Performance of basic kinematic thresholds in the identification of crash and near-crash events within naturalistic driving data

Pérez, Miguel Á.; Sudweeks, Jeremy; Sears, Edie; Antin, Jonathan F.; Lee, Suzanne; Hankey, Jonathan M.; Dingus, Thomas A.

doi:10.1016/j.aap.2017.03.005

Cited by 73 publications

(29 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Short-term driving events (e.g., braking, accelerating, and turning) constitute a major bulk of vehicle control. Naturalistic studies demonstrated that high frequency of intense driving events such as hard braking and sharp turning, in which a kinematic threshold (e.g., speed, acceleration) had been breached, is linked to the risk of crash involvement (e.g., Perez et al, 2017;Toledo, Musicant, & Lotan, 2008). The intensity of driving events was therefore adopted as a risk index in pay-as-you-drive insurance models (Tselentis, Yannis, & Vlahogianni, 2016) and driver training programs (Farah et al, 2014, Musicant & Lampel, 2010Musicant & Lotan, 2016).…”

Section: Introductionmentioning

confidence: 99%

Relationship Between Kinematic and Physiological Indices During Braking Events of Different Intensities

Musicant¹,

Botzer²,

Laufer

et al. 2018

Hum Factors

View full text Add to dashboard Cite

Objective To study the relationship between physiological indices and kinematic indices during braking events of different intensities. Background Based on mental workload theory, driving and other task demands may generate changes in physiological indices, such as the driver's heart rate and skin conductance. However, no attempts were made to associate changes in physiological indices with changes in vehicle kinematics that result from the driver attempts to meet task demands. Method Twenty-five drivers participated in a field experiment. We manipulated braking demands using roadside signs to communicate the speed (km/h) before braking (50 or 60) and the target speed for braking (30 or to a complete stop). In an additional session, we asked drivers to brake as if they were responding to an impending collision. We analyzed the relationship between the intensities of braking events as measured by deceleration values (g) and changes in heart rate, heart rate variability, and skin conductance. Results All physiological indices were associated with deceleration intensity. Especially salient were the differences in physiological indices between the intensive (|g| > 0.5) and nonintensive braking events. The strongest relationship was between braking intensity and skin conductance. Conclusions Skin conductance, heart rate, and heart rate variability can mirror the mental workload elicited by varying braking intensities. Application Associating vehicle kinematics with physiological indices related to short-term driving events may help improve the performance of driver assistance systems.

show abstract

Section: Introductionmentioning

confidence: 99%

Relationship Between Kinematic and Physiological Indices During Braking Events of Different Intensities

Musicant¹,

Botzer²,

Laufer

et al. 2018

Hum Factors

View full text Add to dashboard Cite

show abstract

“…It is also possible that the difference in magnitude seen in this study compared to others, like 100-Car, is because different triggers and kinematic thresholds were used. The kinematic thresholds typically use a combination of parameters related to vehicle speed, lateral acceleration, longitudinal acceleration, steering changes, and yaw rate changes (36). For example, the 100-Car study used a trigger threshold of ≥0.6 g for longitudinal acceleration whereas SHRP2 used ≥0.75 g so there were more candidate events found in 100-Car.…”

Section: Discussionmentioning

confidence: 99%

Distracted Driving and Risk of Crash or Near-Crash Involvement Among Older Drivers Using Naturalistic Driving Data With a Case-Crossover Study Design

Huisingh

Owsley

Levitan

et al. 2018

The Journals of Gerontology: Series A

View full text Add to dashboard Cite

show abstract

“…The developed model was successful in performing a risk assessment for vehicles as well as for roadway segments. SHRP2 NDS data were also used for crash and near-crash identification by establishing kinematic thresholds ( 38 ). Vehicle kinematics such as longitudinal deceleration and acceleration, freeway deceleration, lateral acceleration, swerve, yaw rate, and longitudinal and lateral jerk, among others, were thoroughly analyzed to determine threshold values at which crashes and near-crashes can be identified.…”

Section: Literature Reviewmentioning

confidence: 99%

Development and Comparative Analysis of Advanced Deep Learning Techniques for Crash Prediction in Advanced Driver Support Systems

Osman

Hajij

2021

Transportation Research Record

View full text Add to dashboard Cite

Motor vehicle crashes claimed 38,800 lives and caused 4.4 million injuries in 2019 alone. Studies have shown that 94% of these crashes are because of driver errors. Such a huge contribution of driver errors to crashes indicates that efforts at improving safety should be directed toward both vehicles and drivers through advanced driver assistance systems (ADAS) and vehicular technologies. This study investigates the potential that real-time driver behavior data collected through vehicular technologies offer to predict crashes, as the first line of defense to avoid them. Three deep learning models were developed including multilayer perceptron neural networks (MLP-NN), long-short-term memory networks (LSTMN), and convolutional neural networks (CNN) using vehicle kinematics time series data extracted from the Second Strategic Highway Research Program Naturalistic Driving Study (SHRP2 NDS) dataset. The study builds on the hypothesis that crashes are preceded by turbulences that take place over time (turbulence horizon). If these turbulences are detected promptly they can help predict and avoid crashes. Several values were tested for the turbulence horizon and the prediction horizon (how long before the crash impact it can be predicted) to identify the optimal values. The results showed that the CNN model can predict all crashes with a 100% accuracy and zero false alarms 3 s before the crash impact time when a 6-s turbulence horizon is used. This outstanding performance demonstrates the developed model is a promising tool for implementation in ADAS.

show abstract

Performance of basic kinematic thresholds in the identification of crash and near-crash events within naturalistic driving data

Cited by 73 publications

References 14 publications

Relationship Between Kinematic and Physiological Indices During Braking Events of Different Intensities

Relationship Between Kinematic and Physiological Indices During Braking Events of Different Intensities

Distracted Driving and Risk of Crash or Near-Crash Involvement Among Older Drivers Using Naturalistic Driving Data With a Case-Crossover Study Design

Development and Comparative Analysis of Advanced Deep Learning Techniques for Crash Prediction in Advanced Driver Support Systems

Contact Info

Product

Resources

About