Abstract:
Heart Rate (HR) and Heart Rate Variability (HRV) are indications of many physiological and psychological phenomena such as heart-related diseases, stress and emotional arousal. Therefore, continuous monitoring of HR and HRV is crucial. Although commercially available smart wearables are equipped with Photoplethysmography (PPG) sensors to measure HR and HRV, higher consumption of energy by PPG sensor limits the capability of continuous monitoring for an extended period. This thesis introduces CompRate, a method that enables energy efficient continuous monitoring of physiological signs only using the built-in accelerometer signals of smart wearables. The first part of the thesis describes the stages of building the methodology, evaluation and validation. In particular, data from 12 participants were used to train machine learning (ML) models to predict HR and IBI. Generalisability and cross-device compatibility of ML models were evaluated. The second part of the thesis focuses on the design and implementation of the proof of concept applications with CompRate. Inspired by prior works in the area of stress and student engagement, the thesis demonstrates three proof-of-concept applications using CompRate. 1) self-awareness of fatigue and just-in-time interruption while driving; 2) enabling teachers to be aware of students’ mental effort during a learning activity, and 3) the broadcasting of the location of live victims in a disaster situation.