as Exponential Moving Average (EMA) and Kalman filter to improve the accuracy and robustness of the system in dynamic environments. The research methodology included a thorough literature review, system design, firmware development, software development and a rigorous experimental validation phase conducted in a controlled 90 m² domestic environment. The results demonstrated that the system achieved room-level positioning accuracy with an average error of 1.32 meters. The system showed exceptional energy efficiency, allowing BLE tags a battery life of over six months with a 1000 ms advertising interval and remarkable operational stability, with zero crashes recorded during two weeks of continuous testing. The average zone transition latency was 25 seconds, with an 8% false alarm rate. This research contributes a practical and low-cost indoor positioning solution that effectively balances the requirements of accuracy, energy efficiency, and implementation complexity. Although the system has some limitations in highly mobile scenarios and a sensitivity to electromagnetic interference, its robustness and energy efficiency make it particularly suitable for continuous monitoring applications. Future directions include the integration of inertial sensors to further reduce positioning error, the development of adaptive propagation models based on machine learning, and the exploration of hybrid architectures for greater scalability and precision.