How to put the brakes on truck crashes
Drivers of large trucks and other vehicles involved in truck crashes are 10 times more likely to be the cause of the crash than other factors, such as weather, road conditions, and vehicle performance, according to a study released in March 2006 by the Federal Motor Carrier Safety Administration (FMCSA). This finding underscores the importance of hiring good drivers with nearly clean records. However, as important as it is to start off with a good driver, trucking companies also need to have an established, ongoing safety program to keep driver standards high and crash rates low.
The relationship between lower crash rates and program practices, selection criteria and other management issues with trucking are quantified annually through Liberty Mutual’s truckers’ survey. And every year for the past five years, the survey findings have shown that the trucking companies with better than average crash rates all adhere to a similar safety program structure.
The common characteristics these companies share are: high driver selection standards; driver performance expectations that are communicated to drivers; driver performance monitoring; actual performance versus expected performance comparison; and programs to help change driver performance that does not meet expectations.
The benefits of these characteristics, such as high driver selection standards, include a 42 percent lower crash rate for companies with rollover alert systems in all or some of their trucks and a 32 percent lower crash rate for companies that establish routes and use technology to track off-route miles. However, keep in mind that without high selection standards, any in-cab technology is a waste of money.
According to the FMCSA, there are just over 3,600 authorized for-hire interstate truckers with 50 to 300 power units in the United States. In 2006, Liberty Mutual distributed its trucker survey to 243 of those trucking companies and received an 81 percent response rate making the survey group close to 7 percent of the total truckers in this size range.
The survey established operation benchmarks using information on programs, miles and operations supplied by the trucking companies, Department of Motor Vehicle record summaries, publicly available Safer and SafeStat data, and crash and injury data from Liberty Mutual’s claims database. From this information, the survey compared each trucker to the benchmark frequency for their type of operation and looked at performance and driving records as compared to each program element to determine the relationship between program elements and crash rates.
The survey results show that crash rates varied by type of operation. The “Crash Benchmark Comparison” chart was calculated using the total number of liability crashes with costs of $1,000 or more for the calendar year 2006 valued as of April 1, 2007.
Full-load van operations were the most common company in the survey, which lends its data the most credibility. The logistics and intermodal groups were the smallest groups, offering less credibility. Within each group there are different exposures based on routes, stops, areas of operation and geography. Since many companies have multiple operations, for this analysis, companies were placed into a group if 50 percent or more of their miles were from that type of operation.
Types of crashes
To accurately report crash rates, the survey examined data for both liability crashes and physical damage crashes. Focusing on only liability crashes would have resulted in a slight undercount of total crash rates since some single vehicle crashes incur no liability.
When it compared the crash data against each company’s safety practices, the survey identified 26 program practices that had relationships with lower crash rates.
Some of the identified program practices include the following:
Selection
Companies that conducted background checks for citizenship, past employment, credit, work history gaps and criminal history had a 23 percent lower crash frequency than those only conducting some of these background checks. The difference between these groups was just a slight difference in driver quality.
Survey results showed companies that hire drivers with six or less job changes over a 10-year period had a 14 percent lower crash frequency.
Hiring decisions made by team decision, rather than by only the safety director, had a 29 percent lower crash frequency with just slight differences in driver quality.
Management
Companies providing weather monitoring, weather re-routing, weather communication and paying for weather-related down time had an 11 percent lower crash frequency with equal driver quality. Only half the companies surveyed provide all four of these.
Those companies conducting driver surveys, which was only slightly more than half, had a 25 percent lower crash frequency rate with equal driver quality.
Providing FMCSA compliance training for managers and having FMCSA compliance expectations for managers, resulted in a 32 percent lower crash frequency.
Holding driver managers accountable for enforcing driver hours of service can result in a slightly lower crash rate, even for companies that have twice as many drivers with four or more violations/crashes.
Providing driver managers with available hours of service data from satellite systems can lead to a 12 percent lower crash rate with similar driver quality. Only half the companies surveyed employed this strategy.
Establishing routes and using technology to track off-route miles, showed a 32 percent lower crash rate with slight differences in driver quality.
Measurements
The minority of companies surveyed that measure crash rates, miles per gallon, off-route miles and operating costs per mile had a lower crash frequency rate and better driver quality.
Most companies have an incentive program that use multiple criteria. However, companies using an electric control module (ECM) data as part of their incentive program criteria had significantly lower crash rates with slightly lower driver quality.
Downloading ECM data also had a benefit with a 23 percent lower crash frequency with slightly lower driver quality. Half of the companies downloading ECM data do so each time the vehicle is serviced.
In-vehicle GPS resulted in a 13 percent lower crash frequency with slightly lower driver quality. The most common GPS items companies measured were speed and on-time departure.
Feedback
Slightly more than half of the companies surveyed have their name and contact information on the trucks, or use a toll free monitoring service, or both, which resulted in a 23 percent lower crash frequency.
Most companies conduct road observations. Those not conducting road observations had a lower crash frequency but this is attributed to the fact that these companies had far fewer drivers with serious violations or four or more violations/crashes to begin with. The small group of companies that use radar as part of their road observations had significantly lower crash rates but they also had slightly more drivers with clear driving records but more drivers with four or more violations/crashes.
Interventions
Multiple studies from the Liberty Mutual Research Institute for Safety have shown that having and training drivers on a mirror check stations can help reduce lane change type crashes. However, this is one area where driver quality really matters. Even though close to half of the companies responding to the survey have mirror check stations and have provided training during the past two years, this group had higher overall crash and lane change crash frequency because they had twice as many drivers with four or more violations/crashes during the past three years.
A small number of companies had rollover alert systems in all or some of their trucks. This group had a 42 percent lower crash rate but they also had noticeably better driver quality. And, again, while those companies that provide in-vehicle training for existing drivers had a lower crash rate, they also had higher driver quality.
Training
Three out of four firms provide in-vehicle training for new drivers.
Differences in drive quality also impacted the results between those providing post crash in-vehicle training and those not providing it.
Those using written training plans based on road tests, observations or reviews had a 21 percent lower crash rate. The group using these training plans had significantly better driver quality.
Some training programs worked so well that even companies with lower driver quality achieved good results.
Providing skid pan training for new drivers, existing drivers or drivers that have been in crashes, had a crash frequency rate half that of the group not providing skid pan training.
Using ECM or satellite system data to target drivers for additional training or attention lead to an 11 percent lower crash frequency.
Using driver training simulators had an 11 percent lower crash frequency.
Using satellite or ECM data to conduct driver log audits had a slightly lower crash rate.
Policies
Most companies have adopted a policy of six to eight seconds of following distance. Those using this criteria of following distance had a 22 percent lower “hit other in rear” crash frequency rate.
All but a few companies use governors. Companies setting the governors below 68 had a slightly lower crash frequency with equal driver quality. There were a few companies with governors set below 68 that also adjust governors based on performance. This group had significantly better results.
Almost all companies have written seat belt policies. An insignificant number have technology to prevent starting without an engaged seat belt. Most companies have seat belt enforcement activities; those with high contrast seat belts had significantly lower crash frequencies.
FMCSA Inspection Selection System
There is a strong relationship between ISS-D (Inspection Selection System) scores used by the FMCSA and crash rates for companies with 2.5 million miles or more. The only exceptions to this corollary were noted in the groups above and below the median ISS scores of 49 with a range from 24 to 99.
ISS-D scores vs. crash rates
How a company is run has significant impact on the crash rates regardless of the type of operation. Management must start with selecting drivers that have shown a respect for motor vehicle laws, which shows the company expects drivers to obey the law. Management must also have clearly defined expectations for driving and driver injury prevention. Companies must measure performance using technology and observations to compared performance to expectations and have programs that will change the performance or behaviors that contribute to injury and crashes.
Crash Benchmark Comparison ($1,000 crashes per million miles)
All Types 0.88
Tanker 0.90
Other 1.36
LTL 1.38
Logistics 1.62
Intermodal 1.09
Full Load Van 0.84
Flatbed 0.58
Crash frequency medians are for companies with more than 2.5 millions miles.
Losses for Jan. 1, 2006 to Dec. 31, 2006 valued April 10, 2007.
Liability crashes by type
Type % of Costs % of Frequency
Backing 7% 17%
Hit other in rear 33% 22%
Intersection 17% 3%
Lane change 21% 19%
Rollover 3% 1%
Stationary object 6% 16%
Turning 5% 10%
Totals 92% 87%
Physical damage losses by type
Type % of Cost % of Frequency
Animal collision 3% 9%
Fire 4% 2%
Hit other in rear 9% 10%
Intersection 3% 3%
Lane change 18% 18%
Low bridge 2% 3%
Rollover 48% 27%
Theft 6% 2%
Totals 91% 73%
No. of program practices $1,000 crash frequency per million miles
11 or less program practices 1.00
12 or more program practices 0.84
15 or more program practices 0.73
Median $1,000 crash rate per million miles
Median or below ISS-D score 0.62
Above Median ISS-D Score 0.88
Peter Van Dyne is the technical director of transportation for Liberty Mutual’s Business Market Unit.