Medina, Trimble, Clarke 1

Medina, Trimble, Clarke 1 Linking Descriptive Attributes to Crash Patterns: An Untapped Tool in State Motor Safety Improvement Programs A Comparison of Single Unit Truck and Combination Unit Truck Fatal Crashes Alejandra Medina Flintsch (Corresponding author) Senior Research Associate Virginia Tech Transportation Institute 3500 Transportation Research Plaza, Blacksburg, VA 24061 amedina@vtti.vt.edu Ph: 540-231-1508; Fax: 540-231-1555 Tammy Trimble Research Associate Virginia Tech Transportation Institute 3500 Transportation Research Plaza, Blacksburg, VA 24061 ttrimble@vtti.vt.edu Ph: 540-231-1536; Fax: 540-231-1555 Robert M. Clarke Principal R.M. Clarke Consulting 711 Alyssum Ave, Caswell Beach, NC 28465 rmclarke46@gmail.com Ph: 202-746-9628 Submission date: November 1, 2013 Total word count: (3169 text) + (8 tables) + (6 figures) = 6669 Paper submitted for peer review at the 92nd Transportation Research Board Annual Meeting

Medina, Trimble, Clarke 2 ABSTRACT Historically, the classic paradigm for studying the underlying causes of commercial vehicle crashes has been to focus on driver, vehicle, highway, and environmental factors as the bases for identifying reoccurring or prevalent patterns. There is another, untapped data source: carrier attribute/descriptive data that can be obtained from the Federal Motor Safety Administration s (FMCSA) Motor Identification Report, MCS-150. This study linked data from the MCS-150 file with single-unit-truck (SUT) and combination-unit-truck (CUT) fatal crash data in Virginia from 2005 2009 to yield a more expansive list of factors for consideration and analysis. The results indicate that the fatal crash patterns of were markedly different compared with those involving. Many more of the crashes of them involved carriers domiciled in the State of Virginia. The vast majority occurred on non-interstate highways, whereas the majority of CUT fatal crashes occurred on interstates. The fleet sizes (i.e., number of trucks operated) of carriers operating involved in fatal crashes were generally smaller compared with those of fleets operating involved in fatal crashes. Far fewer of the SUT crashes involved for-hire carriers; proportionally, more were private or other type carriers. The majority of SUT fatal crashes occurred very near to the carriers business location, whereas the crash location distances of ranged more widely. Far fewer involved in fatal crashes were hauling general freight compared with ; generally hauled a much broader range of commodities. The results suggest SUT carriers are small businesses that operate locally/regionally and face significant risks and safety challenges due to the highway environments in which they operate. New ways to assist them need to be explored. Additionally, since the crash and operational patterns of are so different compared with those of, the safety trends of these two distinct vehicle types ought to be analyzed separately, rather than together as large trucks.

Medina, Trimble, Clarke 3 INTRODUCTION This paper presents a summary and compares the descriptive attributes of motor carriers that were involved in single-unit-truck (SUT) fatal crashes with those involved in combination-unit truck (CUT) fatal crashes in Virginia from 2005 2009. This study follows a similar one by the authors that focused only on CUT fatal crashes in the state. The results of this effort like the previous one suggest that the inclusion of carrier attribute data should become a significant part of integrated state and federal commercial motor vehicle (CMV) safety improvement efforts, particularly those focusing on rural, non-interstate roadways. Large trucks make up a relatively small proportion of the overall number of motor vehicles operating in the United States. In 2009, there were 254,212,610 registered motor vehicles in the United States, 10,973,215 (4.3%) of which were large trucks (1). Fatal crashes involving large trucks are a small but significant subset of the overall number of fatal crashes that occur each year, typically composing about 12% of the total (2). In Virginia, between 2005 and 2009, there were 486 fatal crashes involving large trucks. As can be seen in Figure 1, while the majority involved, a signification proportion involved. 10 9 8 7 6 5 4 2 6 4 Proportion of Large Truck Fatal Crashes FIGURE 1 Large truck fatal crashes in Virginia, 2005 2009. should be considered separately because, among other things, licensing requirements for drivers of many are less stringent. Commercial driver s licenses (CDLs) are not required in order to operate a vehicle with a gross vehicle weight rating (GVWR) under 26,001 lbs. The majority of have GVWRs less than 26,000 lbs. In addition, Virginia does not presently require intrastate operators or vehicles with a GVWR of 26,000 lbs. or less to register with the Federal Motor Safety Administration (FMCSA). Thus, a whole other set of information and data are not available for carriers that operate these vehicles. Others have highlighted the need to study separately. A recent American Transportation Research Institute (ATRI) study (3) emphasized that while large truck crash rates

Medina, Trimble, Clarke 4 have decreased in the last 10 years, crash rates for the subset of medium duty trucks, defined in the study as trucks with GVWRs between 10,001-26,000 lbs,(which are almost exclusively ) increased. The study encouraged researchers and policy makers to specifically look at medium duty trucks when trying to reduce large truck crashes. PURPOSE AND SCOPE The proximal cause of the vast majority of all motor vehicle crashes, including those involving commercial heavy trucks, is some type of unintentional human error, usually a lapse in judgment or attention, or a failure to detect, perceive, or respond adequately to a threat (4). Underlying crash risks differ from proximal causes and involve latent propensities that increase the likelihood that crashes can occur. Historically, the classic paradigm for studying underlying crash causes has focused solely on driver, vehicle, highway, and environmental factors. Exposure to crash risks is another important variable that needs to be considered as well. The number of miles a truck is driven, the level of traffic, and the types of roads on which trucks typically travel, as well as the operating patterns of the businesses that use trucks, all contribute to either increasing or reducing the likelihood of crashes. The purpose of this study was to highlight some of those issues. DATA There were four types of data used in this study: SUT crash data, SUT carrier attribute data, CUT crash data, and CUT carrier attribute data. The crash data were obtained from state-specific databases. attribute data were obtained from the FMCSA Identification Report, Form MCS-150, and by contacting individual carriers when necessary. Virginia fatal crash data for 2005 2009 were obtained from the Virginia Department of Motor Vehicles (VDMV) and the Virginia Department of Transportation (VDOT). The Virginia Highway Safety Office (VAHSO), through its Traffic Records Management, Reporting and Analysis Division, manages the state s highway safety traffic records information system, which houses millions of traffic crash records. These data are housed in the Traffic Records Electronic Data System (TREDS). VDOT maintains the VDOT Roadway Network System (RNS) database. As a result of a modification to the police accident report form FR 300, existing database fields were deleted or consolidated and new record fields were created. Attribute Data MCS-150 information is available online for authorized users (including the carrier) via the FMCSA Safety and Fitness Electronic Records (SAFER) System. A Company Snapshot provides a concise electronic record of a company s identification, size, commodity information, and safety record, and includes the safety rating (if any), a roadside out-of-service inspection summary, and crash information. Initial SAFER system searches of the MCS-150 data were conducted using Virginia carrier records and involved secondary analysis when the SAFER search records did not match the provided accident report records (e.g., carrier name, address). In addition, when information was missing, the team used the domicile of the carrier to identify contact information. When this was available, a team member attempted to contact the carrier and fill in the missing fields.

Medina, Trimble, Clarke 5 RESULTS This section of the paper describes the most relevant attributes of the motor carriers that were involved in SUT and CUT fatal crashes in Virginia between 2005 and 2009. The full report (5) contains a more extensive analysis and set of comparisons of SUT and CUT fatal crashes. The vast majority of SUT fatal crashes in Virginia involved carriers domiciled in the state. As can be seen in Figure 2, this pattern is distinctly different than the pattern for fatal crashes involving. VA- s Non-VA- s 10 9 85% 8 Percenet of Fatal Crashes 7 6 5 4 2 15% 41% 59% FIGURE 2 Comparison of SUT and CUT fatal crash patterns in Virginia by carrier domicile, 2005 2009. The overwhelming majority of SUT fatal crashes in Virginia occurred on non-interstate roads (Figure 3). These are typically U.S. and state routes and other secondary roads, many with undivided traffic-ways, unsignalized intersections, driveways and other entrances, left turns across traffic, and vehicle speed differentials all of which create opportunities for conflicts with other vehicles. The pattern is even more pronounced for Virginia-domiciled carriers, both those operating as well as, as can be seen in Table 1.

Medina, Trimble, Clarke 6 Crash Occurred on Interstate Highways Crash Occurred on Non-Interstate Highways 10 9 85% 8 Pecent o Fatal Crashes 7 6 5 4 59% 41% 2 15% FIGURE 3 Comparison of SUT and CUT fatal crash patterns in Virginia by type of highway, 2005 2009. TABLE 1 Percent of SUT and CUT Fatal Crashes in Virginia that Occurred on Interstate versus Non- Interstate Highways, 2005 2009 Crash Location Crash Occurred on Interstate Highways Crash Occurred on Non-Interstate Highways Percent of SUT Crashes VA- Non-VA- Total SUT Percent of CUT Crashes VA- Non-VA- Total CUT 12% 29% 15% 26% 51% 41% 88% 71% 85% 74% 49% 59% Total 10 10 10 10 10 10 The types of collisions in which were involved varied somewhat from those involving, as can be seen in Figure 4. experienced proportionally more fatal collisions with other motor vehicles, principally because their crashes occurred much more often on non-interstate roadways where conflict opportunities with other vehicles are much higher. This pattern is slightly more pronounced for Virginia-domiciled carriers, as can be seen in Table 2.

Medina, Trimble, Clarke 7 Percent of Fatal Crashes 10 9 8 7 6 5 4 2 5% 13% 71% 65% Non-Collision Collision w/ Other Motor Vehicle 16% 17% Collision w/ Fixed Object FIGURE 4 Comparison of SUT and CUT fatal crash patterns in Virginia by type of collision, 2005 2009. 6% 5% Collision with Object Not Fixed 2% Other TABLE 2 Types of Collisions in SUT and CUT Fatal Crashes on All Road Types, by State of Domicile, 2005 2009 Type of Collision VA- Non-Collision 3% Collision w/ Other MV Percent of SUT Crashes Non-VA- Total SUT VA- Percent of CUT Crashes Non-VA- Total CUT 11% 5% 14% 13% 73% 63% 71% 66% 64% 65% Collision w/ Fixed Object Collision with Object Not Fixed 15% 2 16% 17% 17% 17% 7% 3% 6% 7% 4% 5% Other 2% 3% 2% Total 10 10 10 10 10 10 The distribution pattern of the size of fleets operating involved in fatal crashes was skewed proportionally more toward smaller size fleets compared with the pattern for fleets

Medina, Trimble, Clarke 8 operating involved in fatal crashes, as can be seen in Figure 5. The patterns for Virginiadomiciled carriers, both those operating and, were, again, more pronounced in this regard, as can be seen in Table 3. Number of Trucks Operated 1-5 6-10 11-50 51-100 101-500 501-1000 >1000 Percent of Fatal Crashes 35% 25% 2 15% 5% 25% 18% 11% 6% 7% 3% 23% 23% 18% 13% 9% 7% 7% FIGURE 5 Comparison of SUT and CUT fatal crash patterns in Virginia by fleet size, 2005 2009. TABLE 3 Fleet Sizes of s Involved in SUT and CUT Fatal Crashes on All Road Types by Domicile, 2005 2009 Fleet Size Percent of SUT Crashes VA- Non-VA- Total SUT Percent of CUT Crashes VA- Non-VA- Total CUT 1-5 32% 28% 36% 17% 23% 6-10 5% 6% 6% 9% 6% 7% 11-50 32% 9% 25% 25% 14% 18% 51-100 7% 6% 7% 11% 9% 9% 101-500 16% 22% 18% 9% 15% 13% 501-1000 1% 6% 3% 3% 9% 7% >1000 7% 22% 11% 8% 23% Total 10 10 10 10 10 10 Median 17 119 15 155 SUT fatal crashes generally occurred much closer to the business location of the carriers controlling the operation of the truck, compared with the pattern for CUT fatal crashes, as can be seen in Figure 6. Again, this pattern is more pronounced for Virginia-domiciled carriers operating both and that were involved in fatal crashes, as can be seen in Table 4.

Medina, Trimble, Clarke 9 Distance in Miles 1-50 51-100 101-250 251-500 501-750 >750 7 6 6 Percet of Fatal Crashes 5 4 2 22% 19% 12% 14% 16% 17% 12% 13% 8% 4% 2% FIGURE 6 Comparison of SUT and CUT fatal crash patterns in Virginia by distance from carriers business location to crash site, 2005 2009. TABLE 4 Distances from s Business Locations to Crash Site for s Involved in SUT and CUT Fatal Crashes on All Road Types by Domicile, 2005 2009 Distance from Business Location to Crash Site (Miles) VA- Percent of SUT Crashes Non-VA- Total SUT VA- Percent of CUT Crashes Non-VA- Total CUT 1-50 74% 12% 6 51% 2% 19% 51-100 14% 6% 12% 22% 7% 12% 101-250 11% 26% 14% 23% 13% 16% 251-500 2% 29% 8% 2% 25% 17% 501-750 18% 4% 1% 2 13% >750 9% 2% 34% 22% Total 10 10 10 10 10 10 Median 21 281 47 536 s operating involved in fatal crashes were much more likely to be private or other type carriers, compared with carriers operating involved in fatal crashes, the majority of which were for-hire carriers (Figure 7). As before, this pattern is more pronounced for Virginia-domiciled carriers operating both SUTS and involved in fatal crashes (Table 5).

Medina, Trimble, Clarke 10 For-Hire Private Exempt Other 10 Percent of Fatal Crashes 9 8 7 6 5 4 2 44% 46% 1% 82% 5% 3% FIGURE 7 Comparison of SUT and CUT fatal crash patterns in Virginia by type of carrier, 2005 2009. TABLE 5 Types of Operations of s Involved in SUT and CUT Fatal Crashes on All Road Types by Domicile, 2005 2009 Type of Operation Percent of SUT Crashes VA- Non-VA- Total SUT Percent of CUT Crashes VA- Non-VA- Total CUT For-Hire 39% 55% 44% 64% 92% 82% Private 47% 42% 46% 17% 6% Exempt 1% 1% 11% 3% 5% Other 13% 3% 8% 3% Total 10 10 10 10 10 10 When looking at the types of commodities hauled by SUT and CUT carriers involved in fatal crashes, the differences in business operations between the two carrier types become apparent. in general haul a wide variety of other freight types and much less general freight compared to (Figure 8). This other category includes: household goods, metal/sheets/coils/rolls, motor vehicles, drive away/tow away, building materials, mobile homes, machinery/large objects, intermodal containers, oil field equipment, coal/coke, garbage/refuse/trash, U.S. Mail, refrigerated foods, beverages, paper produces, utilities, construction, water well equipment, and all other. This pattern is more pronounced for Virginiadomiciled carriers operating both SUTS and involved in fatal crashes (Table 6).

Medina, Trimble, Clarke 11 Percent of Fatal Crashes 10 9 8 7 6 5 4 2 General Freight Logs Gas/Liquids Agriculture and Related All Other 26% 3% 63% FIGURE 8 Comparison of SUT and CUT fatal crash patterns in Virginia by type of commodity hauled, 2005-2009. 61% 8% 4% 4% 3% 7% 21% TABLE 6 Types of Commodities Hauled by s Involved in SUT and CUT Fatal Crashes on All Road Types by Domicile, 2005 2009 Type of Commodity Percent of SUT Crashes Non-VA- VA- Total SUT VA- Percent of CUT Crashes Non-VA- Total CUT General Freight 18% 45% 26% 39% 73% 61% Logs 4% 3% 19% 3% 8% Gas/Liquids 3% 6% 4% 3% 3% 3% Agriculture and Related 3% 6% 4% 15% 3% 7% All Other 73% 42% 63% 25% 19% 21% Total 10 10 10 10 10 10 CONCLUSIONS The present study highlights the fact that the majority of SUT fatal crashes happened to Virginiadomiciled carriers on non-interstate highways and very close to the carriers place of business. These patterns are very different compared with patterns for. Truck travel is projected to double, again, in the next 15 to 20 years (6), notwithstanding our current economic situation. While a significant portion of that growth will occur on interstates, an even larger portion will occur on non-interstate highways, the principal operating domain of. Projected growth in truck travel makes it very important that detailed, requirementsbased, state-level plans be developed to foster commerce and economic growth by facilitating truck freight flow. These plans should be an integral part of each state s Highway Safety Improvement Plan. They need to include not only roadway design and traffic operations upgrades in areas where truck safety hotspots and bottlenecks are identified but also improvements resulting from a systematic approach to greater highway safety. Highway design

Medina, Trimble, Clarke 12 and traffic operations engineering efforts in this regard should complement and be fully integrated with enforcement and education efforts in these same areas and corridors. Traditional motor carrier crash prevention efforts focus heavily on truck driver licensing and enforcement programs that target carriers and drivers who are found to be in noncompliance based on data and findings from roadside inspections, carriers crash histories, and drivers traffic records. The majority of detailed inspections take place on interstate highways and are principally focused on. This likely results in more inspections of large, out-of-state carriers operating than in-state carriers operating. Efforts to help in-state-domiciled SUT operators improve their safety performance are, therefore, complicated and may need to be different from those currently employed. Because small fleets operate relatively few trucks, it is much less likely that small carriers will be inspected or audited. This is borne out by the recent U.S. Government Accountability Office study (7). This problem is compounded in the non-interstate highway environment where, for many reasons, it is unlikely that traditional in-depth driver/vehicle inspections can be conducted in sufficient numbers to these operators, making the use of this traditional approach problematic. The non-interstate portions of the states highway systems are both extensive and widely dispersed, making it challenging to develop safety countermeasures for travel on these road types. A multidisciplinary approach, particularly for, is needed. Coincidently, a great deal of national-level attention is now being focused on new entrants into the motor carrier industry, many of which are small carriers (8). Training and information exchange are likely to be key elements in efforts to improve small carrier safety and constitute a proactive approach that can be employed to help new entrants and small carriers. Finally, for all the reasons cited above, the results of this study indicate that the safety trends and patterns of and ought to be considered separately, rather than in the aggregate as large trucks. REFERENCES 1. FARS Data Query System [NHTSA Data file]. National Highway Transportation Safety Administration. http://www.fars.nhtsa.gov. Accessed March 2011. 2. Traffic Safety Facts, 2005-2009. National Highway Transportation Safety Administration. Washington, D.C., 2006, 2007, 2008, 2009, 2011. 3. American Transportation Research Institute. Large Trucks Safety Trends. May 2013. 4. Report to Congress on the Large Truck Crash causation study. Federal Motor Safety Administration, U.S. Department of Transportation. Retrieved from http://www.fmcsa.dot.gov/facts-research/research-technology/report/ltccs-2006.pdf, 2011. 5. Medina, A., T. Trimble, and R. Clarke. Phase II Report: Linking Descriptive Attributes to Crash Patterns, An Untapped Tool in State Motor Safety Improvement Programs - Analysis of Single Unit Truck Fatal Crashes. Virginia Tech Transportation Institute, 2012.

Medina, Trimble, Clarke 13 6. 2008 Status of the Nation s Highways, Bridges, and Transit: Conditions & Performance. Report to Congress. Federal Highway Administration and Federal Transit Administration, 2010. Retrieved from http://ntl.bts.gov/lib/33000/33400/33453/2008statussummary.pdf. 7. Motor carrier safety: More assessment and transparency could enhance benefits of new oversight program. U.S. Government Accountability Office, 2010. Retrieved from http://www.gao.gov/new.items/d11858.pdf, 2011. 8. Changes to the New Entrant Safety Assurance Program. U.S. Federal Motor Safety Administration. Department of Transportation. Retrieved from, http://www.fmcsa.dot.gov/documents/changes_to_nesa_012510_508.pdf.