University of Glasgow
Enhancing engineering education through wearable technology: a focus on eye-tracking and multisensory approaches
Abstract
dc:description.abstractThis doctoral research investigates the efficacy of wearable technology with a multisensory approach, particularly eye-tracking and physiological sensors, in capturing and interpreting cognitive engagement among engineering students in higher education. Responding to the growing interest in data-driven, personalised learning, the study develops and validates a multisensory framework to detect attention and mind-wandering during video-based instruction. This study systematically analysed different wearable devices and first identified head-mounted eye-trackers as the most promising tools for monitoring visual attention in educational settings. To evaluate their practicality, a comparative study was conducted using a commercial wearable eye-tracker (Pupil Core) and a custom-built desktop-based solution. The wearable eye-tracker provided greater flexibility for natural head movement and enabled real-time detection of visual attention patterns. It also helped identify segments of the learning material that were skipped or overlooked, offering insights into content that learners found confusing or cognitively demanding. The commercial device outperformed the desktop-based system in both usability and richness of data, validating its utility in dynamic learning environments. To address the limitations of eye-tracking in detecting internal cognitive states, the research implemented a multimodal sensing system by integrating galvanic skin response (GSR) and photoplethysmography (PPG) sensors. Data were collected during learning sessions, and supervised machine-learning models were trained to classify episodes of mind-wandering. The multimodal sensor fusion achieved the highest accuracy of 89% , significantly outperforming unimodal baselines. The experiments in this thesis concluded that the proposed multisensory device, combining eye-tracking with physiological signals (PPG and GCR), provides a robust method for detecting cognitive disengagement in real-time. The outcomes have implications for developing adaptive educational technologies capable of personalising instruction based on learners’ cognitive states.
Degree
thesis:*- Level dc:type.qualificationlevel
- PhD
- Grantor dc:publisher.institution
- University of Glasgow
- Year dc:date.issued
- 2025
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Khosravi, Sara
Subjects
dc:subject × 1Rights
- Language dc:language
- en
Identifiers
dc:identifier.*- DOI dc:identifier.doi
- https://doi.org/10.5525/gla.thesis.85126