BACKGROUND

The Intelligent Transportation System (ITS) program within the U.S. Department of Transportation (US DOT) has sponsored research, development, and field testing of advanced safety systems that can improve transportation operations and address safety problems. One of these problem areas addresses highway intersections and the potential for Intersection Collision Avoidance (ICA) systems to reduce crashes. Intersection crashes account for almost 30 percent of vehicle crashes in the United States.

The US DOT has sponsored research studies of vehicle-based ICA concepts and of infrastructure-based ICA concepts for unsignalized intersections. Since 1993, Veridian, under contract to the National Highway Traffic Safety Administration (NHTSA), has been studying vehicle-based concepts for ICA including the use of vehicle-based systems with map databases and the GPS that recognizes the presence of stop signs or signalized intersections, and vehicle-mounted scanning radars that allow detection and warning of potential conflicts for left turns. These “autonomous” concepts offer the potential for deployment through incorporation of the new technology into passenger vehicles. Under contract to the Federal Highway Administration (FHWA), Raytheon Company developed a prototype Collision Countermeasure System. This infrastructure-only system, which provides active warning signs activated by loop detectors in the roadway, has been pilot tested at an intersection in Prince William County, VA.

Although autonomous vehicle-based concepts and infrastructure-based concepts for ICA have been found to offer considerable potential for alleviating some safety problems at intersections, it has been suggested that systems communicating between vehicles, or between vehicles and the highway infrastructure, could potentially provide even greater safety benefits. These “cooperative” systems would potentially allow vehicle-based systems to utilize information from external sensors or other sources to supplement the information that can be obtained from the vehicle itself.

Because cooperative systems require the deployment of systems both within the vehicle and in the highway infrastructure, deployment involves risk for vehicle manufacturers, drivers who purchase vehicles with cooperative features, and the government agencies who deploy the needed infrastructure components. One potential strategy for overcoming some of the risks of deploying cooperative systems is to first deploy infrastructure-only systems that communicate to drivers through existing traffic control devices, variable message signs, or other means so that drivers can take appropriate actions to avoid crashes. These infrastructure-only ICA systems could then be enhanced and extended later, once motorists and governments are convinced of the potential benefits of these ICA systems and a sufficient population of systems is available to communicate with intelligent vehicles as well.

Accordingly, the primary objective of this project is to define and evaluate infrastructure-only ICA concepts complementary to in-vehicle autonomous and vehicle/infrastructure cooperative concepts aimed at reducing the number of intersection crashes. The scope of this effort is to perform conceptual and analytical work to define and evaluate the feasibility and effectiveness of alternative infrastructure-based advanced technology concepts. In addition to infrastructure-only concepts for all vehicle types, the study will consider infrastructure-vehicle cooperative systems for transit and emergency vehicles and intersection crashes involving pedestrian or pedal-cyclists.

To meet this objective, the following tasks and subtasks are being performed:

Task 1 – Review of Previous Work:

1.1 Accident data studies

1.2 Advanced technology intersection safety concepts

1.3 Human factors studies

Task 2 – Analyze Crashes at Intersections selected from Cooperative States

2.1 Identification of cooperative states

2.2 Identification of high priority intersections

2.3 Intersection accident data analysis

2.4 Selection of case studies for feasibility of advanced technologies

Task 3 – Define and evaluate ICA systems

3.1 Define candidate system concepts

3.2 Survey of existing sensing and processing technology vendors

3.3 Conduct analytic studies

3.4 Assess feasibility and effectiveness of countermeasures

This technical memorandum reports on the results of Task 2.

TASK 2.1 – IDENTIFICATION OF COOPERATIVE STATES

The three Infrastructure Consortium states, Virginia, California, and Minnesota participated in the study. They provided support in the selection of high-priority intersections for the study. They also provided information about each of the candidates including hard copies of reports on crashes occurring at the candidate intersections.

TASK 2.2 - IDENTIFICATION OF HIGH-PRIORITY INTERSECTIONS

The purpose of this subtask was to identify 15 to 20 high priority intersections in each of the three participating states as potential candidates for ICAS. Each of the Infrastructure Consortium’s appointed support staff were contacted for guidance on selecting intersections. BMI requested that either each state select the 20 intersections themselves, or provide access to their state crash database so that BMI could select the intersections based on crash history.

BMI endeavored to select a total set of candidate intersections that were diversified in traffic control and general environment, and included some intersections that had experienced crossing path crashes involving pedestrians, bicyclists, emergency vehicles, and transit vehicles.

BMI requested that the set of sites from each state be contained within one district. This request was made in anticipation that a field review would be conducted of each intersection. Selecting intersections from the same jurisdiction reduces the project resources needed to conduct field reviews. It also facilitates obtaining additional information (e.g. signal timings), with only one district engineer’s office per state from which to obtain additional site information.

California

The California Department of Transportation (Caltrans) staff suggested that candidate intersections be selected from two districts in California, District 3 and District 4. District 3 is a predominantly urban jurisdiction. District 4 is a predominantly rural jurisdiction. Both districts are located in Northern California. Caltrans staff provided a text version of the crash database and the accompanying vehicle database for both District 3 and District 4. The databases contained information on the crashes occurring in the two districts between 1997 and 1999.

The database contained location, roadway, crash, and environment data for each crash. These data included:

Location Information: district, route, county, route prefix, and log post mile

Roadway Information: highway group, access, median type, barrier type, left lanes, right lanes, rural/urban classification, inside/outside urban area, intersection/ramp, and side of highway

Crash Specific Information: day, date, time, year, reporting officer number, primary crash factor, right of way control, type of collision, and number of vehicles involved

Crash Environment: weather, lighting, roadway surface, and roadway condition

BMI used these two databases to select preliminary candidates for the study. Sixty intersections total were selected from the two districts. The preliminary candidates were selected based on the number of total crashes and the number of angle crashes at the intersection. Four intersections were selected as candidates because of the number of pedestrian crashes at the intersection. The preliminary candidates were diversified in their roadway characteristics.

Caltrans staff were contacted to request detailed inventory information (e.g. functional class, traffic control type) on these 60 preliminary sites. BMI used the inventory information to select 21 candidate intersections for the study. The selected intersections represent a variety of traffic control types. The intersections were also selected because of their reasonable proximity to one another. The candidate intersections are briefly described in Table 24 of Appendix A.

Minnesota

Staff members from the Minnesota Department of Transportation (Mn/DOT) indicated that they would select the candidate intersections for the study, as they already had some intersections in mind. Preliminary discussions with Mn/DOT staff revealed that Mn/DOT was interested in concentrating their involvement in the study on crashes at intersections on rural, high-speed corridors with low volume minor road crossings. Based on the interest area of Mn/DOT and the requirements of the study, BMI requested that the selected intersections should have the following characteristics:

· Rural intersections

· Traffic control type: signalized and stop-controlled

· Located on the state trunk system

· Located within one district

BMI requested that all of the intersections be located on the state’s trunk system because the trunk system is classified by functional class. Additionally, the traffic control and other inventory information along those routes are known. These data were only readily available through Mn/DOT for the state trunk system.

Staff members from Mn/DOT selected 20 candidate intersections for the study. Each of the 20 intersections was located along Highway 10, a principal arterial through the state. The majority of the 20 candidates were rural intersections. The candidate intersections are briefly described in Table 25 of Appendix A.

Virginia

During preliminary discussions with staff from the Virginia Department of Transportation (VDOT), it was decided that the staff members would select 20 candidate intersections in Fairfax County, Virginia. Fairfax County was selected because of the proximity to BMI’s office and because information on every traffic signal in the county is readily accessible. The roadway environment of Fairfax County is generally urban. After discussions with VDOT, BMI requested that the set of candidate intersections have the following characteristics:

Traffic Control Type: 20-25 signalized intersections and 5-10 stop-controlled intersections
Crash Frequency: Above average to high crash frequency at the intersection, particularly angle crashes
Pedestrian/Bicycle Crashes: At least one signalized intersection and one stop-controlled intersection with five or more pedestrian or bicycle crashes in the three year period

BMI requested that data on the preliminary candidate intersections be provided in the form of standard crash summaries and include information on crash type, vehicle type, vehicle maneuver, fixed object, weather condition, major factor, total crashes, lighting, surface condition, and vehicle direction. BMI planned to use those standard crash summaries to narrow down the set of 20-35 preliminary intersections selected by VDOT to 15-20 candidate intersections.

Instead of receiving VDOT’s 20 selected intersections, BMI received VDOT’s 1999 Intersection Critical Rate Report. The report provides crash rates and frequencies for every intersection in Fairfax County occurring in 1999 within 0.03 miles of the intersection. Specifically, for each intersection, the report provides the crash rate, critical crash rate, injury crash rate, fatal crash rate, total crashes, total fatal crashes, total injury crashes, and total property-damage-only crashes. Additionally the report notes whether the intersection is signalized or unsignalized, the number of approaches, and the estimated entering AADT (Annual Average Daily Traffic). BMI selected 20 intersections from this list with high crash rates and high crash frequencies to be used as candidates for the study. The set of 20 intersections were varied in the number of approach lanes and the signalization or lack of signalization. The candidate intersections are briefly described in Table 26 of Appendix A.

Selected Candidates

The overall set of 61 candidate intersections were varied in intersection configuration, traffic control type, and surrounding environment. Table 1 displays the distribution of the candidate intersections by rural or urban environments, and signalized or unsignalized traffic control.

Table 1: Distribution of Candidate Intersections by General Environment and Traffic Control

The surrounding environment of the California candidate intersections was categorized from the rural/urban roadway inventory description. This inventory description was contained in the California crash database. The Minnesota crash summaries classify the nature of the environment surrounding the intersections as rural or urban with two variables. One variable notes if each intersection is contained in an urban area and if so, specifies the urban area. Another variable denotes whether the general environment is considered urban or rural. The general environment variable was used to categorize the urban or rural nature of the candidate intersections in Minnesota. For the candidate intersections in Virginia, the surrounding environment was not recorded in the standard crash summaries received. However, all of the candidate intersections in Virginia were from Fairfax County, which is generally urban in nature. Therefore, all intersections in Virginia were considered urban in nature.

TASK 2.3 – INTERSECTION ACCIDENT DATA ANALYSIS

The purpose of this subtask was to analyze the crashes occurring at each of the candidate intersections to identify the types of crossing path crashes that were occurring and potential causes of those crashes. Crossing path crashes are classified into one of the following six categories:

1. Left Turn Across Path – Opposite Direction (LTAP/OD)

2. Left Turn Across Path – Lateral Direction (LTAP/LD)

3. Left Turn Into Path (LTIP)

4. Right Turn Into Path (RTIP)

5. Straight Crossing Path (SCP)

6. Other/Unknown

The crossing path crash types are pictured in Figure 1.

Figure 1: Intersection Crossing Path Crash Types (Source: Najm and Koopmann)

The candidate analysis included a manual review of three years of crash reports from each intersection. When possible, a field review was also conducted of the intersection. Crash diagrams for each candidate intersection were also prepared and can be found in Appendix F.

Crash Reports Review

BMI requested crash data for the three most recent years of available data for each of the candidate intersections in the form of hard-copied police crash reports. In the states of Minnesota and Virginia, copies of reports of all crashes considered to have occurred at or within the influence of the intersection were requested. These would represent the total intersection crashes at each intersection as reported by the state. In the state of California, hard copies of all crashes occurring within the intersection were requested due to the size of the reports.

The crash reports from each candidate intersection were manually reviewed to determine the circumstances involved in the crash. Crashes that were not intersection related or that occurred at another intersection (and were erroneously referenced to the candidate intersection) were sorted out. Minnesota and Virginia consider a crash to have occurred at the intersection based on the proximity of the actual crash to the intersection. BMI defined intersection crashes by the circumstances of the crash. Crashes that were related to the operation of the intersection were considered intersection-related. For example, a rear-end crash that occurs on the immediate approach to the intersection would be considered intersection-related if the vehicle that was rear-ended was stopping because the signal indication had changed to red. The same rear-end crash would not be considered intersection-related if the vehicle that was rear-ended was stopping in order to make a right turn into a driveway on the approach. If the exact circumstances of the crash were not explicitly stated, the intersection was considered intersection-related.

The subset of intersection-related crashes was further classified into crossing path crashes and non-crossing path crashes. The general circumstance, report number and maximum injury sustained of all intersection-related crashes were recorded in a database. Detailed information was recorded in the same database for each crossing path crash. This detailed information included the crossing path crash types, traffic control governing each vehicle, vehicle directions, circumstances involved in traffic signal violations, and the circumstances involved in poor gap acceptance. Any factors contributing to the crash such as alcohol use, excessive speed, or sight obstructions were also noted in the database. Additionally, the general conditions such as lighting, weather, and roadway conditions at the time of the crash were recorded.

The three states differed in the level of detail provided in their crash reports. The level of detail provided determined the level to which the circumstances involved in the crash could be understood. Examples of police crash report from each state are in Appendix D.

California

The California crash report for injury or fatal crashes can be in excess of 20 pages in length. The crash reports often contain detailed information about the crash such as a detailed description and sketch of the intersection, statements of the involved parties, witness statements, a description of the physical evidence, apparent cause as determined by the reporting officer, and recommendations. This information may also be provided for property-damage-only crashes although in some jurisdictions property-damage-only crashes are reported with less detail. The crash sketch, description, witness statements and statements of the involved parties were the basis for BMI’s assessment of the circumstances involving the crash.

In the three-year period from 1997-1999, there were 686 crashes at the 21 candidate intersections in California. Of these 686 crashes, 386 (56%) occurred within the intersection, defined by the extension of the crosswalks on each approach. BMI requested hard copies of the crash reports from these 386 crashes. BMI used the crash reports to separate the 386 crashes that occurred within the intersection into crossing path and non-crossing path crashes. Of the 386 crashes occurring within the intersection, 312 (81%) were crossing path crashes. The frequency of total intersection crashes, crashes within the intersection (non-crossing path), and crossing path crashes for each candidate intersection in California are displayed in Table 27 in Appendix A.

The distribution of the crossing path crash types at the 21 candidate intersections in California is displayed in Table 2. The SCP crash was the predominant crash type at the candidate intersections, followed by the LTAP/OD crash type.

Table 2: Crossing Path Crash Type Distribution at Candidate Intersections in California

Crossing Path Type	Frequency	Percent
LTAP/LD	38	12%
LTAP/OD	119	38%
LTIP	11	4%
RTIP	6	2%
SCP	136	44%
Other	2	1%
TOTAL	312	100%

Minnesota

The Minnesota crash report provides a basic description and sketch of the crash. The crash reports are typically one to two pages in length. The majority of the circumstances involved in the crash are provided through coded fields. The reporting officer can choose up to two contributing factors for each vehicle involved in a crash. These factors include driver maneuvers, driver conditions, vision obstructions, and vehicle defects. Reports of crashes at each of the candidate intersections in Minnesota from 1998-2000 were reviewed. BMI relied on the crash narrative and sketch along with the reporting officer codes of the contributing factors to determine the circumstances involving each crash.

There were 358 crashes considered as intersection crashes at the 20 candidate intersections in the three-year period. The manual review established that 276 of those crashes occurred at the intersection or were intersection-related. Of those 276 intersection crashes, 130 (47%) were crossing path crashes in the intersection. The frequency of total intersection crashes, crashes within the intersection (non-crossing path), and crossing path crashes for each candidate intersection in Minnesota are displayed in Table 28 in Appendix A.

The distribution of the crossing path crash types at the 20 candidate intersections in Minnesota is displayed in Table 3. Similar to California, SCP crash was the predominant crash type, followed by the LTAP/OD crash type.

Table 3: Crossing Path Crash Type Distribution at Candidate Intersections in Minnesota

Crossing Path Type	Frequency	Percent
LTAP/LD	10	8%
LTAP/OD	34	26%
LTIP	7	5%
RTIP	3	2%
SCP	73	56%
Other	3	2%
TOTAL	130	100%

Virginia

The Virginia crash report provides a basic description and sketch of the crash. The crash reports are typically one to two pages in length. The majority of the circumstances involved in the crash are provided through coded fields. Reports of crashes occurring at each of the candidate intersections in Virginia from 1998-2000 were reviewed. BMI relied on the crash narrative along with the reporting officer codes of the driver’s action in the crash, driver vision obstructions, conditions of drivers and pedestrians, and vehicle condition to determine the circumstances involving each crash.

In the three-year period there were 1,005 crashes recorded as occurring at the 20 candidate intersections. Virginia defines intersection crashes as any crash occurring in the intersection or on the approach within 0.03 miles of the intersection. The manual review established that 783 of these crashes were intersection-related. Of these 783 intersection-related crashes, 537 (68.5%) were crossing path crashes. The frequency of total intersection crashes, crashes within the intersection (non-crossing path), and crossing path crashes for each candidate intersection in Virginia are displayed in Table 29 in Appendix A.

The distribution of the crossing path crash types at the 20 candidate intersections in Virginia is displayed in Table 4. The LTAP/OD crash was the predominant crash type, followed by the SCP crash type.

Table 4: Crossing Path Crash Type Distribution at Candidate Intersections in Virginia

Crossing Path Type	Frequency	Percent
LTAP/LD	70	13%
LTAP/OD	244	45%
LTIP	5	1%
RTIP	29	5%
SCP	177	33%
Other	12	2%
TOTAL	537	100%

Field Review

BMI conducted field reviews of the candidate intersections. During the field review the candidate intersections were diagrammed, photographed, and in some cases, videotaped. For each intersection, the field reviewer noted the traffic control device and its operation, any sight obstructions, parking restrictions, lane usage, adjacent land use, channelizations, turning restrictions, adjacent driveways, or any other factors that were relevant to the operation of the intersection. Unfortunately, not all of the intersections were reviewed in the field; in California, only 60% of the sites were reviewed in the field due to time constraints during the site visit. (Table 94 in Appendix D lists the intersections in California that were not field reviewed.)

Crash Diagrams

Crash diagrams were created for each of the candidate intersections. The diagrams were based on the field reviews, aerial photographs, and the crash reports. The diagrams illustrate the types of crashes that occurred at the intersection and the maximum injury sustained in each crash. Only crossing path crashes were included in the crash diagrams. Based on the crash diagrams, the critical crossing path scenario and approach were identified for each of the candidate intersections. The crash diagram for each candidate intersection is included in Appendix F.

Overall Results

The results of this task are the input for Task 2.4—Selection of Case Studies. It is necessary to have the selected case studies be representative of the types of crossing path crashes that are occurring and the circumstances involving those crashes. For example, if the overwhelming majority of the straight crossing path crashes are occurring at signalized intersections, then it is necessary to include a signalized intersection with a straight crossing path crash problem. Therefore, in addition to considering each intersection individually through the crash analysis, field review, and crash diagram, the overall set of crashes must be considered.

Crash Type

The selection of intersections for case studies must include both urban and rural intersections and signalized and unsignalized intersections. The predominant crash type occurring at each intersection is dependent upon both the traffic control and the surrounding environment. Table 5 displays the distribution of crossing path crash types by intersection environment and traffic control for all of the candidate intersections combined. For both signalized and unsignalized urban candidate intersections, the LTAP/OD crash was the predominant crash type. For rural candidate intersections, the SCP crash was the predominant crash type. The LTAP/LD crash type was also a leading crash type at both rural and urban unsignalized intersections. The group of recommended intersections reflects this distribution in its composition. The RTIP and LTIP crash types are not a prevalent crash type for any of the intersection categories. Therefore, the LTIP or RTIP crash types are not specifically targeted in the group of recommended intersections.

Table 5: Distribution of Crossing Path Crash Type by Intersection Environment (Rural or Urban) and Traffic Control (Signalized or Unsignalized) for Candidate Intersections

table 5 shows the distribution of candidate intersections and of total crashes. One intersection was rural and signalized and had 8 crashes, 15 intersections were rural and unsignalized and had 123 crashes, 15 intersections were urban an unsignalized and had 258 crashes, and 30 intersections were urban and signalized and had 588 crashes.

Signalized and unsignalized intersections are represented together in Table 6 . It displays the distribution of crossing path crash types by intersection environment (rural or urban) only for the candidate intersections. The SCP crash type was the predominant crossing path crash type at the rural candidate intersections. The LTAP/OD crash type was the predominant crossing path crash type at the urban candidate intersections.

Table 6: Distribution of Crossing Path Crash Type by Surrounding Environment

Crossing Path Crash Type	Rural frequency	Urban frequency
LTAP/LD	34 (26%)	84 (10%)
LTAP/OD	11 (8%)	386 (46%)
LTIP	7 (5%)	16 (2%)
RTIP	5 (4%)	33 (4%)
SCP	74 (56%)	310 (37%)
OTHER	0 (0%)	17 (2%)
TOTAL	131 (100%)	846 (100%)

Table 7 displays the distribution of crossing path crash types by traffic control (signalized or unsignalized), regardless of surrounding environment, at the candidate intersections. The LTAP/OD crash type was the predominant crossing path crash type for the signalized candidate intersections. The SCP crash type was the predominant crossing path crash type for the unsignalized candidate intersections.

Table 7: Distribution of Crossing Path Crash Type by Traffic Control

Crossing Path Crash Type	Signalized Frequency	Unsignalized Frequency
LTAP/OD	299 (50%)	98 (26%)
SCP	230 (39%)	154 (40%)
LTAP/LD	27 (5%)	91 (24%)
RTIP	14 (2%)	24 (6%)
LTIP	11 (2%)	12 (3%)
OTHER	15 (3%)	2 (1%)
TOTAL	596 (100%)	381 (100%)

Traffic Control Device Violations

The cause of crossing path crashes can broadly be described as either traffic control device violations or insufficient gap crashes. For the set of candidate intersections, traffic control device violations included violations of stop signs, traffic signals, yield signs, and pedestrian crosswalks. Traffic control device violations occurred in 366 of the crossing path crashes. These included:

306 three-phase traffic signal violations,
24 stop sign violations,
9 yield sign violations,
8 pedestrian crosswalk violations, and
4 traffic signal violations while in flashing operation.

Although signalized intersections only represented half of the candidate intersections, violations of traffic signals was the overwhelming type of traffic control device violation. When possible, the reason for this traffic signal violation was ascertained from the crash reports by reading the crash narrative, the statements of the involved parties, and the witness statements. The reason for the traffic signal violation was provided in 139 (44%) of the traffic signal violations. The distribution of the reported predominant causes was:

40% did not see the signal or its indication,
25% tried to ‘beat’ the yellow signal indication,
12% mistook the signal indication and reported they had a green signal indication,
8% intentionally violated the signal,
6% were unable to bring their vehicle to a stop in time due to vehicle defects or environmental conditions,
4% followed another vehicle into the intersection and did not look at the signal indication, and
3% were confused by another signal at the intersection or at a closely spaced intersection.

The remaining 2% of the traffic signal violations were varied in their cause. Care should be taken in interpreting this information as often it is self-reported and cannot be independently verified.

Insufficient Gap

Insufficient gap was the cause of 614 of the crossing path crashes at the candidate intersections. Insufficient gap crashes include those crashes where involved vehicles obeyed any regulating traffic control device at the intersection, but still entered the intersection into the path of another vehicle. These include both signalized and unsignalized intersection crashes. Similar to traffic control device violations, BMI used the crash narratives, the statements of the involved parties, and the witness statements to determine why motorists accepted the insufficient gap in each crash. As with the traffic signal violations, care should be taken in interpreting this information as often it is self-reported and cannot be independently verified.

Pedestrians and Bicyclists

Crossing path crashes involving pedestrians and bicyclists were target crash types identified for this study to address. Six crossing path crashes involving pedestrians and nine crossing path crashes involving bicyclists occurred at the 61 candidate intersections. No intersection had more than one pedestrian crash during the three-year period. However, one intersection in Minnesota had a crossing path crash involving a pedestrian and two crossing path crashes involving a bicyclist. A list of the frequency and severity of these particular types of crashes by site is provided in Table 95, Table 96, and Table 97 of Appendix E for California, Minnesota, and Virginia respectively.

Emergency and Transit Vehicles

A total of four crossing path crashes involving a transit vehicle and seven crossing path crashes involving an emergency vehicle occurred at the 61 candidate intersections over the three-year period. At one site in California, emergency vehicles were involved in two crossing path crashes. A list of the frequency and severity of these particular types of crashes by site is provided in Table 95, Table 96, and Table 97 of Appendix E for California, Minnesota, and Virginia respectively.

TASK 2.4 – Selection of case studies for feasibility of advanced technologies

The purpose of this subtask was to recommend up to 15 intersections from the set of candidate intersections to be used for subsequent tasks in the project. Accordingly, the recommended intersections are those where roadside crash avoidance technology would likely be most amenable and effective.

Selection Criteria for Recommending Intersections

BMI used the results of Task 2.3 to select a set of 15 recommended intersections. The following nine aspects of each intersection were considered in the selection of the recommended intersections:

1. Crossing path crash frequency

2. Injury rate per crossing path crash

3. Traffic control at intersection

4. High concentration of crashes involving one leg of intersection

5. Pedestrian and/or bicyclist crashes

6. Emergency vehicle or transit vehicle crossing path crashes

7. Prevalent crossing path crash type

8. Feasibility of conventional engineering countermeasures

9. Surrounding environment (e.g. urban or rural)

None of the intersections were recommended based on all of the nine aspects. The entire set of recommended intersections is varied. The recommended intersections include both signalized and stop-controlled intersections and urban and rural intersections. Additionally, the prevalent crossing path crash type varies by intersection. Some of nine aspects were only a selection factor in one or two of the intersections. For example, only one of the intersections was recommended because it experienced emergency vehicle crossing path crashes.

These selection aspects helped to identify intersections where intersection collision avoidance systems can offer the most benefits by reducing crossing path crashes. Intersections where conventional engineering countermeasures were more feasible were not given as much consideration for continuation in the study.

Table 8 presents the 15 recommended sites. Five intersections were recommended from each of the three Infrastructure Consortium states. The criteria that compelled BMI’s recommendation for its inclusion in the remainder of the study are noted in the table.

Table 8: 15 Sites Recommended for Application of ICAS

State	Site No.	Intersection	Traffic Control	Environment	Selection Consideration
CA	5	Route 29 & Silverado Trail	Stop	Rural	1,4,7,8
CA	6	Ashby Avenue & Benvenue Avenue	Stop	Urban	1,2,4,7,8
CA	7	Route 84 & South L Street	Stop	Urban	1,2,4,7
CA	9	Route 121 & Soscol Avenue	Signal	Urban
CA	12	Mission Street & B Street	Signal	Urban	1,4,7
MN	2	Route 10 & Route 210	Stop	Rural	4,7
MN	9	Route 10 & St. Germain	Signal	Rural	1,2,7
MN	13	Route 10 & Country Route 52	Stop	Rural	7,8
MN	14	Route 10 and Bradley Road	Stop	Rural	2,4,7
MN	16	Route 10 & CSAH 5	Signal	Rural	1,2,5
VA	5	Braddock Road and Backlick Road	Signal	Urban	1,2,4,7
VA	8	Alban Road and Boudinot Drive	Signal	Urban	1,2,4,7
VA	9	West Ox Road and Eastbound off-Ramp of Route 29 (Lee Highway)	Signal	Urban	1,2,4,7,8
VA	13	Route 50 and East Tripps Run Road	One-way Stop	Urban	1,2,4,7,8
VA	20	Wilson Boulevard and Peyton Randolph Road	One-way Stop	Urban	1,2,4,7

Description of Recommended Sites

The crossing path crash scenario at each intersection was reviewed to suggest potential generalized countermeasures, both conventional engineering countermeasures and intelligent transportation system countermeasures. Detailed engineering studies of each of the recommended intersections were not conducted. Instead, BMI used the manual review of the crash reports and the field reviews to identify potential causes. The potential countermeasures are generalized. Detailed engineering studies would need to be undertaken to validate their applicability.

Each of the recommended intersections is briefly described. The potential causes and generalized countermeasures are also discussed. Appendix C contains tables of crash distributions by time of day, lighting condition, weather, and road surface condition for each of the 15 selected sites.

California Site 5—Intersection of Route 29 & Silverado Trail, Calistoga

Site Description

The intersection of Route 29 and Silverado Trail is a two-way stop-controlled intersection in Calistoga, California. Route 29 is a two-lane, north-south, principal arterial and is uncontrolled at the intersection. The intersection of Route 29 and Silverado Trail/Lake Street is highly skewed. Silverado Trail forms the northwest approach to the intersection. Silverado Trail is a two-lane major collector that is stop-controlled at the intersection directions. Lake Street forms the northeast approach to the intersection. Lake Street is a local collector that is stop-controlled at the intersection. Silverado trail has a channelized right turn at the intersection. The northbound approach of Route 29 has a prominent horizontal curve. Route 29 had an ADT of 4,200 vehicles in 1999, while Silverado Trail had an estimated ADT of 3,800[1] vehicles. The intersection is illustrated in Figure CA5 in Appendix F.

Crash History

In the three-year period from 1997 through 1999 there were 20 crashes at the intersection. Of those 20 crashes, 18 were crossing path crashes. The distribution of fatal, injury, and property-damage-only crashes is presented in Table 9 for both crossing path and non-crossing path crashes at the intersection.

Table 9: Distribution of Fatal, Injury, and Property-Damage-Only Crashes at California Site 5

Crash Severity	Non-Crossing Path Crashes	Crossing Path Crashes	Total
Fatal	0	0	0
Injury	0	8	8
Property-Damage-Only	2	10	12
Total	2	18	20

The crossing path crashes are diagrammed in Figure CA5 in Appendix F. The distribution of crossing path crashes at the intersection, by crash type, was as follows:

8 LTAP/LD crashes,
7 SCP crashes,
2 RTIP crashes, and
One LTAP/OD crash.

The predominant crash type at the intersection was LTAP/LD – all of which involved northbound vehicles on Route 29 and left-turning vehicles on the northwest bound approach of Silverado Trail. In addition, there were also several SCP crashes involving the same two approaches. Of the 18 crossing path crashes, 17 involved the northbound, uncontrolled approach of Route 29.

Possible Causes

A manual review of the crash reports found that only one of the 18 crossing path crashes at the intersection was the result of a stop sign violation. The remaining 17 crossing path crashes were due to motorists failing to yield the right of way to uncontrolled Route 29 traffic. Several motorists reported that they failed to look for oncoming traffic. Others reported not seeing oncoming traffic. In two cases, a turning motorist’s view was obstructed.

Since there is a prominent horizontal curve in the northbound approach, and all but one of the 18 crashes involved the same approach, limited site distance is a likely causal factor. The prominent skew of the intersection also factors in the inability of stopped motorists to see approaching Route 29 traffic.

Conventional Engineering Countermeasures

A detailed engineering study would need to be conducted of the northbound Route 29 approach to determine the appropriate countermeasures for the intersection. The following conventional engineering countermeasures are potential applications to reduce crashes at the intersection caused by stopped vehicles inappropriately turning into the path of northbound traffic:

Signalize the intersection.
Realign approaches to eliminate skew.
Stopped vehicles may benefit from slower northbound through traffic – a speed limit reduction could increase the time available for left-turning vehicles.

ITS Countermeasures

Stopped vehicles, particularly those on Silverado Trail, are unable to see approaching northbound through traffic on Route 29. Accordingly, they need a means of receiving information regarding the presence of oncoming traffic. An ICW system can provide this information and convey it to stopped motorists on Silverado Trail via a roadside sign or with flashing amber signals on the stop signs. The most effective means of communicating this information would have to be studied.

California Site 6—Intersection of Ashby Avenue & Benvenue Avenue, Berkeley

Description of Site

The intersection of Ashby Avenue & Benvenue Avenue in Berkeley, California is an urban, stop-controlled intersection. Ashby Ave is an east-west roadway that is uncontrolled at the intersection with Benvenue Avenue. Benvenue Avenue is a north-south roadway and is stop-controlled at the intersection. There is street parking on all four approaches, both prior to and after the intersection. Pavement markings displaying the message “STOP” accompany the stop signs on both approaches of Benvenue Ave. Ashby Avenue had an ADT of 22,000 vehicles in 1999, while Benvenue Ave had an estimated ADT of 1,100[2] vehicles. The intersection is illustrated in Figure CA6.

Crash History

In the three-year period from 1997 through 1999, there were 19 crashes at the intersection – all of which were crossing path crashes. The distribution of fatal, injury, and property-damage-only crashes is presented in Table 10 for both crossing path and non-crossing path crashes at the intersection.

Table 10: Distribution of Fatal, Injury, and Property-Damage-Only Crashes at California Site 6

Crash Severity	Frequency	Percent
Fatal	0	0%
Injury	12	63%
Property-Damage-Only	7	37%
Total	19	100%

The crossing path crashes are diagrammed in Figure CA6. The distribution of crossing path crashes at the intersection, by crash type, was as follows:

14 SCP crashes,
1 LTAP/OD crash,
1 LTIP crash,
1 RTIP crash,
1 LTAP/LD crash, and
One other crossing path crash type.

The SCP crash type was the predominant crossing path crash at the intersection. Ten of the 14 SCP crashes involved northbound and westbound vehicles.

Possible Causes

The overwhelming cause of crossing path crashes, as determined from the crash reports, was that congestion in the eastbound lanes of Ashby Avenue blocked the stopped motorists’ view on Benvenue of westbound traffic. In two of the crashes, motorists in congested eastbound traffic improperly waved motorists on Benvenue into the intersection. The inability of northbound Benvenue motorists to see past eastbound Ashby Ave motorists queued upstream caused them to improperly advance into the intersection into the path of oncoming traffic.

Conventional Engineering Countermeasures

Potential conventional engineering countermeasure would include installation of a traffic signal to intersection or upgrade to four-way stop control. A detailed engineering study would be needed to ensure that either was warranted. Another potential conventional engineering countermeasure would be to eliminate eastbound congestion. A detailed engineering study would be needed to determine how to eliminate the congestion. Potentially, congestion could be eliminated by restricting parking on both sides of eastbound Ashby Avenue or by progressing signals upstream and downstream of the intersection.

ITS Countermeasures

Potential ITS countermeasures would detect the eastbound congestion and 1) eliminate it through various automated signal timing changes in upstream signals to progress eastbound traffic or 2) alert/inform stopped northbound vehicles of the presence of hidden westbound traffic. The latter countermeasure could take the form of a roadside sign or simply flashing amber signal heads on a stop sign (with an accompanying static warning sign to identify the meaning of the flashers).

California Site 7—Intersection of Route 84 & South L Street, Livermore

Site Description

The intersection of Route 84 and South L Street is a two-way, signalized intersection in Livermore, California. Route 84 is an east-west, four-lane road intersecting perpendicularly with South L Street. Both directions of Route 84 have a shared left turn lane and one through lane. Both approaches of South L Street have a dedicated left turn lane, right turn lane and through lane. There is parking on the far side of the intersection on all four approaches and on the near side for the east and westbound approaches. Route 84 had an ADT of 29,000 vehicles in 1999, while South L Street had an estimated ADT of 17,200[3] vehicles. The intersection is illustrated in Figure CA7-1 in Appendix F.

Crash History

In the three-year period from 1997 through 1999, there were 30 crashes at the intersection. – 18 of which were crossing path crashes. The distribution of fatal, injury, and property-damage-only crashes is presented in Table 11 for both crossing path and non-crossing path crashes at the intersection.

Table 11: Distribution of Fatal, Injury, and Property-Damage-Only Crashes at California Site 7

Crash Severity	Non-Crossing Path Crashes	Crossing Path Crashes	Total
Fatal	0	0	0
Injury	0	14	14
Property-Damage-Only	5	11	16
Total	5	25	30

Of the 25 crossing path crashes, 18 LTAP/OD were and seven were SCP. The crossing path crashes are diagrammed in Figure CA7 in Appendix F. The reporting officer was unable to determine the direction that at least one motorist was traveling in seven of the 25 crossing path crashes. Accordingly, only 18 crashes are diagrammed on Figure CA7. Of the 18 crashes where all motorist directions were known, 13 were LTAP/OD and five were SCP. All of the LTAP/OD crashes involved westbound and eastbound traffic, ten of which involved eastbound left-turning motorists.

Possible Causes

A field review noted no sight distance limitations due to road geometry. For LTAP/OD crashes, a predominant cause listed in the crash reports was a view obstruction by opposing left turn queues. For the seven SCP crashes, a signal violation occurred in three crashes. In one the signal was not functioning, in another the driver was intoxicated, and in the third, the driver was distracted.

Conventional Engineering Countermeasures

One potential conventional engineering countermeasure would be to re-phase the signal timing to protected-only left turn phasing for both east and westbound approaches. This would eliminate the opportunity for left-turning motorists to misjudge the gaps in opposite direction traffic. However, this would require turning the shared left-through lane into a dedicated left turn only lane or operating the signal in split phasing only. Either approach would substantially decrease the flow of traffic through the intersection.

ITS Countermeasures

One potential ITS countermeasure is an Intersection Collision Warning (ICW) sign situated in the left-turning motorists’ line-of-sight of oncoming traffic. The ICW can display information regarding the presence of vehicles obstructed by opposing left turn queues. Vehicle detection equipment could be employed to let left-turning motorists “see” opposite direction traffic hidden by left turn queues. Left-turning motorists can use this information to aid in their decision-making processes whenever their view of opposing traffic is obstructed. Activation of the ICW would be dependent upon the detection of both a left-turning motorist and an opposing left turn queue. The implications of an ICW applying only during congested conditions would need to be explored.

Another potential ITS countermeasure to reduce LTAP/OD crashes would be the use of a dynamic amber phase offset. This phase offset would be on a conditional basis. The countermeasure system would sense that a vehicle trapped in the intersection, waiting to turn left towards the end of the amber phase. If other vehicles were not in the opposite direction left turn lane, the original trapped left-turning vehicle would have its amber phase extend a full second (while the opposite direction has a red ball). This would prevent trapped left-turning vehicles from turning in front of opposite direction through traffic that they assumed would be stopping at the end of the amber phase.

California Site 9—Intersection of Soscol Avenue and Imola Avenue at Route 121 Junction with SR 221, Napa County

Site Description

The intersection of Imola Avenue and Soscol Avenue is a signalized intersection in an urban area. Route 121 constitute the west and north leg of the intersection. At the intersection, Route 121 junctions with SR 221, which constitutes the south leg of the intersection. Eastbound Imola Avenue has one through lane, a dedicated left turn lane, and a channelized right turn lane. Westbound Imola Avenue has one through lane and a dedicated left turn lane. Soscol Avenue is a 4-lane roadway separated by a concrete median to the south and a grass median to the north. Southbound Soscol Avenue has two through lanes, a dedicated left turn lane and a channelized right turn. Northbound Soscol Avenue has two through lanes, two dedicated left turn lanes, and a channelized right turn. Route 121 had an ADT of 36,000 vehicles in 1999. SR 221/Soscol Avenue had an estimated ADT of 40,000[4] vehicles in 1999. The intersection is illustrated in Figure CA9 in Appendix F.

Crash History

In the three-year period from 1997 through 1999, there were 22 crashes at the intersection. Of those 22 crashes, 19 were crossing path crashes. The distribution of fatal, injury, and property-damage-only crashes is presented in Table 12 for both crossing path and non-crossing path crashes at the intersection.

Table 12: Distribution of Fatal, Injury, and Property-Damage-Only Crashes at California Site 9

Crash Severity	Non-Crossing Path Crashes	Crossing Path Crashes	Total
Fatal	0	0	0
Injury	0	8	8
Property Damage Only	3	11	14
Total	3	19	22

The crossing path crashes are diagrammed in Figure CA9 in Appendix F. The distribution of crossing path crashes, by type, is as follows:

11 LTAP/OD crashes,
6 SCP crashes,
1 LTIP crash, and
One LTAP/LD crash.

Ten of the 11 LTAP/OD crashes involved northbound and southbound traffic, with northbound left-turning motorists colliding with southbound motorists in seven of the crashes.

Possible Causes

All but one of the crashes were the result of a signal violation. The crash reports document that at least half of the crossing path crashes occurred at the beginning of the phase. All but one of the “beginning of phase” violations were committed by northbound left-turning vehicles. The crash reports also document that the signal violations were predominantly caused by motorists not seeing the signal indication (5 crashes) and motorists trying to ‘beat’ the yellow signal indication (5 crashes). The field review found adequate sight distance for all four approaches. Additionally, three crashes at the intersection involved alcohol or drugs.

Conventional Engineering Countermeasures

All the crashes are due to signal violations – half of which occurred at the beginning of the phase. A possible conventional engineering countermeasure would be to extend the phase change interval length, either by increasing the duration of the yellow or all-red indications. Another option is to increase the length of the protected left turn phase to allow more vehicles to move through the intersection. This may encourage less people to attempt to turn left on the red phase.

ITS Countermeasures

Because signal violations were the predominant cause of crossing path crashes at this intersection, Red Light Running Photo Enforcement is a potential ITS countermeasure. Another possible alternative is the use of decision zone signaling in the left turn bays. Because, there are numerous examples of left-turning vehicles trying to ‘beat’ the phase change, there may be an opportunity to employ an actuated phase change that ends the northbound approach’s left turn green arrow and begins the phase change interval based on the position of vehicles in the left turn bay. Ideally, the system would sense when there were no left-turning vehicles in the decision zone. This would potentially reduce the crashes caused by vehicles trying to ‘beat’ the phase change.

California Site 12—Intersection of Mission Street & B Street, Hayward

Site Description

The intersection of Mission Street and B Street is a signalized intersection in Hayward, California. B Street is a two-lane, one-way street that runs northeast to southwest. There is parking on both sides of the intersection and both sides of the street. Mission Street is a two-way street that intersections B Street perpendicularly. It has four lanes and has parking on both side of the intersection and street. Northbound Mission Street has a shared left turn lane, while south bound has a shared right turn lane. Mission Street had an ADT of 18,500 vehicles in 1999, while B Street had an estimated ADT of 21,800[5] vehicles. The intersection is illustrated in Figure CA12-1 in Appendix F.

Crash History

In the three-year period from 1997 through 1999, there were 19 crashes at the intersection. Of those 19 crashes, 16 were crossing path crashes. The distribution of fatal, injury, and property-damage-only crashes is presented in Table 13 for both crossing path and non-crossing path crashes at the intersection.

Table 13: Distribution of Fatal, Injury, and Property-Damage-Only Crashes at California Site 12

Crash Severity	Non-Crossing Path Crashes	Crossing Path Crashes	Total
Fatal	0	0	0
Injury	0	6	6
Property-Damage-Only	3	10	13
Total	3	16	19

The crossing path crashes are diagrammed in Figure CA12-1. The distribution of crashes, by type, was as follows:

10 SCP crashes,
4 LTAP/OD crashes,
1 RTIP crash, and
One LTAP/LD crash.

All 10 of the SCP crashes involved westbound traffic; however, in only two of those crashes the westbound traffic was responsible for the crash. A northbound motorist violated the signal in six of the 10 SCP crashes.

Possible Causes

All of the SCP crashes were the result of a signal violation. The review of crash reports revealed that five of the northbound signal violations occurred because a northbound motorist tried to ‘beat’ the yellow signal indication. That is, the motorist saw that the signal phase was changing and still tried to enter into the intersection. Of the remaining SCP crashes, two were hit-and-run crashes, two were due to inattention, and one was the result of defective brakes.

Conventional Engineering Countermeasures

A potential engineering countermeasure to reduce SCP crashes due to motorists trying to ‘beat’ the yellow signal indication is to increase the yellow phase length or increase all red interval. A detailed engineering study would be needed to determine if increasing the phase change would reduce these crashes. A lengthening of the phase change interval would need to be prudently applied so as not to reduce the effectiveness of the signal.

For crashes due to inattention, a possible countermeasure would be the addition of strobes in the red or amber signal heads. The strobes could be activated at the onset of each amber or red signal phase.

ITS Countermeasures

Dilemma zone signaling may provide a means to reduce motorists who try to ‘beat’ a yellow signal phase. Dilemma zone signaling initiates the phase change when there are no vehicles in the decision zone on the approach to the signal. All motorists should have little hesitation about whether to proceed through legally or slow to a stop.

Another potential ITS countermeasure is the installation of a Red Light Running Photo Enforcement system. The system will likely affect motorists who are inattentive or who are inclined to “hit and runs.”

Still another approach is a dynamic warning indication that provides a warning to motorists that have been detected to run a red signal. This type of warning is a mid-phase (after the all red phase) warning designed to alert motorists who run a red light either intentionally or because they did not see the signal due to inattentiveness or obstruction by other vehicles or road geometry. The warning mode may be effective to distracted drivers if it is in-vehicle. In vehicle warning could be visual, audible or haptic (light pulsed braking).

Minnesota Site 2—Intersection of Route 10 & Route 210, Motley

Site Description

The intersection of Route 10 & Route 210 is a stop-controlled T-intersection in Motley, Minnesota. It is the intersection of southbound Route 210, eastbound Route 10 and northbound Route 10. Northbound Route 10 curves to the west immediately before the intersection so that the approach meets the intersection heading westbound. The intersection is illustrated in Figure MN2 in Appendix F. Route 10 is a four-lane roadway. The eastbound approach of Route 10 has a dedicated left turn lane at the intersection. The westbound approach of Route 10 and the southbound approach of Route 210 both have a channelized right turn lane. Route 10 is separated directionally by a median, wide enough to store at least one passenger vehicle. In 1998, the west leg of Route 10 had an ADT of 8,800 vehicles, the east leg of Route 10 had an ADT of 5,900 vehicles, and the north leg of Route 210 had an ADT of 5,100 vehicles.

Crash History

In the three-year period from 1998 through 2000, there were 13 crashes at the intersection. Of those 13 crashes, eight were crossing path crashes. The distribution of fatal, injury, and property-damage-only crashes is presented in Table 14 for both crossing path and non-crossing path crashes at the intersection.

Table 14: Distribution of Fatal, Injury, and Property-Damage-Only Crashes at Minnesota Site 2

Crash Severity	Non-Crossing Path Crashes	Crossing Path Crashes	Total
Fatal	0	0	0
Injury	3	3	6
Property-Damage-Only	2	5	7
Total	5	8	13

Crossing path crashes are diagrammed in Figure MN2. The distribution of crossing path crashes at the intersection, by type, was as follows:

· 6 LTAP/LD crashes,

· 1 LTAP/OD crash, and

· One other crossing path crash type.

All of the six LTAP/LD crashes involved left-turning vehicles from southbound Route 210 and through vehicles on westbound Route 10.

Possible Causes

None of the crashes at the intersection were a result of a stop sign violation. The predominant cause of the crashes was a failure by southbound vehicles to yield the right of way. A field review revealed potential sight distance limitations due to the curve in the westbound approach. The crash diagram confirmed the westbound approach as the critical approach – 6 out of 7 crossing path crashes (86%) involved a motorist in this approach. In some LTAP/LD crashes, a stopped southbound vehicle observed a westbound vehicle yet still proceeded to turn into the intersection. Southbound vehicles may be misjudging the available gap.

Conventional Engineering Countermeasures

Conventional engineering countermeasures would attempt to address crashes caused by the sight distance limitation. Sight distance limitations lead to misjudgment of available gaps in or the inability to see westbound traffic. The following conventional countermeasures may be applicable:

Signalization intersection,
Realign the intersection to eliminate the curve in the westbound approach, and/or
Reduce westbound (Route 10) approach speed limit to assist in southbound (Route 210) in interpretation of available gaps.

ITS Countermeasures

An alternative method to signalization or realignment would be the use of an Intersection Collision Warning sign to inform southbound motorists of the presence of westbound traffic (and presumably, the lack of a sufficient gap). The Intersection Collision Warning message would be displayed on the southbound (Route 210) approach with warning strobes or flashers on the stop sign. Alternatively, a custom informational sign can be placed roadside on the westbound approach facing stopped southbound (Route 210) motorists, to alert them of approaching vehicles.

Minnesota Site 9—Intersection of Route 10 & St. Germain St., St. Cloud

Site Description

The intersection of Route 10 and St. Germain Street is a signalized intersection in St. Cloud, Minnesota. Route 10 is a four-lane, principal arterial running north and south. It is separated by a grass median. Each approach of Route 10 has dedicated left and right turn lanes. Both east and west approaches of St. Germain have a dedicated left turn lane and a shared right/through lane. In 1998, both legs of Route 10 had an ADT of 18,600 vehicles, the south leg of St. Germain had an ADT of 9,800 vehicles, and the north leg of St. Germain had an ADT of 1,600 vehicles. The intersection is illustrated in Figure MN9 in Appendix F.

Crash History

In the three-year period from 1998 through 2000, there were 28 crashes at the intersection. Of those 28 crashes, 16 were crossing path crashes. One of the crossing path crashes resulted in a pedestrian fatality. The distribution of fatal, injury, and property-damage-only crashes is presented in Table 15 for both crossing path and non-crossing path crashes at the intersection.

Table 15: Distribution of Fatal, Injury, and Property-Damage-Only Crashes at Minnesota Site 9

Crash Severity	Non-Crossing Path Crashes	Crossing Path Crashes	Total
Fatal	0	1	1
Injury	5	5	11
Property-Damage-Only	7	9	16
Total	12	16	28

The crossing path crashes are diagrammed in Figure MN9 in Appendix F. The distribution of crossing path crashes at the intersection, by type, was as follows:

· 10 SCP crashes,

· 5 LTAP/OD crashes, and

· One LTAP/LD crash.

The crossing path crashes were varied in crash scenario. There was not a defined critical approach. All four of the approaches were involved in the crossing path crashes.

Possible Causes

The field review found that there were no sight distance limitations due to the roadway geometry. At least ten of the 16 crossing path crashes were caused by signal violations. The cause for the signal violations were varied but included inattention, vehicle defects, and fog. At least one vehicle attempted to ‘beat’ the yellow signal indication. One motorist admitted knowingly violating the red signal.

With regard to the SCP crash that resulted in a pedestrian fatality, the motorist had the right-of-way at the time of the crash. The pedestrian crossed into the path of the vehicle.

Conventional Engineering Countermeasures

The following conventional engineering countermeasures are potentially applicable to reduce crashes at the intersection. They are intended to increase awareness and/or visibility of the signals and their indications:

Larger signal heads to account for lack of visibility.
Pavement markings indicating “SIGNAL AHEAD” upstream of the intersection.
Passive warning sign with continuous flashing yellow balls to alert motorists that a signal is ahead.

Alternative methods of increasing visibility or awareness of signals include the use of strobe lights in the red or yellow signal head. The strobe could be active throughout an approach’s red phase, or during the onset of the red indication. Optionally, a strobe can be placed in the yellow signal lens and activated at the onset of the yellow signal indication.

ITS Countermeasures

Because inattention is the main cause of red light running and there are no geometric sight deficiencies, conventional countermeasures (i.e. strobes, better paint markings, etc.) might prove more effective than ITS countermeasures. Red Light Running Photo Enforcement (RLRPE), however, still may be applicable. The passive warning sign that typically accompanies RLRPE systems may help to increases signal awareness.

Minnesota Site 13—Intersection of Route 10 & Country Route 52, Becker

Description of Site

The intersection of Route 10 and County Route 52 is a stop-controlled intersection in Becker, Minnesota. County Route 52 approaches the intersection from the north and south. Route 10 approaches the intersection heading southeast and northwest. The intersection is skewed on a 135º angle. Eastbound Route 10 has dedicated left and right turn lanes. Westbound Route 10 has a dedicated left turn lane and a channelized right turn lane. Both directions of Route 10 have two through lanes. In 1998, the ADT on the northwest and the southeast legs were 11,000 and 13,100 vehicles respectively. The north leg of the intersection had an ADT of 3,150 vehicles, while the southern leg had an ADT of 300 vehicles.

Crash History

In the three-year period from 1998 through 2000, there were 11 crashes at the intersection. Of those 11 crashes, six were crossing path crashes – all of which involved a vehicle turning from the median and striking a vehicle on Route 10. Five of the six crossing path crashes were SCP crashes. The crossing path crashes are diagrammed in Figure MN13 in Appendix F. The distribution of fatal, injury, and property-damage-only crashes is presented in Table 16 for both crossing path and non-crossing path crashes at the intersection.

Table 16: Distribution of Fatal, Injury, and Property-Damage-Only Crashes at Minnesota Site 13

Crash Severity	Non-Crossing Path Crashes	Crossing Path Crashes	Total
Fatality	0	0	0
Injury	3	2	5
Property-Damage-Only	2	4	6
Total	5	6	11

Possible Causes

All of the crossing path crashes involved a minor road vehicle in the median being struck by a major road vehicle from their right side. The skew of the intersection requires motorists on the minor road to look 135º to their right (as opposed to only 90º for a perpendicular intersection) and over their shoulder. In four of the SCP crashes, the minor road motorist reported that they did not see oncoming vehicles on Route 10.

Conventional Engineering Countermeasures

Conventional countermeasures would attempt to prevent crashes caused by the sight distance limitations and the skew of the intersection approaches. The following countermeasures are potentially applicable:

Signalization of intersection
Convert intersection to four-way stop-controlled
Realignment of the intersection

The first two conventional countermeasures, signalization and conversion to four-way stop-controlled, are likely not warranted at this intersection. The realignment of the intersection is potentially cost restrictive.

ITS Countermeasures

ITS countermeasures should alert motorists in the median to oncoming traffic. A potential countermeasure is the use of an ICW to inform vehicles in the median that there are major road vehicles approaching. An ICW – possibly in the form of a roadside sign -would be placed on both side of the median to account for northbound and southbound motorists that are turning or traversing through the intersection. Other forms of alerting drivers need to be explored by human factors engineers to determine the best mode for alerting motorists of Route 10 traffic, while keeping the integrity of the yield sign.

Minnesota Site 14—Intersection of Route 10 and Bradley Road, Becker

Site Description

The intersection of Route 10 and Bradley Road is a two-way stop-controlled intersection in Becker, Minnesota. Route 10 is a four-lane, principal arterial that intersects perpendicularly with Bradley Road. Both approaches of Route 10 have dedicated left and right turn lanes at the intersection. Directionally, Route 10 is separated by a grass median. A frontage road runs parallel to Route 10 on the north side. Bradley Road is a two-lane road that runs east to west. In advance of the intersection with Route 10, both approaches of Bradley Road curve slightly so that they intersect Route 10 perpendicularly. In 1998, both legs of Route 10 had an ADT of 13,100 vehicles. Both legs of Bradley Road had an ADT of 380 vehicles[6]. The intersection is illustrated in Figure MN14 in Appendix F.

Crash History

In the three-year period from 1998 through 2000, there were 12 crashes at the intersection, 10 of which were crossing path crashes. The distribution of fatal, injury, and property-damage-only crashes is presented in Table 17 for both crossing path and non-crossing path crashes at the intersection.

Table 17: Distribution of Fatal, Injury, and Property-Damage-Only Crashes at Minnesota Site 14

Crash Severity	Non-Crossing Path Crashes	Crossing Path Crashes	Total
Fatal	0	0	0
Injury	0	5	5
Property-Damage-Only	2	5	7
Total	2	10	12

The crossing path crashes are diagrammed in Figure MN14 in Appendix F. In the three-year period, there were eight SCP crashes, one LTAP/OD crash, and one LTIP crash at the intersection. Eight of the 10 crossing path crashes involved the westbound Route 10 approach.

Possible Causes

A review of the crash diagram reveals that the northwest bound Route 10 is the critical approach, involving 9 out of 10 of the crossing path crashes. All of the crashes except one were the result of poor gap acceptance. Of the nine crossing path crashes caused by poor gap acceptance, four motorists claim they did not see the oncoming vehicle, one thought the oncoming vehicle would stop, one reported that their view was obstructed, and three did not provide a reason for the poor gap acceptance.

A field review noted no sight distance limitations due to the roadway geometry. Additionally, there are dedicated right and left turn lanes at the intersection on Route 10. These would reduce the view obstruction of turning vehicles. However, these vehicles may still potentially cause a view obstruction.

The crashes are potentially caused by the approach speed of vehicles on Route 10. Vehicles approaching the intersection on Route 10 are often traveling in excess of 65 mph. Motorists on the minor road may not be looking far enough away on the approach for oncoming vehicles.

Conventional Engineering Countermeasures

The following conventional engineering countermeasures are potential applications to reduce crashes at the intersection:

· Channelization of right turns on Route 10 to reduce occlusion,

· Reduce speed on Route 10 through a lower speed limit and enforcement,

· Signalize or upgrade intersection to four-way stop-controlled.

Signalization of the intersection or the installation of four-way stop-control is probably not warranted at the intersection.

ITS Countermeasures

An ICW may be able to alert minor road vehicles and vehicles in the median about the position of vehicles on Route 10. The ICW can be placed along the critical approach, facing minor road traffic stopped at the stop sign or median. Potential ICW may be in the form of a roadside sign upstream of Route 10 traffic (facing the intersection). The ICW would be especially advantageous when oncoming vehicles on Route 10 are traveling at a high rate of speed.

Minnesota Site 16—Intersection of Route 10 & CSAH 5, Big Lake

Site Description

The intersection of Route 10 and CSAH 5 is a signalized intersection in Big Lake, Minnesota. Route 10 is a principal arterial that intersects perpendicularly with CSAH 5. Route 10 is a four-lane roadway divided by a two-way left turn lane. CSAH 5 is a two-lane road. The intersection is illustrated in Figure MN16-1. In 1998, the east leg of Route 10 had an ADT of 18,200 vehicles, the west leg of Route 10 had an ADT of 15,800 vehicles, the north leg of CSAH 5 had an ADT of 2,900 vehicles, and the south leg of CSAH 5 had an ADT of 380 vehicles.

Crash History

In the three-year period from 1997 through 1999, there were 21 crashes at the intersection. Of those 21 crashes, 10 were crossing path crashes – including three pedestrian crashes. The distribution of crossing path crashes, by type, is as follows:

· 5 SCP crashes,

· 4 LTAP/OD crashes, and

· One LTIP crash.

The crossing path crashes are diagrammed in Figure MN16 in Appendix F. Six of the 10 crossing path crashes resulted in injury. The distribution of fatal, injury, and property-damage-only crashes is presented in Table 18 for both crossing path and non-crossing path crashes at the intersection. One of the crossing path crashes resulted in a pedestrian fatality.

Table 18: Distribution of Fatal, Injury, and Property-Damage-Only Crashes at Minnesota Site 16

Crash Severity	Non-Crossing Path Crashes	Crossing Path Crashes	Total
Fatal	0	1	1
Injury	3	4	7
Property-Damage-Only	8	5	13
Total	11	10	21

Possible Causes (for pedestrian crashes only)

The critical crash scenario at this intersection is pedestrian/bicycle crashes. There were three pedestrian or bicycle crossing path crashes at this intersection in three years. The motorist has a green signal indication in all three pedestrian/bicycle crashes. In two of the crashes, a bicyclist did not have the right of way and entered into the path of a motor vehicle. In the other crash, although the left-turning motorist had a green signal indication, they failed to yield to a pedestrian in the crosswalk who had the right of way. Failure of the pedestrian/bicyclist to recognize right of way, as well as failure of the motorist to see a pedestrian in the crosswalk are both contributing factors of these crashes

Conventional Engineering Countermeasures

The field review noted that the crosswalks were faded; repainting of the crosswalks may increase the visibility of pedestrians and bicyclists. Increased signing may be necessary to deter pedestrians and bicyclists from crossing without the right of way. If the volume of pedestrian traffic warrants it, a separate pedestrian crossing phase could be integrated into the signal phasing.

ITS Countermeasures

Potential ITS countermeasures include lighted crosswalks (small lights imbedded flush with the pavement). Pedestrian sensors aimed at the crosswalks can activate the lights. Another option is to alert motorists, via amber flashers, to pedestrians/bicyclists in the crosswalk. Again, activation of the flashers can be accomplished through pedestrian detectors. A timer countdown could also be added to inform pedestrians of the remaining time available to traverse the crosswalk.

Virginia Site 5—Intersection of Braddock Road and Backlick Road, Fairfax County

Site Description

The intersection of Braddock Road and Backlick Road is a signalized intersection in Fairfax County, Virginia. Braddock Road is an east-west, five-lane road separated by a narrow concrete median. It intersects perpendicularly with Backlick Road. Backlick road is also a five-lane road separated by a narrow concrete median. All four approaches to the intersection have dedicated left turn lanes with protected/permitted traffic signal phasing. Northbound and southbound Backlick Road approaches also have dedicated right turn lanes. Eastbound traffic on Braddock Road has a channelized right turn lane. In 1999, the intersection had an entering average daily traffic (ADT) of 53,200 vehicles. The intersection is illustrated in Figure VA5a.

Crash History

In the three-year period from 1997 through 1999, there were 92 crashes at the intersection. Of those 92 crashes, 65 (71%) were crossing path crashes. The distribution of fatal, injury, and property-damage-only crashes is presented in Table 19 for both crossing path and non-crossing path crashes at the intersection.

Table 19: Distribution of Fatal, Injury, and Property-Damage-Only Crashes at Virginia Site 5

Crash Severity	Non-Crossing Path Crashes	Crossing Path Crashes	Total
Fatal	0	0	0
Injury	15	18	33
Property Damage Only	12	47	59
Total	27	65	92

The crossing path crashes are diagrammed in Figure VA5b. Of the 65 crossing path crashes at the intersection in the three-year period, there were:

· 52 LTAP/OD crashes

· 5 SCP crashes,

· 4 RTIP crashes,

· 1 LTAP/LD crash, and

· 3 other crossing path crashes.

The predominant crossing path crash type at the intersection was overwhelmingly the LTAP/OD. Left-turning motorists from all four approaches were involved in these LTAP/OD crashes although westbound and northbound were over represented. Of the 52 LTAP/OD crashes, 20 involved westbound left-turning vehicles and 22 involved northbound left-turning vehicles.

Possible Causes

The manual review of the crash reports discovered that all of the LTAP/OD crashes were a result of gap misinterpretation, not traffic signal device violations. Because all of the left turn movements at this intersection have both protected and permitted traffic signal phasing, the crashes occurred during the permitted traffic signal phasing. In 30 of the LTAP/OD crashes, no reason for the gap misinterpretation was provided. The following reasons were given in the remaining 22 LTAP/OD crashes:

9 left-turning motorists thought the oncoming vehicle would stop either because the traffic signal indication was yellow or because they thought their left turn was protected;
8 left-turning motorist did not see the oncoming vehicle, often because of a view obstruction; and
4 left-turning motorists misjudged the available gap.

A field review revealed no sight limitations at the intersection due to road geometry for any approach. Those left-turning motorists who reported that their view was obstructed may have been unable to see approaching opposite direction motorists due to the presence of an opposite direction left turn queue.

Conventional Engineering Countermeasures

One conventional countermeasure would be to apply protected-only left turn phasing to both northbound and westbound approaches. However, this signalized intersection is often congested. Providing both protected and permitted left turn phases at the intersection increases the capacity of the left turn lane movements and is important to the operation of the intersection. A detailed engineering study would need to be conducted to determine how changing the signal timing would the impact the operation of the intersection. A variation on this countermeasure would be to provide only protected left turn phasing allowing during peak hours. A study would be required to determine the effects of altering the signal phasing on traffic flow through the intersection.

Another conventional engineering countermeasure would be to realign the lanes so that opposing dedicated left turn lanes are slightly offset from one another. This would allow left-turning vehicles increased visibility of opposite direction traffic. A review of traffic signs at the intersection may also be helpful to determine if left-turning motorists are adequately informed that they must yield to oncoming vehicles during permitted left turn phasing.

ITS Countermeasures

An Intersection Collision Warning (ICW) sign situated in the left-turning motorists’ line-of-sight of oncoming traffic can display information regarding the presence of vehicles obstructed by opposing left turn queues. Vehicle detection equipment could be employed to alert left-turning motorists of opposite direction traffic hidden by left turn queues. Left-turning motorists can use this information to aid them in their decision-making processes whenever their view of opposing traffic is obstructed. Activation of the ICW would be dependent upon the detection of both a left-turning motorist and an opposing left turn queue. An example of an ICW is shown in Figure 6 of Appendix E.

An alternative would be to eliminate permitted left turn phasing on a conditional basis. This is a high volume intersection, especially during peak hours. Accordingly, the protected/permitted phasing was implemented to allow for an increased volume of left-turning vehicles through the intersection. However, if opposite direction traffic is heavy, motorists have fewer opportunities to safely turn left during the permitted phase. Limiting the left turns to a protected-only phasing during unsafe conditions (i.e. high volume opposite direction traffic) would eliminate the opportunity for unsafe left turns. A mode of determining the presence of high volume opposing traffic is required for this type of advanced signaling. A study of the effect of the proposed phasing on the capacity of the intersection would also be required.

Virginia Site 8—Intersection of Alban Road and Boudinot Drive, Fairfax County

Site Description

The intersection of Alban Road and Boudinot Drive is a signalized intersection in Fairfax County, Virginia. Alban Road is a north-south road parallel to Interstate 95. An on-ramp to southbound I-95 forms the fourth leg to the intersection. The northbound approach of Alban Road has a channelized right turn lane to the I-95 ramp, two through lanes, and dedicated left turn lane. Southbound Alban Road has one dedicated through lane, one shared right turn and through lane, and one dedicated left turn lane. In 1999, the intersection had an entering ADT of 22,500 vehicles. The intersection is illustrated in Figure VA8.

The signalized intersection is fully actuated. The signal operates with protected left turn phasing, protected/permitted, and permitted only phases. The signal can also operate with split phasing based on the demand at the intersection.

Crash History

In the three-year period from 1997 through 1999, there were 63 crashes at the intersection. Of those 63 crashes, 50 (79%) were crossing path crashes. The distribution of fatal, injury, and property-damage-only crashes is presented in Table 20 for both crossing path and non-crossing path crashes. There were no fatal crashes, however over half of all crashes resulted in some injury.

Table 20: Distribution of Fatal, Injury, and Property-Damage-Only Crashes at Virginia Site 8

Crash Severity	Non-Crossing Path Crashes	Crossing Path Crashes	Total
Fatality	0	0	0
Injury	3	29	32
Property-Damage-Only	10	21	31
Total	13	50	63

The crossing path crashes are diagrammed in Figure VA8 in Appendix F. Of the 50 crossing path crashes at the intersection, there were:

· 36 LTAP/OD crashes

· 9 SCP crashes,

· 4 LTAP/OD crashes, and

· 1 RTIP crash.

LTAP/OD was the predominant crash type, involving both northbound and southbound left-turning vehicles. Of the 50 crossing path crashes, 36 (72%) were LTAP/OD crashes. The LTAP/OD crashes were evenly divided between northbound and southbound approaches, making both critical.

Possible Causes

A field review revealed limited sight distance for northbound left-turning vehicles. A horizontal and a vertical curve obstruct the view of oncoming southbound vehicles. This sight distance deficiency can be compounded by the presence of a queue in the opposing southbound left turn lane. During the permitted left turn phase, the reduced sight distance may contribute to left-turning vehicles accepting inadequate gaps.

The field review found no apparent sight obstructions for southbound left-turning vehicles. One possible cause of the 18 southbound left-turning crashes may be attributed to vehicles making improper turns at the intersection. In five of those 18 crashes, an officer noted in the crash reports that the left turn was initiated from the through lane, not from the dedicated left turn lane.

Conventional Engineering Countermeasures

A detailed engineering study is required of both the southbound and the northbound approach to determine the appropriate countermeasures for the intersection. The following conventional engineering countermeasures are potentially applicable for reducing crashes at the intersection caused by left-turning vehicles on the northbound and southbound approaches:

The southbound approach may benefit from increased signs guiding vehicles wanting to turn onto the I-95 on-ramp into the inner lane. Advance warning signs may decrease the frequency of improper left turns from the southbound through lane.
Because of the limited sight distance, the northbound approach may benefit from slower southbound through traffic, allowing northbound motorists enough time to turn left.
Altering the signal phasing to protected-only left turn phasing during peak traffic volumes would eliminate the opportunity to misinterpret an available gap in traffic. A study would be required to determine the effects of altering the signal phasing on traffic flow through the intersection.

ITS Countermeasures

Conventional countermeasures would seem more adequate than ITS countermeasures for southbound left-turning motorists turning from an improper lane.

Several ITS options exist for northbound left-turning vehicles. Vehicle detection equipment could be employed to let left-turning motorists “see” opposite direction traffic hidden by left turn queues. An Intersection Collision Warning (ICW) sign situated in motorists’ line-of-sight of oncoming traffic can display information regarding the presence of vehicles obstructed by opposing left turn queues. Left-turning motorists can use this information to aid them in their decision-making processes whenever their view of opposing traffic is obstructed by either the horizontal or vertical curves or opposing left turn queues.

An alternative would be to eliminate permitted left turn phasing on a conditional basis for both northbound and southbound traffic – although is has not been determined that poor gap acceptance for southbound left-turning motorists is the main cause of LTAP/OD crashes in that approach. As mentioned earlier, this is a high volume intersection, especially during peak hours. Accordingly, the protected/permitted phasing was implemented to allow for a higher number of left-turning vehicles through the intersection. However, if opposite direction traffic is heavy, few motorists have the opportunity to safely turn left during the permitted phase. Limiting the left turn phase to permitted only during unsafe conditions (e.g. high volume opposite direction traffic) would eliminate the opportunity for unsafe left turns. A mode of determining the presence of high volume opposing traffic is required for this type of advanced signaling. A study of the effect of the proposed phasing on the capacity of the intersection would also be required.

Virginia Site 9—Intersection of West Ox Road and Eastbound off-Ramp of Route 29 (Lee Highway), Fairfax County

Description of Site

The intersection of West Ox Road and the westbound off-ramp of Route 29 in Fairfax County, Virginia is a signalized intersection. It is one intersection, in a series of at-grade intersections, which comprise the complex interchange of Route 29, the Fairfax County Parkway, and West Ox Road. The intersection is closely spaced to the intersection of West Ox Road and the eastbound off-ramp of Route 29 and the intersection of the westbound off-ramp of Route 29 and the northbound on-ramp for the Fairfax County Parkway.

The westbound off-ramp of Route 29 has two through lanes at the intersection and a channelized right turn lane. Left turns are prohibited at the intersection from the off-ramp. The northbound approach of West Ox road consists of two through lanes. Right and left turns are prohibited for northbound traffic. The southbound approach of West Ox road consists of two right turn only lanes and one through lane. In 1999, the intersection had an entering ADT of 46,000 vehicles.

Crash History

In the three-year period from 1997 through 1999, there were 80 crashes at the intersection. Of those 80 crashes, 71 were crossing path crashes. As shown in Figure VA9b, all but one of the crossing path crashes involved a SCP crash between northbound and westbound vehicles.

The distribution of fatal, injury, and property-damage-only crashes is presented in Table 21 for both crossing path and non-crossing path crashes at the intersection.

Table 21: Distribution of Fatal, Injury, and Property-Damage-Only Crashes at Virginia Site 9

Crash Severity	Non-Crossing Path Crashes	Crossing Path Crashes	Total
Fatality	0	1	1
Injury	2	48	50
Property-Damage-Only	7

Federal Highway Administration

TABLE OF CONTENTS

LIST OF FIGURES

California

Minnesota

Virginia

Crash Type

Traffic Control Device Violations

Insufficient Gap

Pedestrians and Bicyclists

Emergency and Transit Vehicles

Site Description

Crash History

Possible Causes

Conventional Engineering Countermeasures

ITS Countermeasures

Description of Site

Crash History

Possible Causes

Conventional Engineering Countermeasures

ITS Countermeasures

Site Description

Crash History

Possible Causes

Conventional Engineering Countermeasures

ITS Countermeasures

Site Description

Crash History

Possible Causes

Conventional Engineering Countermeasures

ITS Countermeasures

Site Description

Crash History

Possible Causes

Conventional Engineering Countermeasures

ITS Countermeasures

Site Description

Crash History

Possible Causes

Conventional Engineering Countermeasures

ITS Countermeasures

Site Description

Crash History

Possible Causes

Conventional Engineering Countermeasures

ITS Countermeasures

Description of Site

Crash History

Possible Causes

Conventional Engineering Countermeasures

ITS Countermeasures

Site Description

Crash History

Possible Causes

Conventional Engineering Countermeasures

Signalization of the intersection or the installation of four-way stop-control is probably not warranted at the intersection.

ITS Countermeasures

Site Description

Crash History

Possible Causes (for pedestrian crashes only)

Conventional Engineering Countermeasures

ITS Countermeasures

Site Description

Crash History

Possible Causes

Conventional Engineering Countermeasures

ITS Countermeasures

Site Description

Crash History

Possible Causes

Conventional Engineering Countermeasures

ITS Countermeasures

Description of Site

Crash History