blank space

South Dakota Department of Transportation
Project Synopsis
SD1999-05


Title: Identification of Methods for Truck Crash Reduction
Project Researcher: Stephen Tracy, USD
Project Manager: Jon Becker
Research Period: 2/15/1999 - 10/31/1999
Status:
Cost: $50,000.00

Problem Statement:In South Dakota, trucks (as reported in the National Governors Associations SAFETYNET database) involved in fatal crashes represents 8.2% of all vehicles, compared with 3.7% represented in all types of crashes. These statistics reflect national statistics. This number indicates that when a truck is involved in a crash, it is more often a fatal crash than when only smaller vehicles are involved. Clearly, trucks are over represented in fatal crash statistics.

The state of South Dakota maintains an electronic database containing all reported vehicle accidents. This database, called the South Dakota Motor Vehicle Traffic Accident file, does not contain information regarding trucks. If the accident involves a truck meeting certain criteria, a supplemental form is filled out with additional data, which is stored in a national database called SAFETYNET. Personnel can not easily combine the information from the two sources to detect trends or groupings of accidents. The ability to do this is necessary to correct potentially dangerous conditions.

This study would use information from SAFETYNET and the South Dakota Motor Vehicle Traffic Accident file to identify the following accident factors:

  • problem areas-routes with a high number of commercial vehicle crashes

  • problem drivers-by age, driver qualification, experience

  • other accident characteristics-time of day, day of week, weather, highway class, and geometry, etc.

  • motor carriers that posea greater safety risk.

  • These are all factors that contribute to accidents. These data would be analyzed, and the results would be used to develop new programs or strategies for reducing accidents in the state. New programs could include education of high-risk passenger vehicle drivers; focused enforcement in high-risk corridors; and the targeting of enforcement resources on motor carriers that pose a greater safety risk to the general public based on historical safety data. A tool for reassessing causal factors, perhaps on an annual basis, also would be developed as part of the research, so that the programs are not static but respond to changing conditions.



    Findings: 1. The number of deaths associated with accidents involving a commercial motor vehicle declined from 25 in 1996, to 21 in 1997 and 14 in 1998. Part of that decline reflects a decrease in accident severity and the remainder to a decrease in the number of accidents. The accidents per million miles traveled within the state remained fairly constant between 1996 and 1997 but the number of fatal accidents declined by two, which resulted in the number of deaths declining by four. In 1998, the accident rate decreased from 29.5 to 23.3 per million miles traveled. There were fewer accidents, fewer fatal accidents and fewer deaths in that year. 2. Alcohol involvement in commercial vehicle accidents increased in 1997 but fell below the 1996 level in 1998. There were two people killed in alcohol-related accidents in 1996, three in 1997 and one in 1998. The 1998 improvement is desirable but there are no causal relationships to explain the improvement. Further, the number of deaths each year is small and summary statistics are greatly affected by one or two events. 3. Safety restraint usage is the lowest among the younger drivers. It can not be determined from the data if lower safety restraints usage is an issue of attitude, style or if this group is simply uninformed. We recommend additional efforts in convincing young people that safety restraints are necessary and important to safe vehicle operation. Beyond the potential for restraints to reduce injuries, nearly one fourth of all fatalities during the three year period involved partial or total ejection from the motor vehicle. 4. Sites where multiple accidents involving commercial motor vehicles on the Interstate Highway system have occurred are typically at or near an entry or exit point. The accident sites include points near on and off ramps and exit and entries for rest areas. It would appear that there are failure to yield issues at points of entry and exit in that the commercial motor vehicle is most likely to be going straight when the accident occurs. 5. Regression analysis shows that safety and restraint issues figure prominently as factors that are statistically significant in explaining factors that contribute to fatalities in commercial motor vehicle involved accidents. The regression model explained only 16.9 percent of the variation in the data and is considered to have weak explanatory power. The lack of systematic causes for the accidents, a limited number of locations where accidents are common and the lack of explanatory power in the regression model suggest that there is a strong random element in the occurrence of accidents involving commercial motor vehicles. 6. The regression analysis for injuries is also described as being weak in explanatory power, explaining only 16.5 percent of the variation. The factors found to be statistically significant in explaining injury accidents were conditions, type and place of collision and safety factors. As with fatal accidents there does not appear to be a systematic cause, event or issue that explains a large percentage of the accidents involving commercial motor vehicles. 7. Weight ratings of commercial motor vehicles as they relate to accidents is an interesting issue. Department of Transportation personnel told us that their estimate of trucks requiring a permit as a result of weighing more than 80,000 pounds is less than ten percent of all trucks. Some safety officials estimate from their experience that the proportion of trucks weighing more than 80,000 pounds is below five percent. Using the number of South Dakota interstate vehicles registered by South Dakota Division of Motor Vehicles (under the Interstate Registration Plan) in 1999, about 68% are weight rated at or below 80,000 lbs. The issue is that this small group of trucks with GVW ratings of 80,000 to 120,000 pounds represents nearly half of all accidents involving commercial motor vehicles. And, they are involved in fatal accidents at a rate over two and a half (2.7) times that of trucks in the 40,000 to 80,000 category. While no conclusion can be reached, a closer look is indicated. 8. The Interstate Highway system is a highly used and an extremely important route for those in the trucking industry. Because of the frequency of use, the number of crashes on an interstate highway is greater than on other roads. While the percentage per mile driven is not as high, the sheer traffic on interstate highways necessitate special concern. 9. The non-preventable factors, such as weather and light conditions, are not nearly as significant as the more preventable factors. With the exception of Blowing Soil, Dirt and Sand and Fog, Smoke the non-preventable conditions are not significant in the crashes in the dataset. 10. Driver error, from any of the vehicles involved, is very significant. The regression indicates that injuries and fatalities occur when driver error increases. The linked dataset is unable to recognize the vehicle involved that caused the crash, but the results point to the need for continued education in defensive driving. 11. The linked database is limited in its value. Because the dataset includes only the DOT reportable crashes (as defined in Appendix E, Glossary and Acronym List), factors that may contribute to crashes in general may be underrepresented. The current system for indicating the location of the crash is not adequate to develop meaningful geographical locations of crashes. 12. An examination of the frequencies and the rating system in SAFETYNET failed to identify any individual trucker or trucking companies that pose a statistically significant safety risk. As data from subsequent years is collected, future linkages may be able to identify carriers who pose a greater risk.

    Research Objectives:
    1  To develop a software analysis tool to identify accident contributing factors from data available in SAFETYNET and the South Dakota Motor Vehicle Traffic Accident file.
    2  To use crash data from SAFETYNET and the South Dakota Motor Vehicle Traffic Accident file to identify contributing factors.
    3  To recommend methods for reducing the number of truck crashes in South Dakota.

    Research Tasks:
    1  Perform a literature search relevant to reducing truck crashes.
    2  Meet with the technical panel to discuss the research, scope and work plan.
    3  Conduct interviews with SD Highway Patrol, SDDOT, and South Dakota motor carriers (determined by the panel), to document existing accident databases, accident data collection forms, and procedures.
    4  Create a temporary electronic database that establishes a relationship between and combines information from the South Dakota Investigator’s Motor Vehicle Traffic Accident File and SAFETYNET.
    5  Analyze the combined database to identify contributing factors, accident severity and geographic locations.
    6  Provide SDDOT with procedures, system, training, materials, etc. that will identify the contributing factors (high-risk areas, carriers, time of day), either by report or graphical county or state map.
    7  Based on the analysis of the database, recommend crash reduction methodologies.
    8  Prepare final report including methodology, findings, conclusions and recommendations.
    9  Recommend to the panel a method to use current resources to improve the accident data reporting and collection system.
    10  Present any procedures, products and findings to the technical panel.
    11  Make an executive presentation to the Research Review Board.

    Documents Available:
    SD1999_05_final_report.PDF
    blank space