United States Department of Transportation - Federal Highway Administration FHWA HomeFeedback
safety Back to Resources

Innovative Intersection Safety Improvement Strategies and Management Practices: A Domestic Scan

Chapter 2. Safety Management and Comprehensive Safety Processes

This chapter presents findings related to safety management and comprehensive safety processes that were identified and discussed during the scan. It is believed that these processes have a positive influence on intersection safety, although there have been limited attempts to correlate specific crash reductions with the specific processes. It is important to recognize that several of the items discussed in this chapter are broader in scope than just intersection safety. However, it became apparent during many of the interviews and site visits that intersection safety is positively affected by safety management practices of agencies and communities. For these reasons, safety management should be discussed first, before the focus of this report shifts to traffic control, traffic operations, and intersection geometric treatments.

An Uncompromising Commitment to Safety

In terms of intersection safety, there was one public agency that stood out with the scan team due to its concentrated attention to safety. The Road Commission of Oakland County (RCOC), in Michigan is responsible for the design, operation, maintenance, and construction of all 2,700 miles of county roads — about half of its public roads — in this large, rapidly urbanizing county north of Detroit. The notable item about the RCOC is its fundamental commitment to safety. Many years ago, the RCOC management essentially made safety a priority in road decisions. The RCOC created a process in which crash data were to be used to measure the safety of its highways. In addition, RCOC instituted formal documentation of its safety performance goals. Not only that, but the RCOC set about to assure that improvements in safety were the direct result. For example, when the Council of Governments solicits projects as part of the regional constrained long-range transportation plan, the RCOC considers safety as one factor in selecting improvement projects.

In deciding where and how Michigan Transportation Economic Development Fund (TEDF-Category C funds) money is distributed, Oakland County employs a project priority rating that assigns a weight of 30 points out of a possible 103 points for a project's assessed potential to reduce crashes. Table 1 (page 8) presents the factors and their associated weights used in the rating scheme. For the application of Surface Transportation Program (STP) funds, they employ a slightly different project priority rating scale that actually weights crash reduction even higher (35 points out of 103 points). Oakland County's 40 cities and villages, which are eligible recipients of both TEDF (C) and STP funds, also had to agree to the point system (in effect, the Road Commission and the cities/villages compete for the use of these funds). Safety improvements have been taking place on both county roads and city/village streets across the county.

Along the way, the RCOC has created a culture of safety that has allowed significant improvements in highway safety while growing from a county of 300,000 in 1967 to 1.2 million people in 2004. During the scan team's visit, the RCOC indicated that they were able to achieve this safety culture by building safety as a highly weighted factor into federal investment decisions, by requiring safety to be the Number One priority of the agency so that it is driving decisions, and by getting good crash data.

Table 1. Factors and weights used by the Road Commission for Oakland County for project prioritization using Michigan Transportation Economic Development Funds (TEDF-Category C Funds) and Surface Transportation Program (STP) Funds.
Table describes factors and weights used by the Road Commission for Oakland County for project prioritization using Michigan Transportation Economic Development Funds (TEDF-Category C Funds) and Surface Transportation Program (STP).

Performance-Based Safety Systems

To achieve appreciable and meaningful reductions in intersection crashes, several agencies indicated that there is a pressing need for a performance-based safety management system. In order to advance the safety agenda in the United States, especially with respect to intersection safety, systems are needed to ensure that the safety performance can be measured and compared to performance standards. Many highway agencies do not have such a system in place and point to a variety of obstacles and impediments; however, the RCOC has implemented a system. Agency administrators learned years ago that it was not simply enough to claim expected safety benefits from projects. Rather, RCOC learned that it needed to evaluate the effects of its road decisions on safety, specifically crashes. Hence, it was determined that continuously monitoring the safety performance of roads was needed, in terms of reported crash frequency, crash rates, and crash severity. This, in turn, allowed better decisions to be made in roadway investments.

It is because of the systems put in place by the RCOC that the organization can cite the statistics in table 2, which show that over a period of nearly 40 years, despite a four-fold growth in travel in the county, traffic fatalities have been reduced by 64 percent and traffic fatality rates have been reduced by more than 91 percent.

Table 2. Changes in Oakland County's Population, VMT, Crash Fatalities, and Crash Fatality Rate.
 1967 2004
Population 300,000 1,200,000
Annual Vehicle Miles Traveled (VMT) 3.0 billion 13.1 billion
Traffic Fatalities
206 75
Traffic Fatality Rates, Oakland County, Countywide average (fatalities per 100 MVM) 6.8 .57
Traffic Fatality Rates Michigan Statewide Average (fatalities per 100 MVM) No data 1.1
Traffic Fatality Rates United States National Average (fatalities per 100 MVM) 5.3 1.4
MVM = Million Vehicle Miles
Source: TIA and SEMCOG

Timely and Accurate Crash Data

Throughout the scan, many agencies voiced similar sentiments about the importance of timely and accurate crash data. Several agencies indicated that this is recognized as the first fundamental step towards improving traffic and pedestrian safety in their communities. One of the reasons that the RCOC is able to achieve such a positive effect on highway safety is due in a large part to the efforts of the Oakland County Traffic Improvement Association (TIA). This 38-year-old private/public non-profit corporation, which receives funding from public sources, grants, private donations and corporate sponsorship, collects, compiles, and analyzes crash data for all roads in Oakland County.

The TIA established standards for capturing and returning crash data so that almost all of the reported crashes are accurately located, incorporated in an electronic database, and summarized. Data with accompanying detailed individual reports are available to agencies within 60 days of the date of the reported crash. The TIA has established good working relationships with 68 cities and villages, the RCOC and more than 45 police agencies. TIA staff collects or receives reports after the police supervisor, such as a lieutenant, completes a review of the investigating officer’s report. The TIA has developed a cooperative arrangement with the Michigan State Police in which it is permitted to receive crash reports from local police agencies and enter the reports into a database before they are forwarded to the State Police. When one or more officers note a particular quality issue, the TIA can contact the appropriate police agencies and affect positive improvements in reporting practices. This may explain why TIA’s claim of 100 percent accuracy with respect to location, although they frequently must use officers’ estimates recorded on the form.

Distributed Responsibility for Crash Data Entry

It is noted that the Oakland County TIA is somewhat unique in that it is a private, non-profit organization providing services that are typically handled by public agencies. Moreover, TIA is unique in that it can obtain the police crash report forms directly from numerous police agencies. When talking about intersection safety with other agencies during the scan, there were some expressions of frustration with waiting for highly inaccurate crash data from the central unit within the State agency responsible for processing the crash data. Charlotte, North Carolina, and Richardson, Texas, also code their own crash data with their own local agency personnel and maintain their own crash records systems. The benefits resulting from this investment of local agency resources can be significant, depending on the relative backlogs and processing efficiency of the responsible state agency. If the city experiences a relatively manageable number of crashes reported in a year, then it may be possible for one individual to devote a percentage of his/her time to entering crash data. If there are a relatively large number of reported crashes, then one or more people may need to be exclusively dedicated to code crash data. Representatives from Charlotte and Richardson claim that the benefits from coding their crashes were invaluable and well worth the investment. The data entered into their local crash records systems gives them the ability to identify high crash locations, to assess the costs to their communities, to generate yearly comparisons of crashes, and to investigate crashes by potential contributing factors, among other items.

Charlotte developed a series of tools to assist its analysts in locating crashes from the crash reports. This is presented in figure 4. The underlying map to the right in the screen view is a map showing the location of the subject intersection as a red dot. The information to the left is information for various location fields. Charlotte's transportation officials indicated that this is most helpful in resolving uncertainty about whether a crash has been tied to the correct intersection. Like many cities, Charlotte has many of the same challenges with respect to locating crashes. These challenges include streets with multiple street names, two sets of roads that cross twice creating duplicate intersections, route numbers and street names used to indicate the same road, and various errors. Like any system, improvements are made over time.

Screen shot of the Charlotte, NC crash location identification tool.
Figure 4. Photograph of screen showing Charlotte's tool to find crash location.
(Courtesy of Charles Jones, Charlotte DOT)

Screen shot of Charlotte's crash location identification tool showing a location narrative.
Figure 5. Screen view of sketch with narrative for individual crash location reported captured in Charlotte's crash records system.
(Courtesy of Charles Jones, Charlotte DOT)

In addition to this locating-assistance tool, Charlotte has also created other capabilities within its crash record software system to assist the analyst. Specifically, it has devised a way for the user to create sketches of the crash diagram and to enter the narrative such that it becomes part of the crash record stored and therefore retrievable by analyst. Figure 5 depicts the screen showing the narrative and sketch of a specific crash report.

Web-Based Safety Data Systems

During the scanning study, a few agencies indicated that they have created and maintained their own crash records systems and have made them available to others. This could ultimately produce improved intersection safety benefits. The Michigan State Police makes its crash data available to local police agencies by means of a Web-based tool that features additional security features (e.g., passwords). This allows the local police agencies to not only check the data that was extracted from the police crash reports that they prepared, but also to run statistical summaries and execute selected queries of a limited number of fields in the crash records database. The Michigan State Police and the Governor’s Traffic Safety Advisory Council (GTSAC) are expanding this system to make it available to Metropolitan Planning Organizations, State University Transportation Research Centers, road commissions, and others. Currently there are 50 non-law enforcement users, but they project that there could ultimately be as many as 2,500 users.

The Southeast Michigan Council of Governments (SEMCOG) has created a resource tool that can be used by its member agencies and others as well, since it is a Web-based system that runs on SEMCOG’s Web site, www.semcog.org. SEMCOG’s Transportation Data tool allows users access to crash and traffic data. As expressed in the information on the SEMCOG Transportation Planning Web site (http://www.semcog.org/TranPlan/TransportationDataTool.htm):

“By reducing the time needed for data gathering, more time is available for analysis. By providing linkages across many data sets, less time is needed for gathering data, and resources can be maximized for greater insights in analysis. Ultimately, this should result in more timely and informed decisions, and a better transportation system.”

“Up to now, the process of compiling data to analyze a problem has often involved searching for data sources, contacting various data providers to get the data, repeated contacts to get more information about the data, then sorting through the data and integrating data from various sources for analysis. This can be a very labor-intensive process, often taking months. The Transportation Data Tool is an initiative aimed at streamlining this whole process.”

Spatial Analysis Systems and Analytical Tools

Clearly, law enforcement agencies and local governmental units benefit greatly from improved access to crash data. When combined with complementary analysis tools, notably spatial analysis tools, even greater benefits to improved intersection safety are possible. Figure 6 presents a GIS-based map of the location of pedestrian deaths reported over a six-year period. The source data was the Injury Prevention Center of Greater Dallas. Spatial representations of crash locations allow technical personnel, management, and the general public to quickly see crash patterns.

Map of Dallas county with pedestrian deaths identified.
Figure 6. Dallas County pedestrian deaths map.
(Courtesy of NCTCOG)

Similarly, figure 7 presents a crash map for a high crash location in Michigan that was generated from the SEMCOG Web site. This particular 3-D plot represents crashes reported between 1997 and 2002, inclusive, within 150 feet of an intersection. For this map, one chip represents five crashes. The color coding is as follows:

Screen shot of computer generated map with 3-D bars at locations where crashes have occurred.
FIgure 7. Illustrative example of spatial crash summary generated by SEMCOG.

The intersection with the most “stacked chips” is the intersection with the highest number of reported crashes in that county for that five-year period. The crash map shows the crash experience at the other nearby intersections within the grid view. Detailed tabular summaries are also available to the user, but this map visually demonstrates how information shown graphically can present a much more interesting story in far less time than a table of numbers. It is important to note that the data on the SEMCOG Web site can be accessed by the public and that no password is needed. The data elements available do not include any personal information.

Interagency and Intra-Agency Cooperation

Several of the host agencies described programs that were created as a direct result of inter-agency and intra-agency cooperation. Portland, Oregon, identified several notable examples where diverse groups of people were brought in to develop programs that ultimately will produce improvements in intersection safety throughout the city. These included the following:

Charlotte created a process (figure 8) that involves a number of different offices from within the city's Department of Transportation and other agencies. The Charlotte Department of Transportation Safety Commission includes representatives from the traffic safety section, the engineering and operations (i.e., signs, signals, and markings) division, the city engineering and property management division, the planning and design division, the implementation section (which monitors signal installations), a police officer involved with higher-level interdiction, and a representative for the transit system. As a group, efforts are made to identify locations that will need correction by means of major improvements and projects for which low cost improvements are appropriate. This group is also involved in reviewing and scoping major improvement projects to determine if there may be no feasible solutions. Since they represent different groups from within the Department, they also serve as conduits of safety considerations for their respective groups.

Flow chart of Charlotte's safety improvement project selection and evaluation process.
Figure 8. Charlotte's safety improvement project selection and evaluation process.
(Courtesy of Charlotte DOT)

The Michigan State Intersection Safety Action Plan is one example of where interagency and intraagency coordination can truly affect a positive change in intersection safety. The plan was initiated by the Governor's Traffic Safety Advisory Commission (GTSAC). Recognizing that on a national level, 40 percent of all crashes reported in 2002 were intersection-related, the GTSAC identified intersection safety as one of its three main issues to address. The GTSAC created an Intersection Safety Action Team. Using the "National Agenda for Intersection Safety"(3) as a guide, it developed the Michigan Intersection Safety Action Plan. The plan included specific goals to achieve annual reductions in total intersection crashes, total intersection fatalities, and total intersection injuries. The goals are that in 2009 there will be less than 100,000 intersection-related crashes reported in Michigan, less than 3,000 people injured as a result of intersection-related crashes, and less than 300 people who die as a result of intersection-related crashes.

The plan included strategies to address the following:

The agencies involved on the Intersection Safety Action Team included the Michigan Department of Transportation, the Office of Highway Safety Planning (which is an office within the Michigan State Police), the Michigan State Police (specifically the Traffic Services Section of the Special Operations Division), the Federal Highway Administration's Michigan Division Office, SEMCOG, the RCOC, the city of Troy, the Michigan Center for Truck Safety, Wayne State University, and DLZ Michigan, Inc. (a private consulting firm involved in the design of roundabouts). Each agency voluntarily assumed responsibility for aspects of the plan relevant to its purviews. One of the most significant benefits from the plan was the formation of strong bonds between and among various agencies. By sharing in the plan and working closely in various meetings, the level of interagency and intra-agency cooperation achieved made it possible for diverse interests to share common ground.

There is still an ongoing discussion over the next steps for the plan and whether further encouragement is needed or governmental mandates are required. Since its approval in February 2003, the Intersection Safety Action Plan lists several of its many accomplishments to date:

Public-Private Partnerships

In an age of limited resources, public-private partnerships offer substantial promise to create an environment to implement and sustain safety process improvements. One of the most notable examples of a successful public-private partnership is the well-documented road improvement demonstration project sponsored, in part, by the American Automobile Association Club of Michigan (AAA Michigan). The project was implemented jointly with Detroit and Grand Rapids, and resulted in the implementation of specific improvements at selected intersections. Since AAA Michigan is also an auto insurance provider to citizens in Michigan, it too had a vested interest in improving intersection safety. Fewer crashes mean fewer claims and lower payouts for damages that result from crashes. Sharing a mutual goal was sufficient incentive for AAA Michigan to get actively involved. Moreover, this AAA club has a full-time traffic engineer, who is one of three employed in all of the AAA clubs throughout the United States. It was only a natural extension of the engineer's interest and background that AAA Michigan would be the lead in a public-private partnership with traffic engineers working for public agencies to devise a way to implement intersection improvements. The scan team also learned that AAA Wisconsin, which also is an auto insurance provider, is endeavoring to implement a similar project.

The North Central Texas Council of Governments (NCTCOG) is currently considering a means to solicit the involvement of one or more insurance agencies to capture claims data to complement or supplement the crash data that is provided. There has been a significant delay in the processing of crash report forms at the Texas statewide level so that the lag time between when a crash occurs and when it is retrievable from a records system has exceeded two years.

Previous | Table of Contents | Next

back to top

Office of Operations FHWA Safety Home