Crash rates

The incidence of aircraft impacts may be significantly higher in certain areas (e g., in the vicinity or airports). The aircraft crash hazard is site specific and tlie failure is strongly dependent on the kinetic energy of tlie aircraft. Two types of data are needed to analyze for aircraft impact the aircraft crash rate in tlie site vicinity (per unit area per year) and tlie effective target area of tlie vulnerable item. Crash rates for different categories of aircraft can be obtained from state and national autliorities (e.g., FAA). The proximity of the site to airfields must be taken into account because crashes are much more frequent witliin a radius of approximately 3 miles. 

All three wedge barrier designs have a high crash rating, thereby allowing them to be employed for higher-security applications. ... 

No excavation is required in the roadway itself to install crash beams. All three crash beam designs possess a low crash rating relative to other types of barriers, such as wedge barriers, and thus they typically are used for lower-security applications. 

Bollards have an intermediate crash rating, which allows them to be utilized for medium-to higher-security applications. Some maintenance is needed to ensure barrier is free to move up and down. 

The difference between the crash rates per mile in the first two reports is due to the use of a bounding rate (using general data) in the first report and a carrier-specific rate in the second report. The third report derived a truck crash rate per billion ton-miles, based on Federal Motor Carrier Safety Administration data for the number of commercial vehicle hazmat crashes divided by the number of commercial vehicle ton-miles. The commercial vehicle accident rate per ton-mile value was multiplied first by 0.59 and then by 0.013 to account for the number of ton-miles for for-hire commercial vehicles and for Class 6 poisons, respectively. These adjustments to the accident rate are inappropriate without consideration of the corresponding hazmat crashes, for which no data were used. The total accident scenario probabilities may be conservative (too high), however, when the probabilities of fire occurring, impact forces sufficiently high, etc. are considered. These probabilities do not appear to have been addressed, and the committee did not review the reports at that level of detail. 

The intervention process is triggered by one or more deficient I BASICS, a high crash rate, a complaint, or fatal crash. 

Has a crash rate in the top 30 percent of the national average as indicated by FMCSA s Motor Carrier Management Information System (MCMIS) or... 

SMS). The SMS tracks violations (discovered at the roadside or during interventions) and crashes, both of which represent a failure of the carrier s safety management controls. Carriers that have a significantly higher violation and/or crash rate than their peers (as represented by their high scores in the SMS) are the carriers that will be subjected to an inter-I vention under the CSA program. 

The FMCSA tracks carrier on-road performance through a component of the Comphance, Safety, Accountability (CSA) program called the Safety Measurement System, or SMS. If the SMS shows that the carrier s performance was worse than their peers (as evidenced by higher violation or crash rates), the carrier could be selected for a compliance review. In fact, 48 percent of the carriers selected for compliance reviews during 2009 were because of CSA s prodecessor, SafeStat (as of December 2010 CSA s SMS is being used to select carriers for further enforcement action). 

Verifiable crash rate (using verified data fi om the Crash BASIC). 

At this time, FMCSA has not defined how a verifiable crash rate will be calculated. However, the discussion has centered on using a process of verifying the carrier s crash rate per... 

A carrier s SMS scores and its verifiable crash rate (i.e., factors 1 and 2 above), would be applied to a set of Unfit thresholds to determine a BASIC failm e. These thresholds would be based on the carrier s absolute BASIC Measures and crash rate, as opposed to the relative percentile rankings fi-om the SMS. 

Number of Stand Alone BASICS 1) With SMS measure above Unfit threshold or 2) Where essential safety management violations are 10 percent or more of records checked Number of Non-Stand Alone BASICS 1) With SMS measure or verifiable crash rate above Unfit threshold or 2) Where essential safety management violations are 10 percent or more of records checked Fundamental violations discovered SFD Continue Operation, Marginal, or Unfit... 

Driver Fitness, Drugs/Alcohol, Cargo Securement, Vehicle Maintenance, and Verifiable Crash Rate... 

The rationale for this distinction between stand alone and non-stand alone BASICs is that, although each of the BASICs applies to both carriers and drivers compliance and safety, the stand alone BASICs are more directly related to driver behavior that contributes to crashes. In particular, an effectiveness study on the SMS has shown that carriers with past poor performance in the Unsafe Driving or Fatigue Driving/HOS Compliance BASICs were subsequently involved in crashes at a considerably higher rate than the overall crash rate of the motor carrier population. 

Predicting which driver will be involved in a future crash has always been one of the more difficult tasks of the safety pro. The American Transportation Research Institute (ATRl) may have provided some help in this matter. In 2005 ATRl released the results of a study entitled Predicting Truck Crash Involvement Developing a Driver Behavior-Based Model and Recommended Countermeasures. ATRl provides insight into which driving behaviors tend to lead to an increased chance of future crash involvement. The study also determined what companies with low crash rates are doing to prevent behaviors that lead to crashes. 

The other portion of the study that pertains to the safety pro was a survey of carriers that had low crash rates. These carriers maintain low crash rates because they are successfid at intervening and changing driver behaviors before crashes occur. The carriers involved in the study had these traits in common ... 

Another example is the standard on classification of motor vehicle traffic accidents published by NSC. It provides a uniform method for compiling data on highway accidents. The Federal Highway Administration (FHWA) provides formulas for calculating roadway departure crashes from crash rate data. The method helps highway planners select the types and locations for barriers and guardrails and implement other safety improvements. 

A crash rate below the threshold for the current registration cycle compared to the national average. 

Keall, M.F. and Frith, W.J. 2004. Older driver crash rates in relation to type and quantity of travel. Traffic Injury Prevention, 5, 26-36. 

In passing, it may be noticed that the reduction in crash rate was considerably greater than the reduction in the number of young drivers who obtained high-school driver education. Once again, one wonders if this type of education lulls its graduates into an illusion of safety and thus leads them to behave less cautiously (see Section 5.3 in Chapter 5). 

In contrast to the interventions that fail to have the intended effect on crash rates, or those that even backfire, incentive programmes for safety have the effect for which they are intended a more general motivation towards greater safety that is sometimes known as safety culture (Geller, 1997). 

A number of studies have investigated the relationship between absolute driving speed and crash rate (Aarts and van Schagen, 2006). Irrespective of the research method used, almost all studies find that the relationship between speed and crash rate is best expressed as a power function the risk of a crash increases exponentially when speed increases and vice versa (see Figure 22.3). 

FIGU RE 22.3 Diagram of the power function of the relationship between speed and crash rate. 

Casualties = Exposure x Crash rate x Injury rate... 

Crash data indicate that the risk of a crash varies by a factor of over 100 among different countries, and within a given country the rate diminishes greatly when various improvements are made to the infrastructure. If drivers were to adjust for these differences then the crash rates would be similar in all countries and would remain the same over time (Evans, 2004). 

Although monocular vision does not preclude driving in general, many countries restrict commercial driving to people with binocular vision (e.g., Australia, see Horton and Chakman, 2002 USA, see FMCSA, 2001). Critical reviews of research tiiat compared the crash rates of monocular drivers with that of binocular drivers have for the most part concluded that monocular drivers are no worse than binocular drivers (Bartow, 1982 North, 1985 Owsley and McGwin, 1999). In one study that was conducted on California heavy vehicle drivers, Rogers et aL (1987) did find that monocular drivers had more crashes than binocular drivers, but the latter tended to under-report their crashes. This anomaly was due to the fact that monocular California drivers did not drive outside of California, because they did not comply with the Federal vision requirement of binocular vision. The binocular drivers did drive outside of California, but their California license records did not include tiieir out-of-state crashes. 

The elusiveness of such relationships and explanations for our inability to find them was offered in a review of the state of the art in this area by Westlake (2000). It is difficult to establish the relation between visual impairment and crash rates because visually impaired drivers tend to restrict their driving habits and change their behaviour to compensate for their visual loss. Crashes are fortunately rare events with multiple causes, and the effects of a driver s visual impairment are dwarfed by other factors such as the annual mileage driven, the driver s age, inattention, intoxication, and speeding. Furthermore, it is unsurprising that it is difficult to predict crash rates from measures of static visual acuity and the peripheral visual... 

One other significant bit of information that should be noted for the data in Figure 6-6 is that they represent people who reportedly drove approximately 7500 miles/year. This is important because the relationship between the amount of driving and the crash involvement is not a linear one. Drivers who drive very few miles per year have a much greater crash rate per mile, than drivers who drive many miles each year (Forsyth et al. 1995 Hakamies-Blomqvist et al. 2002). One explanation for this is that drivers who drive many miles, drive more of them on inter-urban and limited-access divided highways (freeways, as they are known in the U.S., and... 

---