U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information|
|Publication Number: FHWA-HRT-11-040 Date: November 2012|
Publication Number: FHWA-HRT-11-040
Date: November 2012
The work zone alerts application area is intended to warn drivers of changes in traffic patterns due to construction (e.g., lane closures). It would be targeted to crashes in and approaching work zones. The application area is similar to the work zone warning for reduced speed in work zones but would not be limited to crashes involving speeding.
The work zone alerts application area targets crashes in and approaching work zones, including any crash occurring in or related to a work zone as identified by a work zone variable in the NASS GES database. The distribution of pre-crash scenarios indicates that several pre-crash scenarios are common in work zone related crashes. The pre-crash scenarios road edge departure/no maneuver (28 percent), control loss/no vehicle action (19 percent), and object contacted/no maneuver (14 percent) represented the largest proportions of single-vehicle work zone crashes. For multi-vehicle crashes, the leading pre-crash scenarios were rear-end/LVS (25 percent), changing lanes/same direction (16 percent), and rear-end/LVD (14 percent).
Based on weighted NASS GES data, there were an estimated 86,611 annual national target crashes, with an estimated total annual cost of more than $4.5 billion. Additionally, 29 percent of all the target crashes resulted in fatalities or injuries.
The NASS GES database provides information on the relation of the crash location to the work zone. According to the data, the majority of single-vehicle and multi-vehicle work zone crashes occurred in work or construction zones. Only 2 percent of the single-vehicle crashes and 5 percent of the multi-vehicle crashes did not occur in the work zone, but the first harmful event was related to the work zone. However, the location of a large proportion (43 percent) of work zone-related crashes (i.e., crashes in the work zone or on the approach) was not identified in the database.
The distribution of target crashes by the six vehicle-type categories was considered for both single-vehicle and multi-vehicle crashes, with the latter referring to the first two vehicles involved. While the majority of the crashes (86 percent) involved light vehicles, trucks (single- and combination-unit) represented 8 percent of both single-vehicle and multi-vehicle crashes. Motorcycles accounted for more than 3 percent of single-vehicle target crashes, which was higher than the percentage of motorcycle crashes for all single-vehicle crashes (1.4 percent), indicating that motorcycles were overrepresented in work zone crashes.
The distribution of target crashes by area type was considered for both single-vehicle and multi-vehicle crashes. The number of work zone crashes was greater in urban areas for both single-vehicle crashes (54 percent) and multi-vehicle crashes (63 percent) when compared to those in rural areas.
The distribution of target crashes by posted speed limit was considered for both single-vehicle and multi-vehicle crashes. It is unknown whether the data indicate the normal posted speed limits or the reduced work zone speed limits. A greater proportion of single-vehicle crashes occurred on higher speed limit roads (55 mi/h or greater) and low speed limit roads (25 mi/h or less) when compared to multi-vehicle crashes. At moderate speed limits (30-50 mi/h), multi-vehicle crashes represented a larger proportion of work zone crashes than single-vehicle crashes.
The distribution of target crashes by the number of lanes was considered for both single-vehicle and multi-vehicle crashes. The majority of work zone-related crashes occurred on roadways with three lanes or less (79 percent). Single-vehicle crashes represented a larger proportion of crashes on roadways with fewer lanes compared with multi-vehicle crashes.
The infrastructure pedestrian detection application area is intended to detect pedestrians and allows for changes in the walk phase of a signal (e.g., to allow more time for those with mobility impairments to cross). This application targets pedestrian crashes at signalized crossings including intersections and midblock crossings.
This application area targets single-vehicle and multi-vehicle pedestrian/maneuver and pedestrian/no maneuver pre-crash scenarios. Target crashes were limited to crashes that occurred at intersections and midblock locations with signalized control.
Based on weighted NASS GES data, there were an estimated 17,811 annual national target crashes, with an estimated total annual cost of more than $3.3 billion. Additionally, 99 percent of all the target crashes resulted in fatalities or injuries. Note that the estimated target is based on 506 observed crashes. As a result, it is less reliable than other estimates due to the small sample size.
The relevant distribution of the single-vehicle target crashes is presented in this section. Only single-vehicle target crashes are included because they constitute the majority of the target crashes. The majority of single-vehicle target crashes occurred in urban areas (86 percent) compared to rural areas (14 percent).
According to the relationship to junction description in NASS GES, the target crashes mainly occurred at intersections (48 percent) or were intersection-related (51 percent).(1) According to non-motorist location description in NASS GES, almost half of pedestrian crashes occurred in a crosswalk (49 percent) or in the roadway (43 percent) at the intersection.(1) Crashes at mid-block crossings, (i.e., not located at intersections) represented less than 7 percent of target crashes. It is likely that some of the crashes described as occurring in the intersection or intersection-related may have been signalized midblock crossings that were considered intersections. As a result, midblock crashes may be an underrepresented crash typology.
The distribution of crash types shows that more than half (53 percent) occurred between a pedestrian and a turning or merging vehicle. Light vehicles were involved in the majority (92 percent) of crashes. Only 2 percent of crashes involved transit vehicles, less than 2 percent involved trucks, and 4 percent of crashes involved other vehicles.
The distribution of lighting conditions for target crashes shows that although the majority of crashes occurred during daylight (65 percent), approximately 30 percent occurred during dark conditions (both dark (3 percent) and dark but lighted conditions (27 percent)).
Target crashes occurred most often on moderate speed limit roadways. A total of 20 percent were on roadways posted at 25 mi/h or less, more than 60 percent of crashes were roadways posted 30-35 mi/h, and 19 percent were roadways posted 40-45 mi/h.
Almost half of target crashes (48 percent) occurred on roads with six or more lanes, while less than one-fourth of the crashes were on one- or two-lane roads. This distribution reflects the increased potential for pedestrian crashes along wider, more vehicle-oriented roadways.
This application area is intended to warn drivers of approaching trains, including light and heavy rail trains, for at-grade crossings.
The at-grade rail crossing application area targets single-vehicle and multi-vehicle crashes involving rail vehicles by providing warning of an approaching rail vehicle. This application is not limited by pre-crash scenario. Target crashes for this application include any crashes involving one or more vehicles and a train.
Based on weighted NASS GES data, there were an estimated 1,314 annual national target crashes, with an estimated total annual cost of more than $653 million. Additionally, 41 percent of all the target crashes resulted in fatalities or injuries. Note that the estimated target is based on 23 observed crashes. As a result, it is less reliable than other estimates. All but one of the observed crashes was a single-vehicle crash with a train.
The distribution of traffic control type indicates that single-vehicle target crashes were split nearly evenly between active traffic control devices with gates, flashing lights, or traffic signals (50 percent) and passive devices including stop signs and crossbucks (46 percent). Few target crashes (less than 4 percent total) occurred at rail crossings without control or with other traffic control types.
The distribution of the six vehicle-type categories was considered for single-vehicle target crashes. While the majority of target crashes involved single vehicles (79 percent), 21 percent of crashes involved trucks, including single- and combination-unit trucks.
The distribution of the number of lanes for single-vehicle target crashes indicated that more than 80 percent of target crashes occurred on roadways with two or less lanes. Additionally, almost 90 percent of crashes occurred during favorable weather, while 8 percent of crashes occurred during rainy conditions. Finally, the distribution of land showed that the majority of single-vehicle crashes were in urban areas (80 percent), while only 20 percent occurred in rural areas.