Introduction: Accurate motorcyclist mileage estimates are important because self-evaluation of riding experience is related to riding behavior, the relationship of self-reported to actual or future mileage is necessary in targeting training and considering survey responses, and motorcycle crash statistics require accurate travel data. Method: This study collected real-world data from motorcyclists over the course of two months to two years per rider. This paper explores motorcyclists' self-reported annual riding mileage (obtained via pre-study surveys) and the actual amount of riding during the study (based upon odometer readings and GPS data). Results: Of the 91 riders who had been riding for at least a year before the study, significantly more (73%) rode less the following year than reported for the previous year. The recorded annualized mileage averaged 89% of the reported mileage from the previous year. Analyses based on estimated average annual mileage were similar to those using the previous year estimation, and the pattern held regardless of age group, motorcycle type, or gender. The exception was novice or returning riders, who tended to either significantly underestimate or increase actual mileage as they began (or continued) to ride. Conclusions: Motorcyclists' estimation of riding experience expressed as mileage may not be indicative of current or future mileage. Practical applications: Reliance on self-reported mileage during training to categorize groups, for interpretation of studies, or to develop motorcycle travel data and safety statistics may be unrealistic. Certainly any use of self-reported mileage should incorporate the concept that mileage overestimation seems likely. Because questions about previous year and average annual mileage may elicit similar responses, motorcyclist surveys should be constructed to prompt the most thoughtful responses in terms of mileage estimations. In general, reported mileage should not be relied upon as an accurate predictor of future actual mileage. (C) 2017 National Safety Council and Elsevier Ltd. All rights reserved.

This paper introduces the preliminary categorization of motorcycle crash data collected from 100 riders as they rode for a period between 2 months and 2 years. These riders resided in California, Florida, Virginia, and Arizona, and both video and motorcycle kinematic data were collected for every ride. The videos of incidents were reviewed, and the events were described with motorcycle-specific categories for event severity, event nature, incident type, precipitating event, rider reaction, and postmaneuver control. Within the data set of more than 38,000 trips, 22 incidents defined as crashes (involving 18 riders) were identified and categorized. The majority of the crashes (15) were single-vehicle, low-speed crashes. Of the remaining crashes, one was a single-vehicle (higher speed) crash, two were rear-end collisions, and four involved one vehicle turning into or across the path of another at an intersection. Rider response to the precipitating event for more than half of the crashes involved no visible front braking or lateral input. Most crashes occurred within 20 min of the beginning of a trip, during daylight, and in favorable weather conditions. There was no overriding hypothesis implying a relationship of crash occurrence to a specific demographic group in terms of location, age, gender, or motorcycle class. This research applied motorcycle incident categorization to 22 crashes. The categorization terminology and definitions are also applicable to near-crashes. This information is useful in providing an unbiased understanding of what occurs during such incidents and offering a basis for the structured cataloging of crashes, as well as future categorization of near-crashes and crash-relevant events.

Enhanced Night Visibility Series Phase III, Study 2 (rainy weather) was performed following the same procedures used for Phase III, Study 1 (clear weather). Study 2 served to expand the knowledge of how current vision enhancement systems can affect detection and recognition of different types of objects while driving during adverse weather, specifically during rainy conditions. The empirical testing for this study was performed on the Virginia Smart Road; the rain was controlled by weather-making equipment. Fifteen participants were involved in the study. A 4 by 8 by 3 mixed factorial design was used to investigate the effects of different types of vision enhancement systems, different types of objects on the roadway, and driver's age on detection and recognition distances; subjective evaluations also were obtained for the different vision enhancement systems. The results of the empirical testing suggest that well-designed infrared (IR) systems are consistently associated with often significantly longer detection distances for most types of pedestrian objects during rainy conditions. In particular, the use of the near IR (NIR) systems resulted in earlier detection of nearly all tested pedestrian types than did the use of either far IR (FIR) or baseline halogen (HLB) systems. The exception to this finding is the case in which the pedestrian is on the right side of a right [1,250-m (4,101-ft) radius] curve. In this case, the NIR system was associated with similar or shorter (though not significantly so) detection distances than the FIR and HLB systems. Drivers in this study detected the nonpedestrian object (tire tread) at similar distances regardless of the headlamp system in use (NIR, FIR, or HLB). This indicates that there is no significant loss in detection distance for small, low-contrast objects (such as tire treads) among the types of headlamps tested in this study. All of these findings appear to be applicable regardless of driver age. Subjective comments by the drivers in this study tend to be consistent with the objective results discussed above.

This volume provides an overview of the three studies that compose Phase III of the Enhanced Night Visibility project. The first study compared two prototype near infrared (NIR) vision enhancement systems (VESs), an infrared thermal imaging system (IR-TIS), and three headlamp-only systems in terms of drivers' nighttime detection and recognition of 17 objects. The objects included pedestrians on both sides of straight and curved sections of the road, roadway signs, and obstacles. A subset of the VESs and objects also were tested in rain conditions. The results indicated that both NIR and IR-TIS, if correctly implemented, provided additional detection benefit over headlamps alone for pedestrians in clear conditions. In rain conditions, the NIR also benefited object detection. A disability and discomfort glare study was also conducted with four high intensity discharge lamps and one halogen low-beam lamp. The results indicated that maximum illumination was the best predictor of driver discomfort and disability.