Improving the conspicuity and safety of pedestrians and cyclists on night-time roads

ABSTRACT

The visual limitations of drivers at night are a key contributing factor to the relatively high crash involvement of vulnerable road users including pedestrians, roadworkers and cyclists on night-time roads. Making vulnerable road users more conspicuous (recognisable, rather than simply visible) to oncoming drivers, is one approach to increasing their safety and has been a particular focus of my research. This review highlights the experimental approaches that our multidisciplinary research team have adopted to explore these issues, involving both closed and open road studies at night. One effective strategy to increase night-time conspicuity of vulnerable road users is clothing that includes retro-reflective materials on the moveable joints which, when illuminated in the headlight beam of oncoming vehicles, creates a strong sense of ‘biological motion’ or ‘biomotion’. Our studies demonstrated that this basic visual perception allows drivers to accurately perceive the presence of a person, such as a pedestrian or cyclist, at much longer distances than when retro-reflective materials are positioned on the torso, as in high visibility vests. Subsequent studies demonstrated that the benefits of biomotion clothing are evident in cluttered environments, in the presence of glare, and for drivers of different ages and visual characteristics. Evidence gathered in these studies was instrumental in changing Australian and New Zealand standards governing high visibility clothing for roadworkers to include retro-reflective strips in the biomotion configuration. Ongoing studies are exploring how to make biomotion clothing attractive to vulnerable road users exercising at night, and how to ensure that the limitations of night-time vision and the importance of increasing night-time conspicuity are better understood. This body of research has involved collaborators from a range of disciplines who have been essential to understanding and addressing the visual challenges of night-time roads and assisted in translating this research into tangible benefits for night-time road safety.

KEYWORDS:

• Biological motion
• conspicuity
• cyclists
• night-time driving
• pedestrians

Introduction

It was a great honour to receive the prestigious Barry H Collin Award^* and to join those outstanding “Australia-connected” optometry and vision science researchers who have been previous recipients of this award. My own “Australian connection” involves more than 30 years at Queensland University of Technology, where I have been lucky enough to work in a highly collegial and supportive environment, which encouraged me to stay well beyond the two years of my initial contract.

I have also been extremely fortunate to collaborate with a wide range of researchers, both from Australia and internationally, who come from many disciplines including psychology, ophthalmology, road safety, civil and lighting engineering, design and exercise physiology. The opportunity to collaborate across so many disciplines with such excellent researchers has been a great inspiration to me and without their input and insights my achievements in this field would not have been possible.

The focus of my research has centred around vision, ageing and driving, including how vision impairment and ageing affect driving, as well as night-time driving and visibility.^1,2 My research interests have also extended to investigation of the effects of vision impairment on falls and mobility, as well as exploring the role of vision in academic performance in children. The goal of my research endeavours, regardless of the specific field, has been in the practical application of the findings to the real-world.

One research area where the translation of research findings has been particularly effective is in night driving, specifically in improving the visibility and hence safety of vulnerable road users (including recreational and occupational pedestrians, runners, as well as cyclists), who are at risk of being involved in either an injurious or fatal collision on night-time roads. This research area is the focus of this review which will provide context in terms of the challenges of night-time driving, the experimental approaches used to address these issues, research outcomes, and the translation of findings into practice and standards.

Challenges of night-time roads

Night-time roads are dangerous for drivers, pedestrians, and cyclists. Even though driving exposure is much lower at night, fatality rates are up to 3 times greater than in the daytime adjusted for distances travelled,³ with a recent report based on US crash data indicating that the rate of night-time crashes is increasing.⁴ This increased crash risk at night is exacerbated for collisions between pedestrians and vehicles. Pedestrians are 3–7 times more likely to be involved in a fatal collision with a vehicle at night than in the day,⁵ with 90% of the increase in night-time pedestrian fatalities recorded in the US between 2009 and 2018 occurring in darkness.⁶

It has also been reported that the number of fatal accidents involving roadworkers in construction zones is 5 times higher at night than in the day.⁷ Similarly, collisions involving cyclists and vehicles are 2–6 times more likely to occur at night than in the day.⁸ Importantly, the majority of vulnerable road users are not self-illuminated, with the exception of cyclists using bicycle lighting. Analyses of crash databases demonstrate that the increased risk of injury and fatalities for pedestrians and cyclists on night-time roads arises from the reduction in their visibility, rather than other factors such as driver fatigue or alcohol consumption.^5,9

Experimental approaches

The visibility challenges of night-time driving are not well understood, despite its role in night-time fatalities. The opportunity to work with a series of eminent experimental psychologists, Fred Owens, Rick Tyrrell, and Alex Chaparro, during their sabbatical visits to QUT in the 1990s onwards, provided me with the stimulus to translate my interest in night driving into the development of methodologies to explore factors impacting on the night-time conspicuity of vulnerable road users and to develop ways to improve their conspicuity and hence safety. In this context, conspicuity refers to the ability of vulnerable road users to attract the attention of drivers and be perceivable as a person, rather than being simply visible or detectable as an ambiguous object.

These methodologies have evolved over many years and are largely based on field-based assessments on a closed road circuit, that have been used under daytime conditions since the early 1990s to investigate the impact of simulated and true vision impairment.¹ In more recent years, I have been lucky enough to be joined by Alex Black, a former PhD student of mine, who I work with closely to further develop experimental approaches and designs in this research field.

The closed road circuit includes a series of rural roads including intersections, hills, straight and curved stretches of roadway, as well as standard road markings and signs (Figure 1). The advantage of the closed road is that it is representative of the normal light levels encountered on night-time roads, including the low or high beam headlights of vehicles. Night driving simulators, due to technical limitations associated with digital displays, may not well represent the wide range in luminance values of the night driving environment (which vary from the black of night to the high luminance of an oncoming high beam headlight),¹⁰ although some groups have introduced external glare sources to better represent the effects of oncoming headlights.^11,12

Figure 1. Aerial view of the driving circuit and schematic of the set-up for assessing conspicuity of vulnerable road users at night-time.

The closed road also provides a relatively high level of control. The level of glare can be manipulated to represent the headlights of oncoming vehicles, although this is typically limited to single vehicles for practical and safety reasons, lighting can be added in the form of mobile streetlighting, and specific aspects of driving performance can be assessed in a repeatable way (e.g., pedestrian recognition, lane-keeping, hazard detection). In addition, drivers on a closed road experience a sense of realism and the “high stakes” that may be missing in modern simulators. Obvious disadvantages include having to test during night-time hours, which is a potential negative for both research personnel and participants, as well as the location of the closed road 30 km from the university, which involves significant travel time. However, despite these practical disadvantages, they are a small price to pay for the benefits of fidelity and realism of real-world driving research.

For studies of pedestrian and roadworker conspicuity, we have often assessed the ability of participants as drivers to recognise pedestrians walking in place (i.e. simulating walking but remaining stationary) at the opposite side of the roadway, which ensures the safety of the experimenters, as well as capturing normal walking movements. In these studies, we have simulated complex driving environments with low and high levels of visual clutter,¹³ and construction sites for studies of roadworker visibility.¹⁴ In some studies, the intensity of the drivers headlights¹⁵ and streetlighting¹⁶ was varied, and other studies explored the effects of vision impairments, including simulated optical blur and cataracts.^17–19

One limitation of these previous studies was that for safety reasons pedestrians walked in place at the roadside, rather than physically walking across or along the road, as is the case when pedestrians cross a road under real-world conditions. In recent studies, we addressed this by exploring the factors affecting pedestrian conspicuity and perceptions of the direction of walking when pedestrians actually walked across the roadway, however, in the interests of safety, the participants in these studies as drivers sat in a stationary vehicle.^20,21

In studies of cyclist conspicuity, we have included experimenters cycling in place on bicycles mounted in a training stand to capture the normal movements associated with pedalling a bicycle and have manipulated the effects of cyclist clothing and bicycle lighting, as well as exploring the effect of the characteristics of the drivers on outcome measures.^22,23 In recent studies, we also explored the passing distances selected by drivers as they approached a cyclist from the rear and the impact of bicycle markings on these passing distances.²⁴

Increasing conspicuity of vulnerable road users

The overall goal of improving the conspicuity of vulnerable road users is to increase their contrast against the background of night-time roads. This can include streetlighting, self-illumination (which is typically only adopted on bicycles for cyclists), as well as clothing. Retro-reflective materials, which are designed to reflect incident light directly towards the driver to increase contrast, have traditionally been incorporated into safety vests to increase the conspicuity of vulnerable road users at night. However, while the bright reflections from a vest can be detected at long distances (increased visibility), that does not necessarily imply that the driver will perceive that a pedestrian or cyclist is present (conspicuity) and could alternatively be interpreted as other bright objects in the night-time driving environment, such as road signs, reflective posts or bollards. A particular focus of our research has been to explore the conspicuity benefits of an alternative placement of retro-reflective materials on the moveable joints rather than on the torso, in a configuration known as ‘biomotion’, which creates the sense of biological motion.

Biological motion describes the ability of the human visual system to recognise movement from only a limited array of visual stimuli. Johansson²⁵ showed that when human motion is represented by point-light (PL) displays, with small lights attached to the major moveable joints of actors (ankles, knees, waist, shoulders, elbows, wrists), viewers could easily identify the moving human form, despite only seeing the PL displays against a dark background. Participants were also able to recognise the activities that the PL actors were engaged in, including walking, cycling, and dancing. A significant body of work has further explored the characteristics of biological motion. For example, based only on the motion information available in these PL displays, observers can recognise a PL person’s gender and emotions, estimate the weight of a lifted object from observing the lifting motion of the individual and even identify animals whose movements are represented in PL animations (see review by Blake and Shiffrar).²⁶

The potential applications of the perceptual sensitivity to biological motion include biometrics in surveillance, medicine and security, as well as sport and the military.²⁷ One key application is in increasing the conspicuity of pedestrians and cyclists at night, where positioning retro-reflective strips on the movable joints illuminated in the oncoming vehicle headlight beam, evokes a sensation of biological motion, enabling recognition of the road user at longer distances. This is relevant to the large numbers of people who are killed or injured at night every year and connects what we know about the brain being specialised for biomotion, as a way to apply perceptual research to a major problem facing modern society.

This translation of biomotion to the conspicuity of vulnerable road users at night was explored by Owens et al,²⁸ who were one of the first to explicitly address the possibility that the benefits of biomotion could be leveraged for this application. In this study, videos of runners wearing either no retro-reflective material, retro-reflective material on the torso as a vest, or as biomotion strips on the moveable joints were viewed by observers who could recognise the individuals wearing the biomotion strips sooner than those wearing the other clothing configurations. A potential constraint of this approach was the limited ability of the videos to fully capture the perceptual characteristics of the night-time driving environment. Our research has taken a different but complimentary approach, and significantly extended the scope of this work to real-world night driving conditions. These studies have explored the impact of a range of factors that impact on night-time conspicuity of vulnerable road users, including pedestrians, roadworkers and cyclists, and the ways in which their safety on night-time roads can be improved.

Pedestrians and roadworkers

Our studies exploring the application of biomotion to recreational pedestrians and occupational pedestrians such as roadworkers, involved experimenters walking in place at the roadside of a closed road circuit under night-time conditions. Participants as drivers, drove around the circuit in an instrumented vehicle and tapped a touchpad when they first recognised a pedestrian or roadworker at the roadside, which recorded the recognition distances. A particular focus of these studies was to explore how to maximise conspicuity to oncoming drivers and determine the factors that impacted on conspicuity, including pedestrian clothing, as well as the characteristics of the driving environment and those of the driver themselves.

Pedestrian clothing

In one of our first studies, pedestrians wore either black or white clothing, black clothing and a retro-reflective vest, or black clothing with retro-reflective strips added to the moveable joints, in the ‘biomotion’ configuration.²⁹ By keeping the same total surface area of retro-reflective markings in both the vest and biomotion conditions, the study was able to isolate the role of the placement of retro-reflective markings on conspicuity. Drivers using low beam headlights always recognised pedestrians when wearing biomotion (100%) compared to either vest (60%) or black clothing (40%), where more than half of the drivers never responded to the pedestrian wearing black, even though participants were aware that the study measured their ability to respond to pedestrians.

The distance at which drivers first responded to pedestrians wearing the biomotion configuration was 3 times longer than that for the pedestrians wearing the vest (148.2 m vs 43.4 m), and more than 26 times longer than for the black clothing condition (148.2 m vs 5.6 m) (Figure 2). These benefits were similar for both young and older drivers, although the older drivers responded to pedestrians at much shorter distances than did the younger drivers. The benefits of biomotion were also evident across different headlight beam configurations (low and high beam).

Figure 2. Group mean response distances (SE) as a function of pedestrian clothing and driver age for the low beam headlight condition.

The benefits of biomotion clothing compared to black in recognising the presence of a pedestrian walking across the road ahead are illustrated under real-world conditions in Figure 3. These real-world benefits were also confirmed in an open road study by other researchers,³⁰ where pedestrians were either standing still or walking in place wearing either black clothing or black clothing with one of four different configurations of retro-reflective markings. Balk, et al.³⁰ demonstrated that conspicuity distances were significantly longer (improved pedestrian recognition) when retro-reflective markings were located on the arm and leg extremities, with the effects for biomotion being greatest when pedestrians were moving rather than stationary.

Figure 3. Images of a pedestrian walking rightwards across a suburban street wearing either black clothing (highlighted by a red circle, left) or retro-reflective strips in the biomotion configuration (right).

Another series of closed road studies found that biomotion does not simply increase the distance at which drivers can perceive the presence of a pedestrian, but further, help the driver perceive what the pedestrian’s actions are.^20,21,31 These studies determined that drivers can successfully judge the walking direction of pedestrians when they are wearing biomotion clothing, which is important for a driver to judge whether a pedestrian is likely to be entering the roadway (and therefore requiring evasive action from the driver) or exiting the roadway. In one study, participants driving around the circuit were able to recognise the walking direction of pedestrians (either facing leftwards or rightwards at the roadside) at 2.3 times longer distances when they wore biomotion clothing compared to a vest.³¹ Eye movement data collected in the same study demonstrated that the biomotion clothing attracted the driver’s attention sooner than the retro-reflective vest, and that they were able to recognise that the bright light points resulting from the retro-reflective strips represented a human form in around half the time than for pedestrians wearing a retro-reflective vest.³¹

Recent studies extended these findings and involved more realistic pedestrian motion, with pedestrians walking across the road wearing different clothing configurations, while participants were seated in a stationary vehicle.^20,21 Findings confirmed the superiority of the biomotion configuration for providing information regarding pedestrian walking direction, which was significantly higher than when pedestrians wore retro-reflective materials on the legs and torso, legs only, torso only, or street clothing.²⁰ Figure 4 illustrates the capacity to detect walking direction of a pedestrian when wearing retroreflective strips in a biomotion configuration.

Figure 4. Images of a pedestrian wearing retro-reflective strips in the biomotion configuration enabling recognition of the direction of walking in a leftwards (left) and rightwards direction (right).

Characteristics of the road environment

The characteristics of the road environment in which pedestrians are walking are also important in terms of the ability of drivers to recognise their presence in time to avoid a collision. Factors that have been shown to impact on night-time conspicuity include the level of illumination on night-time roads, both from the headlights and streetlighting, and the complexity (clutter) of the driving environment.

Headlights

The lighting provided by a driver’s headlights has many benefits for forward lighting of the road as well as the road verge and has benefits in terms of detecting the presence of roadside pedestrians with and without retro-reflective clothing. In one closed road study, pedestrians were recognised at longer distances when the driver’s headlights were on the high rather than the low-beam setting, with this improvement in recognition distance being most evident when pedestrians wore black clothing (3.5 times longer recognition distances).²⁹

Similarly, in another study, the intensity of drivers’ headlight beams was reduced using a series of neutral density filters and demonstrated that recognition performance was significantly higher, (increased numbers of pedestrians, road signs and road hazards correctly recognised), for the full high beam condition compared to the filtered conditions.^15,32 The study also demonstrated that drivers failed to reduce their speed sufficiently to compensate for this reduction in recognition ability.¹⁵

However, while headlights are highly beneficial in terms of their capacity to increase lighting on night-time roads, they can also form a glare source to oncoming drivers and decrease their ability to recognise the presence of pedestrians at night. Indeed, glare is the most commonly cited difficulty for those who report problems with night driving,³³ particularly older drivers.³⁴ In closed and open road studies, the presence of oncoming glare has been shown to reduce the ability of drivers to recognise the presence of pedestrians walking at the roadside for studies involving both simulated pedestrians,³⁵ as well as real pedestrians.^18,32,36,37

The negative effects of headlight glare are exacerbated in the presence of vision impairment, including simulated cataracts and optical blur, in closed road,¹⁸ as well as in driving simulator studies of both simulated and real cataracts.¹² The negative effects of glare on night-time pedestrian recognition have also been shown to be more strongly associated with mesopic measures of visual function, including motion sensitivity and visual acuity, than standard photopic measures of visual acuity included in driver licencing.³⁷

Streetlighting

The level of illumination provided by streetlighting varies dependent on the road characteristics, density of traffic, as well as a range of other factors. In recent years, newer streetlight technologies, such as light emitting diode (LED) streetlights, have been introduced which have reduced energy consumption and associated economic benefits compared to older technologies such as high pressure sodium (HPS) lighting. LED streetlighting also provides further benefits, through dimming and adaptive light controls, however, only limited research has considered the safety aspects of LED streetlights, and how dimming might impact on driving performance.³⁸

In closed road studies, the impact of LED streetlight dimming on driving performance, including pedestrian recognition distances was explored for young drivers.¹⁶ Dimming of LED streetlighting to 25% (L25) and 50% (L50) of the brightest level (L100) (average road surface luminance of 1.14 cd/m²), reduced recognition distances for pedestrians wearing non-reflective clothing by 15 m for L25 and 11 m for L50 compared to L100, while dimming to 75% did not significantly affect pedestrian recognition. We are currently extending these studies to include comparisons of pedestrian recognition ability when driving under various levels of LED streetlight dimming compared to standard HPS streetlighting, as well as exploring the interaction with driver age, given the effect of increasing age on pedestrian recognition, as described later in this review.

Visual clutter

The level of visual clutter or complexity within the road environment is an important factor to consider when exploring pedestrian conspicuity, given that pedestrians are typically surrounded by cluttered retro-reflective targets (cones, poles, and lights) on night-time roads. One laboratory-based study, where participants had to recognise pedestrians in video clips of visually cluttered traffic situations, failed to find that biomotion clothing improved night-time pedestrian recognition.³⁹ However, as mentioned in a previous section of this review, a potential limitation of video-based approaches is the potentially low fidelity with which the videos can replicate key elements of the night-time driving environment. Indeed, in closed road studies, we demonstrated that biomotion clothing improved pedestrian recognition distances regardless of the presence or absence of visual clutter that included extraneous points of light produced by retroreflective materials mounted on targets lit up in the headlight beam of the experimental vehicle.¹³

The above findings are highly relevant to roadworker visibility, given that they work in cluttered construction sites and have high levels of exposure on night-time roads, including passing traffic and are frequently involved in crashes, particularly at night.^7,40 We explored this issue in cluttered roadworker sites in both open road settings as well as simulated on the closed road. The open road studies were conducted at suburban and freeway roadwork sites and involved close collaboration with local transport authorities.¹⁴ Given that data collection was undertaken under in-traffic conditions, participants were seated in stationary vehicles at three different distances and made judgements regarding the relative conspicuity of roadworkers walking in place, wearing one of four clothing conditions. The clothing conditions included: (i) standard roadworker vest alone, or in combination with retro-reflective strips on the (ii) thighs, (iii) ankles and knees or in a (iv) biomotion configuration; judgments about relative conspicuity were made using three standardised scales. Findings demonstrated the conspicuity advantages of biomotion strips added to a standard retro-reflective vest for roadworkers in both the suburban and freeway road construction sites.¹⁴

These findings were supported by a subsequent closed road study where conspicuity distances could be measured under controlled conditions for roadworkers located within simulated roadwork sites.⁴¹ Results demonstrated that visibility distances were three times longer when roadworkers wore the standard retro-reflective vest together with retro-reflective strips in the biomotion configuration compared to the vest alone.⁴¹ The evidence provided by these studies was instrumental in changing the standards for high visibility clothing to incorporate biomotion markings,⁴² as described in the final section of this review.

Characteristics of the driver

The characteristics of the driver, including their age and visual status, are also important in terms of their capacity to recognise pedestrians at night and need to be considered when designing interventions to ensure that they are effective for all road users.

Age

Driver age has been shown to have a negative impact on the ability to recognise night-time pedestrians and to some extent reflects the problems reported by older drivers, where one in three older adults report vision-related night driving difficulties, with 20–50% restricting or ceasing night driving.^43,44 In a series of closed road studies, older drivers detected fewer pedestrians and at shorter distances than younger drivers.^{13,17,29,41,45} In one study that explored the effect of cluttered environments, older drivers recognised only 57% of the pedestrians compared to 77% for the younger drivers, and when they did recognise pedestrians, it was at less than half the distance (65 m vs 153 m).¹³ This finding was supported by a subsequent study, where older drivers responded to the presence of a road worker presented within a simulated road work site at less than half the distance of younger drivers.⁴¹ Exploration of the visual predictors of these age-related differences in night-time pedestrian recognition, demonstrated that measures including contrast sensitivity⁴¹ and motion sensitivity⁴⁵ rather than standard measures of visual acuity that are used in driver licencing were the best predictors of these differences.

Visual status

The driver’s visual characteristics also affect their ability to recognise pedestrians at night. In a series of closed road studies, we explored the effect of two common causes of reversible vision impairment, cataract, and uncorrected refractive error, on pedestrian recognition using repeated measures designs. In these studies, participants wore wide aperture goggles containing simulated blur or diffusing lenses to simulate cataracts, and pedestrian recognition was compared to a condition where participants wore their optimal correction.^17,18 Understanding the effects of optical blur are important, given that many individuals drive with uncorrected refractive errors, with one study reporting that uncorrected refractive error was the main reason that drivers did not meet the legal vision standards for driving.⁴⁶

In focus group studies, we also explored why drivers do not wear their distance spectacles for driving and found that many reported not wearing spectacles at night because they felt they made glare worse.⁴⁷ Our closed road studies demonstrated that pedestrian recognition was significantly reduced by refractive blur as low as 0.50 DS,¹⁷ as well as low to moderate levels of uncorrected astigmatism (0.75 to 1.50 DC).⁴⁸ In support of these findings, in a recent study where pedestrians walked across the road wearing either a retro-reflective vest or biomotion strips, the ability to recognise the walking direction of pedestrians was reduced by blur as low as 0.50 DS.²¹ We have also explored the effects of simulated cataracts on night-time driving and found drivers were less likely to recognise the presence of pedestrians and at significantly reduced distances.^18,19 Importantly, pedestrian recognition was impaired to a greater extent for the simulated cataract than the blur condition, even though visual acuity was matched between the conditions and always better than driver licencing standards.

These findings that cataract and low levels of optical blur have detrimental impacts on pedestrian recognition at night, suggests the need for early referral for cataract surgery and the wearing of optimum refractive correction for night-time driving. Our findings also support the suggestion that other measures of visual function including contrast sensitivity and motion sensitivity are important for night-time pedestrian detection.^32,45

Cyclists

Another important group of vulnerable road users are cyclists. However, while the health and environmental benefits of cycling make it an appealing mode of transport, cycling in traffic can be risky, and these risks are particularly evident at night. Indeed, while the distances cycled at night are much lower than in the day,^49,50 cyclist crash rates are relatively high, as indicated in survey-based studies of both drivers and cyclists,^51,52 and crash data.^8,53,54 One study reported a 55% increased risk of cyclist crashes at night compared to the daytime, where data from urban and rural areas were combined,⁵³ with the seriousness of cycling injuries sustained from crash involvement also being higher at night than in the day.⁵⁴

Many factors contribute to the increased injury risk at night, including alcohol and fatigue, as well as low cyclist conspicuity and visibility.⁵⁴ In a survey-based study, we found that poor visibility of cyclists was a key contributor to crashes with vehicles at night,⁵² with other studies reporting that drivers often state that they failed to see the cyclist until it was too late to stop to avoid a collision.^55,56 A driver’s late detection of a cyclist and delay in evasive action, contributes to increased severity of collisions at night, with the probability of fatal cyclist injury doubling on night-time roads without streetlights, compared to those occurring in the daytime or on night-time roads with streetlights.⁵⁷

Crashes are also common in situations where cyclists expect drivers to give them right of way, but drivers fail to yield. In many situations, this occurs because drivers are not aware that a cyclist is present on the roadway, either because they fail to scan the road sufficiently, or because cyclists are not conspicuous enough to drivers at night. Visibility aids either incorporated into cyclist clothing or making the bicycle itself more visible have the potential to improve the ability of drivers to recognise cyclists, providing more time for a driver to take appropriate action to avoid collisions.⁵⁸

Lights mounted on the bicycle (typically including white front and red rear facing lights that are either static or flashing) are used to improve cyclist visibility to other road users at night and are a legal requirement for night-time cycling in many countries. In a survey-based study, we demonstrated that participants who were cyclists, rated the visibility benefits of bicycle lights to be much greater than did the drivers, especially at night.⁵¹ In a closed road study, we explored the effects of retro-reflective clothing and the presence or absence of bicycle lighting on cyclist conspicuity.²² Data showed that a light mounted on the bicycle handlebars (regardless of whether it was flashing or static) did not improve drivers’ ability to recognise that a cyclist was present compared to a condition where no bicycle light was present.²² Interestingly, the position of the bicycle lights also affected cyclist conspicuity. In a recent study, we demonstrated that cyclist conspicuity was maximised when the bicycle light was mounted on the helmet rather than the handlebars,²³ likely because the handlebar-mounted bicycle light acted as a glare source for the leg-mounted reflective strips. Bicycle lights thus may make drivers aware that there is an object on the road ahead but do not necessarily facilitate drivers’ awareness that a cyclist is present (that is, they do not enhance bicyclist conspicuity). Interestingly, in another closed road study, we demonstrated that lights positioned on the handlebars and helmet also increased the passing distances selected by drivers.²⁴ This is highly relevant to safety, as close passing distances contribute to rear-end and side-swipe crashes,⁵⁹ particularly during overtaking manoeuvres.⁶⁰

The possibility that biomotion clothing could enhance the night-time conspicuity of cyclists has also been explored, particularly for leg-mounted retro-reflective strips, given that the primary movement of a cyclist is in the legs rather than the arms and upper body. In a closed road study, where experimenters cycled in place on a bicycle mounted in a training stand facing the driver, clothing that included retro-reflective markers in a modified biomotion configuration (ankles and knees), resulted in increased conspicuity benefits compared to a retro-reflective vest or bicycle lighting alone.²² Cyclists wearing a retro-reflective vest with ankle and knee reflective strips were recognised at almost 6 times longer distances than for black clothing and 3 times longer than the vest alone, both in the absence and to a lesser extent the presence of bicycle lights, possibly because the bicycle light acted as a glare source that reduced the relative efficacy of the biomotion strips.²³ Figure 5 illustrates a cyclist wearing leg-mounted retro-reflective strips riding in the same direction as the driver.

Figure 5. Image of a cyclist wearing biomotion strips on the leg joints cycling away from the driver in a suburban street.

Awareness of the night-time conspicuity problem

The level of awareness regarding which strategies are most effective in increasing the conspicuity of vulnerable road users is important in determining whether they are adopted by vulnerable road users themselves. We explored this issue of awareness in a series of closed road studies of both pedestrians and cyclists. In a study by Tyrrell, et al.⁶¹ participants walking in place at the side of a closed road at night while wearing a range of different clothing configurations, were asked to estimate their own conspicuity to an oncoming vehicle driven by an experimenter. These distances were compared with objective measures of pedestrian conspicuity for the same clothing and experimental conditions.²⁹

Pedestrians overestimated their own conspicuity to oncoming vehicles and also failed to appreciate that different clothing configurations affected their conspicuity, greatly underestimating the benefits of biomotion clothing.⁶¹ These findings are important given that pedestrians are more likely to adopt caution when walking or jogging along night-time roads, and to wear clothing incorporating retro-reflective biomotion markings, if they are aware of their limited conspicuity at night and how to improve it.

In a similar study, participants cycled in place at the roadside, while wearing a range of different clothing configurations, and were asked to estimate their own conspicuity to an oncoming vehicle.⁶² Findings demonstrated that similar to pedestrian studies, participants as cyclists, also overestimated their own conspicuity compared to previously measured recognition distances of oncoming drivers,²² and underestimated the conspicuity benefits of retro-reflective markings on their ankles and knees.⁶² Interestingly, participants also incorrectly estimated that a fluorescent vest that did not include retro-reflective material would enhance their night-time conspicuity, which has been clearly demonstrated not to be the case.⁶² While bicyclists report that they are aware of the need to wear high visibility clothing, and of the existence of visibility aids such as retro-reflective vests, few cyclists use them on a regular basis.^51,63

Collectively, these findings suggest that both cyclists and pedestrians misunderstand the extent of the night-time conspicuity problem and the potential value of conspicuity treatments, including retro-reflective materials located on the moveable joints to take advantage of biological motion. It is therefore imperative to better understand and reduce the conspicuity problems faced by vulnerable road users at night to improve their night-time safety. While strategies should include improving the ability of drivers to recognise the presence of vulnerable road users at night and adopt safer driving behaviours, strategies also need to be developed to encourage pedestrians and cyclists to improve their conspicuity on night-time roads.

Strategies to address the night-time conspicuity problem

A long-term goal of research in night-time conspicuity should be to not only provide greater understanding of the night-time visibility issues but to also provide an evidence base on which to develop interventions that improve safety. For example, while research evidence demonstrating how retro-reflective strips in a biomotion configuration can increase the conspicuity of pedestrians and cyclists is important, the problems of reduced conspicuity will not be solved unless vulnerable road users adopt such clothing.

For high visibility safety garments, our findings on the benefits of biological motion on the conspicuity of night-time pedestrians and roadworkers were used to change the Australia/New Zealand Standards “Committee SF-004-03” to incorporate the ‘biomotion’ retro-reflective strip configuration into AS/NZS 4602.1 2011.⁴² The standard was initially adopted by Queensland Department of Transport and Main Roads, along with Vic Rail (Victorian state railway authority), with others following, as manufacturers of high visibility clothing produced work clothing incorporating the ‘biomotion’ configuration. There has also been significant interest from the mining industry for their workers. The International Standard for High Visibility Clothing ISO 201,471:2013 also refers to the ‘biomotion effect’. Collectively, this evidence highlights the positive road safety benefits of our research over many years (also extending to train drivers and potentially mine workers)

The translation of the benefits of biomotion to other vulnerable road users on night-time roads, including recreational walkers and runners, and cyclists has, however, been more challenging. In order to better understand this issue further, we conducted a series of focus groups of individuals who exercise on night-time roads, to explore their beliefs about conspicuity at night-time and the strategies used to increase their night-time conspicuity and safety.^64,65 While participants were aware of the importance of increased conspicuity at night, they were largely unaware of effective strategies to increase their night-time conspicuity and few acknowledged the importance of retro-reflective materials at night.⁶⁴

One of the barriers to wearing clothing incorporating retro-reflective strips was that participants believed that they would look like construction workers which they wished to avoid, with the design (attractiveness and practicality), functionality (comfort and convenience) as well as promotion (beliefs and costs) of the safety benefits of clothing highlighted as important.⁶⁵ Suggested solutions included clothing that incorporated retro-reflective strips that were either thinner than the strips used in standard roadworker clothing, or patterned to incorporate different logos. Participants also suggested education campaigns highlighting the conspicuity problem at night and the conspicuity benefits of biomotion clothing.⁶⁵

In response to the findings from our focus group studies, we have investigated whether thinner or patterned reflective strips provide the same conspicuity benefits as the standard strips (50 mm) used in high visibility clothing, and which have been used in most research studies. Closed road studies where experimenters as pedestrians either walked in place at the roadside, or physically walked across the roadway, demonstrated that thinner^21,66 and patterned strips⁶⁶ were effective in increasing conspicuity, as long as they were positioned in the biomotion configuration. In addition, we demonstrated that these benefits were evident regardless of the orientation of pedestrians and were much more effective than commercially available night-time sports outfits incorporating retro-reflective materials.⁶⁶ Our ongoing work seeks to further explore ways to encourage recreational vulnerable road users to wear these clothing configurations on night-time roads.

Our findings that the visual status of drivers impacts on their capacity to recognise vulnerable road users at night also has important clinical implications in terms of the management of, and the advice given to those patients who drive at night. As outlined in a previous section of this review, low levels of optical blur resulting from either uncorrected or under-corrected refractive errors, as well as low to moderate levels of uncorrected astigmatism, have a significant negative impact on the capacity to recognise vulnerable road users, yet many individuals choose to drive uncorrected. It is therefore important for clinicians to inform patients of the importance of wearing their current refractive correction to optimise their vision for driving at night.

For presbyopic patients, the type of correction can also make a difference, particularly for presbyopic contact lens designs under night-time conditions when the pupil size is larger. The finding that cataracts have detrimental impacts on the recognition of vulnerable road users at night, also suggests the need for early referral for cataract surgery to maximise road safety. Our findings from the studies reported in this review, as well as in other studies of night driving, also suggest that measures of visual function including mesopic visual acuity, contrast sensitivity and motion sensitivity are important for night-time driving visibility and may assist clinicians in better understanding their patients’ visual limitations for night driving.

Apart from making pedestrians and cyclists more visible through clothing or lighting, another option for improving night-time safety as highlighted in our focus group studies, is in educating drivers and pedestrians and cyclists about the dangers of night-time roads, why night driving is challenging, and strategies to address some of these issues. Studies of interventions that educate pedestrians about the fundamental aspects of night vision and the benefits of biomotion clothing have been shown to improve understanding of conspicuity issues and increase interest in conspicuity-enhancing clothing.^67,68 The potential for such interventions to improve the conspicuity of all vulnerable road users at night is important and is an area we are currently exploring through evidence-based studies.

We do, however, acknowledge that the conspicuity of pedestrians and cyclists is not the only determining factor for safety on night-time roads. Drivers and road infrastructure play important roles in ensuring the safety of vulnerable road users, hence our ongoing research is also focused on these research fields, in order to ensure night-time roads are safe for all road users.

Conclusions

Collectively, our findings across a range of experimental designs highlight factors that negatively impact on the conspicuity of pedestrians and cyclists at night. I am very grateful to have had the opportunity to collaborate with so many excellent researchers in the field and very proud of the advances that we have been able to make over many years. I very much look forward to continuing to work in this field in the future, capitalising on existing collaborations as well as new opportunities, with the goal of translation of research into tangible benefits for night-time road safety.

Acknowledgements

The author thanks Laura Bentley, Alex Black, Leo Carney, Michael Collins and Rick Tyrrell for insightful and constructive comments on a draft of this manuscript. The author also acknowledges current funding support from ARC Discovery grant (DP 190103141).

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

* Professor Joanne Wood was awarded the H Barry Collin Research Medal in 2021 which she received at the Australian Vision Convention (hosted by Optometry Queensland/Northern Territory) on 9 April 2022 at the Gold Coast, Queensland, Australia.