A total of 729 bicyclist fatalities and approximately 50,000 bicyclist injuries were reported in the United States in 2014. Of the bicyclist fatalities, roughly 50% involved crashes that occurred at civil twilight or after dark (National Center for Statistics and Analysis, 2016). Also regarding bicyclist fatalities, Hutchinson and Lindsay (2009) found that the majority of bicyclists’ deaths resulted from collisions where the cyclists were hit from behind, especially at night, after investigating Australian bicycle/motor vehicle crashes. In addition, Bil, Bílováa, & Müller, (2010) found that the highest percentages of cyclists’ fatalities occurred on straight road segments (23%) and curved road segments (16%) after analyzing patterns found in 5428 cyclists/motor vehicle collisions in the Czech Republic. Bicycle taillights have been found to enhance nighttime conspicuity (for a review, see Kwan & Mapstone, 2004). However, there are several gaps in the relevant literature, such as whether there is an optimal taillight mode, such as flashing or steady, and an optimal place to mount a taillight for enhancing a cyclist’s nighttime conspicuity. Although Wood et al. (2012) found no significant difference between the recognition distances of flashing and steady headlight, this comparison has never been studied empirically using bicycle taillights (which are red and less intense than headlights). With regard to taillight placement, studies have consistently found that placing lights or retroreflective material on major joints (e.g., knees, and ankles) can enhance bicyclist conspicuity (e.g., Blomberg, Hale, & Preusser, 1986; Koo & Dunne, 2012; Koo & Huang 2015; Stapleton, Stapleton, Koo, & Koo, 2017; Tyrrell, Fekety, & Edewaard, 2016; Wood et al., 2012). Indeed, there is a growing body of literature that supports the concept of biological motion (“biomotion”; Johansson, 1973); humans’ perceptual sensitivity to discriminating human joint movement patterns (for a review, see Tyrrell, Wood, Owens, Whetsel Borzendowski, and Stafford Sewall, 2016). What remains unclear is whether the bicyclist needs to be physically pedaling in order for biomotion to be perceived. Balk, Tyrrell, Brooks, and Carpenter (2008) found that pedestrians who walked in place elicited longer participant response distances than pedestrians who stood still. While this provided empirical support that biomotion configurations offer more benefits for pedestrians when their extremities are in motion, the parallel comparison for bicyclists (i.e., non-pedaling cyclist vs. pedaling cyclist) has never been made. The present study examined the conspicuity values of four taillight configurations that were systematically varied in terms of taillight mode (flashing or steady) and placement (seat post or heel) on two distinct roadway geometries. Methods. Visually healthy participants (N = 219) were driven along a designated test route and were asked to press a button on a numeric keypad each time they became confident that a bicyclist was present. Participants encountered Cyclist 1 (who was positioned adjacent to a straight road) approximately five minutes into the drive, and approximately two minutes later, participants were driven past Cyclist 2 (who was positioned after a 90 degree curve). Upon each press of the response button, a timer on the computer was activated. The researcher stopped the timer upon passing each relevant cyclist. The time between the participants recognizing a cyclist and the vehicle passing the cyclist was used to calculate each participant’s response distance (Distance = Speed * Time). This technique has been used in numerous on-road pedestrian studies, and its accuracy has been verified (e.g., Fekety, Edewaard, Stafford-Sewall, & Tyrrell, 2016; Whetsel- Borzendowski, Stafford-Sewall, Rosopa, & Tyrrell, 2015). After passing both test cyclists, the participants were informed that the experimental session was finished, and they were then debriefed while they were driven back to campus where they were released. Each experimental session lasted approximately 30 minutes. Response distances resulting from trials where glare from oncoming vehicles interfered with the participants’ view of the bicyclists were excluded from the analyses. A response distance of 0 m was recorded whenever a participant failed to respond or responded after passing a cyclist. Results. Two separate between-subjects analysis of variance (ANOVA) tests were performed to examine each of the four taillight configurations’ distributions for Cyclist 1 and Cyclist 2. For Cyclist 1, the main effect of Taillight Configuration was statistically significant, F(3,168) = 19.21, η2 = .255, p < .001. Bonferroni post-hoc pairwise comparisons revealed that, when steady lights were mounted to the cyclist’s pedaling heels, participants responded from a mean distance that was 1.7 times greater than when a flashing light was mounted to the seat post and 5.5 times than when a steady light was mounted to the seat post or when lights were mounted to the cyclist’s non-pedaling heels. In addition, the mean response distance for the flashing seat post configuration was 3.1 times greater than those for the steady seat post configuration and the non-pedaling heel-mounted lights. For Cyclist 2, the main effect of Taillight Configuration was also statistically significant, F(3,162) = 9.82, η2 = .154, p < .001. Bonferroni post-hoc pairwise comparisons indicated that when either a flashing or steady light was mounted to the seat post or the lights mounted to the cyclist’s pedaling heels the mean response distances were 1.6x greater than the mean response distance for the non-pedaling heel-mounted lights configuration. Discussion. This study investigates the use of bicycle taillights as a way for bicyclists to help approaching drivers to more quickly recognize their presence on roadways of differing geometries at night. The results of this study suggest that, in order to maximize conspicuity while riding at night, it seems best to use two heel-mounted taillights. This configuration capitalizes on drivers’ perceptual sensitivity to the biological motion of other people. However, cyclists with only one taillight should use it on a flashing setting while mounted to their bicycle’s seat post when riding at night in order to enhance their conspicuity to approaching drivers. Bicyclists must also be informed that conspicuity aids are not effective 100% of the time, and therefore, they should always ride defensively. Still, this study empirically demonstrated that highlighting a bicyclist’s movement with lights is effective for maximizing nighttime conspicuity. The results of this study can be useful to designers of bicycle taillights, since these data offer valuable insights into how taillights can be used to maximize bicyclist conspicuity at night.

No abstract

Introduction. According to the National Highway Traffic Safety Administration (NHTSA), 5,376 pedestrians were killed and approximately 70,000 were injured in 2015 in the US. Seventy-four percent of all pedestrian fatalities occurred in low illumination nighttime conditions, which is a two percent increase from the previous year (NHTSA, 2017; NHTSA, 2015). Pedestrians can wear visibility aids such as retroreflective material to increase their own conspicuity to drivers at night. The conspicuity benefit is maximal when the retroreflective material is strategically positioned to highlight the pedestrians’ “biological motion” (e.g., Wood, Tyrrell, & Carberry, 2005). Unfortunately, research indicates that most pedestrians are unaware of their own visibility and tend to overestimate their own visibility to drivers at night. However, upon hearing an educational lecture about nighttime visibility and retroreflection, people are more willing to purchase conspicuity-enhancing clothing (Borzendowski et al., 2014). Thus, it is apparent that education is key to increasing awareness of the conspicuity issues at night and therefore pedestrians’ safety. To our knowledge, typical pedestrians’ understanding of retroreflective and fluorescent materials has not been explored empirically. We hypothesized that typical roadway users do not understand the qualities of the retroreflective and fluorescent material until they see them under specific nighttime conditions. Further, we hypothesized that more in-person interaction with visibility aid materials would lead to more accurate judgments of their utility. The purpose of this study was to quantify observers’ judgments of visibility aids in a simulated nighttime setting before and during a visual demonstration. Methods. Eighty-four (84) participants, each of whom met pre-determined visual performance criteria, were included in the analysis. The Clemson University Institutional Review Board approved all procedures. This study followed a mixed factorial design. In the “pre-demonstration” phase, each observer inspected each of the nine test materials using one of three material presentation modes (Wall, Fixed, and Flexible). For each of the nine materials, the task was to provide a numeric estimate (magnitude estimation) of the brightness that the material would display during the demonstration. Following this, the demonstration began and observers provided new magnitude estimates for each of the nine materials. The 4” x 4” (10.2 cm x 10.2 cm) stimuli were four retroreflective (black retroreflective (hereafter “retro”), blue retro, purple retro, and silver retro) materials, four diffuse reflective (black, blue, purple, & silver) materials and one fluorescent material (yellow fluorescent). Experimenters had, in advance, perceptually matched the chromatic appearance of the retroreflective and paper materials (black, blue, purple, and silver) so that the chromatic appearance of the paper stimuli were similar to the retroreflective stimuli when viewed in room light. During the demonstration, however, each of the retroreflective materials appeared to be white/silver because their appearance chromatically matches that of the light source illuminating them. The observer sat in a chair positioned 20 feet (6 m) from the wall where the material samples were mounted for the demonstration. The observer was presented with the nine material samples using a new random sequence. Presentation of that material had three modes of interaction with the material: placed in the observer’s hand and easily manipulated (Flexible); placed in the observer’s hand and fixed to a rigid piece of cardboard (Fixed); placed on the wall 20 feet (6 m) away with no personal interaction (Wall). Each of the nine materials was presented in a predetermined randomized order for both the pre-demonstration and the demonstration phase. For the pre-demonstration phase, observers were asked to predict how bright each of the nine samples would appear when the room lights in the lab were turned off and only a desk lamp, positioned next to the participant, illuminated the sample. For the demonstration phase, observers were situated in a chinrest while still sitting in a chair positioned 20 feet away from the opposing wall of the lab. Observers then viewed each sample in “demonstration phase” (i.e., mounted on the wall and illuminated by the lamp). Observers provided a second magnitude estimate of the brightness of each material. After the experiment, observers competed a survey of their clothing and visibility aid choices during outside nighttime activities. Results. For brevity, only results that are directly relevant to the hypothesis under investigation are reported here. A mixed ANOVA tested the effects of material type, presentation mode, and presentation phase on the brightness magnitude estimations of each material used. There were three important findings. The most important finding is a significant interaction between presentation phase and material type, F(4.60, 372.79) = 213.77, p < 0.001, partial η2 = 0.73. The second most important finding is the significant three-way interaction between presentation phase, material type, and material presentation mode, F(9.20, 372.79) = 5.50, p < 0.001, partial η2 = 0.12. The third most important finding is the significant interaction between material type and material presentation mode, F(6.86, 372.79) = 5.17, p < 0.001, partial η2 = 0.11. In the survey analysis, 61% of observers reported in the post experimental questionnaire that they were frequently outside at night. Approximately 67% (67.1%) of observers reported that during the past week, they wore light-colored clothing while outdoors at night. Of these, 47% reported that they wore light-colored clothing because it was what they were already wearing during the day while only 25.5% reported wearing light colored clothes for safety reasons. Only one observer from the sample (0.01%) reported wearing a reflective vest outside at night. Seven percent (7.1%) of observers reported wearing fluorescent colored clothing outside at night with 20% of these reporting that this was motivated by safety reasons. In ranking the usefulness at night of the following items: light-colored clothes, retroreflective vest, fluorescent colored clothes, shoes with reflective patch on them, clothes with reflective patch on them, and flashlight, observers ranked the retroreflective vest as being most useful and the light-colored clothes as being least useful. Discussion. This study found that observers are not fully aware of how retroreflective materials function or how they can enhance the ability of drivers to see a pedestrian ahead. We found that naïve observers who closely inspected and touched retroreflective material more accurately judged its value than did observers who saw it from a distance. The methods of this study proved crucial to the study’s results, which indicate that although observers are generally unaware of the benefits of visibility aids they can better ascertain the benefits from touching and studying the material. The post-experimental survey showed that participants are somewhat aware of the appropriate clothing to maximize visibility at night but did not report actually using the appropriate clothing. This highlights a lack of appreciation for the visibility problem at night. Additional studies are needed to further explore observers’ understanding of nighttime visibility problems and to find effective interventions that encourage road users to maximize their own conspicuity at night.

Bicyclists risk being injured or killed in crashes with motor vehicles, even during the daytime. Therefore, cyclists must help approaching drivers detect and recognize their presence. The present study examined the conspicuity benefits of bicycle taillights during the daytime. Participants' eye movements were recorded as they searched for vulnerable road users in videos recorded from a driver's perspective in a moving vehicle. Five of the videos contained a bicyclist who displayed one of five taillight configurations. The distance from which each participant first glanced at the bicyclist was recorded, as was the distance from which the participant pressed a button to indicate that a bicyclist was present. The results indicated that the participants first fixated on the bicyclist at a distance that was 2.7 times greater than the distance at which they responded to recognizing the bicyclist. Additionally, the bicyclist was recognized from significantly greater distances when using a flashing or steady seat post-mounted taillight than when no taillight was displayed. These findings confirm earlier research that bicycle taillights can enhance drivers' ability to recognize bicyclists during daylight.

Bicyclists risk being involved in collisions with motor vehicles, even during daytime. Thus, bicyclists who ride in daylight must enhance their conspicuity. This study assessed the daytime conspicuity benefits of bicycle taillights using eye tracking technology. Participants were driven along an open-road route while wearing an eye tracker and pressed buttons when they detected and recognized a test bicyclist. Participants encountered the bicyclist displaying one of four taillight configurations, and the distances from which they responded to the test bicyclist were recorded. The results revealed that, after participants first glanced at the bicyclist, a significant amount of time was needed to detect and recognize the bicyclist. Further, seat post-mounted lights displayed with or without lights mounted to the heels of the rider’s shoes provided the greatest conspicuity advantage in terms of recognition. This experiment offers useful insights into the optimal light placement options for bicyclists to enhance their daytime conspicuity.