Prediction models for the risk of presenting aberrant driving behaviours and sleep-disordered breathing assessments in occupational drivers
File(s)
Author(s)
Tsai, Cheng-Yu
Type
Thesis
Abstract
Occupational drivers, e.g., bus and taxi operators, play a pivotal role in maintaining transportation efficiency. However, the emphasis on operational efficiencies often overshadows health challenges these drivers face, notably driver fatigue and sleep-disordered breathing (SDB). Such health challenges can escalate the risk of aberrant driving behaviours (ADBs), subsequently endangering both passengers and the wider public. The absence of comprehensive health protocols for managing fatigue and SDB exacerbated this risk. Consequently, this thesis aimed to elucidate the relationship between ADB, fatigue, and SDB, exploring intervention strategies and proposing models to predict the risk of presenting ADBs.
Firstly, to develop models for predicting the risk of presenting ADBs, this thesis utilised a variety of machine learning approaches, including cross-sectional methods, time-series methods, and hybrid models. The dataset for developing these models, collected from urban and highway bus drivers as well as taxi drivers, was predicated on physiological signals—specifically, heart rate variability (HRV)—and sleep disorder indices from preceding days. Furthermore, to investigate the feasibility of mitigating the risk of presenting ADBs through SDB intervention, this thesis analysed data derived from the implementation of various SDB intervention strategies over a three-month period. These strategies encompassed both conservative approaches, such as weight management, and curative treatments, like the use of a continuous positive airway pressure (CPAP) machine. This thesis conducted comparisons to assess changes in both objective metrics, such as driving behaviour and sleep quality, and subjective metrics, including outcomes from various questionnaires.
Regarding the findings, this thesis demonstrated the possibility of mitigating the risk of presenting ADBs through SDB interventions (i.e., using CPAP). Moreover, the hybrid models established, incorporating HRV metrics during driving and sleep disorder indices from the preceding day, demonstrated promise in predicting the risk of presenting ADBs among occupational drivers.
In conclusion, this thesis not only contributed to the existing body of knowledge but also provided actionable insights for the development of comprehensive health and safety awareness in the field of occupational driving for reducing the risk of presenting ADBs.
Firstly, to develop models for predicting the risk of presenting ADBs, this thesis utilised a variety of machine learning approaches, including cross-sectional methods, time-series methods, and hybrid models. The dataset for developing these models, collected from urban and highway bus drivers as well as taxi drivers, was predicated on physiological signals—specifically, heart rate variability (HRV)—and sleep disorder indices from preceding days. Furthermore, to investigate the feasibility of mitigating the risk of presenting ADBs through SDB intervention, this thesis analysed data derived from the implementation of various SDB intervention strategies over a three-month period. These strategies encompassed both conservative approaches, such as weight management, and curative treatments, like the use of a continuous positive airway pressure (CPAP) machine. This thesis conducted comparisons to assess changes in both objective metrics, such as driving behaviour and sleep quality, and subjective metrics, including outcomes from various questionnaires.
Regarding the findings, this thesis demonstrated the possibility of mitigating the risk of presenting ADBs through SDB interventions (i.e., using CPAP). Moreover, the hybrid models established, incorporating HRV metrics during driving and sleep disorder indices from the preceding day, demonstrated promise in predicting the risk of presenting ADBs among occupational drivers.
In conclusion, this thesis not only contributed to the existing body of knowledge but also provided actionable insights for the development of comprehensive health and safety awareness in the field of occupational driving for reducing the risk of presenting ADBs.
Version
Open Access
Date Issued
2023-09
Date Awarded
2023-12
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
Majumdar, Arnab
Publisher Department
Civil and Environmental Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)