Assessing driving behavior upstream of work zones by detecting response points in speed profile: A naturalistic driving study

Abstract

Objective: To assess driving behavior when presented with different work zone features such as signs. Driving behavior upstream of a work zone is of interest since this is the point where drivers need to slow or react to upcoming conflicts such as lane closures, congested traffic, or presence of workers and equipment.

Methods: Using 299 time series traces from four-lane roadways with both shoulder and lane closure scenario from the naturalistic driving study, this study analyzed driving behavior by detecting response points within the advance warning area from the first sign to the start of work zone. Response point was defined as the point where drivers reduced speed by a certain threshold. A mixed effect logistic model was developed to assess the relationship between driver response and work zone characteristics.

Results: Results indicated the first work zone sign a driver encountered in the immediate area upstream of the work zone was not significantly likely to elicit a driver response. The model found lane ends, speed limit, and active changeable message signs (CMS) as statistically significant. Since more than one sign can be legible to the driver at the same time, the effect of overlapping signs was evaluated, but was not found to have significant effect on the driver response. In general, drivers were more likely to show a response to the signs the closer they got to the start of the work zone. Static work zone speed limit and dynamic speed feedback signs were both found to be more likely to elicit a response as compared to normal speed limit signs (non-work zone related). Drivers who were traveling over the posted speed limit were more likely to show response at any given work zone signs with the exception of the first sign. In addition, driver distraction, and driver information like age, gender, experience and other environmental factors were not found to be significant in the model.

Conclusions: Overall, drivers were likely to show a response at the lane ends, work zone speed limit signs, and speed feedback signs.

Keywords:

• Work zone signs
• response point
• speed
• advance warning area
• legibility distance

Acknowledgments

The authors would like to thank Minnesota Department of Transportation for funding this research, the Virginia Tech Transportation Institute for their assistance in obtaining the SHRP2 data, and the Center for Transportation Research and Education (CTRE) at Iowa State University for providing the RID data.