Search

The criticality of a rear end event depends on the brake reaction time (BRT) of the driver. Therefore, distracted driving poses greater threat in such events. Evidence accumulation model (EAM) that uses looming of the lead vehicle as main stimuli has shown significant success in estimating drivers’ BR Ts. It is often argued that drivers collect evidence for braking through peripheral vision, especially during off-road glances, and transition to forward. In this work, we have modeled evidence accumulation as a function of gaze eccentricity for off-road glances while approaching safety critical events. The model is tested with real world crash and near crash event data from SHRP2 naturalistic study. Our model shows that linear relation between gaze eccentricity and evidence accumulation rate during off road glances helps to improve EAM estimation in predicting BRT. We have also shown that brake-light onset does not influence EAM in presence of active looming.

Distractions have been recognised as one important factor associated with pedestrian injuries, as the increasing use of cell phones and personal devices. However, the situation is less clear regarding the differences in the effects of visual-manual and auditory-cognitive distractions. Here, we investigated distracted pedestrians in a one-lane road with continuous traffic using an immersive CAVE-based simulator. Sixty participants were recruited to complete a crossing task and perform one of two distractions, a visual-manual task and an auditory-cognitive task. Moreover, normal and time pressure crossing conditions were included as a baseline and comparison. For the first time, this study directly compared the impacts of visual-manual, auditory-cognitive distractions, and time pressure on pedestrian crossing behaviour and safety in a controlled environment. The results indicated that although pedestrian safety was compromised under both types of distraction, the effects of the applied distractions were different. Whenengaged in the visual-manual distraction, participantscrossed the road slowly, but there was no significant difference in gap acceptance or initiation time compared to baseline. In contrast, participants walked slowly, crossed earlier, and accepted smaller gaps when performing the auditory-cognitive distraction. This has interesting parallels to existing findings on how these two types of distractions affect driver performance. Moreover, the effects of the visual-manual distraction were found to be dynamic, as these effects were affected by the gap size. Finally, compared to baseline, time pressure resulted in participants accepting smaller time gaps with shorter initiation times and crossing durations, leading to an increase in unsafe decisions and a decrease in near-collisions. These results provide new evidence that two types of distraction and time pressure impair pedestrian safety, but in different ways. Our findings may provide insights for further studies involving pedestrians with different distraction components.

To download, select "OSF Storage" below (or some specific item under it) and then click "Download as zip". For more information, please see the README pdfs in each of the two folders "Analysis software" and "Data".

Drivers will be steered round a series of simulated bends. Occasionally a bias will be introduced into the automated trajectory that causes the vehicle to eventually exceed the lane boundaries. The nature of the steering and gaze response to these silent failure will be investigated in the presence or absence of a non-driving related auditory count task of varying levels of difficulty.

A short report series describing a set of pilot experiments conducted within project TRANSITION. We have also uploaded the data and scripts used for the reports.

See README pdf for documentation, and for link to the research data for the paper.

See the README pdf for documentation.

For documentation see the README pdf file.

Simulation of road crossing decision-making models developed in the EU H2020 interACT project.

This set of MATLAB scripts and functions provides an example implementation of a type of visual looming accumulator model of driver brake onset that I and co-authors have used and fitted to data in a number of publications.

Mental health problems among children are prevalent, often found within the socioemotional domains, and tend to persist over time. Early detection could predict the need for interventions for children at risk. This study aimed to study the role of internalizing problems and engagement in a preschool setting with respect to the trajectory of several functional outcomes. Using preschool teacher estimations of a representative cohort of 845 Swedish preschool children (438 boys and 407 girls, age range 13–71 months), this study constructed risk groups based on the emotional subscale of the Strengths and Difficulties Questionnaire (SDQ) and the engagement scale derived from the Children’s Engagement Questionnaire (CEQ) at baseline. These risk groups were associated with the trajectory of WHO International Classification of Functioning, Disability and Health (ICF) (p

To steer a vehicle, humans must process incoming signals that provide information about their movement through the world. These signals are used to inform motor control responses that are appropriately timed and of the correct magnitude. However, the perceptual mechanisms determining how drivers process visual information remain unclear. Previous research has demonstrated that when steering toward a straight road-line, drivers accumulate perceptual evidence (error) over time to initiate steering action (Accumulator framework), rather than waiting for perceptual evidence to surpass time-independent fixed thresholds (Threshold framework). The more general case of steering around bends (with a requirement that the trajectory is adjusted to match the road curvature ahead) provides richer continuously varying information. The current experiment aims to establish whether the Accumulator framework provides a good description of human responses when steering toward curved road-lines. Using a computer-generated steering correction paradigm, drivers (N=11) steered toward intermittently appearing curved road-lines that varied in position and radius with respect to the driver’s trajectory. A Threshold framework predicted that steering responses would be of fixed magnitude and at fixed absolute errors across conditions regardless of the rate of error development. Conversely, the Accumulator framework predicted that drivers should respond to larger absolute errors when the error signal developed at a faster rate. Results were consistent with an Accumulator framework in a manner that supports previous investigations and the computational modelling literature. We propose that accumulation of perceptual evidence captures human behaviour in a variety of steering contexts that drivers face in the real world.