Accident Analysis of Older Drivers at Intersections
(FHWA–RD–94–021)

Federal Highway Administration
6300 Georgetown Pike
McLean, VA 22101–2296

COTR:
Not Specified

Anonymous

1995

5

http://www.tfhrc.gov/safety/hsis/94-021.htm

Crash/Demographics Statistical Analysis

Normal

Not Specified

To examine the specific nature of intersection–related crashes involving elderly drivers through a detailed analysis of crash data from the Highway Safety Information System (HSIS).

The analyses were conducted as part of the FHWA research study, "Traffic Operations Control for Older Drivers." The authors used HSIS data from 1985 to 1987 in Minnesota and Illinois for this research.

• For all of the analyses, comparisons were made among three age groups: (1) "young elderly" (ages 65 to 74), (2) "old elderly" (age 75 and older), and (3) a middle–aged comparison group (ages 30 to 50).
• The crash types at both urban and rural signalized and stop–controlled intersections were examined separately, as well as the type of vehicle maneuver prior to the crash and the investigating officer’s judgment regarding "causal" factors.

Aged Drivers, Intersections, Traffic Accidents, Accident Data, Elderly Drivers, Older Drivers

• The general analyses of crash type in both States indicated that at both urban and rural signalized intersections, elderly drivers were less likely than their middle–aged counterparts to be involved in rear–end collisions, but more likely to be involved in left–turn and angle collisions.
• In both States, right–angle collisions presented a particular problem for elderly drivers at both urban and rural stop–controlled intersections.
• For turning collisions at urban and rural signalized intersections, middle–aged drivers tended to have been going straight, while older drivers were more likely to have been turning left, and were slightly more likely to be turning right and turning right on red (see table below).
• In right–angle collisions at both urban and rural stop–controlled intersections, elderly drivers were more likely than middle–aged drivers to have been starting from a stop.
• In turning collisions, they were more likely to be turning left or right across traffic.
• The examination of the "contributing factors" cited by the officer showed that the middle–aged driver was consistently more likely to have been cited as having exhibited "no improper driving," while the elderly drivers were more likely to have been cited for "failure to yield."

• The crash analyses indicated that both the "young elderly" (ages 65 to 74) and the "old elderly" (age 75 and older) appear to have problems at intersections.
• These problems often involve left–turning maneuvers (at signalized intersections) and turning or "entering" maneuvers at stop–controlled intersections.
• It appears that the problems experienced by elderly drivers involved in crashes either relate to the difficulties in distinguishing target vehicles from surrounding clutter, judging the closing speeds of target vehicles, and/or an inability to use the acceleration capabilities of the cars they are driving.

None

Guidance for Implementation of the AASHTO Strategic Highway Safety Plan, Volume 12: A Guide for Reducing Collisions at Signalized Intersections, NCHRP Report 500

National Cooperative Highway
Research Program
Transportation Research Board
500 Fifth Street, N.W.
Washington, DC 20001

COTR:
Not Specified

Antonucci, N.D., Hardy, K.K., Slack, K.L., Pfefer, R., and Neuman, T.R.

2004

133

http://www.trb.org/publications/nchrp/nchrp_rpt_500v12.pdf

Guidelines

Normal

All

This implementation guide provides guidance to highway agencies that want to implement safety improvements at signalized intersections and includes a variety of strategies that may be applicable to particular locations. While the focus of the strategies discussed in this guide is on reducing fatalities at signalized intersections, the implementation of many of these strategies will probably lead to an overall reduction in intersection crashes.

See Methods.

The strategies in this guide were identified from a number of sources, including recent literature, contact with State and local agencies throughout the United States, and Federal programs. Some of the strategies are widely used, while others are used at a State or local level in limited areas. Some have been subjected to well–designed evaluations to prove their effectiveness. On the other hand, it was found that many strategies, including some that are widely used, have not been adequately evaluated.

The implication of the widely varying experience with these strategies, as well as the range of knowledge about their effectiveness, is that the reader should be prepared to exercise caution in many cases before adopting a particular strategy for implementation. To help the reader, the strategies have been classified into three types, each identified by a letter symbol throughout the guide: Proven (P), Tried (T), and Experimental (E).

Guidance for implementation of the American Association of State Highway and Transportation Officials (AASHTO) Strategic Highway Safety Plan (SHSP) is provided. An overview of an 11–step model process for implementing the program of strategies is presented.

Highway Safety, Signalized Intersections, Intersection Crashes, Collision Reduction, Guidelines

Most of the strategies in this guide are low–cost, short–term treatments to improve safety at signalized intersections, consistent with the focus of the entire AASHTO SHSP. For each of these strategies, a detailed discussion of the attributes, effectiveness, and other key factors is presented. Several higher cost, longer term strategies that have been proven effective in improving safety at signalized intersections are also presented, but in less detail. Safety improvement measures include geometric design modifications, changes to traffic control devices, enforcement, and education.

The table below lists the objectives and related strategies for improving safety at signalized intersections.

P = Proven, T = Tried, and E = Experimental

Source: Guidance for Implementation of the AASHTO Strategic Highway Safety Plan, Volume 12: A Guide for Reducing Collisions at Signalized Intersections, National Cooperative Highway Research Program (NCHRP) Report 500, Transportation Research Board, Washington, DC, 2004, p. V–2. Reprinted with permission.

This report comprises volume 12 of a series of implementation guides addressing the emphasis areas of the AASHTO Strategic Highway Safety Plan, NCHRP Project 17–18(3).

Statistical Models for At–Grade Intersection Accidents,
Addendum (FHWA–RD–99–094)

Office of Safety and Traffic Operations
Research and Development
Federal Highway Administration
6300 Georgetown Pike
McLean, VA 22101–2296

COTR:
Joe Bared

Bauer, K.M., and Harwood, D.W.

March 2000

68

http://www.tfhrc.gov/safety/ihsdm/libweb.htm

Crash/Demographic Statistical Analysis

Normal

All

This report is an addendum to the work published in Statistical Models of At–Grade Intersection Accidents (FHWA–RD–96–125) (Bauer and Harwood, 1996). The objective of both research studies was to develop statistical models of the relationship between traffic crashes and highway geometric elements for at–grade intersections.

While the previously published report used only multiple–vehicle crashes in developing predictive models, this addendum presents models based on all collision types (including both multiple–vehicle and single–vehicle crashes).

• The statistical modeling approaches used in the research included lognormal, Poisson, and negative binomial regression analyses. The models for all collision types are similar to those developed in the previous report for multiple–vehicle crashes.
• The analyses include all collision types (i.e., both multiple– and single–vehicle crashes) using 3–year crash frequencies (1990 to 1992) and geometric design, traffic control, and traffic volume data from a database provided by Caltrans (California DOT).
• The data used for the analyses reported in this addendum are in all respects identical to those used for the previous report, except that all collision types were included in the crash frequencies used as the dependent variable in modeling.
• Statistical modeling results for five specific types of intersections are discussed in this report.

Accident Modeling, Traffic Accidents, Geometric Design, At–Grade Intersections, Poisson Regression,Negative Binomial Regression, Lognormal Regression

• The modeling results for crashes if all collision types are combined are similar to those that were found for multiple–vehicle crashes only.
• Geometric design variables accounted for only a small additional portion of the variability.
• Generally, negative binomial regression models were developed to fit the crash data at rural, three– and four–leg, stop–controlled intersections, and at urban, three–leg, stop–controlled intersections.
• Lognormal regression models were found to be more appropriate for modeling crashes at urban, four–leg, stop–controlled intersections, and at urban, four–leg, signalized intersections.
• The lognormal and negative binomial regression models developed to represent the relationships between crashes of all collision types and intersection geometric design, traffic control, and traffic volume variables explained between 16 and 39 percent of the variability in the crash data.
• In all regression models, the major–road average daily traffic (ADT) and crossroad ADT variables accounted for most of the variability in crash data that was explained by the models. Generally, geometric design variables accounted for only a small additional portion of the variability.
• Because of the overdispersion observed in the crash data, the negative binomial distribution was preferred over the Poisson distribution when using a loglinear model.

• The negative binomial and lognormal distributions appear to be better suited to modeling of crash relationships than the normal distribution.
• The form of the statistical distribution selected for modeling any particular type of intersection should be chosen based on a review of the crash frequency distribution for that type of intersection.
• The models do not include the effects for all geometric variables of potential interest to highway designers,and some of the effects they do include are in a direction opposite to that expected. Furthermore, the goodness of fit of the models is not as high as desired. Therefore, the models presented here are appropriate as a guide to future research, but do not appear to be appropriate for direct application in the field.

None

Statistical Models of At–Grade Accidents (FHWA–RD–96–125)

Office of Safety and Traffic Operations
Research and Development
Federal Highway Administration
6300 Georgetown Pike
McLean, VA 22101–2296

COTR:
Joe Bared

Bauer, K.M., and Harwood, D.W.

November 1996

157

None

Crash/Demographics Statistical Analysis, Field Test

Normal

All

To develop statistical models of the relationship between traffic crashes and highway geometric elements for at–grade intersections.

Statistical models were developed based on document reviews from a number of sources and results from a pilot field study. The review was limited to multiple–vehicle crash data.

Several major technical tasks were performed during the research, including:

• A review of previously published and unpublished literature and ongoing studies concerning the relationship between traffic crashes and intersection geometrics, as well as between traffic crashes and highway geometric design features in general.
• A review of existing policies, guidelines, standards, and practices for design of at–grade intersections.
• A review of existing highway agency files containing geometric design, traffic control, traffic volume, and crash data, including the databases in the FHWA Highway Safety Information System (HSIS). The Caltrans database was used for developing statistical models and testing statistical approaches.
• Statistical models for the relationships between traffic crashes and geometrics were developed. Alternative modeling approaches were investigated based on various assumptions about the distribution of crashes, including the Poisson, lognormal, negative binomial, and logistic distributions. The goodness of fit of these various alternative models and the role of geometric design variables in those models were assessed. Statistical models were developed for five specific types of intersections.
• A pilot field study to collect data on additional geometric design variables and turning–movement volumes was conducted at a sample of the urban, four–leg, signalized intersections in California. Additional statistical analyses incorporating these field data were conducted.
• A review of hardcopy police accident reports was conducted to further investigate the role of geometric design features in the causation of intersection crashes.

Accident Modeling, Traffic Accidents, Geometric Design, At–Grade Intersections, Poisson Regression, Negative Binomial Regression, Lognormal Regression

• Regression models to determine the relationships between crashes and intersection geometric design, traffic control, and traffic volume variables based on the negative binomial distribution explained between 16 and 38 percent of the variability in the crash data.
• Models developed to predict total multiple–vehicle crashes generally performed slightly better than did models for fatal and injury multiple–vehicle crashes.
• In the modeling of crashes for at–grade intersections, overdispersion was commonly observed and,therefore, the negative binomial distribution was preferred.
• In general, the consideration of major–road ADT and crossroad ADT as separate independent variables provided better modeling results than consideration of a single variable representing either the sum or the product of the two ADT variables.
• In negative binomial regression models for three of five specific intersection types, the major–road ADT and crossroad ADT variables accounted for most of the variability in crash data that was explained by the models. Geometric design variables accounted for a very small additional portion of the variability.
• Addition of field data to the existing data set did not increase the proportion of variation in the crashes that was explained by the lognormal regression models.
• The models do not include the effects of all of the geometric variables of potential interest to highway designers, and some of the effects they do include are in a direction opposite to that expected. Furthermore, the goodness of fit of the models is not as high as desired.

The following conclusions were reached as a result of the statistical analysis of the relationships between traffic crashes and the geometrics of at–grade intersections conducted in this research.

• Traditional multiple linear regression is generally not an appropriate statistical approach to modeling of crash relationships because crashes are discrete, nonnegative events that often do not follow a normal distribution.
• The Poisson, negative binomial, lognormal, and logistic distributions appear to be better suited to modeling of crash relationships than the normal distribution. In all cases, the form of the statistical distribution selected for any particular modeling should be chosen based on a review of the data to be modeled.
• Geometric design features explain relatively little of the variability in intersection crash data for at–grade intersections.
• The models presented here are appropriate as a guide to future research, but do not appear to be appropriate for direct application by practitioners.

An addendum to this report,Statistical Models of At–Grade Intersection Accidents, Addendum(FHWA–RD–99–094), was released in March 2000 and is reviewed separately.

Intersection Collision Avoidance Study, Final Report

Office of Safety
Federal Highway Administration
400 Seventh Street, S.W.
Washington, DC 20590

COTR:
Not Specified

Bellomo–McGee, Inc.

September 2003

79

None

Literature Review, Field Test

Normal

Not Specified

To define and evaluate infrastructure–only Intersection Collision Avoidance System (ICAS) concepts aimed at reducing the number of intersection crashes.

System engineering analyses were performed to define and evaluate the feasibility and effectiveness of alternative infrastructure–based advanced technology concepts. These included development of functional requirements and conceptual designs, and the testing of the feasibility of those designs at high–crash intersections in three States.

Literature Review:

• This included a review of crash studies, human factors work related to crash avoidance, and current advanced technology intersection safety countermeasures. Included in the literature review was an examination of technology, sensors, and displays capabilities.

Crash Analysis:

• Crashes were analyzed at selected sites within the Infrastructure Consortium (IC) States: Minnesota, California, and Virginia.
  • Each IC member State identified 20 high–incident intersections for review and analysis.
  • Police reports for 3 years of crashes provided a large database for analysis of crossing–path crashes. This database was used to determine primary crash types and causal factors.
  • A final step of this task was to select two sites from each State that would be candidates for implementing advanced intelligent countermeasures.

Define and Evaluate ICAS Concepts:

• This task included developing several concepts for reducing crossing–path crashes using intelligent vehicle systems and sensors, communication displays, etc.

Feasibility Testing at the Six Candidate Intersections:

• This was performed by collecting field data and applying it to the requirements of the particular concepts.

Intersection, Collision Avoidance, Infrastructure, Intersection Collision Avoidance System

• The project identified certain parameters required for characterizing traffic flow based on current Intelligent Transportation Systems (ITS) applications/concepts for traffic management.
• Information on human factors issues important to the selection and design of infrastructure–based technology was identified. These included driver age, vehicle gap acceptance, and response to emergency events.
• The three successive years of data showed that Left Turn Across Path of Opposite Direction (LTAP/OD), Straight Crossing Path (SCP), and Left Turn Across Path of Lateral Direction (LTAP/LD) crashes were the most frequent types of crash, regardless of whether or not the intersection was signalized.
• Crashes involving signal violation were mostly a result of not seeing the signal or its indication, or trying to "beat" the amber signal.
• Inability to judge available gaps in traffic and not seeing right–of–way vehicle were the main causal factors for crashes that did not involve signal violation.
• Based on the analyses of crashes and casual factors, six intersection collision avoidance concepts were developed. Four of the concepts involve timely communication of information to at–risk motorists, while the remaining two preempt the normal signal operation to prevent a crash.
• Feasibility analysis data showed that at all of the six candidate intersections, the suggested concept was feasible, based on the vehicle data collected at the site.
• The result of the cost–benefit analysis indicated that five of the six candidate intersections showed the potential to quickly recoup the expenses of design and installation of the suggested infrastructure–based collision countermeasure.

• Based on this work, it was determined that implementing an ICAS to address each of the three most prevalent types of intersection crashes was feasible. In addition, the cost–benefit analysis showed a quick recouping of ICAS implementation costs.
• Motorist response to roadside communication devices still requires extensive testing, as this is a critical requirement of several concepts.
• Recommended further studies pertain to increased onsite data collection to validate preliminary findings and human factors testing to meet the functional requirements of the operational concepts. Human factors testing consists of the evaluation of communications modes to inform and warn motorists.

None

Driver Understanding of Protected and Permitted Left–Turn Signal Displays

(Transportation Research Record 1464, pp. 42–50)

Civil Engineering Department
University of Nebraska–Lincoln
Lincoln, NE 68588–0531

COTR:
Not Specified

Bonneson, J.A., and McCoy, P.T.

1994

9

None

Survey

Normal

Not Specified

To determine if some protected and permitted left turn (PPLT) signal designs cause more confusion and operational and safety problems for drivers than others.

Driver comprehension of PPLT signal designs was evaluated by conducting a survey of 1,610 drivers. The survey included a perspective view of an intersection approach and its traffic signal display, followed by multiple–choice questions about the correct driving action.

Survey Questionnaire:

• On each survey, one perspective view of an intersection approach was shown at the top of the page and two multiple–choice questions asked the correct identification of a particular indication type.
• The survey questions focused on the following four display indications in six different PPLT designs:
  • Permitted left turn: Green ball for both the left turn and through movements.
  • Protected left turn only: Left–turn green arrow and through red ball, consistent with the Manual on Uniform Traffic Control Devices (MUTCD) specifications.
  • Overlapped left turn and through: Left–turn green arrow and through green ball.
  • Protected/Modified left turn only: Displayed only the green arrow in the PPLT signal head without the red ball.
• The six PPLT designs varied in terms of the location of the signal head with respect to the lane line, the arrangement of the lenses in the signal head, and the inclusion of an auxiliary sign.

Distribution Method:

• Survey was administered in three of Nebraska’s largest cities: Omaha, Lincoln, and Grand Island.
• Survey was administered in person at the local department of motor vehicles in each city.

Protected and Permitted Left Turn, Signal Design, Intersection Safety

Survey Demographics:

• Only 70 percent of the survey respondents correctly understood the meaning of the PPLT signal design.
• There was a trend toward a decreased understanding of the PPLT designs with increased age and driving experience.
• There was also a trend toward better understanding with more education.

Design Comparisons:

• The results indicated that drivers appear to have the best understanding of the exclusive vertical PPLT design. The difference in the results for this design and the least understood design is about 8 percent (see table).
• None of the differences between each design is significantly different. Although the differences suggest that some designs are better understood, a larger number of responses would be needed to confirm these trends.
• With regard to differences in understanding the various indications, the results indicate that the overlap indication is least understood (only about one–half of the drivers surveyed answered this question correctly).

Signal–Head Location and Sign Use:

• The exclusive head location increased driver understanding by about 4 to 5 percent over the shared head location.
• The results indicated that designs with a sign decrease driver understanding by about 6.5 percent. It was found that the use of a sign tends to confuse more drivers during the overlap and protected phases than it helps during the permitted phase.

From Transportation Research Record 1464, Transportation Research Board, National Research Council, Washington, DC, 1994,
table 2, p. 48. Reprinted with permission.

• The survey results indicated that the exclusive vertical PPLT design is correctly understood by the highest proportion of drivers.
• Of the three indications considered, the overlap indication is understood by the smallest number of respondents.
• The survey results indicate that drivers are better able to understand PPLT designs with any of the following characteristics: Modified protected indication, PPLT head centered over the opposing left–turn lane, and no auxiliary sign.

None

Review and Evaluation of Factors That Affect the Frequency of
Red–Light Running (FHWA/TX–02/4027–1)

Federal Highway Administration
6300 Georgetown Pike
McLean, VA 22101–2296

COTR:
Not Specified

Bonneson, J., Brewer, M., and Zimmerman, K.

September 2001

78

None

Literature Review, Crash/Demographic Statistical Analysis

Normal

Not Specified

To describe how traffic engineering countermeasures can be used to minimize the frequency of red–light running (RLR) and associated crashes at intersections.

This report describes the findings from the first year of a 2–year project. During the first year, studies were conducted on RLR frequency and crash rates at 12 intersection approaches in 3 Texas cities.

Field Data Collection:

• The field study at each site included the collection of a wide range of geometric, traffic flow, traffic control, and operational characteristics.
• These data were collected using a variety of methods, including video recorders, laser speed guns, and site surveys.

Safety Data Collection:

• The safety data collection activity consisted of the acquisition of historical crash records for each intersection included in the field studies.
• To facilitate the analysis, computerized databases were requested from the Texas Department of Public Safety and the appropriate city agencies.
• The request was for the most recent 36 months for which complete information was available and for all four approaches to each intersection. These data were used to quantify the relationship between RLR and crash frequency.

Signalized Intersection, Change Interval, Signal Timing Design, Dilemma Zone

• A review of the literature revealed that the following are influential factors in the RLR process: (1) flow rate on the subject approach, (2) number of signal cycles, (3) phase termination by max–out, (4) probability of stopping, (5) yellow interval duration, (6) all–red interval duration, (7) entry time of the conflicting driver, and (8) flow rate on the conflicting approach.
• A review of the literature also indicated that drivers are less likely to stop when they: (1) have a short travel time to the intersection, (2) have higher speeds, (3) are traveling in platoons, (4) are on steep downgrades,(5) are faced with relatively long yellow indications, and (6) are being closely followed.
• The duration of the yellow interval is generally recognized as a key factor that affects the frequency of RLR. Researchers suggest that the yellow interval should be based on the travel time of the 85^th (or 90^th) percentile driver. The corresponding yellow interval duration should range from 4.0 to 5.5 seconds (s) (with larger values appropriate for higher speed approaches).
• The countermeasures with the greatest potential to reduce RLR (as determined from the literature review) are listed in the table below.

¹Bolded countermeasures were selected for evaluation in this project.

• Analysis of approach volume on RLR frequency revealed that RLR frequency was highly correlated with the flow rate at the end of the phase. Other factors found to be correlated with the frequency of RLR include yellow interval duration and the percentage of heavy vehicles.
• Yellow intervals of less than 3.5 s appear to be associated with a significant number of RLR events per hour.
• The findings from these studies indicate that the frequency of RLR increases in a predictable way with increasing approach volume, increasing heavy–vehicle percentage, and shorter yellow interval durations.
• Crash data analyses indicate that right–angle crashes increase exponentially with an increasing frequency of RLR.
• Models for computing an intersection approach’s RLR frequency and related crash rate are described.

None

Engineering Countermeasures to Reduce Red–Light Running
(FHWA/TX–03/4027–2)

Federal Highway Administration
6300 Georgetown Pike
McLean, VA 22101–2296

COTR:
Not Specified

Bonneson, J., Zimmerman, K., and Brewer, M.

August 2002

122

None

Field Test

Normal

Not Specified

To describe how engineering countermeasures can be used to minimize the frequency of red–light running (RLR) and associated crashes.

This report describes the factors that are associated with RLR, as well as several countermeasures that have been used to reduce its frequency. Initially, there is an examination of the RLR process in terms of the events necessary to precipitate an RLR event. Then, various engineering countermeasures are identified. Next, a before/after study is described.

Field Study:

• During the first year, engineering countermeasures were identified and implemented at 10 intersections in 5 Texas cities.
• Before/after studies of RLR frequency were then conducted at two sites (i.e., approaches) at each of the 10 intersections.
• One or more of the five countermeasures identified were implemented at most of the sites.
• Data collection consisted of a wide range of geometric, traffic flow, traffic control, and operational characteristics.
• The data were collected using a variety of methods, including video recorders, laser speed guns, and site surveys.

Crash Data Analysis:

• The 3–year crash history for each intersection was compared to its observed frequency of RLR.
• Computerized databases were requested from the Texas Department of Public Safety and the appropriate city agencies.

Signalized Intersections, Change Interval, Yellow Interval, Red–Light Running

• Factors that lead to conflict: The following factors are related to the occurrence of RLR: (1) flow rate on the subject approach, (2) number of signal cycles, (3) phase termination by max–out, (4) probability of stopping, and (5) yellow interval duration.
• The results of the field study indicate that more than 10,018 signal cycles were observed at 20 intersection approaches. During these cycles, 586 vehicles entered the intersection (as defined by the stop line) after the change in signal indication from yellow to red. Of the 586 vehicles, 84 were heavy vehicles and 502 were passenger cars. Overall, 0.86 percent of heavy vehicles violated a red indication and 0.38 percent of passenger cars violated the red indication.
• The overall average RLR rates are 4.1 red–light runners per 1,000 vehicles and 1.0 red–light runners per 10,000 vehicle cycles.
• The following countermeasures were implemented at the intersection approaches, with the corresponding percent reduction in parentheses (the only countermeasure found to be statistically significant was the yellow interval duration increase):
  • Add light–emitting diode (LED) lighting to the yellow indication (49 percent reduction).
  • Increase the yellow interval duration (70 percent reduction).
  • Add backplates and increase yellow interval duration (18 percent reduction).
  • Increase cycle length and improve signal operation (uncertain effect).
  • Improve progression and increase cycle length (uncertain effect).
  • Add backplates and add LED lighting to the yellow indications (35 percent reduction).

• The typical intersection approach experiences from 3.0 to 5.0 red–light runners per 1,000 vehicles and 1.0 red–light runners per 10,000 vehicle cycles. An intersection with an RLR rate that is greater than that of the typical intersection should be the primary target of a treatment program.
• A heavy–vehicle operator is twice as likely to run the red indication as is a passenger car driver.
• RLR is more frequent at intersections with platoons arriving near the end of the green indication. Engineers developing signal coordination plans should avoid having platoons arrive near the end of the signal phase. If this situation cannot be avoided, then a longer cycle length should be used.
• About 80 percent of drivers that run red lights enter the intersection within 1.0 s after the end of the yellow cycle. Hence, engineering countermeasures focused on driver recognition of, and response to, the yellow indication are likely to be the most cost–effective.
• In addition to an increase in yellow interval duration, several other engineering countermeasures were identified as having the potential to reduce RLR. Specifically, it was found that the use of backplates would reduce RLR by 25 percent, a 20–s increase in cycle length would reduce RLR by 18 percent, and the use of yellow LEDs may reduce RLR by 13 percent.
• The findings indicate that the frequency of RLR decreases in a predictable way with decreasing approach flow rate, longer clearance path lengths, longer headways, and longer yellow interval durations.
• The crash data analyses indicate that right–angle crashes increase exponentially with an increasing frequency of RLR.

None

Analysis of Fatal Crashes Due to Signal and Stop Sign
Violations (DOT–HS–809–779)

National Highway Traffic Safety
Administration
400 Seventh Street, S.W.
Washington, DC 20590

COTR:
Not Specified

Campbell, B.N., Smith, J.D., and Najm, W.G.

September 2004

159

http://www-nrd.nhtsa.dot.gov/departments/nrd-12/pubs_rev.html

Crash/Demographic Statistical Analysis

Normal

Light Vehicles

This research supports the National Highway Traffic Safety Administration (NHTSA) in developing performance specifications for stop sign/traffic signal violations and insufficient gap warning systems (e.g., left turn across path).

Crash data for the analysis were obtained from the 1999–2000 Fatality Analysis Reporting System (FARS) crash databases. This report identified the crash scenarios, described the crash contributing factors, and characterized the infrastructure where fatal crashes occurred in 1999 and 2000.

• The analysis began with all 1999 and 2000 fatal crashes and then segregated the crashes by the type of traffic control device at the crash site.
• These crashes were then examined to determine whether the driver violated the traffic signal or stop sign and what type of violation occurred.
• Traffic control device violations were classified into two categories: (1) failure to obey and (2) failure to yield.
• Fatal crashes involving light vehicles that violated the traffic signal or stop sign were separated into single–vehicle, two–vehicle, and multiple–vehicle crash categories.

Light Vehicles, Crashes, Contributing Factors, Intelligent Vehicle Initiative, Fatal Crashes, Traffic Signals, Stop Signs, Violations, Precrash

• A total of 9,951 vehicles were involved in fatal crashes at traffic signals in 1999 and 2000—20 percent of these vehicles failed to obey the signal and 13 percent failed to yield the right of way.
• For crashes at stop signs, 13,627 vehicles were involved in fatal crashes —21 percent failed to obey the sign and 23 percent failed to yield the right of way.
• Single–vehicle crashes accounted for 8 percent and 6 percent, two–vehicle crashes accounted for 75 percent and 87 percent, and multiple–vehicle crashes accounted for 18 percent and 7 percent of all light–vehicle violation fatal crashes at traffic signals and stop signs, respectively.
• About 64 percent and 95 percent, respectively, of the "failure to obey" and "failure to yield" single–vehicle crashes at traffic signals were pedestrian crashes. On the other hand, 76 percent of the "failure to yield" crashes at stop signs were pedestrian crashes, while 95 percent of the "failure to obey" crashes at stop signs were other crashes such as run–off–road crashes.
• Single–vehicle traffic signal crashes primarily occurred in urban areas (91 percent), whereas 57 percent of stop sign crashes occurred in rural areas. Most single–vehicle crashes occurred on two–lane roadways regardless of the type of violation.
• Approximately 65 percent and 12 percent, respectively, of the "failure to obey" and "failure to yield" two–vehicle crashes were straight crossing–path crashes and, in contrast, 29 percent and 81 percent, respectively, were left crossing–path crashes.
• Straight crossing–path crashes were 2.24 times more likely than left–turn crossing–path crashes for "failure to obey" violations. In contrast, left–turn crossing–path crashes were 6.55 times more likely than straight crossin–gpath crashes for "failure to yield" right–of–way violations.
• In 1999 and 2000, there were 889 fatal multiple–vehicle crashes that involved violations by light vehicles. About 58 percent occurred at traffic signals, while the remaining 42 percent occurred at stop signs. At traffic signals, drivers failed to obey the signal in 67 percent of the crashes and failed to yield the right of way in the remaining 33 percent of the crashes.
• About 82 percent of multiple–vehicle fatal crashes at traffic signals occurred on urban roadways. Conversely, about 57 percent of multiple–vehicle fatal crashes at stop signs occurred on rural roadways.
• The majority (80 percent) of stop sign crashes occurred on two–lane roadways. On the other hand, half of the traffic signal crashes (50 percent) occurred on two–lane roadways.
• Alcohol was involved in 37 percent of all single–vehicle fatal crashes involving a light vehicle violating the traffic signal or the stop sign.
• Single–vehicle crashes had the highest rate of speeding and inattention, 33 percent and 14 percent, respectively.
• Inattention or distraction was reported for about 11.0 percent of all light–vehicle violations in two–vehicle fatal crossing–path crashes.
• Alcohol was linked to 14 percent of all light–vehicle violations in two–vehicle fatal crossing–path crashes.
• Speeding or racing, including police chase, was related to 10 percent of all light–vehicle violations in multiple–vehicle fatal crashes. This factor was four times more prevalent in traffic signal crashes than in stop sign crashes.
• Inattention or distraction was the second most reported factor, representing about 7 percent of all light–vehicle violations in multiple–vehicle fatal crashes.
• Alcohol was linked to 13 percent of all light–vehicle violations in multiple–vehicle crashes.

• No major differences were found among the crash categories regarding the infrastructure where these fatal crashes occurred.
• The authors concluded that fatal crashes involving a light vehicle violating the traffic signal or stop sign occur in similar locations, regardless of whether they are single–vehicle, two–vehicle, or multiple–vehicle crashes.
• Alcohol, speeding, and inattention are the three most common contributing factors for fatal crashes at traffic signals and stop signs.

None

Examination of Intersection, Left Turn Across Path Crashes and
Potential IVHS Countermeasures (DOT–HS–808–154)

National Highway Traffic Safety
Administration
400 Seventh Street, S.W.
Washington, DC 20590

COTR:
Not Specified

Chovan, J.D., Tijerina, L., Everson, J.H., Pierowicz, J.A., and Hendricks, D.L.

September 1994

52

http://www.its.dot.gov/itsweb/EDL_webpages/webpages/SearchPages/Alpha_Search.cfm

Crash/Demographic Statistical Analysis

Imminent Crash (Intersection Collision Avoidance (ICA))

Light Vehicles

To provide a preliminary analysis of intersection–related, left turn across path (LTAP) crashes and applicable countermeasure concepts for the Intelligent Vehicle–Highway System (IVHS) program. The intent of the report is to increase understanding of the crash avoidance requirements associated with LTAP crashes.

• This report presents the results of a study of the intersection, LTAP type of collision as identified by the NHTSA Office of Crash Avoidance Research (OCAR).
• A total of 154 LTAP crashes selected from the 1992 Crashworthiness Data System (CDS) were analyzed and weighted for severity so that they might more closely approximate the national profile.

• A framework for IVHS crash avoidance concepts regarding LTAP crashes is presented.
• A simple LTAP model is presented in which driver warnings are analyzed in terms of principal other vehicle (POV) time headway. This model incorporates the above framework and is divided into two subtypes based on whether the subject vehicle (SV) comes to a complete stop before entering the intersection.
• Two types of LTAP crashes were identified:
  • Subtype 1, where the SV slows, but does not stop; begins the left turn; and strikes or is struck by the oncoming POV.
  • Subtype 2, where the SV stops, then proceeds with the left turn, and strikes or is struck by the POV.
• The report concludes with a discussion of research needs to support further refinement of the LTAP scenario and other crash avoidance concepts.

Vehicle Crash Analysis, Crash Countermeasures, Intelligent Vehicle–Highway System, Kinematic Models, Crash Circumstances

Causal Factors and Crash Characteristics:

• At both signalized and unsignalized intersections, the LTAP crashes occurred for the following reasons:
  • SV driver was unaware of the crash hazard.
  • SV driver misjudged how fast the POV was approaching.
  • SV driver misjudged how close the POV was to their intersection.
  • Potentially harmful situation was not obvious to the SV driver.
  • SV driver’s view was obstructed.
• SV was more likely to be struck by another vehicle than to strike another vehicle.
• Most LTAP crashes occurred on roadways with posted speed limits of 56 kilometers per hour (km/h) (35 miles per hour (mi/h)) or greater, on dry pavement (80 percent), and under no adverse weather conditions (86 percent).

IVHS Crash Avoidance Concepts for LTAP Crashes:

• A framework for IVHS crash avoidance concepts was presented based on a series of sequential countermeasure steps as follows (see figure A):
  • Driver alerts.
  • Higher intensity driver warnings.
  • Partially automated control crash avoidance maneuvers.
  • Fully automated control maneuvers.

Research Needs:

• Clinical analysis area: Cross–tabulation of causal analysis between subtypes, concordance of parallel analyses,analysis of cases caused by a loss of traction.
• Driver behavior at left turns across path: Higher order responses, correlations, driver decision processes,maximum turn velocities, control intervention, interaction between drivers, alternative alert displays, transition from preplanned to emergency maneuvers, driver acceptance of LTAP collision avoidance systems (CAS), headway time prediction, driver reaction time.
• LTAP algorithm research needs: Additional CAS concepts, CAS set points, impact of acceleration profiles on robustness, false alarms, warning familiarity, evasive maneuvers, POV turning.
• Further modeling research needs: Multiple–vehicle interactions, inclusion of variables, speed profiles, indicators of intent, normal driving behavior.

None

Examination of Unsignalized Intersection, Straight Crossing–
Path Crashes, and Potential IVHS Countermeasures
(DOT–HS–808–152)

National Highway Traffic Safety
Administration
400 Seventh Street, S.W.
Washington, DC 20590

COTR:
Not Specified

Chovan, J.D., Tijerina, L., Pierowicz, J.A., and Hendricks, D.L

August 1994

72

http://www.its.dot.gov/itsweb/EDL_webpages/webpages/SearchPages/Alpha_Search.cfm

Crash/Demographic Statistical Analysis

Imminent Crash (ICA)

Light Vehicles

To provide a preliminary analysis of unsignalized intersection, straight crossing path (UI/SCP) crashes and applicable countermeasure concepts for the IVHS program. The intent of the report is to increase the understanding of crash avoidance requirements associated with UI/SCP crashes.

• This report presents the results of a study of the UI/SCP type of collision as identified by the NHTSA Office of Crash Avoidance Research (OCAR).
• 100 UI/SCP crashes selected from the 1992 Crashworthiness Data System (CDS) were analyzed and weighted for severity so that they might more closely approximate the national profile.

• An analytic model of intersection negotiation behavior at unsignalized intersections was presented to indicate possible sources of driver actions that might contribute to such crashes.
• Two types of UI/SCP crashes were identified as follows:
  • Subtype 1, where the SV ran the stop sign.
  • Subtype 2, where the SV stopped, then proceeded against cross traffic.
• The two crash subtypes were examined for the following characteristics: Speed distribution, POV travel direction, SV’s role in the crash event.
• Crash avoidance concepts regarding UI/SCP crashes were discussed, and partially automatic control systems and fully automatic control systems were presented as control intervention schemes.
• The report concluded with a discussion of research needs to support further refinement of the UI/SCP scenario and other crash avoidance concepts.

Vehicle Crash Analysis, Crash Countermeasures, IVHS, Kinematic Models, Crash Circumstances

Crash Causal Factors:

• UI/SCP crashes occurred for the following reasons:
  • Driver unawareness caused by inattention, failure to see, and obstructed vision.
  • Driver misjudgment of POV velocity/gap.
  • Deliberate violation of sign.

Crash Countermeasure Concepts:

IVHS crash countermeasure concepts, specific to UI/SCP crash subtypes, were devised in three different categories to address the major causal factors as follows (see figure A):

• In–vehicle alert: Subtype 1—Intersection detection alert, Subtype 2—In–vehicle display of approaching POV.
• Driver warning: Subtype 1—Graded warnings to SV driver, Subtype 2—Gap acceptance aid that warns the SV when it is unsafe to enter the intersection.
• Control intervention: Both subtypes—CAS–controlled soft braking, moderate braking, or graded braking with or without driver override (see figure B).

Research Needs:

• Clinical analysis area: Increase sample size in analysis, concordance of parallel analysis.
• Driver behavior at unsignalized intersections: Higher order responses, correlations, drivers’ decision processes, control intervention, interaction between drivers, alternative alert displays.
• UI/SCP algorithm research needs: Additional CAS concepts, error modeling of algorithm data, CAS set points, impact of velocity profiles on algorithm robustness.
• Further modeling research needs: Multiple–vehicle interactions.

None

Safety Impact of Permitting Right–Turn–on–Red: A Report to Congress by the National Highway Traffic Safety Administration (DOT–HS–808–200)

National Highway Traffic Safety
Administration
400 Seventh Street, S.W.
Washington, DC 20590

COTR:
Not Specified

Compton, R.P., and Milton, E.V.

December 1994

47

None

Literature Review, Crash/Demographic Statistical Analysis

Normal

Not Specified

To provide a brief summary of State laws and the safety impacts of permitting right and left turns at red lights.

This report presents a brief summary of the current status of State implementation of laws permitting right and left turns at red lights, a brief review of previous research, and the results of analyses of currently available data assessing the safety impact of permitting a right turn on red (RTOR).

Two sources of data were used in completing this report:

• Fatality Analysis Reporting System (FARS): FARS includes a code for an RTOR vehicle maneuver. However, FARS does not include information on whether a vehicle was turning right on red at the time of the crash, only that the vehicle was turning right at the time of the crash at an intersection where RTOR is permitted.
• Data from four State crash data files (Illinois, Indiana, Maryland, and Missouri): The four State files include on their crash report form either a code for an RTOR vehicle maneuver or other codes that make it possible to determine that an RTOR maneuver was executed. With one exception, data used in the analysis cover the years from 1989 through 1992. From Illinois, only 1989 through 1991 data were available.

Right Turn on Red (RTOR), Left Turn on Red (LTOR), Safety Impact, Intersection Crashes

Analysis of FARS data showed the following:

• Approximately 84 fatal crashes occurred per year during the time period involving a right–turning vehicle at an intersection where RTOR is permitted.
• During this same time period, there were 485,104 fatalities. Thus, less than 0.2 percent of all fatalities involved a right–turning vehicle maneuver at an intersection where RTOR is permitted. FARS, however, does not discern whether the traffic signal indication was red. Therefore, the actual number of fatal RTOR crashes is somewhere between zero and 84 and may be closer to zero.
• Slightly less than half of the fatal RTOR crashes involve a pedestrian (44 percent); 10 percent a bicyclist; and, in 33 percent of the crashes, one vehicle striking another vehicle (see figure).

The results of the data analysis from the four State crash files suggest the following:

• RTOR crashes represent a very small proportion of the total number of traffic crashes in the four States (0.05 percent).
• RTOR injury and fatal crashes represent a fraction of 1 percent of all fatal and injury crashes (0.06 percent).
• RTOR crashes represent a very small proportion of signalized intersection crashes (0.4 percent).
• When an RTOR crash occurs, a pedestrian or bicyclist is frequently involved. For all States, for all years of the studies, the proportion of RTOR pedestrian or bicyclist crashes to all RTOR crashes was 22 percent.
• RTOR pedestrian and bicyclist crashes usually involve injury. Some 93 percent of RTOR pedestrian or bicyclist crashes resulted in injury.
• Only 1 percent of RTOR pedestrian and bicyclist crashes resulted in fatal injury. However, less than 1 percent of all fatal pedestrian and bicyclist crashes result from RTOR vehicle maneuvers.
• Most RTOR crashes occur between 6:00 a.m. and 6:00 p.m.

• A relatively small number of deaths and injuries each year are caused by RTOR crashes.
• These represent a very small percentage of all crashes, deaths, and injuries.
• Because the number of crashes resulting from RTOR is small, the impact on traffic safety has also been small.
• Insufficient data exist to analyze LTOR.

None

Safety Evaluation of Red–Light Cameras
(FHWA–HRT–05–048)

Federal Highway Administration
6300 Georgetown Pike
McLean, VA 22101–2296

COTR:
Michael Griffith

Council, F.M., Persaud, B., Eccles, K., Lyon, C., and Griffith, M.S.

April 2005

8

http://www.tfhrc.gov/safety/pubs.htm

Field Test

Normal

Not Specified

To determine the effectiveness of red–light camera (RLC) systems in reducing crashes.

The study involved Empirical Bayes (EB) before/after research using data from seven jurisdictions across the United States to estimate the crash and associated economic effects of RLC systems. The study included 132 treatment sites and specially derived rear–end and right–angle unit crash costs for various severity levels.

• The choice of jurisdictions to be included in the study was based on an analysis of sample size needs and the data available in potential jurisdictions.
• The jurisdictions chosen were: El Cajon, San Diego, and San Francisco, CA; Howard County, Montgomery County, and Baltimore, MD; and Charlotte, NC.
• Data were required not only for RLC–equipped intersections, but also for a reference group of signalized intersections that were not equipped with RLCs, but were similar to the RLC locations.

Red–Light Camera, Empirical Bayes, Crash Evaluation, Economic Analysis, Signalized Intersection

• There was a significant decrease in right–angle crashes, but there was also a significant increase in rear–end crashes (see table A).
• The economic estimates, with property damage only (PDO) crashes excluded, show a positive aggregate economic benefit of more than $18.5 million over approximately 370 site–years, which translates into a crash–reduction benefit of approximately $50,000 per site–year (see table B).

Note: A negative number indicates a decrease.

Note: A negative number indicates a decrease.

• Crash effects detected were consistent in direction with those found in many previous studies (a decrease in right–angle crashes and an increased in rear–end crashes).
• There was a modest aggregate crash cost benefit of RLC systems.
• A disaggregate analysis found that the greatest economic benefits are associated with factors of the highest total entering annual average daily traffic (AADT), the largest ratios of right–angle to rear–end crashes, and the presence of protected left–turn phases.
• There were weak indications of a spillover effect that point to a need for a more definitive, perhaps prospective, study of this issue.

None

Red Light Violations and Crashes at Urban Intersections

(Transportation Research Record 1734, pp. 52–58)

Federal Highway Administration
6300 Georgetown Pike
McLean, VA 22101–2296

COTR:
Not Specified

Datta, T.K., Schattler, K., and Datta, S.

2000

7

None

Field Test

Normal

Not Specified

To determine if any difference existed between red–light violation characteristics among intersections with properly designed clearance intervals and intersections that did not have appropriate yellow change intervals and, more importantly, an all–red interval.

A study was performed in Detroit, MI, to compare the red–light violation characteristics of intersections with properly designed all–red intervals and those intersections without all–red intervals. In the absence of "before" violation data, a comparative parallel experimental study was used. An evaluation of before/after crash frequencies was also performed to determine the effectiveness of implemented improvements on right–angle crashes and injuries.

• Five signalized intersection sites in Detroit were studied: Three treatment (test) intersections, two intersections in the same area were selected as control sites.
  • Treatment sites: All treatment intersections had clearance intervals (yellow and all–red intervals) that were calculated based on site–specific criteria such as approach speed, vehicle deceleration rates for stopping, and intersection geometry.
  • Control sites: These sites had a yellow interval only.
• Red–light violations were monitored through a series of onsite field observations. A total of 16 h of field data were collected at each of the five sites during off–peak periods.
• Trained field personnel observed all traffic movement through each intersection and recorded the frequency of red–light violations based on the directional movement of travel.

Red–Light Violations, Intersection Safety, Yellow Change Intervals

• In performing the effectiveness evaluation, after–improvement crashes were compared with the 3–year averages of crash data for the same months of the "before" period.
• The results show a significant reduction in red–light violation rates for the treatment sites. The average red–light violations per hour for the treatment sites was 3.6, while the control sites had an average of 8.08.
• The before/after comparison of right–angle, injury, and total crashes at all three treatment sites shows that the crash frequencies were significantly lower after the treatment (see tables below).

^aRepresents an annual average of 24–month data (June 1997 to May 1999).

^aRepresents an annual average of 21–month data (June 1997 to May 1999).

^aRepresents an annual average of 29–month data (June 1997 to May 1999).

From Transportation Research Record 1734, Transportation Research Board, National Research Council, Washington, DC, 2000, table 3, p. 57. Reprinted with permission.

• Analysis indicated significantly lower red–light violations at the treatment sites.
• Analysis also indicated an extraordinary reduction in right–angle and injury crashes.
• Study demonstrated that substantial benefits, in terms of reducing red–light violations and right–angle crashes, can be achieved by introducing a well–designed, all–red interval.

None

Guidance for Using Red Light Cameras

Federal Highway Administration
6300 Georgetown Pike
McLean, VA 22101–2296

COTR:
Not Specified

Federal Highway Administration and National Highway Traffic
Safety Administration.

March 2003

60

http://www.tfhrc.gov/safety/intersect.htm

Guidelines

Normal

Not Specified

The guidance in this report is intended to provide critical information for State and local agencies on relevant aspects of red–light camera (RLC) systems in order to promote consistency and proper implementation and operation.

FHWA and NHTSA have developed this guidance for the use of State and local agencies on the implementation and operation of RLC systems. This guidance can be used by State and local agency managers, transportation engineers, and law enforcement officials to identify and properly address safety problems resulting from red–light running (RLR) within their jurisdiction.

The document is divided into the following sections:

• Understanding of the problem.
• Problem identification.
• Countermeasures and their applications.
• RLC program implementation.

Red–Light Running, Red–Light Cameras, Intersections

• An engineering study may identify the following conditions that may be present at a signalized intersection and contribute to RLR by motorists: Grade, poor visibility, temporary roadside obstructions, line of sight, sign reflectivity, traffic volumes, signal timing, and weather.

Problem Identification:

• The following steps are recommended for investigating intersection safety: Data collection; RLR violation data; intersection crash data; driver behavior observations; traffic–, signal–, and intersection–related data; and motorist complaints and comments.

Countermeasures and Their Applications:

• Engineering countermeasure solutions to be considered include: Modifying traffic signal timing, improving signage and marking, improving sight lines, modifying grades and/or grade separation, adjusting the prevailing speeds, changes in surface treatments, altering lane configurations, and replacing the traffic signal with some other form of traffic control device or intersection type.
• Education: A well–designed public information and education campaign should provide information and data that explain what RLR is, why RLR is dangerous, and what actions are currently being undertaken to reduce the incidence of RLR.
• Enforcement by law enforcement officers: Officers in patrol cars or using motorcycles can be a cost–effective solution to reduce RLR at problem intersections. However, unless an observer and a stopping team are used, officers also must pass through the intersection on a red signal indication.
• Red–light cameras: If engineering, educational, and traditional enforcement countermeasures are proven to be unsuccessful, RLR camera technologies, if authorized by law, may be considered.

RLC Program Implementation:

• Early planning and startup: The following are the key elements required for the early planning and startup of an RLC program.
  • Establishment of an oversight committee: This should be inclusive of all stakeholders (engineers, educators, law enforcement, prosecutors, judges, and, most importantly, private citizens).
  • Establishment of program objectives: The oversight committee should define, as clearly as possible, the RLC program objectives as an early step for moving forward. Program objectives should address specific operational needs.
  • Identification of the legal requirements: In particular, concerns and issues related to privacy, citation distribution, and types of penalties need to be thoroughly addressed and resolved prior to the startup of an RLC program.
• Engineering design of RLC systems: Plans should address the placement of the RLC system equipment and related components, including camera equipment, supporting structure, intersection lighting, vehicle detection system, communications, pull boxes and conductor schedule, electrical service, and warning signs.

See Key Results above.

None

Intersection Angles and the Driver’s Field of View

Arkansas State Highway and
Transportation Department
P.O. Box 2261
Little Rock, AR 72203

COTR:

Not Specified

Gattis, J.L., and Low, S.T.

November 1997

37

None

Field Test

Normal

Various Types

To identify the constraints on the angle of a left-skewed intersection, as affected by the vehicle body limiting a driver’s line of sight to the right.

In this research project, the angles at which drivers’ lines of sight were obstructed by the body of their vehicles were measured. Two driver positions ("sit back" and "lean forward") were used. A 13.5-degree vision angle was selected to represent an intermediate position (between the "sit back" and the "lean forward" positions).

Design Vehicle:

• The following vehicle design types were located and arrangements were made to allow measurements to be taken: Ambulance, dump truck, motor home, school bus, small bus on a van chassis, single-unit truck mounted with container, and truck tractor (cab of an 18–wheeler).

Driver Position:

• "Sit back" position: Driver was in a fully leaned-back position, with his/her back touching the seatback. In this position, the driver relies mainly on head and neck movement to get the maximum viewing angle to his/her right. This position permits the driver to remain comfortably seated against the seatback.
• "Lean forward" position: Driver leaned forward so that the driver’s eyes were over the juncture where the steering wheel is attached to the column. In addition to using head and neck movements, the driver leaned his/her upper body far forward to get a greater viewing angle to the right. In such a position, the driveer’s chest was often pressing the driver’s arms against the steering wheel, thus confining the movement of the driver’s arms.

Field Measurements:

• The lengths of both the front and rear axles were measured. The difference between these two widths was divided by two. This "half of the width" difference was added to the front axle width and this dimension was marked on the parking lot surface to the outside of the right-front tire. The "right-edge parallel line" was determined by connecting a line from this point to the edge of the right-rear tire.
• Next, the researchers constructed a perpendicular line projecting from the driver’s eyes with the driver in the "sit back" position and another perpendicular line projecting from the "lean forward" position.
• A surveying range pole with an attached level was placed on the right-offset line, within the seated driver’s field of view. As it was slowly moved backward, the person in the driver’s seat signaled when a vehicle body obstruc tion caused him/her to lose sight of the pole. This position was marked. This procedure was performed three times for each position.

Intersection Angle, Sight Distance, Geometric Design

Effects on Sight Distance at Intersections:

• With a 5.4-meter (m) (17.7-foot (ft)) setback and the driver in the intermediate "lean forward" position, the resulting available sight distances for 60, 65, 70, and 75 degrees were found to be 40, 55, 96, and 408 m (131, 180, 315, 1339 ft), respectively (see table A).
• The currently recommended minimum intersection angle, 60 degrees, has a resulting available sight distance equal to the stopping sight distance (SSD) for 37-km/h (23-mi/h) travel on the major roadway.
• Designers should recognize that some drivers will position themselves so that they are less than 5.4 m (17.7 ft) from the edge of the through-road traveled way. Table B lists the angular sight distance (ASD) and design speeds calculated with E = 4.4 m (14.4 ft).