Home Skip to main content

Research Projects (2009-10)

Identifying Number


Project Title

Reliability-based Safety Risk and Cost Prediction of Large Trucks on Rural Highways


Colorado State University

Principal Investigator

Suren Chen
E-mail: suren.chen@colostate.edu

Description of Project Abstract

The rural roads of the nation serve as a critical link in the transportation system around the country, which provides access from urban areas to the nation's heartland. In America, there are about 60 million people in rural areas, where rural roads provide the primary routes of travel, agricultural production, forest services, commerce and tourism (The Road Information Program 2005). Despite the fact that only 39.4% miles traveled by all vehicles are in rural areas, about 68.4% crash fatalities occur in rural highways (FHWA Highway Statistics 1998). According to the fatality analysis reporting system (FARS) by the National Highway Traffic Safety Administration (NHTSA), there are 57.8% of fatal accidents happened in the United States in 2005 are single-vehicle accidents and about 53.6% of fatal accidents happened on rural roads. According to the Large Truck Crash Causation Study (LTCCS) database, nearly 27% of all crashes involved with large trucks were non-collision single-vehicle crashes. Understanding and mitigating single-vehicle safety risks for large trucks is an important topic for rural transportation and is the focus of the proposed study.

Among various causes of single-vehicle accidents of large trucks on rural roads, many researchers suggest that the dominant causes are excessive speed and adverse environments (The Road Information Program 2005). Nationally, according to FARS, when the speed limit is 45-50 mph for both urban and rural roads, the fatal accident numbers are close (urban road 50% and rural roads 44% of total fatal accidents). While when the speed limits become 55 mph for both roads, fatal accidents for the rural road increase dramatically to 77.89% while those on urban roads only count for 18.46% of total fatal accidents (data from FARS database for 2005). In mid-west mountainous states, single-vehicle accidents are more complicated due to the uniqueness of the adverse environments. Specifically, inclement windy weather, rough road surface profiles and hilly topographic conditions (e.g. sharp curves and steep grades) among others, are often blamed for these accidents in mountainous areas. The inclement weather in mountainous areas includes strong gust, rain and snow storm. The road surface covered by snow and ice in winter season can further challenge drivers. These driving conditions, coupled with (negative) driver behavior, significantly increase the risk for single-vehicle accidents for large trucks on rural roads. It has been reported that many large trucks have experienced serious accidents due to sudden wind gusts, hilly terrain and rapidly-changing weather, such as I-25, I-70 and I-80 to Wyoming. These accidents of large trucks not only threaten the lives of truck drivers, but also endanger other vehicles and drivers nearby. The uniqueness of adverse environments on rural roads in mountainous states warrants an investigation in order to reduce crashes associated with trucks and protect truck drivers from injury. Despite serious risks of trucks in mountainous states, there is very limited resource of assessing the risks and associated costs of vehicle accidents on rural highways which can be used for the transportation management agencies and truck industry to assess the risks associated with fleet, plan the trip and make reasonable decisions when adverse conditions exist.

The present study plans to develop a framework to 1) predict the large truck accident risks with the combination of given weather, topographical, road and vehicle information; 2) based on the risks assessed, a cost analysis model is developed to quantify the potential costs and benefits associated with the trip; and finally 3) all the risks and cost information will be put on a GIS map on the Internet, which can be updated with different conditions for demonstration and educational purposes.

Project Objectives

The proposed research will provide a framework to assess the single-vehicle accident risks and associated costs of large trucks in mountainous states such as Colorado considering unique adverse environments. By applying the reliability theory, the proposed framework can consider the uncertainties associated with the coupled problem of environmental conditions and adverse driver behavior.

Project Approach/Methods

The research plan is introduced as follows:

1) Literature review and collection of historical accident data of large trucks
Comprehensive literature review will be conducted on the related studies of adverse weather, truck safety and local data of mountainous states. The data will be primarily from two databases:

  • Fatality Analysis Reporting System (FARS) http://www-nrd.nhtsa.dot.gov/departments/nrd-01/summaries/FARS_98.html
  • The Large Truck Crash Causation Study (LTCCS) http://ai.fmcsa.dot.gov/ltccs/default.asp

In addition to the two large databases mostly about fatal accidents, more non-fatal accident database will be accessed through working with Colorado Department of Transportation (CDOT) on some vulnerable rural highways (e.g. I-25 and I-70). From all the databases, several most accident-prone rural roads with speed limits between 40-60 mph (with highest fatal accident numbers in Colorado according to FARS) for large trucks in Colorado will be selected as candidate roads. Detailed historical data of those roads will be extracted for further study.

In addition to the above data, some field testing environmental data on I-70 in a project sponsored by MPC in 2007(DTRT07-G-0008) will also be used for the data analysis in Task 2.

2) Data analysis and probability description of critical variables
The data about large trucks will be analyzed using cross-tabulations and modeling such as binary logit and ordered probit models to correlate the vehicle conditions (e.g. types, weight, year), driver operational conditions (e.g. driving conditions), adverse environment (including both inclement weather and complicated topographic conditions) and single-vehicle accidents (e.g. rollover and side slipping) probability and severity. The adverse environment will be carefully characterized: the inclement weather is characterized with wind speed, rain volume, snow depth and ice-covered or not on the highway; the topographic features use typical grade angle, camber angle and curvature radius. Driver operational conditions are characterized by the driving speed and vigilance levels of the driver based on the crash records. In order to supplement the adverse environmental data which may not be complete from the database, more detailed local weather and highway topographic features at the scene will be collected from national weather service and the Colorado transportation maps, respectively. Besides, the PI will collect traffic accident data on I-70 from the Colorado Department of Transportation (CDOT). With the data analysis in this section, critical factors which contribute to the accidents most significantly are identified. Once critical variables to the accident risks of large trucks are identified, the probabilistic descriptions of some of these random variables are decided from existent studies and statistical results from the historical data in Task 1. For example, existent studies on wind speed and directions and driver behavior differences.

3) Framework of analytical framework of truck accident risk and cost assessment
A reliability-based framework of numerical assessment of accident risks of large trucks will be extended from some existent studies by the PI on a deterministic basis (Chen and Cai 2004; Chen et al. 2007). Compared to the existent study, probabilistic descriptions of those identified critical factors will be made and incorporated. Besides, more advanced vehicle dynamic model will be adopted which can also consider the adverse environmental conditions appropriately. Based on the observations from existent studies of large trucks on rural roads, rollover accidents and side slipping accidents will be primarily evaluated based on the probabilistic description of those identified critical factors. The accident risks are numerically related to the vehicle response in both rolling and lateral directions. Risk index of accidents can be numerically obtained under any combination of environmental, vehicle and topographic conditions along the rural road. Those locations with relatively higher accident risks under common combinations of conditions will be identified as "truck accident vulnerable location" (TAVL).

Based on the safety risks model, the cost assessment will be made based on existent cost model associated with accidents. This model is important to assist on the planning and decision. The research team will also try to work together with trucking companies in Colorado to develop the model.

4) GIS-Internet-based platform for planning, management and education
All of the analytical data and results will be reported in the project website. It is planned to make the risk assessment model as well as the cost-benefit analysis model available on an Internet GIS map with topographic conditions embedded. Through displaying the data on a GIS-based map, different accident risks indices can be easily displayed and compared on the GIS map. As a result, the information will be helpful for transportation management agencies or trucking companies to plan the trip, manage the fleet and educate novice drivers and people who plan to drive the studied highway about safe driving.

MPC Critical Issues Addressed by the Research

  1. High-risk rural roads
  2. Rural transportation operations
  3. Heavy vehicles and commercial trucks

Contributions/Potentials Applications of Research

Large trucks, as a type of heavy vehicles, are very prone to various single-vehicle accidents, especially under adverse environments. This is especially true for trucks moving on rural roads, where more fatalities and injuries are found than their urban counterparts. If the project is successfully conducted, the large truck safety and associated cost risks can be rationally predicted based on limited information of those critical factors. Colorado Motor Carriers Association has been contacted and they showed strong interests on related studies as the important stake holder.

It is expected that this framework will help the transportation authorities, truck industries and even emergency management agencies to better understand the risks, decide on the prevention policies and educate the drivers and general public especially under inclement weather. Firstly, the truck industry and large-truck drivers from anywhere who will drive through investigated highways can assess the risk, plan and prepare for the trip based on the forecasted weather information interactively; secondly, the study can help the transportation management agencies decide what traffic management(e.g. restriction) enforcement should be conducted under some adverse conditions; finally, the results will provide a powerful tool for truck industry as well as transportation agencies to educate and train large-truck drivers on reducing the accident and injury risks. Besides, it will also help general public and the whole society to realize the importance of safety risks exposed to large truck drivers and other drivers sharing the same highway.

Based on this study, a more comprehensive study will be actively pursued to seek federal funding on rural highway safety.

Technology Transfer Activities

Once the research is finished, the results can easily be transferred to a software or GIS-based risk prediction system, which can be used by many stakeholders. Through communicating with some stakeholders, the PI believes the prospective of technology transfer of this study is very promising.

Time Duration

July 1, 2009 through June 30, 2010

Total Project Cost


MPC Funds Requested


NDSU Dept 2880P.O. Box 6050Fargo, ND 58108-6050