Two-wheeler vehicles (i.e., bicycles, mopeds, and motorcycles) are becoming increasingly popular in congested cities because of their small dimensions, low cost of use compared to cars, and their contribution to a healthy lifestyle. Even though the use of two-wheelers offers benefits, their low conspicuity, instability, and vulnerability of the users create safety risks. Due to their small size, two-wheelers tend to be overseen by other road users, especially at intersections. Furthermore, the stability of two-wheelers is easily affected by disturbances such as an uneven road surface. Moreover, the unprotected state of two-wheeler users contributes to a high risk of serious injuries once an accident happens. A better understanding of how crashes occur in the rider-vehicle-road system is needed....

Research shows that the ability to anticipate safety-critical situations is predictive of safe performance in traffic. Thus far, hazard anticipation training has been developed mainly for car drivers. These training programs may not be appropriate for cyclists who are exposed to different types of hazards. This study aimed to develop a PC-based hazard anticipation training for experienced cyclists, and evaluate its short-term effectiveness using hazard anticipation tests. Sixty-six electric bicycle users completed either a hazard anticipation training or a control intervention. The hazard anticipation training consisted of videos divided into two modules (instructions and practice) and was designed using various evidence-based hazard anticipation educational methods such as a ‘What happens next?’ task, expert commentary, performance feedback, and analogical transfer between hazardous traffic situations. The evaluation of the training showed that cyclists from the training group identified hazards faster compared to the control group cyclists, but no significant difference was found in the number of detected hazards between the two groups. The training had a small positive effect on cyclists’ prediction accuracy at safety-critical intersection situations. No effect was found on perceived danger and risk in hazardous traffic situations. Our results suggest that experienced cyclists’ hazard anticipation skills can be improved with the developed PC-based training. Future research should evaluate the retention and transfer of learned skills.

Powered two-wheeler riders are frequently involved in crashes at intersections because an approaching car driver fails to give right of way. This simulator study aimed to investigate how riders perform an emergency braking maneuver in response to an oncoming car and, second, whether longitudinal motion cues provided by a motion platform influence riders' braking performance. Twelve riders approached a four-way intersection at the same time as an oncoming car. We manipulated the car's direction of travel, speed profile, and its indicator light. The results showed that the more dangerous the situation (safe, near-miss, impending-crash), the more likely riders were to initiate braking. Although riders braked in the majority of trials when the car crossed their path, they were often unsuccessful in avoiding a collision with the car. No statistically significant differences were found in riders' initiation of braking and braking style between the motion and no-motion simulator configurations.

Objective: It is well established within the traffic psychology literature that a distinction can be made between driving skill and driving style. The majority of self-report questionnaires have been developed for car drivers, whereas only limited knowledge exists on the riding skill and style of cyclists. Individual differences in cycling skills need to be understood in order to apply targeted interventions. Methods: This study reports on a psychometric analysis of the Cycling Skill Inventory (CSI), a self-report questionnaire that asks cyclists to rate themselves from definitely weak to definitely strong on 17 items. Herein, we administered the CSI using an online crowdsourcing method, complemented with respondents who answered the questionnaire using paper and pencil (n = 1,138 in total). Our analysis focuses on understanding the major sources of variance of the CSI and its correlates with gender, age, exposure, and self-reported accident involvement as a cyclist. Results: The results showed that 2 components underlie the item data: Motor–tactical skills and safety motives. Correlational analyses indicated that participants with a higher safety motives score were involved in fewer self-reported cycling accidents in the past 3 years. The analysis also confirmed well-established gender differences, with male cyclists having lower safety motives but higher motor–tactical skills than female cyclists. Conclusions: The nomological network of the CSI for cyclists is similar to that of the Driving Skill Inventory for car drivers. Safety motives are a predictor of self-reported accident involvement among cyclists.

Introduction: Many bicycle–car crashes are caused by the fact that the driver fails to give right of way to the cyclist. Although the car driver is to blame, the cyclist may have been able to prevent the crash by anticipating the safety-critical event and slowing-down. This study aimed to understand how accurate cyclists are in predicting a driver's right-of-way violation, which cues contribute to cyclists' predictions, and which factors contribute to their self-reported slowing-down behavior as a function of the temporal proximity to the conflict. Method: 1030 participants were presented with video clips of nine safety-critical intersection situations, with five different video freezing moments in a between-subjects design. After each video clip, participants completed a questionnaire to indicate what the car driver will do next, which bottom-up and top-down cues they think they used, as well as their intended slowing-down behavior and perceived risk. Results and conclusions: The results showed that participants' predictions of the driver's behavior develop over time, with more accurate predictions (i.e., reporting that the driver will not let the cyclist cross first) at later freezing moments. A regression analysis showed that perceived high speed and acceleration of the car were associated with correctly predicting that the driver will not let the cyclist cross first. Incorrect predictions were associated with believing that the car has a low speed or is decelerating, and with reporting that the cyclist has right of way. Correctly predicting that the driver will not let the cyclist cross first and perceived risk were significant predictors of intending to slow down in safety-critical intersection situations. Practical applications: Our findings add to the existing knowledge on cyclists' hazard anticipation and could be used for the development of training programs as well as for cycling support systems.

Hazard anticipation skills are crucial for safe performance in traffic. Most hazard anticipation training programs have been developed for car drivers [1]. Traffic situations used in these programs may not be appropriate for cyclists because cyclists are subjected to specific types of hazards. In line with research evidence that hazard anticipation skills are suboptimal even in experienced road users, we aimed to develop and evaluate PC-based hazard anticipation training program for experienced adult cyclists. The first evaluation of this training program was conducted among electric bicycle users who seem to be more likely to be involved in a crash that requires treatment at an emergency department than people riding a conventional bicycle [2]. Sixty-six participants, randomly sampled into training and control groups (mean age of 58 and 57 years, respectively), completed either a PC-based training or a placebo intervention. The training intervention consisted of two modules: instructions and practice. In each module, participants saw video-clips of seven hazardous traffic situations. The screen was set to black just before a hazard developed, and participants had to choose one of the four responses to two questions: “What is the location of the hazard” and “What happens next?”. The placebo intervention consisted of the video-clips as well, but participants were asked questions about traffic rules and behaviors of other road users. Immediately after the training and placebo interventions, two video-based hazard anticipation tests were administered to the participants. During the first test participants saw video clips of hazardous intersection scenarios and were asked questions about (1) the prediction of the driver’s behavior, (2) the cyclist’s (own) slowing down behavior, and (3) perceived risk. In the second test, we measured hazard perception response time while participants watched video-clips in which hazards gradually developed. Participants were asked to press the space bar to indicate that the hazard was detected. After each video-clip, participants answered a question about how dangerous the viewed situation was. Results showed that participants assigned to the training group were more accurate in predicting what a car driver will do next in hazardous situations at intersections, and they reported more frequently that they would slow down in these situations compared to the control group. Cyclists who completed the training program outperformed cyclists in the control group regarding hazard perception response time (t (64) = 3.028, d = 0.745, p = 0.004). No significant differences between two groups were observed in perceived riskiness and dangerousness of the hazardous situations. Our preliminary results suggest that accuracy of predicting driver behavior at intersections as well as hazard perception response time among electric bicycle users could be improved with the developed hazard anticipation training program.

Research indicates that crashes between a cyclist and a car often occur even when the cyclist must have seen the approaching car, suggesting the importance of hazard anticipation skills. This study aimed to analyze cyclists’ eye movements and crossing judgments while approaching an intersection at different speeds. Thirty-six participants watched animated video clips with a car approaching an uncontrolled four-way intersection and continuously indicated whether they would cross the intersection first. We varied (1) car approach scenario (passing, colliding, stopping), (2) traffic complexity (one or two approaching cars), and (3) cyclist’s approach speed (15, 25, or 35 km/h). Results showed that participants looked at the approaching car when it was relevant to the task of crossing the intersection and posed an imminent hazard, and they directed less attention to the car after it had stopped or passed the intersection. Traffic complexity resulted in divided attention between the two cars, but participants retained most visual attention to the car that came from the right and had right of way. Effects of cycling speed on cyclists’ gaze behavior and crossing judgments were small to moderate. In conclusion, cyclists’ visual focus and crossing judgments are governed by situational factors (i.e., objects with priority and future collision potential), whereas cycling speed does not have substantial effects on eye movements and crossing judgments.

In the Netherlands, 30% of fatal crashes between 2010 and 2015 involved a cyclist [1], with a large portion of these crashes occurring at intersections in urban areas. Contributing factors to driver-cyclist collisions at intersections are not only inadequate visual search, but also incorrect expectations about the other’s intentions [2]. Research also suggests that crashes between drivers and cyclists often happen even when the cyclist must have seen the approaching car [2].
The ability to anticipate future events is crucial for safe performance in traffic [3]. Recently, research has started on hazard anticipation in cycling. For example, an experiment using a hazard perception test has found that adult cyclists detect hazards earlier than children [4]. Furthermore, results from an eye-tracking experiment using animated video clips showed that cyclists are more likely to look at an approaching car (e.g., a car on a collision course) than to a car that has stopped before the intersection or a car that has passed the intersection [5]. However, it is unknown at which point in time and based on which visual cues a cyclist can predict that a perceived hazard becomes an actual hazard (i.e., that the car driver will not yield to a cyclist).
We developed a video-based survey with the aim to gain an understanding of cyclists’ predictions in hazardous intersection situations. The following research questions were addressed herein:
(1) How do cyclists’ predictions of the behavior of a car change in the moments prior to a crash or near miss with that car?
(2) Is there a difference in cyclists’ predictions of the car’s behavior between crash and near miss scenarios?

This study investigated cycling performance of middle-aged (30–45 years old; n = 30) versus older (65+ years; n = 31) participants during low-speed tasks for which stabilization skills are known to be important. Additionally, participants’ self-ratings of their cycling skills and performance were assessed. Participants rode once on a conventional bicycle and once on a pedelec, in counterbalanced order. Three standardized tasks were performed: (1) low-speed cycling, (2) acceleration from a standstill, and (3) shoulder check. During Tasks 1 and 3, the mean absolute steering angle (a measure of the cyclist's steering activity) and the mean absolute roll rate (a measure of the amount of angular movement of the frame) were significantly greater for older participants than for middle-aged participants. These large lateral motions among older cyclists may indicate a difficulty to control the inherently unstable system. Comparing the conventional bicycle and the pedelec, participants reached a 16 km/h threshold speed in Task 2 sooner on the pedelec, an effect that was most pronounced among the older participants. Correlations between skills assessed with the Cycling Skill Inventory and actual measures of cycling performance were mostly not statistically significant. This indicates that self-reported motor-tactical and safety skills are not strongly predictive of measures of actual cycling performance. Our findings add to the existing knowledge on self-assessment of cycling skills, and suggest that age-related changes in psychomotor and sensory functions pose hazards for cycling safety.

The aim of this chapter is to illustrate to driving instructors how science contributes to cumulative knowledge on road safety. We do this by reviewing a scientific study for each of the three classical Es of road safety: (1) education, (2) enforcement, and (3) engineering.
Regarding education, we review the DeKalb experiment from the 1980s, which was a largesample randomized controlled trial that studied the effect of driver education on postlicense crash rates. The DeKalb experiment showed that participants who were assigned to a state-of-the-art driver education program performed better on theory and road tests, and became licensed sooner than control participants who did not receive formal driving instruction. Although the state-of-the-art education improved these target outcomes, there is no consistent evidence that it reduced crash risk. The recent consensus is that theoretical knowledge and skillful maneuvering alone are not sufficient for safe driving. Drivers should also have postlicense on-road experience and the lifestyle and attitudes that contribute to a safe driving style.
Regarding enforcement, we describe a UK study from the late 1990s on the statistical reliability of the formal road test. In this study, driving test candidates were asked to retake the test with a different examiner. The results showed surprisingly low consistency between the two tests, indicating that an assessment of a 30-minute drive might not be trustworthy. We provide several recommendations (such as increasing the test duration and implementing standardized routes and checklists) for improving the reliability of road testing. Furthermore, the value of computerized testing (e.g., hazard perception testing) and long-term data collection (e.g., in-vehicle driver state monitoring) is addressed.
Regarding engineering, the growing prevalence of active safety systems in vehicles has raised the question of how to treat such technologies in driver education curricula. A study on electronic stability control (ESC) was reviewed to illustrate how advances in technology improve road safety and affect elements of on-road training. In the case of ESC, skid training has become less relevant, but it is unknown whether learner drivers should experience critical driving situations during which the ESC gets activated. This may foster their overconfidence.