Warranty data and other time-to-failure data is frequently analyzed through the estimation of the parameters of one or more lifetime distributions including: two-parameter Weibull, three-parameter Weibull, Lognormal, Logistic, Gumbel, Gamma. The form of some lifetime distributions is derived from the nature of the failure process, e.g., the Gumbel is used to model the maximum of a number of samples from the same distribution. Other lifetime distributions, e.g., Weibull can be used to fit a wide variety of failure mechanisms. When the form of the lifetime distribution is not specified by the failure mechanism, and there is no historical precedent demonstrating that a particular family of distributions is a good fit for the data in question, goodness of fit methods can be used to choose the form of the lifetime distribution that best fits the available data. The lifetime distribution chosen may have a large impact on the inferences drawn from the data, particularly when the subject of interest is future reliability or reliability projected beyond the range of the currently available data. Maximum Likelihood is one of the most popular distribution selection methods. The robustness of Maximum Likelihood for choosing the ‘best’ lifetime distribution, and the impact of choosing the wrong distribution was evaluated by using Monte Carlo simulation to generate samples from four specified lifetime distributions. The Reliasoft® Weibull++ software and SAS were used to estimate the best fitting distribution and to estimate the B50, age at 50% failure, B90, the age at 90% failure. Results of the research include the proportion of Monte Carlo runs that the specified lifetime distribution was correctly chosen as the best fitting distribution, and comparison of the accuracy of estimates of B50 and B90 when calculated from the best fit distribution as compared to the B50 and B90 when calculated with estimated parameters for the Actual distribution. In this study, the Lognormal distribution was usually correctly selected while the Weibull and the Logistic distribution were frequently misidentified as another lifetime distribution. Suggestions for addressing the practical implications of misidentified lifetime distributions are discussed.

Warranty data with an implicit censoring at the end of the warranty period is often used in combination with commercially available "Weibull Analysis" software to estimate the rate of product failure after the end of the warranty period. Warranty databases frequently contain some "goodwill" or other special type of claims that are made after the end of the warranty period. Inclusion of such after warranty claims with the usual commercial software violates the assumption of independence of the time-to-failure process and the censoring process. This independence assumption is built into the algorithms used by Reliasoft Weibull++, Win SMITH, and SAS. The research reported here uses Monte Carlo simulation to investigate the robustness (i.e. whether or not the accuracy and precision of the estimation algorithms are maintained) when the independence assumption is violated. The proportion of after warranty claims which are included in the analysis, while censoring non-failed product at the end of the warranty, and the distance between the end of the warranty and the product age for the relevant projection for the failure rate will determine the magnitude of the bias of the projected failure rates. The magnitude of the bias is characterized here as the percent of projection estimates that are at or above the actual failure rate at the relevant age. A method for including after warranty claims along with other information about after warranty failures, e.g. part sales, is proposed and found to be less biased than simply including after warranty claims with no modeling of the censoring process that produced the after warranty claims.

To prevent wheel lock up (and possible loss of control and capsize) during hard braking motorcycle manufacturers have equipped motorcycles with Antilock Brake Systems (ABS) either as an option or as standard equipment. Several studies utilizing real-world crash data have been published which estimate the effectiveness of motorcycle ABS in reducing the risk of a crash based on varying assumptions. These investigations have reported mixed results. The present investigation relies upon the Fatality Analysis Reporting System (FARS) and the Florida police-reported crash databases to further investigate the effectiveness of motorcycle ABS by expanding upon and refining previous approaches. Notably a case-control approach is used whereby crashes involving ABS- and non-ABS-equipped motorcycles are divided into five groups with a varying likelihood that ABS will affect the risk of crashes in that group. The group of crashes with the least likelihood of being influenced by ABS is considered the control group and used as a measure of exposure to crashes. This methodology attempts to reduce any selection biases that might exist in the two motorcycle classes. The results support the hypothesis that ABS is effective in reducing the crash risk in some crash types. However, it was found that the case-control approach does not incorporate all factors that might influence the overall effectiveness of ABS, for example, motorcycle class and operator age. Accounting for these additional factors would likely require the use of regression analyses and would benefit significantly from additional data.

This paper explores tire placement with given tread depths on vehicles from two distinct perspectives. The first area explored is an analysis of crash data recently reported by the National Highway Traffic Safety Administration (NHTSA). In this report, thousands of tire-related crashes were investigated where the tread depth and inflation pressure were logged for each tire and assessments were made as to whether tire condition was a factor in the crash. The analysis of the data shows that in regards to accident causation, it is not statistically significant which axle has the deepest tread. What is significant is that a tread depth at or below 4/32. anywhere on the vehicle leads to an increased rate of crashes. To understand the physics implied by the NHTSA data, a study was performed on how the placement of tires of various tread depths affects the steering, handling, and braking performance of a modern sport utility vehicle. The test vehicle was instrumented with on board video equipment and a computer with transducers to measure driver inputs as well as vehicle responses during the testing. The vehicle was tested on a uniform wet test surface at a test track specifically designed for this purpose. Specific repeatable tests were performed to study the wet surface steady-state and transient handling performance as well as the straight line stopping distance during limit braking. These tests included an SAE J266 one hundred foot circle test, a closed loop single lane change, a slowly increasing steer test, and a limit ABS brake stop. These tests were each performed three times on the vehicle with tires of various makes and manufacturers and different levels of real world customer wear. A pair of "shaved" tires was also evaluated. The results show that placement of tires has a definite effect on the vehicle dynamic performance of the utility vehicle tested and that deeper tread depth on the rear suspension is not always the best configuration for overall vehicle performance.

Police accident reports (PARs) of motor vehicle collisions typically include information regarding occupant restraint use. It has been suggested that PARs overestimate restraint use. Previous studies comparing PAR restraint usage with that determined during a NASS/CDS in-depth investigation found agreement in approximately 90% of cases. The accuracy of PAR-reported restraint usage for outboard vehicle occupants was compared to that determined by NASS/CDS investigators as a function of injury severity and crash type. Restrained occupants were more likely to be identified correctly in the PAR, and unrestrained occupants were more likely to be accurately identified as injury severity increased. Differences in the accuracy of PAR-reported restraint usage rates for different crash types were small.

Fires originating in large trucks can be significant in terms of both the potential for personal injury or death and the potential for substantial economic loss of the vehicle and its cargo. This analysis examines the large trucks involved in fire incidents and the causes of the fires by examining the National Automotive Sampling System (NASS/GES), the Large Truck Crash Causation Study (LTCCS), and the Fatality Analysis Reporting System (FARS).In this report we compare the rate of post-collision fire observed in these databases, analyze the reasons for differences in and describe the circumstances of large truck fires as reported in the LTCCS.

Inadvertent vehicle movement incidents, in which a vehicle rolls away after the driver has exited, may occur in automatic transmission vehicles as a result of environmental, vehicular, and/or driver factors. Some explanations have focused on claimed potential malfunctions or design flaws in the vehicle's console shift mechanism or in the automatic transmission itself. However, growing evidence suggests that driver errors unrelated to vehicle design may in fact be the primary cause of many inadvertent vehicle movement incidents. The present research extends previous work on driver gear-shifting behaviors and vehicle egress by conducting more in-depth analyses of data collected by Harley et al. (2008). First, timing characteristics of shifts into Park measured under hurried conditions were not affected by the presence of driver distraction, further evidence that gear shifting is not a visually-guided process, but rather a ballistic movement that is preplanned and, once initiated, cannot be modified or terminated until the planned action has been completed. Second, a noteworthy mode of vehicle exit reported by Harley et al. (2008) wherein the driver's foot remains on the brake during egress, was found to be associated with rapid vehicle exit times. Such egress patterns may combine with other driver errors (e.g., failure to shift into Park, failure to set the parking brake, and failure to remove the key from the ignition before exiting the vehicle) to allow inadvertent vehicle movements in which the driver is unable to safely regain control of the vehicle.

Abstract Comparative risk assessment is an analytic process of evaluating and ranking the different factors that contribute to a particular outcome, such as human disease. The term comparative risk assessment is most frequently used in the context of environmental risk, but the principles and methodology are applicable in other contexts. Comparative risk assessment is both a scientific and political process. Scientific methods are used to gather data, classify the type and severity of risks, and identify and verify relationships between outcomes and risk factors. As it is a guide to decision making, comparative risk assessment also involves the political functions of defining the scope of the assessment, involving stakeholders in the process, and putting the results into a context that can be used to develop consensus concerning appropriate action. This article reviews the history of comparative risk assessment for environmental risk, discusses methods of risk ranking used in environmental risk, reviews methods of communicating risk, considers a few examples of comparative risk assessment in nonenvironmental areas, and touches on United States Environmental Protection Agency (USEPA) and other regulations as they relate to comparative risk assessment.

Weibull analysis is a powerful predictive tool for studying failure trends of engineering systems. [1] One noted shortcoming is that traditional techniques require the size of the susceptible population to be known. The method described in this paper allows for estimation of the size of the susceptible population using only failure data and no assumptions about total population size or susceptible portion.In the analysis of failures of mass-produced products, a large amount of failure data may be available, but all the conditions that define the susceptible population may never be known. For example, units with a particular usage condition may be expected to fail over time following a Weibull model, but the number of units subjected to that usage condition may never be known. To assume that the entire population is susceptible to the failure mode would greatly over-predict future failures, and the model could not be used to guide decision-making.By doing a least squares fit to the trend of failures versus time, a Weibull model can be fit to the data and then used to estimate the total number of susceptible units expected in the population. The ability to accurately estimate the size of the susceptible sub-population from failure data will be explored as a function of the size of the data set used, for known sets of failure data. For example, for a failure distribution that has increased, peaked, and then decreased to zero, almost the entire population has failed, so an estimate of the size of the susceptible population from this data is likely to be accurate. On the contrary, for only a few data points that show an increasing failure rate over time, little can be determined. Monte Carlo simulations will be used in order to estimate the error associated with this technique.Our analysis will show that predictions of total susceptible populations become similar to the actual susceptible populations when the predicted mean time to failure (MTTF) from the observed data is shorter than the observation time. In effect, predictions become accurate when it is clear to the observer that the number of failures per unit time has peaked.

Abstract Accelerated testing is a set of methods that attempts to replicate and predict the quantitative effect on product life of what are assumed to be the most important degradation mechanisms determining product life in actual service. In all accelerated testing approaches, one or more physical degradation methods are applied to promote failure. The history and limitations of accelerated testing and mathematical tools for modeling product lifetime as a function of acceleration mechanisms are discussed.

The potential human health risks associated with consuming fish containing hazardous substances are related to the frequency, duration, and magnitude of exposure. Because these risk factors are often site specific, they require site-specific data. In anticipation of performing a risk assessment of the lower 6 miles of the Passaic River in New Jersey (Study Area), a year-long creel/angler survey collected such site-specific data. The lower Passaic River is urbanized and industrialized, and its site conditions present unique survey design and sampling challenges. For example, the combined population of the municipalities surrounding the Study Area is nearly 330,000, but because the Study Area is tidal, state law does not require fishing licenses for anglers to fish or crab in the Study Area. The sampling challenges posed by the lack of licensing are exacerbated by the industrialization and lack of public access in the lower half of the Study Area. This article presents a survey methodology designed to overcome these challenges to provide data for accurately estimating the Study Area's angling population and the fish and crabs they catch, keep, and eat. In addition to addressing the challenges posed by an urban and industrial setting, the survey methodology also addresses the issues of coverage, avidity, and deterrence, issues necessary for collecting a representative sample of the Study Area's anglers. This article is a companion to two other articles. The first companion article describes the analytical methodology designed to process the data collected during the survey. The second presents, validates, and interprets the survey results relating to human exposure factors for the lower Passaic River.

This article describes a unique analytical method employed to characterize angler activities on the lower 6-mile stretch of the Passaic River in New Jersey. The method used data collected by a creel/angler survey that was designed to capture the information necessary to calculate the exposure factors needed to characterize the fish consumption pathway for recreational anglers in a human health risk assessment for the river. The survey used two methods to address the challenges of conducting a creel/angler survey in an urban and industrial setting with limited river access. While unique, the analytical method described in this article is based upon accepted methods of interpreting survey data and basic laws of probability. This article was written as a companion to two other articles, also in this issue and cited here, of which one describes in detail the survey methodology designed for the lower Passaic River creel/angler survey to meet various challenges unique to conducting such a survey in urban and industrialized rivers, and the other presents, validates, and interprets the results of the lower Passaic River work relating to human exposure factors using the methodology described in this article.

The results of an analysis of site-specific creel and angler information collected for the lower 6 miles of the Passaic River in Newark, NJ (Study Area), demonstrate that performing a site-specific creel/angler survey was essential to capture the unique characteristics of the anglers using the Study Area. The results presented were developed using a unique methodology for calculating site-specific, human exposure estimates from data collected in this unique urban/industrial setting. The site-specific human exposure factors calculated and presented include (1) size of angler population and fish-consuming population, (2) annual fish consumption rate, (3) duration of anglers' fishing careers, (4) cooking methods for the fish consumed, and (5) demographic information. Sensitivity and validation analyses were performed, and results were found to be useful for performing a site-specific, human health risk assessment. It was also concluded that site-specific exposure factor values are preferable to less representative “default values.” The results of the analysis showed that the size of the angling population at the Study Area is estimated to range from 154 to 385 anglers, based on different methods of matching intercepts with anglers. Thirty-four anglers were estimated to have consumed fish; 37 people consumed fish from the river. The fish consumption rate for anglers using this area was best represented as 0.42 g/day for the central tendency and 1.8 g/day for the 95th percentile estimates. Anglers fishing at the river have relatively short fishing careers with a median of 0.9 yr, an average of 1.5 yr, and a 95th percentile of 4.8 yr. Consuming anglers tend to fry the fish they caught. The demographics of anglers who consume fish do not appear to differ substantially from those who do not, with no indication of a subsistence angling population.

Both state police and the National Automotive Sampling System/Crashworthiness Data System (NASS/CDS) keep automotive collision statistics to varying levels of detail. Some of these details (e.g., collision relative velocity and driver height and weight) are reported in the NASS/CDS, but not in state databases. This article explores whether these details are confounding factors that would bias conclusions based on analysis of state data. To determine this, a methodology was created to predict overall risk based only on the distribution of possible confounding factors experienced by each manufacturer. Relative impact velocity (Delta-V) and driver height and weight are found not to be true confounding factors. Although the distribution of possible confounding factors varies somewhat among manufacturers, the impact on overall expected risk is minimal, and therefore evaluation of risk based on datasets that do not contain information for Delta-V and driver height and weight appears appropriate. The accuracy of match between police reporting and NASS/CDS was also explored and found consistent across manufacturers. Therefore, inaccuracies between police reports and NASS/CDS should not bias comparisons of risk between manufacturers.