Limits...
Novel measure of driver and vehicle interaction demonstrates transient changes related to alerting.

Brooks JR, Kerick SE, McDowell K - J Mot Behav (2014)

Bottom Line: Here we develop a measure that examines how drivers turn the steering wheel relative to heading error velocity, which the authors call the relative steering wheel compensation (RSWC).In contrast, more traditional variables are dynamic over longer time scales consistent with previous research.The results suggest drivers alter their behavioral output (steering wheel correction) relative to sensory input (vehicle heading error velocity) on a distinct temporal scale and may reflect an interaction of alerting and control.

View Article: PubMed Central - PubMed

Affiliation: a Army Research Laboratory, Human Research and Engineering Directorate, Aberdeen Proving Ground , Maryland.

ABSTRACT
Driver behavior and vehicle-road kinematics have been shown to change over prolonged periods of driving; however, the interaction between these two indices has not been examined. Here we develop a measure that examines how drivers turn the steering wheel relative to heading error velocity, which the authors call the relative steering wheel compensation (RSWC). The RSWC transiently changes on a short time scale coincident with a verbal query embedded within the study paradigm. In contrast, more traditional variables are dynamic over longer time scales consistent with previous research. The results suggest drivers alter their behavioral output (steering wheel correction) relative to sensory input (vehicle heading error velocity) on a distinct temporal scale and may reflect an interaction of alerting and control.

Show MeSH

Related in: MedlinePlus

Example regression fits for single subjects across in each time interval. Points in each block represent a single trial within that 5-min time interval. The slope of these regression lines was our independent variable, relative steering wheel compensation. dHE = derivative of heading error.
© Copyright Policy - open-access
Related In: Results  -  Collection


getmorefigures.php?uid=PMC4376238&req=5

f0003: Example regression fits for single subjects across in each time interval. Points in each block represent a single trial within that 5-min time interval. The slope of these regression lines was our independent variable, relative steering wheel compensation. dHE = derivative of heading error.

Mentions: We applied regression analyses between SWRM and dHE/HE at the time of response to better understand the relationship between measures of vehicle-environment kinematics and driving behavior. As the values that went into this comparison were the absolute value and nonnormally distributed, we utilized a nonparametric correlation measure to ascertain the degree of relatedness between the HE or dHE and SWRM. The average Spearman's correlation coefficient between the peak SWRM and the dHE and the HE, across all subjects was 0.59 and 0.35 respectively. For each subject independently, the correlation between SWRM and dHE was significant while for only two subjects the correlation between HE and SWRM was significant. Across all subjects, the larger correlation between the SWRM and dHE was confirmed statistically by performing a one-tailed paired t test on the Fisher's Z transformed correlation coefficients (p < .003). We extended this analysis by also comparing the root mean squared error (RMSE) for these regressions. The regressions were fit for each 5-min interval in the experiment and utilizing a paired, one-sided t test we found a significantly larger RMSE for the regressions between the SWRM and the HE than with the dHE (p < .001). From these analyses we were confident that the dHE in our experiment was more closely related to the SWRM than the HE. This finding was important because it provided confirmation that we should focus subsequent analyses on the most meaningful relationship between vehicle kinematics and driver responses. Consequently, we focused on the relation between SWRM as a linear function of the dHE. In FigureĀ 3 we show an example subject's regression fit between the SWRM and dHE for all nine 5-min blocks over the 45-min drive.FIGURE 3.


Novel measure of driver and vehicle interaction demonstrates transient changes related to alerting.

Brooks JR, Kerick SE, McDowell K - J Mot Behav (2014)

Example regression fits for single subjects across in each time interval. Points in each block represent a single trial within that 5-min time interval. The slope of these regression lines was our independent variable, relative steering wheel compensation. dHE = derivative of heading error.
© Copyright Policy - open-access
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC4376238&req=5

f0003: Example regression fits for single subjects across in each time interval. Points in each block represent a single trial within that 5-min time interval. The slope of these regression lines was our independent variable, relative steering wheel compensation. dHE = derivative of heading error.
Mentions: We applied regression analyses between SWRM and dHE/HE at the time of response to better understand the relationship between measures of vehicle-environment kinematics and driving behavior. As the values that went into this comparison were the absolute value and nonnormally distributed, we utilized a nonparametric correlation measure to ascertain the degree of relatedness between the HE or dHE and SWRM. The average Spearman's correlation coefficient between the peak SWRM and the dHE and the HE, across all subjects was 0.59 and 0.35 respectively. For each subject independently, the correlation between SWRM and dHE was significant while for only two subjects the correlation between HE and SWRM was significant. Across all subjects, the larger correlation between the SWRM and dHE was confirmed statistically by performing a one-tailed paired t test on the Fisher's Z transformed correlation coefficients (p < .003). We extended this analysis by also comparing the root mean squared error (RMSE) for these regressions. The regressions were fit for each 5-min interval in the experiment and utilizing a paired, one-sided t test we found a significantly larger RMSE for the regressions between the SWRM and the HE than with the dHE (p < .001). From these analyses we were confident that the dHE in our experiment was more closely related to the SWRM than the HE. This finding was important because it provided confirmation that we should focus subsequent analyses on the most meaningful relationship between vehicle kinematics and driver responses. Consequently, we focused on the relation between SWRM as a linear function of the dHE. In FigureĀ 3 we show an example subject's regression fit between the SWRM and dHE for all nine 5-min blocks over the 45-min drive.FIGURE 3.

Bottom Line: Here we develop a measure that examines how drivers turn the steering wheel relative to heading error velocity, which the authors call the relative steering wheel compensation (RSWC).In contrast, more traditional variables are dynamic over longer time scales consistent with previous research.The results suggest drivers alter their behavioral output (steering wheel correction) relative to sensory input (vehicle heading error velocity) on a distinct temporal scale and may reflect an interaction of alerting and control.

View Article: PubMed Central - PubMed

Affiliation: a Army Research Laboratory, Human Research and Engineering Directorate, Aberdeen Proving Ground , Maryland.

ABSTRACT
Driver behavior and vehicle-road kinematics have been shown to change over prolonged periods of driving; however, the interaction between these two indices has not been examined. Here we develop a measure that examines how drivers turn the steering wheel relative to heading error velocity, which the authors call the relative steering wheel compensation (RSWC). The RSWC transiently changes on a short time scale coincident with a verbal query embedded within the study paradigm. In contrast, more traditional variables are dynamic over longer time scales consistent with previous research. The results suggest drivers alter their behavioral output (steering wheel correction) relative to sensory input (vehicle heading error velocity) on a distinct temporal scale and may reflect an interaction of alerting and control.

Show MeSH
Related in: MedlinePlus