Understanding high-speed curve negotiation: the interplay between driving expertise and cognitive load

Based on the current literature in driver behaviour and skill acquisition, experience is considered a key factor in the development of driving skills, shaping both vehicle control and cognitive abilities. However, the sheer amount of driving experience is not a sufficient predictor of skill, as drivers commonly accumulate large amounts of experience through repeated exposure to highly similar situations (e.g., routine commuting on familiar routes), rather than a diverse range of qualitatively different driving contexts. This does not appear to be the case for formally trained expert drivers, whose expertise emerges from prolonged periods of structured learning and extensive deliberate practice across diverse conditions.
Nevertheless, both driving experience and expertise remain relatively under-explored, as many of the proposed relationships are grounded in theory rather than supported by empirical evidence. One such assumption is that highly skilled drivers should incur smaller performance decrements when engaged in secondary, non-driving tasks that compete for cognitive resources and increase drivers’ cognitive load. Yet, the role of driving expertise in moderating the effects of cognitive load on driving performance has received little scientific attention. Even less is known about the mechanisms through which expertise might confer an advantage across both baseline and cognitively demanding conditions, whether by enabling more efficient allocation of attention, superior anticipation of upcoming road, or reliance on compensatory strategies that mitigate the effects of this type of distraction. Furthermore, it remains unclear whether expertise alone is sufficient to buffer drivers against cognitive load, or whether external factors, such as environmental cues and road design, play an equally important role in supporting safe performance under such demanding conditions.
Addressing these research gaps is crucial for a comprehensive understanding of skilled performance in driving and the roles of both experience and expertise in challenging driving situations. The aim of this thesis, therefore, was to explore differences between expert and experienced drivers in sharp curves and to investigate whether greater driver skill can buffer the adverse effects of cognitive load on driving performance and gaze patterns. A combination of real-world driving and controlled laboratory approach was employed to address these aims.
The real-world study compared non-expert and expert racing drivers as they drove on a racetrack while simultaneously performing cognitively demanding secondary tasks. Overall, non-expert drivers were not more vulnerable to performance decrements due to cognitive load than expert racing drivers, although the relationship between driving expertise and cognitive load appeared to be task-dependent. Moreover, the effects of cognitive load varied by curve type, being most pronounced in sharp hairpin curves, highlighting curve geometry as a factor requiring greater attention in future research.
The driving simulator study built upon this work by further examining the performance and gaze patterns of expert police drivers and experienced non-expert drivers when approaching and negotiating sharp hairpin curves with and without cognitive load. A curve preview was introduced to test whether such an intervention could aid drivers’ performance under cognitive load. The findings revealed that differences between expert and non-expert drivers were evident primarily in longitudinal vehicle control, whereas cognitive load exerted its strongest influence on lateral control. Curve preview was shown to support anticipatory behaviour among expert drivers and to improve steering performance in curves regardless of cognitive load. In terms of gaze behaviour, experts and non-experts adopted different look-ahead fixation strategies during curve negotiation; however, cognitive load disrupted these patterns and narrowed drivers’ horizontal gaze distribution. Importantly, curve preview mitigated these negative effects by broadening gaze dispersion in both groups. Together, these findings highlight the task-specific nature of driving expertise while also suggesting that anticipatory cues, such as curve preview, can play a crucial role in offsetting the detrimental effects of cognitive load.
Overall, this thesis provides a comprehensive behavioural and physiological account of the relationship between driving expertise and cognitive load. This work contributes to research within the automotive domain by informing the development of systems designed to monitor drivers’ state and to support their perceptual and motor performance when concurrent cognitive activities tax their resources.