Continuous Alertness Assessments
Benjamin Tag, Andrew Vargo, Aman Gupta, George Chernyshov, Kai Kunze, Tilman Dingler
Abstract
Open-access reader
Benjamin Tag, Andrew Vargo, Aman Gupta, George Chernyshov, Kai Kunze, Tilman Dingler
Abstract
Open-access reader
As the day progresses, cognitive functions are subject to fluctuations. While the circadian process results in diurnal peaks and drops, the homeostatic process manifests itself in a steady decline of alertness across the day. Awareness of these changes allows the design of proactive recommender and warning systems, which encourage demanding tasks during periods of high alertness and flag accident-prone activities in low alertness states. In contrast to conventional alertness assessments, which are often limited to lab conditions, bulky hardware, or interruptive self-assessments, we base our approach on eye blink frequency data known to directly relate to fatigue levels. Using electrooculography sensors integrated into regular glasses' frames, we recorded the eye movements of 16 participants over the course of two weeks in-the-wild and built a robust model of diurnal alertness changes. Our proposed method allows for unobtrusive and continuous monitoring of alertness levels throughout the day.
OpenAlex reports 61 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
As the day progresses, cognitive functions are subject to fluctuations. While the circadian process results in diurnal peaks and drops, the homeostatic process manifests itself in a steady decline of alertness across the day. Awareness of these changes allows the design of proactive recommender and warning systems, which encourage demanding tasks during periods of high alertness and flag accident-prone activities in low alertness states. In contrast to conventional alertness assessments, which are often limited to lab conditions, bulky hardware, or interruptive self-assessments, we base our approach on eye blink frequency data known to directly relate to fatigue levels. Using electrooculography sensors integrated into regular glasses' frames, we recorded the eye movements of 16 participants over the course of two weeks in-the-wild and built a robust model of diurnal alertness changes. Our proposed method allows for unobtrusive and continuous monitoring of alertness levels throughout the day.
Key concepts: Alertness, Computer science, Electrooculography, Cognition, Circadian rhythm, Audiology, Psychology, Cognitive psychology