iihs comment to nhtsa concerning ncap update for rearview

July 25, 2013 The Honorable David L. Strickland Administrator National Highway Traffic Safety Administration 1200 New Jersey Avenue, SE Washington, DC 20590

Request for Comments; Planned Update to U.S. New Car Assessment Program; 49 CFR Part 575, New Car Assessment Program (NCAP); Docket No. NHTSA-2013-0076

Dear Administrator Strickland:

The National Highway Traffic Safety Administration (NHTSA) has requested comments on a proposed update to the U.S. New Car Assessment Program (NCAP) to provide information concerning rearview video systems on vehicle models listed on the agency’s website, safercar.gov. NHTSA plans on using a two-phase approach. Initially, rearview video systems will be listed in the “Safety Features” section of the website for vehicle models with the feature. In the second phase, vehicle models with rearview video systems that meet the performance requirements described in the agency’s proposed amendment to Federal Motor Vehicle Safety Standard No. 111 (Office of the Federal Register, 2010) will receive credit in the “Recommended Advanced Technology Features” section of the website. The Insurance Institute for Highway Safety (IIHS) has supported NHTSA’s efforts to promote countermeasures to increase driver awareness of objects to the rear of the vehicle (IIHS, 2009, 2011), and we agree that promoting rearview video systems through NCAP is a useful step toward addressing the backover crash problem.

Backing crashes represent only a small proportion of all highway crashes, but can be tragic because injuries or fatalities in backover crashes typically involve young children and families. The potential of rearview video systems to address the backover crash problem is well established. Rearview video systems greatly increase visibility behind the vehicle. IIHS recently measured rear visibility in 21 2010-13 vehicles and the improvements in visibility provided by rearview video systems and areas behind the vehicle detected by rear parking sensors (Kidd and Brethwaite, 2013). Rearview video systems decreased blind zones, defined as the areas not visible in side mirrors, rearview mirrors, or using direct glances over the right shoulder of a 50th percentile male, by 72-99 percent across the vehicle sample. Experimental studies have found rearview video systems help drivers avoid an unexpected stationary and moving obstacle in the backing path when the system is used while backing (Kim et al. 2012; Mazzae, 2010; Mazzae and Barickman, 2008). Additionally, rearview video systems are the most capable technological solution available, as other technologies like ultrasonic sensors have not be shown to be reliable or as effective at enhancing awareness of objects behind the vehicle (Mazzae and Garrott, 2006).

Past research suggests rearview video systems should have a measurable effect on backing crashes, but, to date, there is little evidence to suggest these systems are preventing crashes and reducing loss. In 2011, the Highway Loss Data Institute (HLDI) compared insurance claim frequencies for physical damage to the at-fault vehicle (collision coverage) and physical damage to a struck vehicle or other property (property damage liability coverage) in select Mazda and Mercedes-Benz vehicle models with a rearview video system with the same vehicle models without the system (HLDI, 2011, 2012). As shown in Table 1, changes in claim frequencies were directionally inconsistent across coverage types and statistically significant reductions in claim frequencies were not observed. In a naturalistic driving study of backing maneuvers conducted by NHTSA, 6 backing crashes were observed in a sample of 37 drivers of 2007 Honda Odyssey minivans during 4 weeks of driving (Mazzae and Barickman, 2008). Of the 6 crashes, 5 involved drivers whose vehicle had a backup camera system. Four of these 5 drivers’ vehicles also had a parking sensor system.

David L. Strickland July 25, 2013

Table 1 Percent change in insurance claim frequency per insured vehicle year

for select Mazda and Mercedes-Benz vehicles with rearview video systems compared with those without the systems by coverage type

Make Coverage type

Collision Property damage liability Mazdaa +3.1c -2.3 Mercedes-Benzb +0.5 -0.5 a HLDI (2011), b HLDI (2012), c p<0.05

These preliminary findings suggest the real-world effectiveness of rearview video systems may not be as beneficial as expected, but more extensive evaluations are required. NHTSA’s proposed update to the NCAP program may accelerate the penetration of rearview video systems in the U.S. passenger vehicle fleet. Having more passenger vehicles equipped with rearview video systems on the roadways will allow HLDI, NHTSA, and others to better estimate the real-world effectiveness of rearview video systems.

Rearview video systems currently might be the best means of increasing drivers’ views of the areas behind their vehicles, but many drivers still rely on mirrors or rearward glances to maintain awareness of their surroundings while backing (Mazzae and Barickman, 2008). In our evaluations of rear visibility, we were surprised to see several instances where smaller passenger cars had larger blind zones than larger trucks or minivans (Kidd and Brethwaite, 2013). As IIHS has expressed in previous comments, we encourage NHTSA to consider implementing direct rear visibility requirements for all vehicles (IIHS, (2009, 2011). Having an available rearview video system on a vehicle model should not justify design choices that restrict direct visibility around the vehicle.

In summary, the tragic nature of backover crashes justifies NHTSA’s continuing efforts to promote rearview video systems. IIHS supports the inclusion of rearview video systems in NCAP and will continue to monitor the effects of these systems in the real world. Rearview video systems might only be a partial solution to the backover crash problem, so we encourage NHTSA to consider additional measures to address this issue such as regulating the size and position of directly viewable areas behind vehicles.

Sincerely, David G. Kidd, Ph.D. Research Scientist

Attachment

Kidd, D.G. and Brethwaite, A. 2013. Visibility of children behind 2010-13 model year passenger vehicles using glances, mirrors, and backup cameras and parking sensors. Arlington, VA. Insurance Institute for Highway Safety.

References

Highway Loss Data Institute. 2011. Mazda collision avoidance features: initial results. Bulletin 28(13). Arlington, VA.

David L. Strickland July 25, 2013 Page 2 Highway Loss Data Institute. 2012. Mercedes-Benz collision avoidance features: initial results. Bulletin 29(7). Arlington, VA.

Insurance Institute for Highway Safety. 2009. Comment to the National Highway Traffic Safety Administration concerning proposed amendments to rearview mirrors standard and no-blind-spot requirement to reduce deaths and injuries associated with backover crashes (Docket no. NHTSA-2009-0041). Arlington, VA.

Insurance Institute for Highway Safety. 2011. Comment to the National Highway Traffic Safety Administration concerning modification of Federal Motor Vehicle Safety Standard 111 Rearview Mirrors to require systems to enhance visibility of the areas directly behind the vehicle (Docket no. NHTSA-2010-0162). Arlington, VA.

Kidd, D.G. and Brethwaite, A. 2013. Visibility of children behind 2010-13 model year passenger vehicles using glances, mirrors, and backup cameras and parking sensors. Arlington, VA. Insurance Institute for Highway Safety.

Kim, R.; Rauschenberger, R.; Heckman, G.; Young, D.; and Lange, R. 2012. Efficacy and usage patterns for three types of rearview camera displays during backing up. SAE 2012 World Congress & Exhibition (Paper no. 2012-01-0287). Warrendale, PA: Society of Automotive Engineers.

Mazzae, E.N. 2010. Drivers’ use of rearview video and sensor-based backing aid systems in a non-laboratory setting. Docket no. NHTSA-2010-0162. Washington, DC. National Highway Traffic Safety Administration.

Mazzae, E.N. and Barickman, F. 2008. On-road study of drivers’ use of rearview video systems (ORSDURVS). Report no. DOT HS-811-024. Washington, DC. National Highway Traffic Safety Administration.

Mazzae, E.N. and Garrott, W.R. 2006. Experimental evaluation of the performance of available backover prevention technologies. Report no. DOT HS-810-634. Washington, DC. National Highway Traffic Safety Administration.

Office of the Federal Register. 2010. Federal Register, vol. 75, no. 234, pp. 76186-76250. National Highway Traffic Safety Administration, Docket no. NHTSA-2010-0162, 49 CFR Parts 571 and 585: Federal Motor Vehicle Safety Standard, Rearview Mirrors; Federal Motor Vehicle Safety Standard, Low-Speed Vehicles Phase-in Reporting Requirements. Washington, DC: National Archives and Records Administration.

1005 N. Glebe Rd., Arlington, VA 22201 Tel. 703/247-1500 Fax 703/247-1588 www.iihs.org

Visibility of Children Behind 2010-13 Model Year Passenger Vehicles Using Glances, Mirrors, and Backup Cameras and Parking Sensors

July 2013

David G. Kidd Insurance Institute for Highway Safety

Andrew Brethwaite Insurance Institute for Highway Safety

1

ABSTRACT

It is estimated that about 18,000 people are injured and 292 are killed annually in backover

crashes. A little more than 10 percent of these injuries involve children younger than 5, but older children

also are frequently injured in backing crashes. Visibility of the areas behind the vehicle influences how

well drivers can detect and avoid obstacles while reversing. Previous research has measured rear

visibility in an assortment of vehicles for a 1-2-year-old child as a worst case scenario. The current study

measured rear visibility for 12-15 month-olds, but also examined visibility for 30-36- and 60-72-month-old

children. The purpose was to identify the areas behind vehicles where younger and older children are not

visible and determine the extent to which vehicle technologies like backup cameras improve visibility and

parking sensors detect objects in areas that are not visible.

Rear visibility in 21 2010-13 model year vehicles with a backup camera system or a backup

camera and rear parking sensor system was assessed in an area that was 20 feet wide and extended 70

feet from the rear bumper. A 50th percentile male observer was used to make judgments about the

visibility of targets simulating the height of a child 12-15 months old (30.2 inches tall), 30-36 months old

(36.8 inches tall), and 60-72 months old (42.7 inches tall). Judgments were made using the left and right

side mirrors, rearview mirror, glances over the right shoulder through the rear window, backup camera,

and parking sensors (if available). The area behind the vehicle where each target was not visible (blind

zone) and the average sight distance from the rear bumper to the target were used to characterize rear

visibility for each vehicle. The size of the image in the backup camera display from the midpoint of the

observer’s eyes also was assessed.

The average blind zone for a 12-15-month-old child (240.24 ft2) was twice as large as the

average blind zone for a 60-72-month-old child (116.67 ft2). Average sight distance from the rear bumper

to a 12-15-month-old child-size object (27.3 feet) also was twice as far as it was for a 60-72-month-old

child-size object (13.2 feet). On average, large SUVs had the worst rear visibility, i.e., largest blind zone

and longest rear sight distance, and small cars had the best visibility. Generally, the size of the vehicle

was positively associated with the size of the blind zone and average sight distance from the bumper.

However, there were exceptions to this pattern. Some cars, like the Infiniti M37 and Hyundai Sonata, had

worse rear visibility for a 12-15-month-old child than larger pickup trucks and minivans.

On average, backup cameras and backup cameras combined with parking sensors reduced the

blind zone for the different target heights by around 90 percent. Furthermore, these systems drastically

reduced average rear sight distance from the bumper to the different child-size objects to less than 2 feet.

On average, increases in rear visibility provided by backup cameras were larger than the non-visible

areas detected by parking sensors, but parking sensors detected objects in areas near the rear of the

vehicle that were not visible using direct glances, mirrors, and backup cameras. Finally, the image size of

each vehicle’s backup camera display met the federal government’s proposed requirements.

This study was the first to assess rear visibility for taller child-size objects among vehicles in

various classes. Although rear visibility typically was better for 60-72-month-old child-size objects than for

2

12-15-month-old child-size objects, a 42.7-inch tall object was not visible up to 13 feet behind the vehicle.

Backup cameras and parking sensors nearly eliminated the blind zone behind each vehicle. Based on

their functionality, vision enhancing technologies can mitigate or completely prevent backing crashes.

However, behavioral studies have found that drivers still may collide with child-size objects when

reversing even when using backup cameras and parking sensors. Additional research is needed to better

understand how drivers use the technologies, to identify which technology features optimize their use,

and to devise ways to get drivers to use them optimally. Crash avoidance technology that automatically

brakes when a backup collision is eminent also is a promising alternative solution.

INTRODUCTION

Backover crashes can result in severe and fatal injuries to pedestrians or people standing behind

the vehicle. Based on data from the Not-in-Traffic Surveillance (NiTS) system, the Fatality Analysis

Reporting System, and the National Automotive Sampling System General Estimates System, an

estimated 18,000 injuries and 292 fatalities occur each year due to backover crashes (Austin, 2008).

About 2,000 of these injuries were estimated to involve children younger than 5. Children are at a higher

risk of being involved in a backover crash because their shorter stature makes them harder to see.

One factor that contributes to backover crashes, especially those involving children, is vehicle

rear visibility. Rear visibility typically is worse in larger vehicles like trucks and SUVs compared with

passenger cars. Consumer Reports (2012) measured the distance from the vehicle’s rear bumper to the

location where a cone 28 inches tall was first observed by drivers 5 feet, 1 inch and 5 feet, 8 inches tall

using glances over the right shoulder. The 28-inch cone was used to approximate the height of a 1-year-

old child. For the 5-foot, 1-inch driver, the average rear sight distance was longest for pickups and

shortest for minivans. Pickups also had the longest rear sight distance for the 5-foot, 8-inch driver, and

midsized sedans had the shortest.

The National Highway Traffic Safety Administration (NHTSA) assessed rear visibility in passenger

vehicles using a human driver (Mazzae & Garrott 2008) and a laser-based measurement system (;

Mazzae, 2013; Mazzae & Barickman 2009). The primary measure in these studies was blind zone,

defined as the area behind the vehicle that was not directly visible using glances over the shoulder or

through side windows, indirectly visible using side mirrors and the rearview mirror, or some combination

of these fields of view. The blind zone for a 1-year-old child-size object typically was larger for large

pickup trucks, cargo vans, and SUVs than it was for passenger cars.

Given that rear visibility is poorer in larger vehicles than smaller vehicles, it is not surprising that

larger vehicles are more frequently involved in backover crashes involving children. Pinkney et al. (2006)

examined police crash reports in Utah from 1998 to 2003 involving children younger than 10 struck by a

motor vehicle on a residential driveway. The children were 53 percent more likely to be injured by a

pickup truck than a passenger car, relative to registrations of each vehicle type, and 141 percent more

3

likely to be injured by a minivan than a passenger car. They also were more likely to be injured by an

SUV than a passenger car, but this difference was not statistically significant.

Mazzae and Garrott (2011) performed a Monte Carlo simulation using data from a naturalistic

study of backing maneuvers to estimate the probability of a reversing vehicle striking a pedestrian moving

at a constant speed and direction behind the vehicle. The probability of a backover crash was highest

when the pedestrian’s starting position was immediately behind the rear bumper and decreased as the

starting location moved away from the rear bumper. However, a crash occurred in 40 percent or more of

the simulated trials where the pedestrian’s starting position was located 15 feet or less directly behind the

rear bumper. This area is typically not visible to a 50th percentile male driver using mirrors or glances

over the shoulder in most vehicles.

Backing technologies such as ultrasonic parking sensors and backup cameras can enhance

visibility or detect objects behind the rear of the vehicle and help prevent backover crashes. Ultrasonic

parking sensors detect objects behind the vehicle by emitting ultrasonic sound waves from the rear of the

vehicle. If an object is present, the wave reflects off the object and returns to the vehicle where it is

detected by a sensor. Distance and location information is relayed to the driver with a visual display or

auditory tone. A disadvantage of parking sensor systems is that they have limited detection range. One

study reported that owners typically turned sensor systems off because they were not reliable (Mazzae &

Garrot 2006). Drivers who reverse too fast also may exceed the functional capabilities of sensor systems,

rendering them useless (Llaneras et al., 2005). Finally, one study found that sensor systems had

difficulty detecting pedestrians, especially moving children (Mazzae & Garrott, 2006).

Backup cameras display the area directly behind the vehicle on a screen generally located in the

vehicle center console or rearview mirror. Backup cameras are passive and do not alert drivers about

objects behind the vehicle. Thus, the driver must look at the display to detect and respond to obstacles.

Some newer vehicles combine sensor-based systems and backup cameras.

Several organizations are working to encourage auto manufacturers to fit vehicles with backing

technology. The Insurance Australia Group (IAG, 2013) rates vehicle rear visibility using a reversing

visibility index. Ratings are based on the area where a laser positioned at the approximate eye location

of a driver 70.1 inches tall and directed through the rear window is observed on a cylinder 23.6 inches tall

in an area 5.9 feet wide and 49.2 feet long behind the vehicle (Paine et al., 2003). Vehicles are awarded

up to 5 stars based on the area of the test grid where the child-sized cylinder is visible, and an extra 0.5

star is awarded to vehicles with backing technology.

In the United States, NHTSA has proposed rulemaking to require manufacturers to provide

drivers a way to see seven cylinders 32 inches tall placed along the perimeter of an area 10 feet wide and

20 feet long directly behind the vehicle when the vehicle is placed in reverse (Office of the Federal

Register, 2010). NHTSA’s proposed rule does not require a specific technology, or any technology at all,

but backup cameras are currently the only technology that will meet the proposed minimum visibility

requirements.

4

To date, assessments of rear visibility have been limited to shorter objects reflecting the height of

a 1- or 2-year-old child. Young children represent the “worst case scenario” for rear visibility, but older

children also are involved in backover crashes. A study using the Canadian Hospitals Injury Reporting

and Prevention Program database found that, between 1993 and 2004, about 12 percent of pedestrian-

motor vehicle collisions involving children younger than 5 were backovers (Nhan et al., 2009). About 4

percent of pedestrian-motor vehicle collisions involving 5-13 year-olds were backovers.

The improvement in rear visibility provided by different technologies in different vehicles has not

been systematically evaluated. The purpose of the current study was to characterize rear visibility for 12-

15-, 30-36-, and 60-72-month-old children in 2010-13 model year passenger vehicles, and measure the

additional visibility provided by backup cameras for each age group and areas where each age group was

not visible but detected by parking sensors. The image size of vehicles’ backup camera displays also

was examined to determine whether current camera systems meet the functional requirements proposed

by NHTSA.

METHOD

Vehicle Sample

Twenty-one 2010-13 model year passenger vehicles with a backing camera system or a backing

camera and rear parking sensor system were evaluated (Table 1). Candidate vehicles were limited to the

manufacturers and models available at dealerships in the Charlottesville, Virginia, area. Efforts were

made to include 2012 models with the largest sales volumes. At least two vehicles in each vehicle class

were included.

Target and Measurement Field

The visual target was a cylinder 42.7 inches tall and 4.5 inches wide (Figure 1). Three different

color bands were painted onto the cylinder. The distance from the bottom of the cylinder to the top of

each color band corresponded with the 50th percentile standing heights of a 12-15-, 30-36-, and 60-72-

month-old child (Tilley, 2002). The standing heights simulated by the visual target were 30.2, 36.8, and

42.7 inches for each age group, respectively. The height of each color band corresponded with the

average head width of a 50th percentile child in each age group (Tilley, 2002). The heights of the color

bands simulating the head heights of a 12-15-, 30-36-, and 60-72-month-old child were 5, 5.3, and 5.4

inches, respectively.

Visibility from various direct and indirect fields of view was measured using a measurement field

20 feet wide that extended longitudinally 5 feet in front of the rear bumper and 70 feet behind the rear

bumper of each vehicle (Figure 2). The measurement field was divided into 1- by 1-foot squares using 1-

inch wide tape adhered to a flat concrete surface. Six different fields of view were measured and

included the (a) left side mirror, (b) right side mirror, (c) rearview mirror, (d) glances over the right

shoulder, (e) areas visible in the backup camera display, and (f) areas detected by the rear parking

5

sensor system. Visibility for the three target heights using each field of view was assessed at each

square in the measurement grid. Only visibility judgments made at or behind the rear bumper (0-70 feet

in Figure 2) are reported in the current study.

Observers and Inter-Observer reliability

Visibility judgments were made by two male observers who were similar in stature and seated

eye height to each other and a 50th percentile male (Table 2; Gordon et al., 1989). The observers were

Insurance Institute for Highway Safety (IIHS) employees and were assigned to different vehicles based

on the employee’s availability. The observers were selected from four candidate observers based on

inter-observer agreement data collected in a pilot study.

In the pilot study, observers made visibility judgments for the three visual target heights in up to

three different vehicles. The percent of agreement between observations made by pairs of observers’ for

each target height using each field of view were analyzed. Cohen’s Kappa also was calculated as a

second measure of inter-observer agreement (Cohen, 1960). Observers TV and DK showed a high

percent of agreement between their visibility judgments for different target heights across nearly every

field of view in three different vehicles. Their percent agreement was above 90 percent for judgments

made using side and rearview mirrors, but was lower for judgments made using glances over the right

shoulder (60-90 percent). Cohen’s Kappa of the visibility judgments made for different target heights with

mirrors showed substantial (0.61-0.8) or almost perfect (>0.8) agreement. Visibility judgments made

using glances over the right shoulder also showed substantial or almost perfect agreement for all three

target heights in two of the three pilot study vehicles and the shortest target height in the third vehicle.

Visibility judgments for the 36.8- and 42.7-inch target heights in the third vehicle showed low agreement

(0.01-0.2). Overall, the results indicated a high degree of correspondence between the visibility

judgments made by observers TV and DK, and the agreement between these observers was higher than

for the other pairs of observers. Thus, observers TV and DK were used in the full study.

Vehicle Preparation

The following procedure was used to prepare each vehicle to enhance repeatability of visibility

judgments. First, the vehicle tires were inflated to the manufacturer’s recommended inflation pressure

and the vehicle’s dimensions were measured. The vehicle’s fuel tank was either filled to capacity or

weight was added to approximate the weight of the fuel needed to fill the tank to capacity; 6.1 pounds of

weight were added per gallon of fuel (Marathon, 2010). The vehicle’s rear passenger windows were

cleaned and closed. The driver and front passenger windows were slightly open to allow for better

communication between the observer and the researcher recording measurement data. If a window

sticker obscured the view of the observer, then the sticker was removed or the window with the sticker

was rolled down. The driver and front passenger seats were positioned for a 50th percentile male using

the University of Michigan Transportation Research Institute (UMTRI) anthropometric test device (ATD)

6

positioning procedure (IIHS, 2004), which is used by IIHS to position ATDs in frontal crash tests. The

vertical position of the driver seat head restraint was even with the topmost point of the observer’s head

or set at the highest vertical height of the head restraint, whichever was lower. All other head restraints

were in upright positions, set to the lowest vertical height. If fore-aft adjustment of the head restraint was

possible, then the driver head restraint was set to the forward-most position and the right front passenger

and rear passenger head restraints were set to the midpoint position. If the second and third row seats

were adjustable, the seat pan was set to mid-track position and the seat back angle was set to 25

degrees. Every front and rear passenger seat belt was buckled, including those originating from the

headliner for center seating positions.

Once the vehicle was prepared it was positioned on the measurement field. The vehicle’s rear

bumper was positioned flush with the 20-foot lateral axis of the measurement field at the origin of the

longitudinal axis (i.e., 0-foot line). The vehicle’s centerline was positioned in the center of the lateral axis

allowing for lateral measurements ranging between 10 feet to the right and left of the vehicle centerline.

Next, the observer entered the vehicle, sat in the driver’s seat, and buckled the seat belt. Then

the observer positioned the side and rearview mirrors. The position about the vertical axis for the left and

right side mirrors was set so that the side of the vehicle was slightly visible at the most inward point of the

mirror. This setting was found to be the most common position for side mirrors among a representative

sample of drivers (Reed et al., 2001; Reed et al., 2000). The position about the lateral axis for the side

mirrors was set so the horizon was vertically centered in the mirror. Lastly, the observer adjusted the

rearview mirror to be able to see through the entire rear window.

Measurement Procedure

The visual target was placed on individual squares in the measurement field. The observer was

instructed to indicate the shortest target height that was visible using different fields of view. To be

visible, the full width and height of the color band corresponding with a target height had to be seen. The

target was considered visible to a vehicle’s rear parking sensor system if a visual or auditory warning was

provided. If the target was detected by the vehicle’s rear parking sensor system at a given location, then

it was assumed that all three target heights were detectable at that location. When making visibility

judgments using the mirrors, the observer was instructed to maintain a normal body position, keep his full

body weight against the seat pan and seat back, and keep his head centered in the head restraint. For

glances over the right shoulder, the observer was instructed to make a normal glance over the shoulder

and not to exaggerate his normal turning motion to increase visibility. Some vehicles’ backup camera

systems supported multiple viewing angles. Only the default view that was first available after the gear

selector was placed in reverse was assessed.

After completing the visibility measurements, researchers collected information to assess the

image quality of the backup camera display. A target 32 inches tall and 12 inches wide was placed 20

feet behind the rear bumper. The target was centered at the vehicle’s centerline. A measurement tape

7

was placed across the backup camera display and a photograph was taken for later analysis of the

backup camera image (see Figure 3). Researchers also measured the distance from the midpoint

between the observer’s eyes to the center of the backup camera display. The height and width of the

display also was recorded along with the visual and auditory warning characteristics of the parking sensor

system, if equipped.

Dependent Measures

Several measures were used to characterize the rear visibility of each vehicle. The first measure

was blind zone, defined as the area of the measurement field where a specific target height was not

visible using a single field of view or multiple fields of view. Blind zone without technology and blind zone

with technology was calculated for each vehicle. Blind zone without technology was the area not visible

using the left and right side mirrors, the rearview mirror, and glances over the right shoulder. Blind zone

with technology indicated the area not visible using the mirrors, glances over the right shoulder, backup

cameras, and not detected by parking sensors (if equipped). Visibility maps showing the individual

contribution of each field of view for the three target heights for al 21 vehicles can be found in the

Appendix.

Average minimum sight distance without and with technology also was calculated for each

vehicle. Minimum sight distance was the longitudinal distance from the rear bumper of the vehicle to the

first of at least two consecutive 1- by 1-foot squares where the target was visible or detected by parking

sensors for vehicles with this system. Minimum sight distance was recorded at 1-foot increments laterally

across an 8-foot span along the rear of each vehicle — 4 feet to the left and right of the vehicle’s

centerline. The average minimum sight distance was the mean of these eight values. If the target was

not visible or detected anywhere in the measurement field, then the maximum distance in the

measurement field was recorded (i.e., 70 feet) as the average minimum sight distance. The image size

for each backup camera display was assessed by calculating the subtended visual angle of the target.

This procedure was similar to that used by Mazzae and Garrott (2011) and the proposed procedure to

test image size in the recent proposed rulemaking (Office of the Federal Register, 2010). The primary

difference between the current effort and previous work is that the subtended visual angle was only

calculated for a single object behind the vehicle instead of multiple objects.

Subtended visual angle was calculated by combining information about the distance from the

midpoint between the observer’s eyes to the display with the width of the visual target in the backup

camera display. First, a scaling factor was calculated to determine the number of pixels per inch for each

photograph of a backup camera display. The X and Y pixel coordinates were recorded for the start point

and end point of a 2-inch section of the measurement tape across the camera display. The Euclidean

distance between these two points was the scaling factor. Next, the pixel coordinates at the left and right

edges of the target in the display were recorded. This distance, the scaling factor, and the distance from

8

the display to the midpoint between the observer’s eyes were used to calculate the subtended visual

angle of the target using formula 1.

Θi = sin-1

(pi / (di * si)) (1)

Formula 1 states the subtended visual angle of the target for a given display i in radians (Θi) is

equal to the inverse sine of the quotient produced by dividing the width of the target object in pixels (pi) by

the product of the distance from the midpoint between the observer’s eyes to the center of the backup

camera display (di) and the scaling factor (si). The subtended visual angle then was transformed from

radians to minutes of arc to compare with requirements proposed in NHTSA’s rear visibility test procedure

(Office of the Federal Register, 2010). A photograph of each vehicle’s backup camera display can be

found in the Appendix.

RESULTS

Rear Visibility without Technology

The blind zone for each vehicle and average minimum sight distance without technology and with

technology was calculated to characterize rear visibility for each target height. This section describes the

results for blind zone and sight distance without technology (Table 3). The average blind zone for the

30.2-inch tall object (240.24 ft2) across all vehicles was twice as large as the average blind zone for the

42.7-inch tall object (116.67 ft2). The Acura MDX had the largest blind zone for the 30.2-inch tall target

(360 ft2), and the Cadillac Escalade had the largest blind zone for the 36.8- and 42.7-inch tall targets (241

and 171 ft2, respectively). The Ford Focus had the smallest blind zone for each target height.

The average minimum sight distance for a 30.2-inch tall object (27.3 feet) was more than twice as

long as the average distance for a 42.7-inch tall object (13.2 feet) (Table 3). The Audi A6 had the

shortest average minimum rear sight distance for the 30.2-, 36.8-, and 42.7-inch tall targets among all

vehicles in the sample (15.5, 9.9, and 5.5 feet, respectively). Average minimum sight distance for the

30.2- and 36.8-inch tall targets was longest for the Mazda CX-9 (41.3 and 32.6 feet), and the GMC

Acadia had the longest sight distance for the 42.7-inch tall target (19.5 feet).

Figures 4 and 5 show the average blind zone and minimum rear sight distance by vehicle class

for each target height. Average blind zone and average sight distance for each vehicle class decreased

as target height increased. The smallest blind zone and average sight distance for each target height

was observed among small cars. Large SUVs had the largest blind zone and average sight distance for

each target height, followed closely by midsize SUVs.

Benefits of Cameras and Sensors

Table 4 shows the blind zone and average minimum sight distance for the three target heights

using indirect glances, direct glances, and backing technology. The average blind zone with technology

across all vehicles was 21.5, 15.4, and 13 ft2 for the 30.2-, 36.8-, and 42.7-inch tall target heights,

9

respectively. The average minimum sight distance to each target across the entire vehicle sample was

1.1 feet or less. The Ford Focus and Ford F-150 did not have a blind zone behind the rear bumper for

any target height after including the areas visible using technology. After including technology, there was

no blind zone for the Chevy Equinox with the 36.8- and 42.7-inch tall targets and for the GMC Acadia with

the 42.7-inch tall target. The Mercedes GLK had the largest blind zone after including technology for

each target height. Average minimum sight distance after including technology was less than 2 feet for

every vehicle.

Figure 6 shows the average blind zone with backing technology by vehicle class for each target

height. Pickups had the smallest blind zone for each target height after including technology. Small cars

had the second smallest blind zone for the 30.2- and 36.8-inch tall targets, and large SUVs had the

second smallest blind zone for the 42.7-inch target height. The largest blind zone with technology was

observed for minivans. The 30.2-, 36.8-, and 42.7-inch tall targets were not visible for minivans in an

average area of 30.5, 21, and 20 ft2, respectively.

Tables 5A and 5B show the reduction in blind zones due to increased visibility provided by

backup cameras, detection of non-visible areas by parking sensors, and the combination of these

technologies for all 21 vehicles. On average, backing technology reduced the blind zone by 91 percent

for the 30.2- and 36.8-inch tall target heights and 88 percent for the 42.7-inch tall target height. The

technology provided the least reduction in blind zone for the 30.2-, 36.8- and 42.7-inch tall targets for the

Ford Escape (83 percent), Mercedes E class (82.6 percent), and Chevrolet Cruze (76.3 percent),

respectively.

Eight vehicles were equipped with a backup camera system and a parking sensor system.

Among these eight vehicles, the average percent reduction in blind zone for each target height provided

by backup camera systems was 2-8 times larger than the percent reduction in blind zone provided by

parking sensor systems alone. Backup cameras reduced blind zone by 72-99 percent, and parking

sensor systems reduced blind zone by 12-48 percent. The area detected by parking sensor systems was

not completely redundant with the area visible using backup cameras. For vehicles with a parking sensor

and backup camera system, the average percent reduction in blind zone was 2-3 percentage points

greater when using both technologies compared with reductions provided by backup camera system

alone.

Backup Camera Display Image Size

Table 6 lists the display characteristics of each vehicle’s backup camera display and warning

characteristics of the rear parking sensor system (if equipped). In 19 of the vehicles, the backup camera

display was located in the center console and in two vehicles it was inset in the rearview mirror. Every

display in the center console was 6-8 inches wide except for the 5-inch wide display in the Honda CRV.

The two rearview mirror displays were smaller than the center console displays and measured just more

than 3 inches wide.

10

The average distance from the midpoint between the eyes of a 50th percentile male observer to

the display across all 21 vehicles was 31.1 inches. The shortest distance was observed in the Toyota

Tundra (23.7 inches) and the longest in the Honda CRV (39.8 inches). The distance from the midpoint of

the observer’s eyes to the display was used to calculate the subtended visual angle of a 12-inch wide

target viewed by the observer in the display. The average visual angle across all vehicles was 12.9

minutes of arc. The largest visual angle was observed in the Mercedes E class (17.5 minutes of arc) and

the smallest visual angle was observed in the Honda CRV (5 minutes of arc). The visual angle was less

than 10 minutes of arc for three vehicle displays, 10-14.9 minutes of arc for 15 vehicle displays, and 15

minutes of arc or greater for three vehicle displays.

Eight vehicles in this study were equipped with rear parking sensor systems. Each had a graded

auditory warning that increased in pulse rate as the target object moved closer to the rear bumper. Six of

the eight systems also had graded visual warnings that provided distance and location information for the

target relative to the vehicle’s rear bumper. Three of these vehicles provided this information using a

schematic of the vehicle, and the other three overlaid this information on the backup camera display. The

Ford Focus and Ford F-150’s rear parking sensor systems only provided auditory warnings.

DISCUSSION

Adequate visibility around and behind the vehicle is necessary for the driver to maintain

awareness of the vehicle’s surroundings and to reduce the probability of a reversing vehicle striking or

running over objects. Previous studies have characterized rear visibility for an object simulating the

height of a 1-2-year-old child, but older, taller children also are frequently involved in backover crashes

(Nhan et al., 2009). This study characterized rear visibility for a 12-15-, 30-36-, and 60-72-month-old

child-size object in 21 vehicles and evaluated how backup cameras and parking sensors improve visibility

for children of these ages.

As expected, taller children were more visible than shorter children. The blind zone for a 12-15-

month-old child-size object was, on average, more than twice as large as it was for a 60-72-month-old

child-size object. On average, a 12-15-month-old child-size object was not visible to a 50th percentile

male driver within about 27 feet of the rear bumper. A recent naturalistic study of backing maneuvers

found that the average backing distance out of a parking space was 21 feet (Mazzae & Barickman, 2008);

a 12-15-month-old child would never be visible during such a maneuver. Although visibility for a 60-72-

month-old child-size object was better than for a 12-15-month-old child-size object, on average, a 60-72-

month-old child-size object was not visible to a 50th percentile male driver up to about 13 feet behind rear

bumper. Mazzae and Garrott (2011) estimated that the risk of backing into a pedestrian was greatest

when the pedestrian starts within 15 feet directly behind the rear bumper. Based on the current results,

children younger than 7 are not easily seen in this area and would be at high risk for being struck by a

reversing vehicle.

11

Consistent with previous research, larger vehicles typically had poorer rear visibility for child-size

objects than smaller vehicles (Consumer Reports, 2012; Mazzae & Barickman 2009; Mazzae & Garrott

2008). On average, large SUVs had the largest blind zones and longest average minimum sight

distances among all vehicle classes. Small cars had the smallest blind zones and shortest minimum sight

distances. Average rear visibility in minivans and pickup trucks was better than in SUVs even though

minivans and pickup trucks are over-represented in backover crashes with pediatric injury. Pinkney et al.

(2006) found that children younger than 10 were significantly more likely to be injured in a driveway

backover crash involving a pickup truck or minivan compared with a passenger car, but the likelihood for

pickups and SUVs was similar. The Pinkney et al. study was based on police crash reports from 1998 to

2003, so the vehicles in that study were much older than the ones measured in the current study.

Differences in rear visibility among newer vehicles observed in the current study may differ substantially

from the differences among the vehicles in Pinkney et al.

In the current study, rear visibility was not always worse in larger vehicles compared with smaller

ones. As illustrated in Figure 7, the blind zone for a 30.2-inch tall object in the Hyundai Sonata (midsize

car) extended beyond the blind zone in the Ford F-150 (pickup truck). The Sonata had an extremely

sloped rear window and tall rear trunk lid that limited visibility directly behind the vehicle. Although the

dimensions of the F-150 were larger than in the Sonata, the F-150 had large side mirrors with auxiliary

mirrors that increased visibility behind the vehicle and made up for its larger size.

An extremely sloped rear window and tall rear trunk lid also reduced visibility behind the rear

bumper of the Infiniti M37 (Figure 8). The blind zone for the 30.2-inch tall object in the Infiniti M37 (large

car) was slightly greater than the larger Dodge Grand Caravan (minivan), and the average minimum sight

distance also was about 6 feet longer. These two examples illustrate how design features can restrict

rear visibility. Highly sloped rear windows and taller rear trunk lids are intended to make the vehicle

aerodynamic but decrease the area visible through the rear window. Large rear head restraints and large

B- and C-pillars also impacted rear visibility of vehicles in this study. Auto manufacturers should consider

eliminating or reducing the size of vehicle design features that block large rear areas from unassisted

views in addition to installing technology to enhance rear visibility.

A number of manufacturers are installing backup cameras to improve rear visibility and parking

sensors to detect objects in areas near the rear of the vehicle that are not visible. The results from the

current study illustrate the large improvements in visibility and detection of objects behind the vehicle

provided by these technologies. Across all vehicles, backing technologies reduced the blind zone by

about 90 percent for each target height. They also nearly eliminated the differences in blind zones

observed among different vehicle classes. For example, on average, the blind zone for large SUVs was

141.5 ft2 larger than the area for small cars for the shortest target height; however, the difference was

only 5 ft2 when factoring in technology. Enhancements in rear visibility and detection of objects behind

the vehicle provided by backing technologies can help prevent property-damage only collisions where a

vehicle reverses into a fixed object. They also can help drivers avoid backing into another vehicle, which

12

can lead to costly repairs especially if the bumpers of the two vehicles are not compatible and do not line

up (IIHS, 2010). Most important, backup cameras and parking sensors enhance visibility and detection of

objects in the area directly behind the vehicle where pedestrian backover crash risk is the highest

(Mazzae & Garrott, 2011). When examined for a span 4 feet to the left and 4 feet to the right of the

vehicle center line, average minimum sight distance was less than 2 feet for every vehicle after including

the areas visible with technology.

Increases in rear visibility provided by backup cameras were greater than areas detected by

parking sensors that were not visible using other fields of view. Across all target heights, backup

cameras reduced the blind zone by 72-99 percent compared with 12-48 percent reductions observed for

parking sensors. However, the areas detected by parking sensors were not always visible using backup

cameras. The average percent reduction in blind zone for the eight vehicles with a backup camera and

parking sensors was 93-94 percent. The percent reduction in blind zone was 90-92 percent for these

eight vehicles when only including the additional areas visible using the camera. Two to three percentage

points is not a dramatic functional increase in rear visibility, but it is of practical importance. Parking

sensors detect objects near the rear of the vehicle where backover crash risk is greatest. Plus, the

alerting function of the sensor systems can draw drivers’ attention to potential hazards.

Technology can provide drivers with a more comprehensive view behind the vehicle, but these

technologies may not completely eliminate backing crashes. Studies show that while cameras and

parking sensors significantly increase the likelihood of avoiding an unexpected obstacle behind the

vehicle, they do not prevent all backing crashes (e.g., Kim et al., 2012; Llaneras et al., 2011; Mazzae,

2010). Drivers may not look at the backup camera display at the appropriate time to prevent a collision,

or they may not look at the display at all when reversing (e.g., Mazzae & Barickman, 2008). They also

may ignore or disable parking sensors if the system is prone to false alarms (e.g., Llaneras et al., 2011),

or reverse too quickly so as to exceed the functional capability of the system (Llaneras et al., 2005).

Backing crashes still can occur in vehicles with both backup cameras and parking sensors, and there is

some evidence that the combination of these systems may not perform as well as a backup camera alone

(Mazzae, 2010).

One possible solution for preventing backing crashes is to have the vehicle intervene on behalf of

the driver by automatically applying the brakes when a possible collision is detected. Kochhar et al.

(2012) examined driver responses to a reverse braking aid that issued an auditory warning and applied

the brakes to stop the vehicle. All the drivers put their foot on the brake pedal after automatic braking

occurred in the presence of an obstacle behind the vehicle, and half looked immediately at the camera

display following the braking intervention. Preliminary evaluations of frontal collision avoidance systems,

systems that intervene autonomously rather than relying on driver intervention, indicate autonomous

interventions may prove to be more effective at mitigating frontal collisions than warning systems alone

(IIHS, 2012). Autonomous braking backup systems might prove similarly more effective at reducing the

13

risk of backover crashes than systems that display the area behind the vehicle or sensor systems that

alert drivers of potential hazards behind the vehicle.

To reduce backover crashes, NHTSA has proposed rulemaking to enhance the rear visibility of

passenger vehicles. Backup cameras currently are the only technology that will meet the proposed

requirements. The size of the image displayed may influence the use of backup camera displays to

detect obstacles behind the vehicle. With small displays, drivers may have trouble interpreting or

identifying images. Satoh et al. (1983) found that subjective judgments of image visibility using a

periscope mirror system were not possible for images with a subtended visual angle less than 5 minutes

of arc. Based on this work, NHTSA’s proposal would require the visual angle of three objects located 20

feet behind the vehicle to each exceed 3 minutes of arc and the average visual angle of these objects to

exceed 5 minutes of arc. The visual angle measured in every vehicle in the current study exceeded 3

minutes of arc and all but one vehicle exceeded 5 minutes of arc (Honda CRV). The CRV display was

among the smallest of all the study vehicles and positioned the farthest from the midpoint between a 50th

percentile male driver’s eyes. It is unclear if the Honda CRV’s smaller image would lead to significant

differences in object identification and collision avoidance compared with larger displays like that in the

Ford Escape (16.5 minutes of arc). This should be a topic for future research. However, in general, the

backup cameras in many of the study vehicles easily surpassed the proposed requirements for image

size.

One limitation of the study was that the vehicle sample was small. The results across vehicle

classes may not accurately reflect the relative rear visibility of all the 2010-13 model year vehicles in

different vehicle classes. However, it is worth noting that the patterns of results in this study are

consistent with other recent studies of rear visibility and studies of backover crash risk (Consumer

Reports, 2012; Mazzae & Garrott, 2008; Mazzae & Barickman, 2009).

Another limitation of the current study is that it relied on two human observers to make visibility

judgments. It is unlikely that the two observers had identical body motions and glances or even that

either observer had the same body motions and glances across the study trials. Although similar, the

body dimensions of the observers were not identical, and body dimensions like height are known to

influence rear visibility (Consumer Reports, 2012). A pilot study of three vehicles provided strong

evidence of high inter-observer reliability between the observers, suggesting their observations were

consistent with one another despite differences in body dimensions. The only field of view where inter-

observer reliability was an issue was for glances over the right shoulder. Observer DK had a limited

range of neck motion, which affected his ability to completely turn his head over his right shoulder.

In conclusion, the results showed that rear visibility typically is worse in larger vehicles compared

with smaller vehicles, but there were exceptions to this general pattern. A taller object simulating the

height of a 60-72-month-old child was easier to see than a smaller object simulating a 12-15-month-old

child, but the taller object still was not visible up to 13 feet behind the rear bumper. Backup cameras

nearly eliminated the blind zone behind vehicles, and parking sensors provided information that further

14

enhanced awareness of objects behind the vehicle. Hence, backup cameras, as well as the combination

of backup cameras and parking sensors, have the potential to substantially decrease the estimated

18,000 injuries and 292 fatalities that occur annually in backover crashes if drivers use the technology

appropriately.

ACKNOWLEDGEMENTS

The authors would like to thank all of the individuals at the Insurance Institute for Highway Safety

who helped obtain the vehicles, prepare vehicles for measurement, set up the measurement field, and

assisted with data collection. We express our gratitude to dealerships in the local Charlottesville, Virginia,

area for cooperating with us on this study. We especially want to thank Troy Vann and Dan Kellerman for

making visibility judgments. This work was supported by the Insurance Institute for Highway Safety.

REFERENCES

Austin, R. (2008). Fatalities and injuries in motor vehicle backing crashes report to congress (DOT HS 811 144). Washington, DC: National Highway Traffic Safety Administration.

Cohen, J.A. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37-46.

Consumer Reports. (2012). Best and worst rear blind zones. Retrieved from the Consumer Reports website: http://www.consumerreports.org/cro/cars/car-safety/car-safety-reviews/mind-that-blind-spot/best-and-worst/index.htm

Gordon, C.C., Bradtmiller, B., Churchill, T., Clauser, C.E., McConville, J.T., Tebbetts, I., & Walker, R.A. (1989). 1988 Anthropometric survey of U.S. Army personnel: Methods and summary statistics (Technical Report NATICK/TR-89/044). Natick Massachusetts: U.S. Army Natick RD&E Center.

Insurance Australia Group. (2013). Reversing visibility. Retrieved from the Insurance Australia Group website: http://iagresearch.com.au/index.php?option=com_content&task=view&id=279&Itemid=263

Insurance Institute for Highway Safety. (2004). Guidelines for using the UMTRI ATD positioning procedure for ATD and seat positioning (Version V). Arlington, VA. Accessed at http://www.iihs.org/ratings/protocols/pdf/ATD_positioning_procedure.pdf

Insurance Institute for Highway Safety. (2010). Bumper mismatch is still a problem. Status Report 45(12), 1-7.

Insurance Institute for Highway Safety. (2012). They’re working: insurance claims data show which new technologies are preventing crashes. Status Report 47(5), 1-3.

Kim, R., Rauschenberger, R., Heckman, G., Young, D., & Lange, R. (2012). Efficacy and usage patterns for three types of rearview camera displays during backing up. SAE 2012 World Congress & Exhibition (Paper no. 2012-01-0287). Warrendale, PA. : Society of Automotive Engineers.

Kochhar, D. S., Talamonti, Jr., W.J., & Tijerina, L. (2012). Driver response to unexpected automatic braking/haptic warning while backing (pp. 2211-2214). Proceedings of the Human Factors and Ergonomics Society 56th Annual Meeting. Santa Monica, CA: Human Factors and Ergonomics Society.

15

Llaneras, R. E., Green, C. A., Kiefer, R. J., Chundrlik, Jr., W. J., Altan, O. D., & Singer, J. P. (2005). Design and evaluation of a prototype rear obstacle detection and driver warning system. Human Factors, 47(1), 199-215.

Llaneras, R. Neurauter, M., & Green, C. (2011). Factors moderating the effectiveness of rear vision systems: What performance-shaping factors contribute to drivers’ detection and response to unexpected in-path obstacles when backing?. SAE 2011 World Congress & Exhibition (Paper No. 2011-01-0549). Warrendale, PA. : Society of Automotive Engineers.

Marathon. (2010). Marathon Premium Unleaded Gasoline Material Safety Data Sheet. Findlay, OH. Accessed at http://www.marathonpetroleum.com/brand/content/documents/mpc/msds/0126MAR019.pdf

Mazzae, E. N. (2010). Drivers’ use of rearview video and sensor-based backing aid systems in a non-laboratory setting. Washington, DC. National Highway Traffic Safety Administration.

Mazzae, E. N. (2013). Direct rear visibility measurement data: 2010-11 passenger cars and 2008-10 low-speed vehicles. Washington, DC. National Highway Traffic Safety Administration.

Mazzae, E. N. & Barickman, F. (2008). On-road study of drivers’ use of rearview video systems (ORSDURVS) (DOT HS 811 024). Washington, DC. National Highway Traffic Safety Administration.

Mazzae, E. N. & Barickman, F. (2009). Direct rear visibility of passenger cars: Laser-based measurement development and findings for late model vehicles (DOT HS 811 174). Washington, DC. National Highway Traffic Safety Administration.

Mazzae, E. N. & Garrott, W. R. (2006). Experimental evaluation of the performance of available backover prevention technologies (DOT HS 810 634). Washington, DC. National Highway Traffic Safety Administration.

Mazzae, E. N. & Garrott, W. R. (2008). Light vehicle rear visibility assessment (DOT HS 810 909). Washington, DC. National Highway Traffic Safety Administration.

Mazzae, E. N. & Garrott, W. R. (2011). Vehicle rearview image field of view and quality measurement (DOT HS 811 512). Washington, DC. National Highway Traffic Safety Administration.

Nhan, C., Rothman, L., Slater, M., & Howard, A. (2009). Back-over collisions in child pedestrians from the Canadian Hospitals Injury Reporting and Prevention Program. Traffic Injury Prevention, 10, 350-353.

Office of the Federal Register. (2010). Federal Register, vol. 75, no. 234. Pp. 76186-76250. National Highway Traffic Safety Administration, Docket No. NHTSA-2010-0162, 49 CFR Parts 571 and 585: Federal Motor Vehicle Safety Standard, Rearview Mirrors; Federal Motor Vehicle Safety Standard, Low-Speed Vehicles Phase-in Reporting Requirements. Washington, DC: National Archives and Records Administration.

Paine M, Macbeth A, Hnderson M. (2003). The danger to young pedestrians from reversing motor vehicles. Proceedings of the 18th International Technical Conference on the Enhanced Safety of Vehicles (Paper no. 466). Washington, DC: National Highway Traffic Safety Administration.

Pinkney K.A., Smith, A., Mann, C., Mower, G.D., Davis, D., & Dean, J.M. (2006). Risk of pediatric back-over injuries in residential driveways by vehicle type. Pediatric Emergency Care, 22(6), 402-406.

Reed, M.P., Lehto, M.M., & Flannagan, M.J. (2000). Field of view in passenger car mirrors (UMTRI-2000-23). Ann Arbor, MI: University of Michigan Transportation Research Institute.

16

Reed, M.P., Ebert, S.M., & Flannagan, M.J. (2001). Mirror field of view in light trucks, minivans, and sport utility vehicles (UMTRI-2001-1). Ann Arbor, MI: University of Michigan Transportation Research Institute.

Satoh H, Yamanaka A, Kondoh T, Yamashita M, Matsuzaki M, Akizuki K. (1983). Development of periscope mirror system. JSAE Review, 12, 66-72.

Tilley, A. R. (2002). The measure of man and woman: Human factors in design. New York, NY: John Wiley & Sons.

17

Table 1 Dimensions, available backing technology, and observer for each vehicle.

Backing technology

Dimensions (in.) Backup Rear

parking Vehicle Length Width Height camera sensor Observer Small car 2013 Chevrolet Cruze 4dr 181.0 70.7 58.1 x DK 2012 Ford Focus 5dr 171.6 71.8 57.7 x x DK

Midsize car 2013 Honda Accord 4dr 191.4 72.8 57.7 x TV 2013 Hyundai Sonata 4dr 189.8 72.2 57.9 x DK 2012 Toyota Camry 4dr 189.2 71.7 57.9 x TV

Large car 2012 Audi A6 Quattro 4dr 4WD 193.9 73.7 57.8 x x TV 2012 Infiniti M37 4dr 2WD 194.7 72.6 59.1 x TV 2011 Mercedes E Class 4 dr 2WD 191.7 75.9 57.9 x DK

Small SUV 2013 Ford Escape 4 dr 4WD 178.1 72.4 66.3 x DK 2013 Honda CR-V 4dr 2WD 178.3 71.6 64.7 x TV

Midsize SUV 2011 Acura MDX 4 dr 4WD 191.6 78.5 68.2 x DK 2013 Chevrolet Equinox 4dr 4WD 187.8 72.5 66.3 x x DK 2012 Mazda CX-9 4dr 4WD 200.2 76.2 68.0 x TV 2013 Mercedes GLK Class 4WD 178.3 74.3 66.9 x DK 2012 Toyota 4Runner 4WD 189.9 75.8 71.5 x DK

Large SUV 2013 Cadillac Escalade 4dr 4WD 202.5 79.0 75.9 x x DK 2012 GMC Acadia 4WD 200.7 78.2 72.8 x x TV

Minivan 2013 Dodge Grand Caravan 2WD 202.8 78.7 67.9 x x DK 2010 Honda Odyssey 202.1 77.1 68.8 x x DK

Large pickup 2012 Ford F-150 Super Cab Pickup 4x4 243.9 79.2 79.2 x x DK 2012 Toyota Tundra CrewMax 4WD 228.7 79.9 76.0 x TV

Table 2 Stature and seated height of candidate observers and a 50th percentile male.

Observer Stature (in.) Seated eye height (in.) TV* 70.5 33.8 DK* 69.7 31.8 JS 69.1 32.3 DP 69.1 31.8 50th percentile male 69.1 31.2 *Observers used in the full study

18

Table 3 Blind zone and average minimum sight distance without backing technology as a function of target height.

Blind zone (ft2) Average minimum sight distance (ft)

Target height (in.) Target height (in.) Vehicle 30.2 36.8 42.7 30.2 36.8 42.7 Small car Chevrolet Cruze 230 157 97 24.1 15.3 8.1 Ford Focus 118 80 50 15.5 10.8 7.0 Class mean 174 118.5 73.5 19.8 13.0 7.6

Midsize car Honda Accord 175 105 57 18.4 12.0 6.1 Hyundai Sonata 263 187 121 30.3 21.0 13.4 Toyota Camry 201 147 97 20.4 14.8 8.6 Class mean 213 146.33 91.67 23.0 15.9 9.4

Large car Audi A6 138 89 54 15.5 9.9 5.5 Infiniti M37 265 193 120 31.3 22.9 13.8 Mercedes E class 191 132 76 22.6 15.3 7.3 Class mean 198 138 83.33 23.1 16.0 8.8

Small SUV Ford Escape 223 158 111 24.5 17.9 12.4 Honda CRV 264 190 130 30.0 22.5 15.1 Class mean 243.5 174 120.5 27.3 20.2 13.8

Midsize SUV Acura MDX 360 226 156 38.1 25.6 17.9 Chevrolet Equinox 277 200 131 34.3 25.8 17 Mazda CX-9 323 236 151 41.3 32.6 18.8 Mercedes GLK 279 203 148 25.8 18.5 12.6 Toyota 4Runner 239 158 96 26.4 16.8 11.5 Class mean 295.6 204.6 136.4 33.2 23.9 15.6

Large SUV Cadillac Escalade 326 241 171 32.5 26.0 19.1 GMC Acadia 305 226 153 38.6 29.0 19.5 Class mean 315.5 233.5 162 35.6 27.5 19.3

Minivan Dodge Grand Caravan 238 181 148 25.1 18.3 14.6 Honda Odyssey 242 172 118 29.1 21.0 14.4 Class mean 240 176.5 133 27.1 19.6 14.5

Large pickup Ford F-150 185 169 140 23.8 21.9 18.3 Toyota Tundra 203 170 125 25.3 21.5 15.9 Class mean 194 169.5 132.5 24.5 21.7 17.1

Overall mean 240.24 172.38 116.67 27.3 20.0 13.2

19

Table 4 Blind zone and average minimum sight distance with backing technology as a function of target height.

Blind zone

with technology (ft2) Average minimum sight distance (ft)

Target height (in.) Target height (in.) Vehicle 30.2 36.8 42.7 30.2 36.8 42.7 Small car Chevrolet Cruze 29 26 23 1.1 ≤1 ≤1 Ford Focus 0 0 0 ≤1 ≤1 ≤1 Class mean 14.5 13 11.5 ≤1 ≤1 ≤1

Midsize car Honda Accord 18 14 13 1.3 ≤1 ≤1 Hyundai Sonata 27 25 21 1.3 1.3 1.3 Toyota Camry 29 20 19 1.1 1.1 1.1 Class mean 24.67 19.67 17.67 1.2 1.1 1.1

Large car Audi A6 14 11 11 ≤1 ≤1 ≤1 Infiniti M37 19 16 16 ≤1 ≤1 ≤1 Mercedes E class 23 23 17 1.8 1.8 1.4 Class mean 18.67 16.67 14.67 1.2 1.2 ≤1

Small SUV Ford Escape 38 22 19 1.8 ≤1 ≤1 Honda CRV 22 12 11 ≤1 ≤1 ≤1 Class mean 30 17 15 1.4 ≤1 ≤1

Midsize SUV Acura MDX 31 16 14 1.3 ≤1 ≤1 Chevrolet Equinox 9 0 0 ≤1 ≤1 ≤1 Mazda CX-9 17 14 9 1.5 1.5 ≤1 Mercedes GLK 42 30 28 1.5 1.3 1.3 Toyota 4Runner 24 18 12 1.4 1.1 ≤1 Class mean 24.6 15.6 12.6 1.3 1.2 ≤1

Large SUV Cadillac Escalade 36 28 14 ≤1 ≤1 ≤1 GMC Acadia 3 1 0 ≤1 ≤1 ≤1 Class mean 19.5 14.5 7 ≤1 ≤1 ≤1

Minivan Dodge Grand Caravan 32 26 26 ≤1 ≤1 ≤1 Honda Odyssey 29 16 14 ≤1 ≤1 ≤1 Class mean 30.5 21 20 ≤1 ≤1 ≤1

Large pickup Ford F-150 0 0 0 ≤1 ≤1 ≤1 Toyota Tundra 9 6 5 ≤1 ≤1 ≤1 Class mean 4.5 3 2.5 ≤1 ≤1 ≤1

Overall mean 21.5 15.4 13.0 1.1 ≤1 ≤1

20

Table 5A Blind zone with different combinations of backing technologies.

Blind zone without

technology (ft2)

Blind zone including parking sensors only (ft2)

Blind zone including backup camera only (ft2)

Blind zone with all backing technologies (ft2)

Target height (in.) Target height (in.) Target height (in.) Target height (in.) Vehicle 30.2 36.8 42.7 30.2 36.8 42.7 30.2 36.8 42.7 30.2 36.8 42.7 Small car Chevrolet Cruze 230 157 97 — — — 29 26 23 29 26 23 Ford Focus 118 80 50 94 56 26 4 4 4 0 0 0 Class mean 174.0 118.5 73.5 94.0 56.0 26.0 16.5 15.0 13.5 14.5 13.0 11.5

Midsize car Honda Accord 175 105 57 — — — 18 14 13 18 14 13 Hyundai Sonata 263 187 121 — — — 27 25 21 27 25 21 Toyota Camry 201 147 97 — — — 29 20 19 29 20 19 Class mean 213.0 146.3 91.7 — — — 24.7 19.7 17.7 24.7 19.7 17.7

Large car Audi A6 138 89 54 106 57 28 18 15 15 14 11 11 Infiniti M37 265 193 120 — — — 19 16 16 19 16 16 Mercedes E class 191 132 76 — — — 23 23 17 23 23 17 Class mean 198.0 138.0 83.3 106.0 57.0 28.0 20.0 18.0 16.0 18.7 16.7 14.7

Small SUV Ford Escape 223 158 111 — — — 38 22 19 38 22 19 Honda CRV 264 190 130 — — — 22 12 11 22 12 11 Class mean 243.5 174.0 120.5 — — — 30.0 17.0 15.0 30.0 17.0 15.0

Midsize SUV Acura MDX 360 226 156 — —

31 16 14 31 16 14 Chevrolet Equinox 277 200 131 244 170 102 14 3 2 9 0 0 Maxda CX-9 323 236 151 — — — 17 14 9 17 14 9 Mercedes GLK 279 203 148 — — — 42 30 28 42 30 28 Toyota 4Runner 239 158 96 — — — 24 18 12 24 18 12 Class mean 295.6 204.6 136.4 244.0 170.0 102.0 25.6 16.2 13.0 24.6 15.6 12.6

Large SUV Cadillac Escalade 326 241 171 274 193 135 41 33 15 36 28 14 GMC Acadia 305 226 153 263 184 117 4 2 1 3 1 0 Class mean 315.5 233.5 162.0 268.5 188.5 126.0 22.5 17.5 8.0 19.5 14.5 7.0

Minivan Dodge Grand Caravan 238 181 148 201 151 118 38 29 29 32 26 26 Honda Odyssey 242 172 118 208 142 88 33 18 16 29 16 14 Class mean 240.0 176.5 133.0 204.5 146.5 103.0 35.5 23.5 22.5 30.5 21.0 20.0

Large pickup Ford F-150 185 169 140 149 133 104 2 2 2 0 0 0 Toyota Tundra 203 170 125 9 6 5 9 6 5 Class mean 194.0 169.5 132.5 149.0 133.0 104.0 5.5 4.0 3.5 4.5 3.0 2.5

Overall mean

240.2 172.4 116.7 192.4 135.8 89.8 23.0 16.6 13.9 21.5 15.4 13.0

21

Table 5B Blind zone with different combinations of backing technologies and the percent reduction in blind zone provided by each.

Percent reduction in blind zone provided by parking

sensor system alone

Percent reduction in blind zone provided by backup

camera alone

Percent reduction in blind zone provided by both technologies combined

Target height (in.) Target height (in.) Target height (in.) Vehicle 30.2 36.8 42.7 30.2 36.8 42.7 30.2 36.8 42.7 Small car Chevrolet Cruze — — — 87.4 83.4 79.3 87.4 83.4 79.3 Ford Focus 20.3 30.0 48.0 96.6 95.0 92.0 100.0 100.0 100.0 Class mean 20.3 30.0 48.0 92.0 89.2 85.7 93.7 91.7 89.7

Midsize car Honda Accord — — — 89.7 86.7 77.2 89.7 86.7 77.2 Hyundai Sonata — — — 89.7 86.6 82.6 89.7 86.6 82.6 Toyota Camry — — — 85.6 86.4 80.4 85.6 86.4 80.4 Class mean — — — 88.3 86.6 80.1 88.3 86.6 80.1

Large car Audi A6 23.2 36.0 48.1 87.0 83.1 72.2 89.9 87.6 79.6 Infiniti M37 — — — 92.8 91.7 86.7 92.8 91.7 86.7 Mercedes E class — — — 88.0 82.6 77.6 88.0 82.6 77.6 Class mean 23.2 36.0 48.1 89.3 85.8 78.8 90.2 87.3 81.3

Small SUV Ford Escape — — — 83.0 86.1 82.9 83.0 86.1 82.9 Honda CRV — — — 91.7 93.7 91.5 91.7 93.7 91.5 Class mean — — — 87.4 89.9 87.2 87.4 89.9 87.2

Midsize SUV Acura MDX

91.4 92.9 91.0 91.4 92.9 91.0 Chevrolet Equinox 11.9 15.0 22.1 94.9 98.5 98.5 96.8 100.0 100.0 Maxda CX-9 — — — 94.7 94.1 94.0 94.7 94.1 94.0 Mercedes GLK — — — 84.9 85.2 81.1 84.9 85.2 81.1 Toyota 4Runner — — — 90.0 88.6 87.5 90.0 88.6 87.5 Class mean 11.9 15.0 22.1 91.2 91.9 90.4 91.6 92.2 90.7

Large SUV Cadillac Escalade 16.0 19.9 21.1 87.4 86.3 91.2 89.0 88.4 91.8 GMC Acadia 13.8 18.6 23.5 98.7 99.1 99.3 99.0 99.6 100.0 Class mean 14.9 19.3 22.3 93.1 92.7 95.3 94.0 94.0 95.9

Minivan Dodge Grand Caravan 15.5 16.6 20.3 84.0 84.0 80.4 86.6 85.6 82.4 Honda Odyssey 14.0 17.4 25.4 86.4 89.5 86.4 88.0 90.7 88.1 Class mean 14.8 17.0 22.8 85.2 86.8 83.4 87.3 85.3 81.5

Large pickup Ford F-150 19.5 21.3 25.7 98.9 98.8 98.6 100.0 100.0 100.0 Toyota Tundra 95.6 96.5 96.0 95.6 96.5 96.0 Class mean 19.5 21.3 25.7 97.3 97.7 97.3 97.8 98.3 98.0

Overall mean

16.8 21.8 29.3 90.4 89.9 87.0 91.1 90.8 88.1

22

Table 6 Backup camera display characteristics and rear parking sensor system warning characteristics.

Backup camera

Display Distance

from eye to Visual angle of

12-inch wide target Rear parking sensor system Vehicle Display location size (in.) display (in.) (minutes of arc) Auditory warning Visual warning Small car Chevrolet Cruze Center console 7 32.5 15.4 — — Ford Focus Center console 8 30.0 14.1 Graded, pulse rate increased

when object was closer —

Midsize car Honda Accord Center console 8 35.5 11.7 — — Hyundai Sonata Center console 7 30.3 13.8 — — Toyota Camry Center console 6.1 29.1 13.4 — — Large car Audi A6 Center console 8 31.2 14.3 Graded, pulse rate increased

when object was closer Graded color scheme (red imminent, yellow

cautionary); schematic of vehicle with distance and location specific information

Infiniti M37 Center console 8.0 31.1 14.3 — — Mercedes E class Center console 7 30.9 17.5 — — Small SUV Ford Escape Center console 8 31.0 16.5 — — Honda CRV Center console 5 39.8 5.0 — — Midsize SUV Acura MDX Center console 8 39.0 11.1 — — Chevrolet Equinox Center console 7 33.0 14.1 Graded, pulse rate increased

when object was closer Graded color scheme (red imminent, yellow cautionary); distance and location specific cues overlaid on backup camera image

Mazda CX-9 Center console 7 30.6 13.8 — — Mercedes GLK Center console 7 29.9 12.4 — — Toyota 4Runner Center console 6.1 31.0 13.7 — — Large SUV Cadillac Escalade Center console 8 32.8 13.2 Graded, pulse rate increased

when object was closer Graded color scheme (red imminent, yellow cautionary); distance and location specific cues overlaid on backup camera image

GMC Acadia Center console 6.5 30.8 13.3 Graded, pulse rate increased when object was closer

Graded color scheme (red imminent, yellow cautionary); distance and location specific cues overlaid on backup camera image

Minivan Dodge Grand Caravan Center console 6.5 29.9 12.2 Graded, pulse rate increased

when object was closer Schematic of vehicle with distance and location specific information

Honda Odyssey Center console 8 33.5 14.3 Graded, pulse rate increased when object was closer

Text cue and schematic of vehicle with distance and location specific information

Large pickup Ford F-150 Rearview mirror 3.3 24.4 7.9 Graded, pulse rate increased

when object was closer —

Toyota Tundra Rearview mirror 3.2 23.7 8.0 — —

23

(a) (b) Figure 1. (a) Visual target and (b) dimensions

Figure 2. Measurement field, 20 by 75 feet

24

Figure 3. Example backup camera display with target to evaluate image size

Figure 4. Blind zone as a function of target height for different vehicle classes.

0

50

100

150

200

250

300

350

Small car Midsize car Large car Small SUV Midsize SUV Large SUV Minivan Large pickup

Blin

d zo

ne a

rea

(ft2 )

30.2"

36.8"

42.7"

Target height

25

Figure 5. Average minimum sight distance as a function of target height for different vehicle classes.

Figure 6. Blind zone with backing technology for different vehicle classes as a function of target height.

0

5

10

15

20

25

30

35

40


Ave

rage

min

imum

sig

ht d

ista

nce

(ft)

30.2"

36.8"

42.7"

0

5

10

15

20

25

30

35

40


Blin

d zo

ne a

rea

(ft2 )

30.2" 36.8" 42.7"

Target height

Target height

26

Figure 7. Blind zone for a 30.2-inch tall object in the Hyundai Sonata (orange) and Ford F-150 (purple).

Figure 8. Blind zone for a 30.2-inch tall object in the Infiniti M37 (blue) and Dodge Grand Caravan (brown).

APPENDIX

REAR VISIBILITY MAPS FOR EACH TARGET HEIGHT AND BACKUP CAMERA DISPLAY IMAGE

FOR EACH VEHICLE IN ALPHABETICAL ORDER

A-1

2011 Acura MDX 4 dr 4WD: Visibility map for 30.2-inch tall target


A-2


A-3

2011 Acura MDX 4 dr 4WD: Backup camera display

A-4

2012 Audi A6 Quattro 4dr 4WD: Visibility map for 30.2-inch tall target


A-5


A-6

2012 Audi A6 Quattro 4dr 4WD: Backup camera display

A-7

2013 Cadillac Escalade 4dr 4WD: Visibility map for 30.2-inch tall target


A-8


A-9

2013 Cadillac Escalade 4dr 4WD: Backup camera display

A-10

2013 Chevrolet Cruze 4dr: Visibility map for 30.2-inch tall target


A-11


A-12

2013 Chevrolet Cruze 4dr: Backup camera display

A-13

2013 Chevrolet Equinox 4dr 4WD: Visibility map for 30.2-inch tall target


A-14


A-15

2013 Chevrolet Equinox 4dr 4WD: Backup camera display

A-16

2013 Dodge Grand Caravan 2WD: Visibility map for 30.2-inch tall target


A-17


A-18

2013 Dodge Grand Caravan 2WD: Backup camera display

A-19

2013 Ford Escape 4 dr 4WD: Visibility map for 30.2-inch tall target


A-20


A-21

2013 Ford Escape 4 dr 4WD: Backup camera display

A-22

2012 Ford F150 super cab pickup 4x4: Visibility map for 30.2-inch tall target


A-23


A-24

2012 Ford F150 super cab pickup 4x4: Backup camera display

A-25

2012 Ford Focus 5dr: Visibility map for 30.2-inch tall target


A-26


A-27

2012 Ford Focus 5dr: Backup camera display

A-28

2012 GMC Acadia 4WD: Visibility map for 30.2-inch tall target


A-29


A-30

2012 GMC Acadia 4WD: Backup camera display

A-31

2013 Honda Accord 4dr: Visibility map for 30.2-inch tall target


A-32


A-33

2013 Honda Accord 4dr: Backup camera display

A-34

2013 Honda CR-V 4dr 2WD: Visibility map for 30.2-inch tall target


A-35


A-36

2013 Honda CR-V 4dr 2WD: Backup camera display

A-37

2010 Honda Odyssey: Visibility map for 30.2-inch tall target


A-38


A-39

2010 Honda Odyssey: Backup camera display

A-40

2013 Hyundai Sonata 4dr: Visibility map for 30.2-inch tall target


A-41


A-42

2013 Hyundai Sonata 4dr: Backup camera display

A-43

2012 Infiniti M37 4dr 2WD: Visibility map for 30.2-inch tall target


A-44


A-45

2012 Infiniti M37 4dr 2WD: Backup camera display

A-46

2012 Mazda CX-9 4dr 4WD: Visibility map for 30.2-inch tall target


A-47


A-48

2012 Mazda CX-9 4dr 4WD: Backup camera display

A-49

2011 Mercedes E class 4dr 2WD: Visibility map for 30.2-inch tall target


A-50


A-51

2011 Mercedes E class 4dr 2WD: Backup camera display

A-52

2013 Mercedes GLK class 4WD: Visibility map for 30.2-inch tall target


A-53


A-54

2013 Mercedes GLK class 4WD: Backup camera display

A-55

2012 Toyota 4Runner 4WD: Visibility map for 30.2-inch tall target


A-56


A-57

2012 Toyota 4Runner 4WD: Backup camera display

A-58

2012 Toyota Camry 4dr: Visibility map for 30.2-inch tall target


A-59


A-60

2012 Toyota Camry 4dr: Backup camera display

A-61

2012 Toyota Tundra CrewMax 4WD: Visibility map for 30.2-inch tall target


A-62


A-63

2012 Toyota Tundra CrewMax 4WD: Backup camera display

iihs comment to nhtsa concerning ncap update for rearview

Documents