U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information|
Publication Number: FHWA-RD-99-079
A Safety Evaluation of UVA Vehicle Headlights
Chapter 1: Literature Review
Nighttime driving is one of the motorist's most difficult tasks. The risks of having a nighttime accident on the road is 2 to 3 times greater than during the daytime. Since the basic difference between night and day driving is the absence of light at night, the increase in the accident rate in periods of darkness can be mainly attributed to poor visibility conditions. During nighttime driving, the visibility distance largely depends upon the availability of artificial light, the source of light being the vehicle headlamps or fixed overhead lighting. High object visibility is an essential characteristic of traffic control devices and a significant factor in highway safety. Researchers have studied this problem, looking for ways of making objects and pedestrians more visible at night. The use of ultraviolet (UVA) headlamps combined with low beams has been suggested as a promising solution to this problem.
The purpose of this literature review was to identify information that might document how UVA headlamps can enhance nighttime visibility. A comprehensive search was made of journals and reports in appropriate subject areas, and a listing was made of articles and reports with potential relevance to this project. Emphasis was placed on the application of UVA technology for highway visibility enhancements. Because this is a recent technology, the amount of existing research was limited. Nevertheless, a number of significant articles and papers have been written on the subject. This section highlights the most relevant sources.
The vehicle's headlamps are by far the main source of illumination available to the driver. The condition and potency of these lamps will directly impact the driver's ability to see objects at night.
The main factors that play a role in nighttime accidents and fatalities were examined by Vanstrum et al. The study revealed that those which predominate at night are alcohol, inadequate visibility, and driver fatigue. Of those factors, the one that can be enhanced through safety engineering is visibility.
Olson et al. studied the effect of headlamps on nighttime visibility. The study concluded that low beams do not provide adequate visibility under many driving conditions. The major limitations to improved headlamp design are due to certain characteristics of the human vision system as well as practical problems associated with headlamp mounting and conditions of use. Limitations arising from the human visual system are related to problems of brightness contrast detection at night, problems of glare, sharpness of central vs. peripheral vision, and the sensitivity of the eye to a broad range of lighting conditions.
Vaswani, for example, concluded that inadequate visibility of road signs and pavement markings at night contributed to wrong-way driving. He described a concept termed the "keg of legibility," which delineates the limits of nighttime visibility under low-beam headlights. The application of the keg-of-legibility concept to the placement of signs, markings, and additional devices that help guide the motorist through the intersection of a four-lane divided highway and another road was discussed.
Ultraviolet light has a shorter wavelength than ordinary light and is invisible to the human eye. However, when ultraviolet light is reflected in certain materials, it is returned on longer wavelengths and becomes visible. This phenomenon, known as fluorescence, makes objects more visible and therefore offers a large potential for improving safety.
Most of the research in the application of UVA headlamps for enhancing highway visibility has been conducted in Europe, particularly by the Swedish National Road Administration under the ARENA Program.
ARENA is a Swedish National Road Administration (SNRA) project that uses field tests to ascertain how road traffic and safety can be improved by advanced technology. The project, based in Gothenburg in 1992, carries out field tests with the cooperation of industry and other organizations. Under the area of traffic safety, a project called UV Light was conducted. UVA headlamps, with emission spectra between 320 nm and 400 nm, and fluorescent road markings were used to create a full light effect in conditions of poor visibility without blinding the oncoming traffic. Pedestrians were seen much more clearly and the path of the roadway could be seen far beyond oncoming vehicles. This work has been documented in various papers.
Barrie described the Swedish trials on UVA vehicle headlamps. The ultraviolet light alone did not enable drivers to see where they were going. However, they were able to highlight anything that contained fluorescent pigments. These included road signs, road markings, and protective clothing. They greatly improved visibility, allowing drivers to identify objects 200 m (656 ft) away, compared with a low-beam range of 50 m (164 ft). They also worked well in fog.
UVA lamps are currently being tested by Saab and Volvo on various cars and a bus. Since 1990, these Swedish car manufacturers have been involved in a joint development company called Ultralux. Work on the dangers and problems of ultraviolet light are being carried out by the Swedish Road and Traffic Research Institute. They have found that UVA beams from headlamps do not constitute any health hazard. By using the filters, all the harmful UVB and UVR rays are eliminated. The remaining UVA intensity is very low compared with normal daylight. Because of the filters, the lamps appear black in the daylight and glow faintly blue when switched on in the dark. Saab, Volvo, and Philips are responsible for the technical development of a viable lamp. In the United States, the Ford Motor Company, in its Contour concept car, has incorporated prototypes of a High-Intensity Discharge (HID) lighting system. This system, scheduled to be in production in about 2 years, also emits UVA light.
In addition to the use of filters, the Swedish Road and Traffic Research Institute has proposed that sensors be placed in the cars to determine when the car is moving at speeds in excess of 48 km/h (30 mi/h). The UVA lighting system will be designed to operate only at speeds greater than 48 km/h (30 mi/h) to minimize the health threat to pedestrians.
Ultralux states that fluorescent road markings can be seen at a distance of 150 m (492 ft) with UVA light, compared with 60 m to 70 m (197 ft to 230 ft) with low beams. The corresponding visibility distance for roadside posts was even better. The posts were visible at more than 200 m (656 ft) with UVA light.
The percentage improvement in safety as a result of greater visibility varies between different studies. According to Road Transport Research, Organization of Economic Cooperation and Development (OECD), the percentage reduction in accidents associated with improved lighting is 20 percent on average. In some cases, the number of accidents is reduced by up to 56 percent.
Behavioral changes at higher speeds and with more aggressive driving is minimal, according to the same source.
Ultralux also found that pedestrians and other unprotected road users can be seen more easily when they are illuminated by UVA lighting. Different clothes, depending on the material and color, have different levels of fluorescence. Jeans could be seen at approximately 100 m (328 ft), for example, while white cotton clothes and synthetic fabrics could be seen at even greater distances. Dark clothes such as black wool, however, were no more visible with UV light than with normal low beams. They also found that washing can improve the fluorescent properties of garments due to the optical whiteners present in many detergents.
The detergents used for washing dishes and clothing generally have a number of additives—for example, bleaches, brighteners, and abrasives. Bleaches whiten fabrics by destroying dirt and colors. Brighteners are chemicals that convert normally invisible ultraviolet light into visible light. Because of the brighteners, additional light reflects back from the fabric, making it seem more vivid, or "whiter."
An important finding in this study was that, when combined with low beams, UVA light gives a unique possibility of increasing the visibility of fluorescent objects on the road without dazzling drivers in oncoming cars. Even fabrics of relatively low fluorescent efficiency, such as jeans, could be detected at a distance of about 100 m (328 ft), even in the presence of dazzling lights from oncoming cars.
When there was no dazzling light to reduce the visibility, detection distances of more than 150 m (492 ft) were achieved, even for clothes with low fluorescent efficiency. Since UV light is invisible to the eye, drivers in oncoming cars did not become dazzled.
Ultralux also studied how motorists experience UV light in traffic. The National Swedish Road Administration has equipped some 100 km (62 mi) of road with fluorescent properties. Some 40 drivers were interviewed after driving cars with UVA lamps on public roads and on test tracks. The results provided an "extremely positive picture of the effect UVA light has on driving in the dark." They found that the visibility of road markings more than doubled using UVA light. The test drivers found that the fluorescent markings improved visibility by more than 40 percent when driving with low beams only. UVA beams in combination with low beams received a higher rating than low beams alone in all cases. The drivers experienced a greater improvement in visibility on freeways and main roads than in city traffic with street lighting.
No fewer than 78 percent of the drivers in the Ultralux survey found that the new road markings could be seen more easily in the daylight.
Ultralux's Peder T. Fast also documented how with supplementary high-beam near-ultraviolet headlamps, fluorescent road markings and clothes could be seen at much greater distances. He utilized the 100-km (62-mi) road equipped by the Swedish National Road Administration with fluorescent road markings and delineator posts for his study. The headlamp system design is presented briefly in his paper, which also described the tests run to evaluate the headlamps. Some of these were on open roads and some were on a closed test circuit. The test methods used and the visibility ratings recorded are presented and discussed.
One of the most comprehensive studies of detection distances to obstacles on the road when using UVA headlights was done by Helmers et al. The study investigated whether low-beam illumination supplemented by UVA headlamps could provide long and safe detection distances. The detection distances were measured in a full-scale simulated opposing situation between two vehicles on a straight, level, two-lane road closed to traffic. The opposing car was stationary in the lane for opposing traffic. The subject's car was driven in the driving lane towards the stationary car. The task of the subject, as well as the driver, was to detect obstacles to the right in the driving lane. Upon detection, they were immediately to push a silent handheld switch. The obstacles were flat plates in the form of a square with a side of 0.4 m (1.3 ft). They were covered with cloth and had three reflectances: black, light gray, and white. The result showed a non-detectable or minor increase in the detection distances for the black and the light gray targets. However, the detection distances for the white targets were twice as long when the ordinary low-beam illumination was supplemented by UVA lighting. The relationship between the reflectance of clothes and the power to emit visible light in UVA light was also studied. The increase in luminance related to the increase in whiteness of the garments was found to be approximately 30 times greater in UVA lighting than in ordinary headlight illumination. Measurements showed a relatively strong positive connection and, furthermore, the ability to emit visible light in UVA lighting increased much more rapidly than the ability to reflect ordinary light when the reflectance of the clothes increased.
A positive effect on the visibility of pedestrians and on road design elements was demonstrated by Staehl et al., who tested two systems developed to enhance visibility during nighttime driving (the Volvo ultraviolet light system and the Jaguar night-vision system). It was apparent that using these systems would give older drivers more confidence when driving at night and should improve both their own safety and that of other vulnerable road users such as pedestrians.
The function and types of application of headlamps were described by Schoon et al. Subsequently, the technical lighting aspects of HID lamps, which have an influence on road safety, were described. Due to the intense light produced by HID lamps and the presence of ultraviolet radiation, two possible areas of application are described: (1) polarized light, and (2) the use of fluorescent material to enhance the visibility of objects. The health hazard presented by UV radiation is not ignored, however. The light emission of a low beam as stipulated by Economic Commission for Europe (ECE) regulation R20 is dealt with extensively. The paper also presents the results of study into HID lamps in the United States and by European research through EUREKA "VEDELIS." It is concluded that: (1) the proposed light emission of the low beam for HID lamps is higher than the current ECE standard for virtually all measurement parameters, and (2) the orientation of the light-beam emission should pay greater attention to the position of vulnerable road users.
The IIT Research Institute conducted a research study to develop a state-of-the-art assessment of current technology to produce fluorescent pigments suitable for use in retroreflective traffic control devices. In the study, Kozak identified materials likely to produce stable, long-life devices and what specific problems must be addressed before any implementation is undertaken. He determined that two distinct classes of materials would satisfy the base criteria for traffic control devices and allow for their use with retroreflectors: organic dyes and lanthanide-doped glasses. Toxicity and carcinogenicity were examined for the selected materials. A model was produced to demonstrate how UVA-activated fluorescent traffic controls might appear to the eye.
Only recently have scientists begun to understand the full effects of ultraviolet light on living organisms. Human exposure to longer wavelength ultraviolet radiation is necessary for the production within the body of vitamin D, a substance that helps promote and maintain proper bone development. Ultraviolet radiation also causes the production of melanin in skin cells, resulting in a suntan. However, scientists have established a strong correlation between exposure to shorter wavelength ultraviolet radiation, genetic mutation in basal skin cells, and skin cancer. It is common today for sunbathers to apply special lotions that block or absorb the high-energy and harmful portions of ultraviolet light.
Fortunately, the shortest wavelengths of ultraviolet radiation are absorbed by gases such as ozone in the Earth's atmosphere. This ionizing radiation has the potential to cause serious tissue damage to many life forms. The effects of chemical pollution on the gases in the ozone layer are being closely monitored in an attempt to preserve these protective layers. The blocking effects of the atmosphere are crucial to the survival of many organisms.
In general, small objects can be seen most clearly when wavelengths of light as small as or smaller than the objects themselves are reflected from them. Since ultraviolet wavelengths are extremely small, objects not readily seen in visible light can be more easily and clearly detected.
The work summarized in the previous sections seems to indicate that ultraviolet lighting has an enormous potential to enhance nighttime driving visibility. The one issue that remains is the safety issue. How harmful is the exposure to ultraviolet radiation (UVR)?
Ultraviolet radiation can produce direct and indirect effects upon the human body. The direct effects are limited to the surface skin because the rays have low penetrating power. Direct effects include sunburn, suntan, and progressive adaptation to heavier doses. Ultraviolet burns can be mild, causing only redness and tenderness, or they can be so severe as to produce blisters, swelling, seepage of fluid, and sloughing of the outer skin. The blood capillaries, which are tiny blood vessels, in the skin dilate with groups of red and white blood cells to produce the red coloration. A suntan occurs when the pigments in cells in the deeper, tissue portion of the skin are activated by ultraviolet radiation, and the cells migrate to the surface of the skin. When these cells die, the pigmentation disappears. The degree of pigmentation is directly related to the length of ultraviolet exposure and the body's inherent ability to produce pigments. Tanning is a body's natural defense to help protect the skin from further injury.
Frequent overexposure to sunlight induces thickening of the skin, more rapid skin aging, and a much higher frequency of skin disorders, including cancer, particularly in persons with fair skin. There is an increase in skin temperature, skin respiration, and skin cholesterol after ultraviolet radiation. Similarly, there is a decrease in pain sensitivity, perspiration, and mineral levels in the body's tissues.
The indirect effects of ultraviolet radiation are, for the most part, caused when the damaged skin cells release histamine, causing swelling. The respiratory tract becomes more vulnerable to bronchitis and pneumonia, and calcified scar tissue may form in the lungs after overexposure to ultraviolet radiation. Histamine stimulates the stomach to produce more secretions and a stronger acid concentration than normal; this, in turn, can lead to inflammation of the stomach lining, or ulcers.
Furthermore, there is usually a fall in blood pressure, but an increase in the quantity of red blood cells, white blood cells, and clotting proteins. There may be loss of weight, an increase in appetite, and a reduction in the respiration rate. Despite all of these harmful effects, ultraviolet radiation is generally not lethal to the human body, but it can kill individual tissue cells and such organisms as bacteria.
Two main papers have studied the safety issue. One evaluated the risks from unintentional exposure to UVR, while the other analyzed the safety aspects of UVA headlamps.
Sliney's work applied to acute exposure to ultraviolet radiation. It should be indicated that the amount of radiation emitted by UVA headlamps is far from acute. Ultraviolet radiation is present in sunlight, so we are all exposed to it. The question then is "how much is safe?" He states that occupational exposure limits (ELs) do exist for acute exposure to ultraviolet radiation, but that they may not be sufficient to ensure total protection from the delayed effects of chronic, repeated exposure to ultraviolet radiation. Sliney states that the development of realistic guidelines for chronic exposure would present a challenge for many reasons. According to Sliney, safety standards for ultraviolet radiation emitted by lamps have not been developed, mainly because they have been used for generations with few serious injuries being reported.
He provides a good summary of ultraviolet lighting's known biological hazards (adverse effects) to the eye and skin, which have been considered in the development of ELs:
Sliney goes on to describe the concept of "threshold," which involves the development of exposure limits below which an adverse effect does not occur. He states that all the aforementioned biological effects, except for carcinogenesis, are acute effects and are also generally considered to have a threshold, although the threshold would vary depending on skin pigmentation and other factors. Finally, Sliney discusses the importance of the geometry of exposure.
Sliney and Fast documented the safety aspects of UVA headlamps using the prototype headlamps of the Swedish program. They stated that most individuals do not experience a strong visual stimulus from the UVA light, unless standing directly in front of the source, so they investigated whether direct exposure to ultraviolet radiant energy was potentially dangerous to the eye or skin, even at close range.
They mention that only one group, the American Conference of Governmental Hygienists (ACGIH), has recommended ELs for visible radiation (i.e., light). The ACGIH refers to its ELs as "Threshold Limit Values," or TLVs, and these are issued yearly, so there's an opportunity for revision. The current ACGIH TLVs to protect the cornea and skin (180 nm to 400 nm) and aphakic retina (310 nm to 760 nm) from UVR have been largely unchanged. The TLVs have an underlying assumption that outdoor environmental exposure to visible radiant energy are normally not hazardous to the eye except in very unusual environments, such as snow fields and deserts.
They state that, at the time of the study, there were no device-specific ultraviolet safety standards for ultraviolet lamps or illumination systems, except for sunlamps and tanning beds. They also provide equations to compute permissible exposure duration.
After studying the results of their testing, they recommended the following safety requirements to limit potentially hazardous exposure:
Sliney and Fast et al. concluded that risk to the eye or skin does not exist. Indeed, the UVR exposures from the UVA prototype headlamp were "virtually the same" as that experienced when exposed to a conventional white-light headlamp.
It was the goal of the research to measure the increased detectability and recognizability of roadway markings and pedestrian conditions using the UVA headlamps. The task was determined to be somewhat an issue of conspicuity. Webster defines the term "conspicuous" as: (1) obvious to the sight or mind, manifest; (2) attracting attention, striking. There are a number of operational definitions of conspicuity. Engel views conspicuity as that combination of properties of a visible object in its background by which it attracts attention. Cole and Jenkins define a conspicuous object as one that will be seen with certainty for a given background within a short observation time (250 msec), regardless of the location of the object in relation to the line of sight. For this particular study, the concept of a short observation time was used to give an indication of the conspicuity of the roadway markings and pedestrian scenes.
The task of detecting pedestrians and roadway markings at night was observed to be more than just an issue of conspicuity. The task of detecting a right curve, for instance, requires inputs from a variety of locations, such as the center lines and edge lines both near and far, as opposed to a single point. The pedestrian scenes were also complex in nature and required a visual search to discern the stimulus. Therefore, it was decided that a search time of greater than 250 msec, as recommended by Cole and Jenkins, would be required. The task would then be somewhat of a tradeoff between conspicuity and visual search.
Some preliminary pilot work was done using 35-mm slides projected from a xenon source in a laboratory setting. Slides of pedestrian scenes similar to the ones used in the field experiment were presented to three subjects using various exposure times. An exposure time of 2 sec was found to produce data in a meaningful range. The exposure time was then used in the pilot testing for the field experiment. Analysis of the pilot data indicated that the data was not clustered at the farthest distance nor close to the stimulus. It was concluded, therefore, that the exposure time of 2 sec would be adequate.
A review of the literature was done to determine if a similar exposure time had been used in previous research. Zwahlen and Schnell conducted a detection and recognition task dealing with fluorescent targets at various peripheral angles. The subjects in the study were seated in a stationary vehicle, and the targets were displayed for 2 sec. Although the study was done during the daytime, the author agreed that the same should work for the nighttime static situation.
The Ford visibility model assumed that 2 sec of visual input was required in a driving situation. The sources of information for the model agreed that driver performance did not change appreciably when a preview time of more than about 2 sec is provided.
Zwahlen and Schnell collected spatial driver eye-scanning behavior and driving speeds along four rural two-lane road test sites under low-beam conditions at night. The study compared the eye-scanning behavior for two conditions, the first with low-visibility temporary pavement markings and the second with new fully restored markings. The 50-percentile preview time ranged from 0.8 sec to 1.2 sec for the first condition and 1.2 sec to 1.8 sec for the second condition.
The existing research showed that the 2 sec had worked in a previous study. The past research also showed that drivers generally use a preview time of about 2 sec when driving at night, which can be related to the shutter exposure time. Based on the pilot testing and the information from past research, a shutter timing of 2 sec was deemed adequate.
Quality and durability are always important aspects in the development of automotive components. Consumer expectations and competitive pressures require reliable and rapid testing to determine end-use performance. One of the most important areas of test development is in exterior durability predictions for polymeric materials.
Fischer used Spearman's rank correlation coefficients to determine the ability of several accelerated testing devices to predict the weathering performance of flexible polymer films used in exterior automotive graphic design. Commonly used cycles in carbon arc and fluorescent ultraviolet-condensation test equipment exhibited generally unacceptable correlation levels for these materials. Three causes for this poor correlation were identified: (1) test device variability can be a significant problem and certain precautions have to be taken to ensure uniform environmental stress application to all test specimens; (2) similar predetermined exposure periods for all samples have proven to be misleading—there is an optimum exposure period for best correlation for each material and each cycle; and (3) for maximum correlation, the stresses induced by the accelerated testing device should be characteristic of the natural environment—the effect on correlation by varying these stresses (UV spectral distribution and condensation-cycle duration) can be significant.
Rokosz discussed two methods for extinguishing a lamp's arc tube when the outer envelope is broken. One method utilized a mechanical switch, and the other method utilized an oxidizable fuse. It was felt that the mechanical switch approach was the more reliable method because it extinguished the arc immediately when the bulb was removed. The design does not depend on the electrical characteristics of the arc tube or ballasts, has no effect on the lamp characteristics, and is probably the least expensive approach.
McNaught et al. evaluated beads under field conditions on longitudinal paint stripes applied by a standard state paint truck and crew. Types tested were the state's standard bead, uniformly graded flotation beads, state standard beads with flotation treatment, and polyester beads with and without anti-ultraviolet light treatment. All were applied at three different rates to both asphalt and portland cement concrete pavements. Performance was rated using a vehicle-mounted photocell device, subjective visual ratings, and close-up photography until substantial failure occurred after 3 months. The uniform gradation flotation beads were found to be brighter than the others tested, having about the same brightness at a 0.5-kg/L (4-lb/gal) application rate as the New York standard beads applied at 0.7 kg/L (6 lb/gal). All other types have less, but essentially similar, brightness. The application rate did not appear to influence brightness to the degree expected, but this was attributed to bead distribution problems caused by a poorly designed dispenser.
Arens et al. assessed the possible improvement in driver recognition of traffic sign colors to be achieved by changing from the Federal Highway Administration (FHWA) highway colors to the American National Standards Institute (ANSI) safety colors. Sign color appearance under daylight and various nighttime viewing conditions was investigated. Although some conditions increased correct color recognition when viewing the ANSI safety colors, these improvements were marginal and applied only to some of the signing material colors and types.
The legibility of retroreflective road markings and, in particular, of signs depends on many parameters, notably the luminance contrast between the message and the background of the sign. For a given level of illumination, the luminance contrast itself depends on the retroreflection coefficients (R') of the materials used. There are a number of retroreflective products that have different values of R'. Colomb et al. studied how variations in R' affect visibility distance (d) at night for a driver at the wheel of his vehicle, the headlights of which illuminate the sign. In the first stage, a small-scale experiment (approximately 1/10th scale) was conducted in the laboratory to identify the variables. Legibility thresholds were determined by presenting different combinations of the alphabets and colors used on road signs to observers at different luminance contrast levels. In the second stage, working from these results, simulations were carried out to quantify the influence of the various parameters, in particular R' and d. It was found that R' is only one of the parameters that affect d. Most often, multiplying R' by 3 increases d by only 30 percent. But the dimensions of the letters used for the messages play a preponderant and limiting role because of their direct relationship to visual acuity. According to earlier experiments, other parameters that are harder to quantify (incident illumination of the signs by vehicle headlamps; dirt on the headlamps, signs, and windscreens; and weather conditions) are important and may require a correction of d ranging from 0 to 100 percent.
The majority of the studies and research described above indicate that UVA headlamps can considerably increase detection distances, particularly those to pedestrians, compared with the use of only ordinary low beams. Even with clothes of relatively low fluorescent efficiency, the detection distance may increase from 50 m (164 ft) to approximately 100 m (328 ft) in the presence of glare from oncoming cars, as demonstrated by Fast et al.
Those findings lead to the conclusion that the use of UVA headlamps in automobiles may significantly increase highway safety by increasing detection distances, even in the presence of glare from oncoming cars.