Safer Roads Through Effective Traffic Safety Management

Empowering safety engineers

U.S. efforts to improve traffic safety have delivered considerable progress over the last five years. From 2005 to 2009, traffic fatalities have declined over 21%. The fatality rate has dropped from 1.46 fatalities per 100 million miles traveled, to 1.13 – the lowest rate since 1954. While this success can be attributed to a variety of factors, the focus on safety by State Departments of Transportations (DOT) and State Offices of Public Safety certainly deserves some of the credit.

I believe there are four key areas where GIS has, and will continue to assist safety engineers in reducing traffic crashes and fatalities:

  • Data Collection: It is still the case that capturing the accurate location of crash information remains a challenge for many states. It’s not uncommon for there to be a multi-year backlog for taking textual descriptions from police accident reports, and assigning accurate locations to the crash record. And yet, the collection of crash location data is paramount to the downstream analysis that helps safety managers understand the causes of these crash hotspots. GPS and GIS can provide the tools for the responding officer to accurately collect this information at the scene, and have it seamlessly incorporated into their crash reports. We must be more aggressive in adopting and implementing this kind of technology in the field.
  • Data Management: It is not enough to have our crash data geo-coded and displayed on a one-dimensional map. Safety analysts need access to a wealth of other information, such as traffic volumes, roadway characteristics, pavement and weather conditions, and even video-logs, all assembled into multi-year crash databases for use by analysts. GIS can supply the integrative framework to bring that data and its telling information into a comprehensive database for complex analysis.
  • Spatial Analysis: While a number of states have implemented impressive crash database solutions, less have taken full advantage of spatial statistics to perform more comprehensive crash analyses. GIS has the capability of uncovering spatial relationships in crash data that cannot be discovered through traditional statistical techniques. Today, many spatial statistical tools are now built into GIS software, and available to analysts to build complex models in a relatively straightforward fashion.
  • Data Dissemination: State DOTs have a statutory responsibility to submit an annual report describing the top five percent of their highway locations with the greatest safety needs. Almost all states meet that requirement with a descriptive table, but only three state DOTs include a simple map to accompany their reports. GIS technology can breathe life into these reports by demonstrating current safety improvement projects and initiatives, viewable by the public via the web. In this manner, GIS provides a powerful way for communicating the DOT’s current initiatives and successes to the traveling public. And the public itself can use this information to alter its own driving habits and behaviors.

These are ways that safety managers can achieve even greater success in the future.

What are the ways that GIS and other geospatial technologies are working to improve safety in your organization?

Terry Bills

About Terry Bills

Terry Bills is the global transportation industry manager at Esri. He has served as a Principal Planner with a regional transportation planning agency, and as President of a transportation-related GIS consulting services company. He has more than 20 years of experience in transportation planning and policy, information technology and GIS.
This entry was posted in Industry Focus and tagged , . Bookmark the permalink.

Leave a Reply


  1. Dana Steil says:

    In Alabama, we have found that giving our users (mainly patrol officers and their supervisors) a shared web portal has not only been quite useful information to them, but it has also improved the accuracy of the State’s location data for roadway events such as crashes, citations, commercial vehicle weights, and other law enforcement patrol activities. The Alabama Dashboards for Visualization, Analysis and Coordinated Enforcement (ADVANCE) provides basic statistical analysis of officers’ activities, the ability to view events on a map as individual points or as aggregated heat maps, an improved reporting capability, and an automated vehicle location (AVL) tool. Crash and citation data are collected using electronic crash and citation data entry techniques operating under the Mobile Officers’ Virtual Environment (MOVE). Once this data is entered from the field, it is immediately available to any officers with a mobile Internet connection. Otherwise, the data is uploaded by MOVE at the next opportunity.

    The live AVL view of officer location and recent activity is very popular, and it has obvious officer safety benefits. Dispatchers and supervising officers can view the locations all of the officers in their agency. Selecting an individual officer reveals the crashes and citations that the officer has responded to, along with a color-coded path of the officer’s movement for a specified time period. Clicking on the crashes or citations allows you to view the associated document and an historical AVL view is also available. In 2009, prior to the deployment of ADVANCE, only 32% of the electronic traffic citations issued by officers in Alabama’s largest law enforcement agency contained coordinates. In 2010, the same agency collected coordinates on 70% of their citations. We believe that this improvement is a direct result of giving officers and their supervisors immediate access to the data they are collecting.

    Improved deployment strategies come from viewing the aggregated data by point or heat map. Traffic safety engineers and patrol planners can view any combination of crashes, citations, and historic officer presence on a single map. This allows planners to see how frequently officers are patrolling those areas that have a high number of crashes and to determine if they are issuing citations in those places. Recently, we have been doing research and development activity for an automated patrol route planner, named Patrol-Sim, that can recommend complementary patrol rounds for multiple officers in a defined area by leveraging these historical location data.

  2. Brian Smith says:

    Safety management usually involves a number of agencies collecting, validating, and analyzing data. Data is usually spread through multiple referencing systems, road networks, and data types. The one constant is latitude and longitude. The use of spatial data will be, if it is not already, essential in the development of safer roadways.

    Traffic engineers in transportation agencies have used a number of methods to collect, review, and analyze crash data. Hard copy crash reports, out-of-date Excel spreadsheets, and individually created GIS datasets have been a standard for traffic engineers. These datasets could take days or weeks to assemble and analyze, but now, through the use of GPS and enterprise data, traffic engineers have the ability to display, map, and analyze crash data through simple mapping interfaces. This allows traffic engineers to make quicker and more accurate decisions, thus creating better decisions for safer roadways.
    As traffic engineers continue to develop the use of GIS in everyday decision-making, new and innovative uses will develop. Integrating BIM and 3D analysis will allow traffic engineers to see how designs affect traffic patterns, creating more informed engineers and safer roadways.

    Safety managers have had similar challenges as traffic engineers: long waits for data, hard copy datasets, self-created spreadsheets, and GIS data. But with the use of the same types of Internet-based mapping applications, safety managers have the ability to analyze data in as little as a one-day turn-around.

    Imagine a time when data is at your fingertips – an accident happens and a user can simultaneously find the frequency of accidents at that intersection and react to the situation. Better yet, imagine an application that provides users with limited GIS knowledge the ability to locate average accident rates and safety risk locations. These scenarios are not fantasy; this is reality with the use of GIS-based applications for data collection, validation, and analysis. The time is now!

  3. It would be great if ArcGIS included a symbolization tool that would offset ‘stack’ cases in particular location, or within a parametrized radius. When multiple crashes are logged at the same milepost or XY, only the last case read is visible. Simple tool. Big payoff for visualization.

  4. Dan Gieseman says:

    It is true that roads have been made safer through effective traffic safety management. Better data has led us to a better understanding of the causal factors contributing to traffic fatalities. However, the pace of technological change and communications integration leads me to believe that traffic safety professionals and policy makers can do even more to further the state of the art. My experience developing data collection tools suggests an integrated vision for data collection that meets the challenge of continuous improvements in the delivery of informative information to safety professionals and policy makers.

    Today, many law enforcement agencies have the capability to collect, validate, and disseminate incident data in real-time, directly from the roadside. Roadside data collection eliminates the need for paper forms and redundant data entry at the various consumer agencies. Mobile applications support point-and-click geocoding of incidents and spatial joining of roadway inventory information directly by an incident first responder. These two facets eliminate backlog and expert spatial systems ensure an accurate location is assigned to the event. GPS serves to enhance the point-and-click process, providing further location accuracy feedback to the entire data collection system.

    What is the vision? The emergence of ubiquitous, high-bandwidth mobile data communications guides systems designers toward even further integration of the various vehicular and roadside systems with ever increasing real-time capabilities. The squad car or emergency response vehicle in effect becomes an endpoint in a pipeline of data transportation from incident location to centralized data repositories and analysis functions. The utility of this communications infrastructure is only now beginning to be leveraged.

    An example of an integration opportunity is the development of a framework for vehicular systems to communicate crash event data immediately to first responders. For instance, vehicle electronic data recorders could be designed to transmit data to law enforcement and emergency response devices near the scene of a crash event. First responders would know the number of occupants, restraint devices employed, and vehicle dynamics data (e.g. deceleration g-forces). Arriving medical responders would already have insight regarding expected injury severities. Weather data from the vast network of roadside weather information sensors can be integrated with crash and vehicle systems data. If this was coupled with hospitalization and ambulance transport data, we could continue forming an ever more complete portrait of each pre-, during-. and post-crash event that can contribute to better response, identification of causal factors, and mitigation design.

    A challenge to this vision is the problem with too much data. We sometimes hear the phrase “drowning in data and starving for information.” Schemes for the effective triage of immediate roadside from data that is valuable to post-event analysts should be a research focus. To achieve the integration vision, raw event data must be quickly transformed into information. Over time, this information will lead to improved crash knowledge, and ultimately to the wisdom to reduce the carnage still so painfully evident on our nation’s roadways.

  5. Scott Sandusky says:

    Matthew, There are a few ways that we can handle “stacked” points within a GIS. Depending on the situation some of these solutions may be more appropriate than others.

    Within a web map, we can use dynamic clustering and displacement options from the web API’s like Flex ( and Silverlight ( to handle this. As you zoom in the point clusters disperse and you can click on a cluster to retrieve information of individual points.

    In the ArcGIS Desktop environment you can use cartographic representations ( There is a Disperse Markers ( tool that allows you to display overlapping points in several ways.

    Also on the desktop we can use the Maplex Label Engine ( to manage the display when the number of stacked points is below around 6 to 8. The technical details are described in a Mapping Center blog post (

  6. Dr Arshad Mahmood says:

    GIS / ARC VIEW is very important tool for the Data management ind its interpretation with spatial analysis, I have tried a lot to learn in it, ( Arc view 3.2) It is perhaps my bad luck that no advance trainings are provided here in Pakistan, if there are I never knew about these and no financial support is provided for such trainings. Cab any agency / donor manage some trainings for those who have a background of GIS work and are really interested.Government sector has no budget for such trainings.

  7. Doug Mowbray says:

    The challenge for adopting GIS spatial analysis in traffic safety management is not necessarily one of technology or motivation, or even understanding and agreeing with the benefits; it is the age-old challenge of organizations that are slow to adopt change, whether it is due to budget restrictions, legal restrictions, or just plain vanilla inertia (cured, somewhat, by attrition, i.e., retirement), to name a few. And the budget restrictions are felt now more than ever.

    While it may be cost-effective, and rather relatively inexpensive, to implement the latest GIS spatial analysis tools, it is often at odds with an organization’s (let’s just say it: the government’s) ability to upgrade all the legacy systems that would be affected, and doing so in a way that does not create risks for ongoing applications.

    Additionally, government agencies are now operating with sizeable funds from the federal government, and many of these funds are being spent on contractual services, because the State simply has not kept up in investing in its infrastructure and skilled workers. It is very costly to outsource the work, and the procurement process for software and hardware is a very slow process, so adopting and implementing new technologies can be a glacial exercise, something that keeps the State far behind private industries in this area. And when it comes to release of the data, the public is far more savvier, and highly expectant, toward access and transparency and accessibility.

    Government institutions have not, generally speaking, kept up with these distribution and accessibility questions either. There is movement, there are highly skilled people in government, and there is hope, but it seems that the expectation and potential of the technology are way far out ahead of the government’s ability, statewide, to adopt them. A particular agency, small and nimble and well-funded, may be able to get out ahead with spatial analysis, but when budgets are cut, divisions are re-organized, and the Statewide standards mandate uniformity, it becomes a very slow process to create statewide systematic management processes that are flexible and progressive.