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-HRT-04-143
Date: December 2005
Enhanced Night Visibility Series, Volume XII: Overview of Phase II and Development of Phase III Experimental Plan
PDF Version (631 KB)
PDF files can be viewed with the Acrobat® Reader®
CHAPTER 3—PHASE III DEVELOPMENT
Not one of the vision enhancement system configurations evaluated in the Phase II studies was clearly beneficial across all of the conditions tested; therefore, at this stage of the project, an onroad field study using a given configuration would have been premature. As a result, the implementation portion of the Phase I work plan (activity 5) was eliminated along with the following activities:
Additional Smart Road testing was recommended rather than conducting an onroad field study. Ongoing changes in night vision enhancement technology presented new opportunities to gain valuable information from expanded testing focusing on comparisons between conditions that were tested in Phase II and new VESs. Thus, the Phase III work plan was developed with the primary objective of improving visibility of the road environment. The Phase III work plan added the following four activities, detailed in this chapter and labeled 6 through 9 for consistency, to replace the eliminated tasks from the revised statement of work:
During the last 10 years, significant advancements have been made in new headlamp technologies that provide greater visibility than traditional halogen headlamps. This was evident in the results of Phase II of this research effort. Some other advantages of these newer headlamps (e.g., HID) include a greater beam-spread, which may not only use the available light more efficiently but may also increase the visibility of objects in the roadway periphery. Some disadvantages of these lights may include discomfort and disability glare effects for oncoming drivers. Empirical studies designed to investigate possible advantages and disadvantages of headlight technology often do not occur until after the technology has appeared on U.S. roadways. A more proactive approach involving communication between researchers and car manufacturers is needed to initiate testing on what may be available in the near future.
It was proposed that up to three new VESs—differing with respect to technology, spectrum, and beam pattern—be tested as part of the protocols suggested in activities 8 and 9. The experimental designs used in the Phase II Smart Road studies had sufficient flexibility to allow these technologies to be evaluated and compared to other technologies tested in Phase II. The goal was an advanced evaluation of technologies that automotive manufacturers were considering for implementation in the near future, as well as a better understanding of their possible advantages (visibility of objects in the periphery) and disadvantages (disability glare effects).
In the first portion of the investigation, the contractor identified and contacted automotive manufacturers and headlamp suppliers who had innovative headlamp technologies intended for market distribution in the near future. The criteria for selection of the new technologies were the following: (1) the new technologies should be different from the technologies that had already been tested, and (2) these new technologies should be testable at the Smart Road testing facility.
Three HID headlamps were selected for inclusion in the disability glare study (activity 8), and two of these headlamps were selected for object detection and recognition testing in clear weather as well as for potential off-axis benefits (activity 9).
The research in Phase II aided in the understanding of VESs such as UV–A, HLB, HID, and hybrid headlamps as well as other technologies including IR–TIS. The IR–TIS showed significant benefits in Phase II for detecting pedestrians in clear weather conditions. Recall that IR–TIS uses the difference between the thermal signature of objects and that of the surrounding driving environment to aid in object detection. Several OEMs and suppliers are developing IR–TISs as well as near IR (i.e., active IR) technologies, which both have potential to greatly improve visibility during nighttime driving. The newer IR–TISs may be more sensitive to temperature differences, making it possible to detect and identify more objects (e.g., pedestrians) or to detect objects at greater distance. Several OEMs and suppliers are also developing near IR to provide a more detailed view of the driving environment including lane markings. These systems use IR emitters to act similarly to headlamps when viewed through the IR camera and its associated display. Unlike IR–TISs, near IR systems show many details of the forward roadway scene such as headlamp light, pavement markings, and signs. Because IR will not generate glare, near IR systems have the potential to increase visibility distances beyond those of conventional headlamps without negative effects for oncoming traffic.
Using methodology similar to the headlamp search described in activity 6, the researchers contacted OEMs and suppliers to determine which systems were going to be available in the near future. Three IR systems were selected for additional testing: two prototype near IR systems and the same IR–TIS from the Phase II studies. At the time, no new IR–TISs could be obtained. All three systems obtained were to be placed on SUVs.
The testing methodology for this activity was designed to determine if the evolving IR technologies further improve detection and recognition of objects in the roadway and what could be done to the roadway infrastructure to provide the greatest possible integration and benefit from these systems. To determine possible detection and recognition benefits, the IR technologies were to be tested using a methodology similar to that used in Phase II. Development work was required to determine how to provide infrastructure that would benefit from these systems. Roadway infrastructure components designed for integration with IR VESs do not currently exist; however, as IR systems become more prevalent in the marketplace, these components could be designed to increase driving safety by providing more information from the infrastructure to the driver. For example, heat-retaining roadway delineators might be visible at greater distances, potentially improving the ability of the driver to reconcile the view through the IR system (enhanced view) with the forward scene from the windshield. Materials such as route management signage, temporary road markings, and safety vests should be capable of reflecting emitted IR as well as visible light to ensure that information remains conspicuous in both formats. In critical roadway sections (e.g., crosswalks, complex intersections, and roadway areas during incident management), roadway IR emitters could be used to increase conspicuity.
This activity devoted a great deal of effort to investigating infrastructure alternatives both by designing potential prototypes and contacting suppliers. After substantial pilot testing, it was determined that resources would be better spent determining how well existing infrastructure interacted with IR systems. During the pilot testing, it was determined that the near IR appeared to have some potential problems with blooming when exposed to certain road signs. On the other hand, it also showed promise in providing drivers forewarning of a traffic sign. As a result, it was determined that assessing existing signage and road markings would provide the most benefit because this infrastructure is already in place and will likely remain for a significant period of time.
During the pilot testing, it appeared there was sufficient time to include activity 9 (off-axis testing) in this study using the same test participants. The purpose of activity 9 is described in more detail in the activity 9 section. Combining these studies would provide several advantages:
A methodology similar to what was used in the Phase II object detection and recognition studies was planned for this activity. Initially, a clear-condition study was planned using a 6 (VES) by 3 (Age) by 17 (Object) mixed-factor design. VES, a within-subject factor, was to include the HLB headlamp used in the Phase II studies, the three IR systems (i.e., one IR–TIS and two near IRs), and two HIDs. Age was to be the only between-subjects factor. Phase III was also to use the same gender-balanced age group criteria used in the Phase II visual performance studies. For the objects within-subject factor, a total of 17 different objects were to be presented, including signs, directional markings, and some of the objects used in Phase II (e.g., pedestrian dressed in black and the tire tread). Phase III was also to use pedestrians in curves and off-axis positions. After completion of the clear study, a rain study was to be conducted with a subset of the objects to determine the merit of the Phase III systems in rain.
Public concern and press coverage about glare associated with new headlamps has been an increasingly prevalent topic in recent years, especially since the introduction of HID headlights.(4) HID headlamps provide more luminous flux than conventional HLB headlamps. This trait has made them excellent candidates for vehicular applications, and they have already been implemented as standard components in some automobiles; however, limited research exists on the possible negative effects of these headlights on the vision of oncoming drivers.
Public opinion about the glare problem has revolved around drivers’ perceived increase in discomfort glare when approaching a vehicle equipped with HIDs. Although driver comfort is very important and may ultimately decide whether or not a new technology is universally adopted, disability glare is more likely to affect safety.
Disability glare is a result of light scattering in the ocular media. Light from a glare source, such as the headlights of an oncoming vehicle, enters the eye and scatters, creating a uniform luminance, or veiling luminance, over the retina. Regardless of whether an object is brighter or darker than its background, veiling luminance will decrease the contrast of the object. As a result, the object is less likely to be seen.
Recall that Phase II of the ENV project included a discomfort glare evaluation of 11 different headlamp configurations, including HLB, HID, and UV–A headlamps (ENV Volume VII). The primary focus of this Phase II study was on rating the discomfort glare of UV–A as compared to other VESs. However, it is difficult to fully understand the effects of these VESs on safety without a direct disability glare evaluation. The two types of glare have different physiological origins, and factors that affect one type often do not affect the other;(5) therefore, a disability glare evaluation in combination with a discomfort glare evaluation was needed to determine what effect the newer headlight technologies have on oncoming drivers. As part of this evaluation, a literature review was to be conducted and included in the report.
The disability glare study was planned as a 5 (VES) by 2 (Driver’s Adaptation Level) by 2 (Pedestrian Location) by 3 (Age) mixed-factor design. As mentioned in activity 6, three additional HID headlamps were selected for comparison to the baseline HLB and HID headlamps used in the Phase II studies. This provided the following headlamp intensities and patterns of oncoming glare:
Driver age was the only between-subjects variable. It included the same three gender-balanced age ranges used in the majority of Phase II studies: a younger group (18 to 25 years old), a middle-aged group (40 to 50 years old), and an older group (65 years and older).
At night, a driver’s eye will adapt to the ambient lighting condition. This adaptation level will change the ability of the driver to perceive objects as well as the driver’s glare sensitivity. For this study, driver light adaptation level was a within-subjects variable including a low level of 0.15 lux (lx) and a high level of 0.45 lx. A dimmable light source inside the vehicle (on top of the instrumentation panel) was to allow experimenters to control the driver’s light adaptation level.
Pedestrian location was also a within-subjects variable. The location of pedestrians in the roadway significantly affects their visibility to drivers in the presence of glare. Two locations were chosen for this study, one near the centerline and the other near the right edgeline. Both locations were set 15.2 m (50 ft) behind the oncoming glare headlamps, and both pedestrians were to wear white clothing and stand facing the glare vehicle.
The human visual system consists of two types of photoreceptors: rods and cones. These two photoreceptor types have different characteristics and different visual functions. Cones, mostly located in the fovea of the eye, are sensitive to higher (photopic) lighting conditions. They provide the finest detail of visual acuity as well as perception of color. The periphery of the eye has very few cones. Rods, entirely located in the periphery of the eye, are sensitive during lower (scotopic) lighting conditions. They provide most of the detection of motion and objects but very little acuity. At the extremes of lighting conditions (photopic and scotopic), either rods or cones are active; but at the lighting levels used for roadways, both rods and cones are active, which is called mesopic vision. Because the sensitivity and effectiveness of the peripheral visual field is affected by the photoreceptor in use, the adapted luminance level influences the ability to perceive objects. Typically, the lower the adaptation luminance level the lower the visibility of objects.
Another aspect of rod photoreceptors is their difference in spectral sensitivity. Cone sensitivity is characterized by the photopic sensitivity function, shown in a bell-shaped sensitivity curve peaking at 555 nanometers (nm). Rods sensitivity is characterized by the scotopic sensitivity function, a bell-shaped curve peaking at 507 nm. This means that at night, the peak eye sensitivity changes from green toward blue colors.
HID lamps have a greater blue spectral component, which is more closely related to the scotopic sensitivity of the human peripheral visual field. Some HID headlamps also have a wider beam-spread than conventional halogen headlamps. These characteristics can effectively increase the visibility of objects that are eccentric to the drivers’ line of sight. This increased visibility creates a potential safety benefit by allowing earlier detection of pedestrians, animals, and other objects that could enter the roadway.
While the Phase II studies evaluated detection and recognition of objects in the roadway including the edgelines, this portion of the Phase III testing was intended to focus on establishing the benefit of headlamps with wider beam-spreads for the visibility of pedestrians beyond the roadway edge. This testing was to be accomplished by presenting off-axis objects to drivers and recording their visual performance by measuring detection distance (as previously defined for Phase II studies). Recall that this study was to be conducted in activity 7, the IR evaluation, in which pedestrians were to be positioned 9.5 m (31 ft) to the left or the right of the travel lane. Pedestrians were also to be positioned either to the left or right of the road in a curve to the left or a curve to the right. These pedestrians were to be included in the counterbalance of the 17 objects tested in activity 7 studies to avoid having participants anticipate the next-appearing object.