Crash Data Analyses for Vehicle-To-Infrastructure Communications for Safety Applications

Chapter 8. Unaddressed Crashes

Crashes identified in the NASS GES data that are not targeted by one or more of the current V2I application areas mentioned previously are considered unaddressed crashes. These crashes can be explored to identify other potential applications. This chapter provides information on these crashes.

8.1 Distribution by Pre-crash Scenarios

Table 9 and table 10 present the distribution and cost information of total and unaddressed crashes from NASS GES data sorted by pre-crash scenario for single-vehicle and multi-vehicle crashes, which are presented separately. The second row of table 9 indicates that for vehicle failure, unaddressed crashes represent 8 percent of total unaddressed crash costs, and 98 percent of the total crash cost for this scenario remains unaddressed. Pedestrian/no maneuver is the leading unaddressed single-vehicle pre-crash scenario, followed by cyclist/no maneuver and animal/no maneuver. Collectively, these three pre-crash scenarios represent 56 percent of the costs of unaddressed single-vehicle crashes. Unaddressed pedestrian, bicyclist, and animal pre-crash scenarios represent 67 percent of the unaddressed crash costs for single-vehicle crashes.

The distribution of unaddressed multi-vehicle crashes differed from the distribution of unaddressed single-vehicle crashes. Rear-end/LVS was the leading multi-vehicle pre-crash scenario followed by straight crossing path at non-signal and left turn across path/opposite direction at non-signal. Collectively, these three pre-crash scenarios represented 50 percent of the costs of unaddressed multi-vehicle crashes. The five rear-end pre-crash scenarios represented nearly 50 percent of the unaddressed multi-vehicle crashes and 44 percent of the unaddressed multi-vehicle crash costs. The Volpe report on frequency of target crashes suggests V2V and AV applications as the primary strategies for rear-end crashes.⁽⁹⁾

The annual cost of unaddressed crashes was $136.9 billion, which includes all single-vehicle and multi-vehicle crashes. The leading unaddressed pre-crash scenarios are discussed in detail in the following sections.

Table 9. Single-vehicle unaddressed crashes by pre-crash scenario.

Note: Pre-crash scenarios for which there were no annual observed crashes were omitted from this table.

Table 10. Multi-vehicle unaddressed crashes by pre-crash scenario.

N/A = Not applicable.

Note: Pre-crash scenarios for which there were no annual observed crashes were omitted from this table.

8.2 Characteristics of Leading Unaddressed Crashes by Pre-crash Scenario

8.2.1 Single-Vehicle Pedestrian Crashes

A total of 40 percent of unaddressed costs for single-vehicle crashes were represented by two pedestrian pre-crash scenarios, pedestrian/maneuver and pedestrian/no maneuver, with the latter being the more prevalent of the two. The estimated total annual cost of unaddressed single-vehicle pedestrian crashes was more than $17.8 billion. As stated in the Volpe report, these two crash scenarios are primarily targeted by AV applications, with the V2I applications only addressing crashes that occur at crosswalks.⁽⁹⁾

The locations of these unaddressed pedestrian crashes are presented in table 11. There were no unaddressed pedestrian crashes at signalized intersections because those crashes are targeted by one of the currently identified applications and were previously discussed in this report. Intersections represented 35 percent of unaddressed single-vehicle pedestrian crashes, while the majority (58 percent) were non-junction crashes. The remaining crashes occurred at driveways/alleys (6 percent) and other locations. Additionally, 10 percent of the unaddressed pedestrian crashes occurred at stop-controlled intersections, which could potentially be targeted through the extension of the driver gap assist at stop-controlled intersections application as discussed in section 4.4 of this report.

Notably, the segment crashes represented the more severe crashes. Although segments represented 58 percent of the crashes, they represented 97 percent of the fatalities.

Table 11. Distribution of crash location for annual national (weighted) unaddressed pedestrian crashes.

In total, 77 percent of these crashes occurred in urban areas. The majority of crashes also occurred on moderate speed roadways, with two-thirds of those crashes occurring on roadways with posted speeds limits of 25-35 mi/h. Crashes mainly occurred on two-lane roadways (41 percent).

8.2.2 Single-Vehicle Bicycle Crashes

In total, 17 percent of the total unaddressed costs for single-vehicle crashes were represented by two bicycle pre-crash scenarios, cyclist/maneuver and cyclist/no maneuver, with the latter being the more prevalent of the two. The estimated total annual cost of unaddressed single-vehicle bicycle crashes was more than $7.5 billion. The primary strategies proposed for addressing cyclist crashes are AV applications, as reported in the Volpe study.⁽⁹⁾

Table 12 presents the number and percentage of crashes based on location of unaddressed single-vehicle bicycle crashes. Intersection-related crashes represented 64 percent of unaddressed single-vehicle bicycle crashes, 21 percent of crashes were non-junction crashes, and 15 percent occurred at driveways/alleys.

Table 12. Distribution of crash location for annual national (weighted) unaddressed bicycle crashes.

Over 70 percent of the unaddressed bicycle crashes occurred in urban areas. The majority of crashes also occurred on moderate-speed roadways, with 70 percent of the crashes occurring on roadways with posted speeds limits of 25-35 mi/h. Additionally, 40 percent of the crashes occurred on two-lane roadways.

8.2.3 Single-Vehicle Animal Crashes

A total of 9 percent of the total unaddressed single-vehicle crashes were represented by two animal pre-crash scenarios, animal/maneuver and animal/no maneuver, with the latter being the more prevalent of the two. While 91 percent of the unaddressed animal crashes resulted in no injury, the estimated total annual cost of unaddressed animal crashes was more than $4.1 billion due to the large frequency of these crashes. Similar to cyclist crashes, AV applications are proposed as strategies for targeting animal crashes.⁽⁹⁾

Over half of the unaddressed animal crashes (57 percent) occurred in rural areas, while the remaining 43 percent occurred in urban areas, indicating that animal crashes do not occur in just one area type (see table 13). The distribution of animal crashes by location (tangent and curve) and area type indicated that the crashes predominately occurred on tangent sections for both urban and rural locations (88 and 90 percent, respectively).

Table 13. Distribution of area type for annual national (weighted) unaddressed animal crashes.

The distribution of vehicle type for single-vehicle animal crashes indicated that light vehicles were involved in 95 percent of the crashes. Other vehicles accounted for 2 percent of the crashes. Each of the remaining vehicle types accounted for 1 percent or less of the total animal crashes.

8.2.4 Rear-End Crashes

A total of 44 percent of the total unaddressed multi-vehicle crashes were represented by five multi-vehicle rear-end pre-crash scenarios: rear-end/striking maneuver, rear-end/LVA, rear-end/ LVM, rear-end/LVD, and rear-end/LVS. The latter pre-crash scenario was the most prevalent of the five. The estimated total annual cost of unaddressed rear-end crashes was more than $40.9 billion. As previously mentioned, the majority of rear-end crashes are addressed by V2V and AV applications, which was a main cause for the high number of unaddressed read-end crashes in the Volpe report.⁽⁹⁾ However, there were a substantial number of rear-end crashes targeted by the speed zone, work zone, and speed work zone V2I applications.

Table 14 presents the distribution of location for unaddressed crashes with urban and rural locations combined. Approximately half of the crashes occurred at intersection locations, and half occurred on roadway segments. The distribution by area type, shown in table 15, indicates that 76 percent of the crashes occurred in urban areas.

Table 14. Distribution of crash location for annual national (weighted) unaddressed rear-end crashes.

Table 15. Distribution of area type for annual national (weighted) unaddressed rear-end crashes.

The distribution of vehicle type for multi-vehicle rear-end crashes indicated that crashes involving only light vehicles represented 93 percent of the crashes. Trucks, including single-unit and combination-unit vehicles, constituted 4 percent of rear-end crashes.

8.2.5 Straight Crossing Path Crashes at Non-Signals

A total of 15 percent of the total unaddressed multi-vehicle crashes were represented by the pre-crash scenario straight crossing path crashes at non-signals. The estimated total annual cost of these target crashes was more than $14 billion.

Table 16 presents the distribution of locations of unaddressed crashes. Notably, 53 percent of unaddressed straight crossing path at non-signal crashes occurred at driveways/alleys. The Volpe report proposes V2I applications for straight crossing path crashes at intersections and V2V applications for the straight crossing path crashes at non-intersections such as driveways.⁽⁹⁾ Table 17 shows the distribution of these crashes by area type.

Table 16. Distribution of crash location for annual national (weighted) straight crossing path crashes at non-signals.

Table 17. Distribution of area type for annual national (weighted) straight crossing path crashes at non-signals.

The distribution of vehicle type for multi-vehicle, straight crossing path crashes at non-signals indicated that crashes only involving light vehicles represented 91 percent of crashes. Trucks, including single-unit and combination-unit vehicles, were involved in 4 percent of the crashes.

8.2.6 Left Turn Across Path/Opposite Direction Crashes at Non-Signals

The pre-crash scenario left turn across path/opposite direction at non-signals represented a total of 12 percent of the total unaddressed multi-vehicle crashes. The estimated total annual cost of these target crashes was more than $11.2 billion. The Volpe report proposes V2I applications only address intersection crashes for this pre-crash scenario, while V2V applications address all remaining crashes.⁽⁹⁾

Table 18 presents the distribution of locations of unaddressed crashes. While the majority (60 percent) of the unaddressed left turn across path/opposite direction at non-signal crashes occurred at intersection or intersection related locations, 37 percent occurred at driveways/alleys.

Table 18. Distribution of crash location for annual national (weighted) unaddressed left turn across path/opposite direction crashes at non-signals.

A total of 18 percent of the crashes occurred on non-level roadways (i.e., on a grade, hillcrest, or sag). Over two-thirds (71 percent) of these unaddressed crashes occurred in urban areas, and two-thirds of the crashes occurred on roadways with a speed limit between 35 and 45 mi/h. The distribution of vehicle type for unaddressed multi-vehicle left turn across path/opposite direction crashes at non-signals indicated that crashes only involving light vehicles represented 96 percent of the crashes. Trucks, including single- and combination-unit vehicles, were involved in less than 3 percent of the crashes.

8.3 Characteristics of Unaddressed Crashes by Vehicle, Area, and Location Type

8.3.1 Vehicle Type

The distribution of unaddressed single-vehicle and multi-vehicle crashes by vehicle type is presented in table 19. For single-vehicle crashes, single-unit and combination-unit trucks appear to be overrepresented, accounting for 8 percent of the vehicles. Other vehicles also are overrepresented when compared to multi-vehicle crashes, representing another 6 percent of the unaddressed crashes.

Table 19. Distribution of vehicle type for unaddressed crashes.

8.3.2 Area Type

Table 20 presents the distribution of unaddressed crashes by area type for single-vehicle and multi-vehicle crashes. More than 70 percent of the unaddressed crashes were in urban areas. Urban locations represented a greater proportion of the unaddressed crashes for multi-vehicle crashes than for single-vehicle crashes.

Table 20. Distribution of area type for unaddressed crashes.

8.3.3 Location Type

Table 21 presents the distribution of unaddressed crashes by intersection and segment locations. For both single-vehicle and multi-vehicle crashes, the majority of the unaddressed crashes were on tangent segments. This was expected because many of the current applications target intersections and curved segments. Specifically, 75 percent of single-vehicle crashes compared to 51 percent of multi-vehicle crashes occurred on tangents. Furthermore, the difference in percentage between single-vehicle and multi-vehicle crashes for crashes at intersections and segments was expected because multi-vehicle crashes are generally more prevalent at intersections. Additionally, 21 percent of the unaddressed multi-vehicle crashes were at signalized intersections. Signalized intersections could be a target for new applications or extensions to existing application areas.

Table 21. Distribution of crash location for unaddressed crashes.