1.1 Getting Started

Practitioners can draw on a variety of sources to assess the safety of non-motorized road users. These sources can be formal, such as crash or citation data collected by local law enforcement, or informal, such as a conversation with someone over a cup of coffee. Practitioners should use the highest quality data that are available but understand that formal data describing non-motorized road user problems may be a challenge to obtain. Engaging a group of stakeholders from all four disciplines (engineering, enforcement, education, and emergency services) can help to both identify the problems and facilitate the sharing of ideas in order to reach consensus and garner support for implementing effective strategies and measures.

Local practicioners should consult their State's Strategic Highway Safety Plan (SHSP), which can provide a comprehensive framework for reducing fatalities and serious injuries on all public roads. An SHSP is a data-driven, comprehensive, coordinated safety plan that provides a framework for reducing fatalities and serious injuries on all public roads within a State. SHSPs are developed by the State Department of Transportation (DOT) in a cooperative process with local, State, and Federal input, as well as that from other relevant safety stakeholders. The plan establishes statewide goals, objectives, and key safety emphasis areas that integrate the four E's—engineering, education, enforcement and emergency services. Very often, a State's SHSP will include local or rural roads as an emphasis area for safety improvements. Local practitioners may consult the State's SHSP to determine whether there are emphasis areas, data, or other programs that provide opportunities for coordination to address non-motorized roadway user safety.

Other methods that a local practitioner may investigate when addressing non-motorized safety within the local agency include the following:

• Leveraging opportunities to address non-motorized safety through other projects.
• Using data to identify non-motorized road user safety problems.

Leveraging opportunities through existing projects may include implementing non-motorized user safety measures as part of resurfacing, maintenance, or other programmed projects. This may also include projects initiated to address the safety of other users. Data, in particular crash data involving non-motorized users, can also initiate a process to address non-motorized safety. Local agencies may also consult the State DOT, Local Technical Assistance Program (LTAP), or Tribal Technical Assistance Program (TTAP) for information and assistance in initiating efforts to improve non-motorized safety.

1.2 Assembling Data

Quality data is the foundation for good decisions. Data obtained through existing sources or collected in the field can facilitate the identification of the factors affecting non-motorized safety. While crash data may be the primary means to identify safety issues involving non-motorized roadway users, other data may be critical to provide an understanding of the causes of safety problems. Data needed for improving non-motorized users safety fall under two categories: quantitative and qualitative.

Quantitative Data

Crash Data

Detailed crash data provide an excellent source of information to use in understanding the effects of roadway features and roadway user behaviors. Typically, at least three years of crash data are necessary to be able to identify trends; however, since crashes involving non-motorized users tend to be dispersed, a minimum of five years of crash data is recommended. This larger sample size of crash data may increase the likelihood that severe crash locations or crash trends involving non-motorized users will be identified, although the dataset's size and quality may be questionable since roadway and travel patterns may have changed due to modifications in land use patterns and roadway geometry, as well as possible changes in data collection methodology Practitioners should be aware of possible deficiencies in crash data that may result in the inaccurate identification of safety issues. When assembling and analyzing crash data, the following points should be considered:

• Not all crashes and near-misses may be recorded. Studies have shown that pedestrian and bicycle crashes are underreported, which may be due in part to crashes occurring outside of the public right-of-way (such as at driveways or on shared use paths), lack of injury or sufficient property damage to warrant a police report, lack of police involvement for fear of legal consequences, and other factors.(11)
• Many non-motorized crashes do not involve a motorized vehicle. Collisions may occur between non-motorized users, such as a crash between two bicyclists or between a pedestrian and a bicyclist. Crashes may also occur where a pedestrian or bicyclist is severely injured due to interaction with a roadway or roadside elements. For example, a cyclist running off the road and striking a tree is non-motorized crash that may go unreported and may only be identified through local knowledge.
• Safety issues may still be present in locations with minimal or no recorded crashes or fatalities. Mitigation of those safety issues is not dependent on recorded crashes or fatalities but can be used as a preventive measure.

Data regarding all fatal crashes in the U.S. are maintained in the Fatality Analysis Reporting System (FARS). Local practitioners can query the FARS database to obtain information on fatal crashes involving motorized and non-motorized users.(12) Some local jurisdictions may maintain a crash records database. In many States, the DOT maintains crash records and can assist a local agency with obtaining crash data. A local practitioner can also contact the LTAP or TTAP representative to determine the availability of crash data.

Traffic Data

Traffic volume data are useful to assess the potential for conflict with non-motorized users and are helpful in selecting appropriate countermeasures. However, non-motorized user volumes are seldom collected in rural areas. On local roads, the Average Daily Traffic (ADT) may be the only available measure of vehicle traffic volume. Public land and recreational facility managers may have trail counts (pedestrian and bicycle volume counts on trails and shared use paths) but local agencies may have only limited traffic volume data. In situations where non-motorized traffic safety is a concern, counting non-motorized traffic may be useful or necessary to determine the extent of the safety issue. Some examples of non-motorized traffic generators that might be collected include schools, retail stores, recreational trails, entertainment facilities such as fairgrounds, and public transit stops.

Counts should cover peak-period non-motorized traffic at a minimum (which could range from two to four or more hours) and may occur during peak "recreational" hours, such as weekend mornings. The counts should be categorized by adult/child pedestrian, adult/child cyclist, or other non-motorized user when appropriate. Other key data elements include the following:

• Direction of travel if walking or riding on the side of the road.
• Whether cyclists and pedestrians are traveling in the right or wrong direction (the correct direction of travel for cyclists is with traffic, and for pedestrians, it is against traffic).
• Whether cyclists and pedestrians are crossing within or outside of crosswalks.

Figure 5 depicts a two-lane road through a rural village divided into six traffic count zones. A person is assigned to each zone to count pedestrians and cyclists entering, exiting, or crossing the road within the zone in 15-minute intervals for a two-hour period. All six zones extend through the length of the rural village to capture the area of highest pedestrian and bicycle activity. Each zone is between 200 feet and 300 feet in length; that is long enough where one observer can see and accurately record all pedestrian and bicycle activity in the zone. Agencies conducting similar non-motorized user counts may position observers differently based on the following:

• Non-motorized user volumes.
• Ability to observe non-motorized users.
• Available observers.

Alternatively, automated bicycle/pedestrian counting technologies that perform 24-hour counts can be utilized. An example of summarized counts is shown in Table 1. The data are helpful in assessing volumes of non-motorized traffic and behavioral patterns.

If staff resources are limited, volunteers from stakeholder or community organizations (e.g., pedestrian or biking groups, members of the local chamber of commerce, etc.) can assist in collecting non-motorized traffic volumes. Additional information about conducting traffic counts can be obtained from the Handbook of Simplified Practice for Traffic Studies.(13)

Figure 5. Non-Motorized Traffic Count Locations.

Table 1. Example Non-Motorized Traffic Count Summary Table.

Speed Data

Considering vehicle speeds within the context of the roadway environment may be helpful in determining the risk to non-motorized users. Working with local law enforcement or automated data collection equipment, vehicle speed data can be collected in the field. Additional information about conducting speed studies can be obtained from the Handbook of Simplified Practice for Traffic Studies.(14)

Inventories

Other data that may be helpful are inventories of specific facilities for non-motorized users: shoulders, paths, trails, greenways/byways, sidewalks, or crossing conditions. Typically, the location, extents, condition, width, operating characteristics, and significant issues affecting non-motorized modes associated with these facilities should be recorded. Knowledge of the location and functionality of these facilities is helpful to fully understand potential problems for non-motorized users.

Quantitative Tools

Quantitative tools provide numeric measures of non-motorized user safety. The Bicycle Level of Service (BLOS) and Pedestrian Level of Service (PLOS) models provide a numeric score and grade of bicycle and pedestrian facilities based on geometrics and other conditions, including lane width, vehicle speed, and traffic volume.(15) The 2010 version of the Highway Capacity Manual (HCM) also provides a multimodal level of service (LOS) approach that can be used to quantify the performance of pedestrian and bicycle facilities. These tools can help to evaluate the suitability of a roadway for non-motorized use and can be utilized to assess how different improvements (either individually or cumulatively) will likely affect bicycle and/or pedestrian travel. In conjunction with other information—such as non-motorized trip and crash data—the results can be used to identify potential problem areas.

Some State DOTs have BLOS/PLOS/LOS scores and data maps available, allowing for identification of routes, corridors, and/or spot locations that are in need of further evaluation and improvement. A local agency may wish to contact the State DOT or LTAP/TTAP representative to see if this information is available or may look online or in the HCM for BLOS/PLOS/LOS calculators.(16) One caution is that the models may not yield useful results for roadways with relatively low traffic volumes. For example, the PLOS model may not be useful for roadways ranging from those with an ADT below 5,000 vehicles per day (vpd) and a posted speed limit of 25 mph to those with an ADT below 2,000 vpd and a posted speed limit of 50 mph.

Qualitative Data

Behavioral Data

Behaviors of road users may contribute to safety problems. Common behaviors that have been identified as causal factors in crashes involving non-motorized users include the following:

• Walking or bicycling while intoxicated or impaired.
• Walking in the same direction as adjacent traffic.
• Bicycling in the opposite direction of adjacent traffic.
• Walking or bicycling at night with no lighting or reflective material.
• Driving without scanning for non-motorized users or awareness of potential conflicts.
• Motorized vehicles passing too closely to non-motorized users.
• Speeding or driving too fast for conditions to allow adequate stopping distance.

Behavioral data can be assembled from crash or citation data or from field observations. Field assessments, which aid in the evaluation of roadway user behaviors, are discussed in Section 2.1. Citation data may reveal behavioral trends that contribute to problems for non-motorized users.

Field Assessments

Field assessments can be used to collect other relevant data directly. A field assessment could be as simple as one person conducting site visits of locations to collect targeted data elements, such as the presence and quality of facilities for non-motorized users or as detailed as teams of reviewers conducting comprehensive reviews with formal checklists.

A cyclist is riding on the shoulder of a rural road against traffic. The roadway only has a paved shoulder on one side of the road. This promotes not only riding against traffic but also walking with traffic; both practices/behaviors are significant contributing factors in crashes.

The location of schools, stores, bus stops, train stations, and other non-motorized trip generators is also important to understand potential route choice for non-motorized users and the safety implications associated with those routes. Awareness of future development and roadway improvements may also lend insight into programmed measures addressing non-motorized mobility and safety.

1.3 Engaging Stakeholders

Successfully addressing non-motorized safety in rural areas requires the involvement and support of a variety of people, examples of whom are provided in Table 2. The following are among the key considerations regarding stakeholders:

• Stakeholders should be involved early and regularly to be effective.
• Stakeholders should provide a connection to the local community.
• Stakeholders can become advocates and build local support for initiatives.
• Stakeholders can bring specific knowledge of the local conditions and behaviors.

Stakeholders can provide valuable insights on the safety concerns that are present along rural roadways affecting non-motorized users. They represent the "eyes on the street" and have a keen knowledge of many of the conditions and behaviors present that affect safety. There is no prescribed size for the stakeholder group. The practitioner will determine which stakeholders are invited to the table based on the issue.

Law enforcement personnel are particularly important stakeholders, as they have firsthand knowledge of conditions and behaviors that may contribute to the crash risk on local rural roads. Law enforcement collect data on violations and crashes that help to identify safety concerns related to non-motorized users. Also, law enforcement personnel will be able to provide information on potential enforcement countermeasures and provide useful input in determining engineering or education strategies.

Additionally, LTAP and TTAP centers, universities, and other research institutions that work on transportation safety issues may also be available to provide assistance.

Bringing the right agencies or individuals together will help foster coordination and collaboration, the sharing of resources , and momentum in making a commitment to improving non-motorized safety.  Some of the ways to engage stakeholders include the following:

• Conduct a kickoff meeting and present data to stakeholders.
• Obtain stakeholder input through community meetings, interviews, surveys, or onsite discussions.
• Conduct a safety audit/assessment.
• Present the preliminary findings to stakeholders for additional input.

The continued participation of stakeholders will help guide an approach to find effective treatments that can encompass the 4 E's of safety.

Table 2. Local/Rural Road Non-Motorized Safety Stakeholders.

1.4 Organizing Data

Data are most beneficial when they can be easily understood by the user. Organizing and presenting data in a clear and concise manner aids in the process to define and understand the factors affecting safety. Summary tables and annotated maps are two effective methods to organize and present data .

Summary tables are a simple way to present data by location, crash data (e.g., crash types, crash severities, pavement conditions, lighting conditions, etc.), and other data (e.g., vehicular traffic volumes and speeds). Tabular crash data, as shown in the example in Table 3, can be simple by showing only the total number of crashes per location, or they can be more detailed (if available), showing the percentage breakdown for each crash attribute. Charts and graphs can also illustrate how various factors contribute to safety concerns.

Annotated maps and crash diagrams present a geographical perspective on the occurrence of crashes and other data within a study area. Annotated maps can range from a simple "push pin" map that identifies crash locations, as shown in Figure 6, to more detailed maps illustrating crash characteristics, anecdotal information from a conversation, and other data. For example, a map or aerial image can be annotated manually or electronically with crashes or other conditions affecting safety; these other conditions may include traffic volume data, speed data, and roadway condition data, such as severe curves and hills, areas of limited sight distance, and the presence of roadway debris or standing water on the roadway. Figure 7 illustrates a roadway condition diagram that a local practitioner can create to display available data. The diagram shows a schematic representation of the corridor of concern and displays pedestrian and bicycle crash information such as location and direction of travel, as well as motorized traffic volume data. Other data can also be displayed on a condition diagram, as needed.

Table 3. Example Summary Table.

Figure 6. Example Regional "Push Pin" Map Showing Non-Motorized Crash Locations.

Maps can also be used to engage stakeholders. Stakeholders can be invited to annotate concerns on a map during a meeting, which can then become part of the information used to assess the conditions facing non-motorized users.

Summary tables, annotated maps, and crash diagrams are useful when identifying trends in location (e.g., crash clusters or the lack of sufficient facilities between key points) and contributing crash factors, and they can help to define the extent of the problem.

However, given the previously mentioned limitations in non-motorized crash data, it may not be possible to readily identify trends. It is important to understand these constraints and realize that a review of crash data alone is generally not sufficient to comprehensively identify and address non-motorized safety concerns.

Figure 7. Condition Diagram Example in a Rural Town.