USA Banner

Official US Government Icon

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure Site Icon

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

U.S. Department of Transportation U.S. Department of Transportation Icon United States Department of Transportation United States Department of Transportation

Public Roads - September/October 2014

September/October 2014
Issue No:
Vol. 78 No. 2
Publication Number:
Table of Contents

Your Go-To Data Source for Roadway Safety Research

by Mark Fitzgerald

The HSIS just might hold the answer to your next big question about the safety of highway design and operations—from the Safety EdgeSM to LEDs. Check it out today for your next project.


This sloped pavement edge, known as the Safety EdgeSM, makes it easier for motorists who inadvertently drive onto the shoulder to reenter the roadway safely. Using HSIS data helps determine whether countermeasures like this can reduce the potential for run-off-road crashes.


Highway safety data can help solve safety problems through improved design and operating decisions. Take, for instance, a serious safety problem that can be caused by pavement edge dropoffs. Suppose a driver veers too far to the side of the highway, and the vehicle’s tires drop off the paved edge. When the driver tries to steer back onto the road, the pavement edge is too sharp and steep, and the sides of the tires rub along the paved edge — a situation known to engineers as tire scrubbing. The driver then overcompensates, turning the steering wheel hard to achieve a greater angle for reentry. But the vehicle is moving too fast, so the tires slam into the pavement edge, and the vehicle careens across the road into oncoming traffic, overturns, and causes a fatal crash.

Although national-level data on crashes related to pavement edge dropoffs are scarce, the data reported by some States suggest that the extent of the problem is considerable. In Iowa, for example, a study funded by the Federal Highway Administration (FHWA) and the AAA Foundation for Traffic Safety found that pavement edge dropoffs may have contributed to 18 percent of rural run-off-road crashes on paved roads with unpaved shoulders. In Missouri, pavement edges may have been a contributing factor in nearly 25 percent of run-off-road crashes. The study in Iowa also found that dropoff crashes are four times more likely to result in a fatality than other types of crashes on similar roads.

These data raise questions about actions that States and localities could take to mitigate the dangers posed by pavement edge dropoffs and improve roadway safety. Because of the need to stretch transportation budgets, many State departments of transportation are looking for solutions that will enhance safety at a low price tag and reduce operating costs in the long term. Thankfully, a number of resources are available to analyze these kinds of problems and inform decisionmaking related to the design and operation of U.S. roadways.


Evolution of the Database

According to Forrest Council, a senior research scientist at the University of North Carolina Highway Safety Research Center, the need for a national database merging both crash and roadway inventory data spurred creation of the Highway Safety Information System (HSIS).

“Prior to the establishment of HSIS, those who wanted highway safety data had to go out and get their own data,” he says. “Usually they had to get the data from one State, and each research study meant paying for the data collection that was needed.”

The creation of the database enabled FHWA to answer questions for in-house research projects and provided the larger highway safety research community with a better understanding of how roadway design can reduce crashes.

In 1987, the HSIS project team selected Illinois, Maine, Michigan, Minnesota, and Utah to be included in the database. In 1995, they added California, North Carolina, and Washington. Ohio was added in 2002, and in 2011, they added data from the urban area of Charlotte, NC.


One of those resources is FHWA’s Highway Safety Information System (HSIS), a multi-State database that contains statistics on crashes, roadway inventories, and traffic volumes. Researchers and policymakers can use these data to analyze such factors as the geometric design of roadways, the selection and placement of roadside hardware, the use of traffic control measures, the size and performance capabilities of vehicles, and the needs and abilities of highway users.

“We’ve used the database fairly extensively on a number of projects that involved analyzing infrastructure-related features,” says Eric Donnell, associate professor of civil engineering at Pennsylvania State University. “We’re currently using data from North Carolina and Ohio to determine a crash modification factor for the Safety EdgeSM , a countermeasure designed to address roadway departure crashes at locations with pavement edge dropoffs.”

Applying HSIS Data to Pavement Edge Dropoffs

The Safety Edge is a pavement edge treatment that makes it easier for drivers to safely reenter the roadway after inadvertently driving onto the shoulder, reducing the likelihood of crashes. The angled edge is formed by a device affixed to the paving equipment that creates a 30-degree slope on the pavement edge. In addition to the safety benefits, this design also creates a more durable edge that resists raveling — the loosening or falling off of pavement materials. 

Researchers who were part of the HSIS team conducted a 3-year study that considered sites both with and without the Safety Edge treatment. They found that for two-lane highways in Georgia and Indiana, the Safety Edge treatment reduced crashes by approximately 5.7 percent. Using cost information collected from the participating States, they also determined that the Safety Edge is relatively inexpensive, approximately $536 to $2,145 per mile ($335 to $1,340 per kilometer) for application on both sides of a roadway.

Under another FHWA-funded study currently in progress, Development of a Crash Modification Factor for the Application of the Safety EdgeSM on Two-Lane Rural Roads, researchers are using HSIS data to determine the crash modification factor (CMF) of the Safety Edge. A CMF helps estimate the safety effectiveness of various countermeasures and provides information that engineers can use to analyze benefits and costs associated with a given countermeasure. This information, in turn, helps with the selection of appropriate treatments to improve safety. In the case of the Safety Edge, the researchers expect the CMF to increase understanding of the potential crash reductions for various geometries and provide a greater justification for widespread implementation.

According to Jerry Roche, a transportation specialist with FHWA’s Office of Safety, who previously worked with the agency’s Iowa Division and helped with that State’s policy decision to use the Safety Edge routinely, crash reduction and cost are significant factors in the decisionmaking process. “Say there are two countermeasures designed to address the same type of crash and one has a higher CMF,” he says. “You’ll analyze the cost of each countermeasure over its life cycle compared to how many crashes they are expected to reduce over that lifetime before making a decision on which to use, or whether to use both.”

Thanks to FHWA’s Every Day Counts (EDC) initiative — which aims to identify and deploy innovations that can shorten project delivery, enhance roadway safety, and protect the environment — awareness and implementation of the Safety Edge has increased significantly across the country.

“Through the EDC program, we were able to show the value of the Safety Edge to States through demonstration projects,” says Cathy Satterfield, who leads the Safety Edge team in FHWA’s Office of Safety. “More than 40 departments of transportation have since adopted the Safety Edge as part of their paving practices. The large number of miles of roadway paved with the Safety Edge in the past few years is saving lives and also provides us with an opportunity to measure progress. Now we’re collecting data on many newer installations and combining it with older data, which will result in a strong and statistically robust set of CMFs.”

Using HSIS Data for LED Conversion Studies

Researchers also have used HSIS data to evaluate the safety benefits derived from converting traffic signals from incandescent bulbs to LEDs, which are significantly more energy efficient and have a longer lifespan than incandescent bulbs. Although LEDs can reduce energy consumption and decrease maintenance, according to “LED Traffic Signal Retrofits: Implications for Intersection Safety,” a paper sponsored by the Transportation Research Board’s (TRB) committee on Traffic Control Devices (AHB50) and presented at the organization’s 91st Annual Meeting in 2012, they might pose some other problems. For example, because they do not generate as much heat, LEDs are unable to melt snow that accumulates on them, and some drivers report that LEDs cause visual discomfort due to glare when viewed at night.

In fact, the TRB paper reports that a study conducted in Memphis, TN, found that the conversion of incandescent bulbs to LEDs actually increased crashes by 47 percent. In another study, titled “Analyzing the Effects of LED Traffic Signals on Urban Intersection Safety,” conducted on urban roads in Middletown, OH, researchers found that converting incandescent bulbs to LEDs resulted in a 71 percent increase in crashes.


A commercially available “shoe” device installed on a paving machine, as shown here, shapes the edge of the pavement to a 30-degree slope, resulting in a more durable, safer edge.


But according to Raghavan Srinivasan, a senior transportation research engineer at the University of North Carolina Highway Safety Research Center (HSRC) and a member of the HSIS team, there were some methodological issues with these studies that call for cautious interpretation and application of the results. “Both studies had limited sample sizes,” he says. “The studies used data from only eight intersections where incandescent bulbs were replaced by LEDs and data from two comparison, or reference, sites.”

In a recent project, described in the HSIS summary report Safety Evaluation of Converting Traffic Signals from Incandescent to Light-Emitting Diodes (FHWA-HRT-13-070), researchers used a larger sample of treatment sites to determine whether the conversion of traffic signals from incandescent bulbs to LEDs could affect crashes. Another objective was to quantify the effect (if there was one) in terms of a CMF. A CMF greater than 1 would suggest that the treatment could lead to an increase in crashes, while a CMF less than 1 would imply that the treatment could lead to a reduction in crashes.

The project involved using HSIS data provided by the Charlotte Department of Transportation to investigate signalized three- and four-leg intersections that underwent conversion from incandescent bulbs to LEDs. In 2008, the agency contracted with a firm to convert the majority of the city’s signalized intersections. About 90 percent of the signalized intersections were converted over the following year.


Researchers are using HSIS data to evaluate how incandescent bulbs (top) compare to LEDs (bottom) in traffic signals. Although LEDs use less energy to operate and require less maintenance, researchers want to learn more about their safety effectiveness.


Using data from 282 treated intersections and more than 3,000 comparison intersections, researchers employed an empirical Bayes evaluation that compared the actual number of crashes that occurred after installation of the LEDs with an estimate of the expected number of crashes in the after period if the LEDs had not been implemented.

They found differences in the effectiveness of LEDs and wide variations in individual CMFs across the sites. For three-leg intersections, the CMFs for various crash types and times of day ranged from 1.016 to 1.177, indicating a possible increase in crashes after the conversion to LEDs. However, the CMFs were not statistically different from 1.0 at the 0.05 significance level. It could not be determined whether the treatment could lead to an increase or a reduction in crashes, as the increase in crashes could be partially due to other modifications, such as addition of protective left-turn phasing, or other random factors.


The HSIS Workroom

The Highway Safety Information System (HSIS) workroom, located at the Turner-Fairbank Highway Research Center in McLean, VA, provides a site where researchers can advance the operation and maintenance of the database. The workroom also facilitates research to support FHWA’s mission and strategic goals, and provides data to highway safety researchers. The workroom offers an environment where FHWA staff and visiting researchers can access a variety of analytical tools in the study of highway safety. The HSIS workroom is staffed by two full-time analysts and two graduate research assistants.

Equipped with advanced computer hardware and software, the workroom is designed to handle data collected by States on more than 5 million crashes and inventory and traffic volume data for approximately 165,000 miles (265,500 kilometers) of highway. The workroom contains the only national linkable database that has common identifiers on crashes, roadway inventory, and exposure, enabling researchers to associate the risk of crashes with roadway and traffic variables.


For four-leg intersections, the CMFs ranged from 0.827 to 1.091, and five out of the eight CMFs were less than 1.0. But for rear-end crashes during all time periods, the CMFs were lower than 1.0 and statistically significant, indicating a safety benefit from the conversion to LEDs for this crash type.

“The safety effects of the LEDs varied significantly across the large number of sites,” Srinivasan says. “Additional future research may be able to better explain the differences across many locations, or provide a more nuanced understanding of the effects. In the meantime, LEDs continue to offer a more energy-efficient alternative to incandescent bulbs, with overall comparable safety performance, if not some safety benefits in certain scenarios.”

Using HSIS Data to Develop Guidance for Interchange Projects

According to James A. Bonneson, a senior principal engineer with Kittelson & Associates, Inc., the HSIS database has been useful for researchers in answering questions State highway officials might have about safety influences and crash trends.

“The database helps researchers answer questions that previously could not be answered, such as whether roads with narrow shoulders are associated with more frequent crashes than roads with wide shoulders,” he says. “For instance, if some sections of roadway with sharp curvatures have more crashes, then patterns in crash frequency start to develop and can be quantified using the HSIS data.”


Shown here is a section of West Dodge Road in Omaha, NE, built in the 1990s by the Nebraska Department of Roads. The elevated roadway increased capacity and mobility, reducing traffic coming off I–680 and West Dodge Road and improving northbound and southbound access into and out of this busy metropolitan area.


In the early 1990s, Bonneson conducted research for the Nebraska Department of Roads (NDOR) using HSIS data on that State’s rural expressways. The project, Interchange Versus At-Grade Intersection on Rural Expressways, focused on determining the traffic volume levels at which an interchange would be more cost effective than a conventional signalized intersection. To achieve their objectives, the researchers quantified the benefits to road users based on consideration of traffic safety and operational costs.

One option was to make improvements to the existing intersection by installing a traffic control signal, additional through lanes, and left and right turning bays. Another option was to construct a multimillion-dollar interchange. Bonneson’s team used the HSIS data to develop models for estimating the crash frequency for each option. His team then used the results of the research to develop guidelines that NDOR could use when evaluating the intersections on several planned expressways.

“We provided NDOR with a guideline document,” says Bonneson. “We also provided charts and graphs that they could consult on a case-by-case basis so they could answer certain questions about the most cost-effective option.”

Based on the guidance and further analysis, NDOR decided to build an interchange at some locations, which improved roadway safety, operations, and service. “The main consideration was the [traffic volumes] on the cross streets and main lanes,” notes Bonneson. “The interchange typically offered considerable benefits in safety and operations, reducing crashes and reducing delays.”

The researchers described in the study conditions for rural expressways where conversion from an intersection to a diamond interchange is justified based on a benefit-cost analysis.

Data Requests and Support

Researchers seeking access to HSIS data first need to complete a request form that asks them to describe the specific research question they plan to address and how the results of their research will be used. They also are asked to define a preliminary set of variables of interest based on the following file types in the database: accident (the term used on the form), vehicle, occupant, roadway inventory, traffic volume, intersection/interchange, curve, and grade. The reasoning behind the questionnaire is this: FHWA designed HSIS to provide data to be used in research conducted in the public interest and intended for publication in a scientific journal or other national publication. By submitting a request form, the user agrees to follow the HSIS guidelines.


Data Users and File Types

Data files in the Highway Safety Information System are stored in a relational database that enables FHWA staff to provide data to users in various formats, such as Microsoft® Excel® and Access®, dBase, and American Standard Code for Information Interchange (ASCII). These files also can be converted to SAS® format for analysis.

Primary users of the database include researchers conducting studies for national or State-funded projects. These researchers pursue studies in the public interest and their findings often appear in scientific journals and other national publications.

Frequent users include staff with FHWA’s Office of Safety and Office of Safety Research and Development and university researchers, either conducting their own research or research for State departments of transportation. Other major users include researchers involved in efforts funded by the National Cooperative Highway Research Program, sponsored by the American Association of State Highway and Transportation Officials. Other users include metropolitan planning organizations, private research companies, and international researchers.

FHWA conducts extensive quality control checks on the data it receives from States. Each year, HSIS analysts examine new data files and compare them to the previous year’s data. Then they develop metrics to measure differences in the data. If more than a very small difference is found between the values of a given year and the previous year, the analysts check the variable to understand the difference. The goal is uniformity and consistency year after year.


Except for HSIS contractors and FHWA staff, outside researchers seldom work onsite in the workroom at FHWA’s Turner-Fairbank Highway Research Center. Virtually all data are exchanged electronically. The user requests specific data variables by reviewing HSIS file descriptions online and then completing a request form on the HSIS Web site, The user receives the requested files via email, an FTP site, or compact disc.

According to the original agreements with the participating States and the need to document each use of the HSIS data, FHWA staff cannot send researchers HSIS files with every variable. However, they can extract files containing subsets of variables. If a user wants FHWA staff to link the different file types for them, such as linking crashes and roadway inventory data, the user is asked to specify whether the data record needed is crash based (where roadway descriptors are added for each crash) or roadway segment based (where specific types of crash counts are added to each roadway segment record).

FHWA staff conducts quality control checks on all data added to the HSIS, ensuring that the data are uniform and that documentation is available. Through informative conversations, mentoring, and technical support, staff also provides guidance to researchers on using the data. For example, researchers might not know whether to request a roadway-based file with certain crashes appended to it, or a crash-based file with roadway data appended to it. Based on the specific research question being asked, HSIS staff helps them understand what kinds of data are available and how the file will be structured.


Research Products

FHWA produced a series of HSIS summary reports like this one that include brief descriptions of the issues addressed, data used, methodologies applied, significant results, and practical implications.

HSIS data are applicable to a wide variety of research efforts, and some of the products of these studies include HSIS summaries, research reports, safety analysis tools, published articles, and technical papers in professional journals.

FHWA’s Interactive Highway Safety Design Model (IHSDM), a suite of software tools that expedites the safety evaluation of existing highways and proposed design alternatives, is one example of a product that resulted from using the HSIS database. Overseen by FHWA’s Geometric Design Laboratory, the IHSDM can be used to evaluate two-lane rural highways, multilane rural highways, and urban and suburban arterials. The software supports the American Association of State Highway and Transportation Officials’ (AASHTO) Highway Safety Manual, Part C: Predictive Method.


“It’s always difficult to say we did research that led to a decision that was made 5 years after the research was done,” says Forrest Council, a senior research scientist at the University of North Carolina HSRC. “However, the Highway Safety Manual has tools that offer a lot of new ways for States and other users to analyze their data. It’s based on a series of predictive formulas and has methodologies for picking out sites to treat and also for looking at alternative designs to see which one is going to produce more crashes. A lot of the tools in the manual were based on research using HSIS data.”

HSIS data also proved instrumental in contributing to the development of Safety Analyst and FHWA’s Inter-change Safety Analysis Tool (ISAT). Safety Analyst is a set of software tools that State and local agencies can use to help inform decisionmaking with regard to highway safety. It is an AASHTO product whose development was managed under an FHWA contract and made possible through a pool of money supplied by States. The tools in Safety Analyst implement state-of-the-art analytical procedures to identify a program of site-specific improvements to enhance highway safety in a cost-effective manner.

Through the development of ISAT, researchers used HSIS data in addressing issues related to interchange design and traffic control. The ISAT tool helps engineers assess the safety effects of geometric design and traffic control features at existing interchanges and on adjacent roadway networks. Engineers also use the ISAT to predict the safety performance of design alternatives for new interchanges and for reconstructing existing interchanges. Primary outputs from an ISAT analysis include the number of predicted crashes for the entire interchange, as well as the number of crashes by interchange element type, year, and collision type.

Continuing Benefits

For more than 20 years, HSIS has served as an important resource for making informed decisions about the design and operation of U.S. highways. The database is made possible through collaboration between FHWA and the State departments of transportation in California, Illinois, Maine, Minnesota, North Carolina, Ohio, and Washington State. (Historical data from Michigan and Utah also are included.) Data from these States empower researchers to answer a multitude of questions regarding highway safety.

The database is growing continually and being refined, offering information that is accessible and ready to use. Because it is so versatile, the database enables researchers to pursue both long- and short-term studies.




Although other States might be willing to share data, adding a new State to the database is expensive because it involves learning about the new dataset, setting up new procedures, and formatting data in specific ways to meet the needs of users. For the near future, FHWA officials do not plan to add any more States to the database, but they do intend to increase the inventory of files from the States currently participating.

“With States in the north, south, east, and west, the database offers a pretty good representation of the country,” says Bonneson, of Kittelson & Associates, Inc. “The files can be merged together to help [an agency] understand what physical attributes make road segments more risky or less risky for drivers. So some very good insights can be obtained from the database.”

Mark Fitzgerald is a senior writer at Woodward Communications. He teaches writing at the University of Maryland, College Park. Before joining Woodward, he was the editor of several trade magazines and worked at the American Society of Civil Engineers. He has a B.A. in English from Franklin & Marshall College and an M.F.A. in creative writing from George Mason University.

For more information, visit or contact Mark Fitzgerald at 202–493–3995 or