Click for next page ( 18


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 17
17 the largest angle at which they could detect the warning lights Data Analysis using their peripheral vision. During the nighttime session, sub- jects again rated the conspicuity of the warning lights. Subjects The data analysis was undertaken in four parts: analysis of were next asked to identify what function a vehicle might be the halogen, LED and strobe panel lights; analysis of the rotat- performing with the viewed warning-light pattern. Once all ing beacons; analysis of passive retroreflectivity; and analysis 41 nighttime conditions were seen and rated, the vehicle was of the position of the beacons with respect to the daylight. moved closer to the truck containing the lighting panel, and The four dependent (measured) variables of daytime participants were asked to rate each condition according to a attention-getting, daytime peripheral detection angle, night- discomfort glare scale. time attention-getting, and nighttime glare were considered in It should be noted that attention-getting ratings were taken each of these analyses; the other variable, recognition scale, was 400 ft away from the lights during both daytime and night- considered separately. For the lamp type factor, the halogen, time sessions, while the glare ratings were taken at 150 ft from LED, and strobe sources could not be directly compared be- the lighting panel during the nighttime sessions. The test truck cause the halogen and LED were each presented to only half of was positioned at the top of a hill during the daytime session the participants; lamp types were compared in pairs: halogen to support the "daylighting" tests and at the bottom of the hill and LED, strobe and halogen, and LED and strobe. during the nighttime session to maximize the dark background and to minimize distraction to passing vehicles. Figure 3 shows Summary of Results the physical arrangement of the test vehicles for the daytime and Lamp Type, Intensity, nighttime sessions. and Color Comparisons Half of the participants completed sessions that used LED panel-mounted lights, while the other half viewed halogen The first factors considered in these analyses were lamp panel-mounted lights. The division was necessary owing to type, intensity, and light color. The results are summarized limited time and resources. Upon completion of testing, sub- below, with only the significant findings discussed: jects were asked to provide comments on the warning lights, and then were paid at a rate of $20/h. Peripheral Detection Angle Overall, the amber and the white light sources typically performed equivalently for detection with peripheral vision; the red and the blue performed worse. The paired comparisons that included a strobe (i.e., strobe/halogen and strobe/LED comparisons) performed better than the halogen/LED comparison. Experimental Vehicle Location The lamp-type-by-color interaction was significant for the halogen and LED comparison. The halogen light sources performed significantly worse than the LEDs for all colors except white (in which halogen and LED per- formed virtually identically). In this comparison, the red LED was also the best performer for peripheral detection Participant Vehicle Location for angle, which is different than what was found in all other glare portion of the nighttime protocol relationships, where amber and white performed better. The purity of the red LED source may have resulted in a stronger response than for the red halogen source. Attention-Getting Rating There was a statistically significant intensity difference for both the daytime and nighttime attention-getting rating for the halogen and strobe comparison. This dif- ference showed that a higher effective intensity results in Participant Vehicle Location a significantly higher attention-getting rating. for the daytime and nighttime non-glare protocol There was a significant lamp-type-by-color interaction in the daytime, with strobes having a higher attention- getting rating in all colors except for the red LED. This Figure 3. Testing area layout for nighttime result could be due to two factors: the LED configuration and daytime testing. was a steady light source as opposed to a flashing strobe,

OCR for page 17
18 and the red LED configuration may thus have appeared Other comparisons indicate that the flashing behavior to be similar to vehicle brake lights. of the lights and the number of sources are more im- The nighttime attention-getting rating of the light sources portant than the effective intensity of the sources, with was higher than the daytime rating for the same lights. flashing being better than steady. The asynchronous pat- Discomfort Glare Rating tern also seems to provide an additional benefit over the Like the attention-getting scale, only intensity was sig- synchronous condition. nificant for the halogen and strobe comparison, with a Attention-Getting Rating higher effective intensity producing a higher glare rating. The daytime attention-getting rating results showed that It was also noted that a higher attention-getting rating the flash patterns have a higher attention-getting rat- also corresponds to a higher glare rating (in other words, ing than the steady condition. However, there seems to it is difficult to find a light source that is more attention- be very little difference between the flash patterns and getting without also having higher glare). frequencies. Summary Higher intensities resulted in a higher attention-getting Through all of these comparisons of the halogen and LED rating, regardless of flash pattern. sources, only peripheral detection angle showed a differ- Glare Rating ence between these two light sources. In this case, the color The steady condition had a higher glare rating, indicating was the most significant, with the amber and the white that a greater amount of glare was experienced by the light sources performing better than the blue and red. observer. However, as seen in the previous compari- For attention-getting and glare, the most critical aspect son, it was also the worst condition for attention-getting. of these comparisons is the effective intensity of the light Therefore, it seems that the flashing light provides a source. However, these results must be balanced because way of getting attention without causing a higher glare a higher effective intensity provides a higher attention- experience for the observer. getting rating, but also causes a higher glare rating. The characteristics of the flash pattern seem to be rel- One aspect that is worth noting is that the LED and atively unimportant as there was no statistical differ- halogen comparisons to the strobe compared the steady ence among the asynchronous and synchronous flash halogen or LED systems to the flashing strobe systems. patterns. This aspect is an artifact of the research design. However, the strobe and the flash characteristics were investigated Strobe Lights in two other comparisons. For the strobe lamps, two analyses were conducted. One analysis compared the results with different flash patterns (dou- Flash Characteristics ble versus quad flash) and intensities. The other analysis con- Other comparisons investigated the impact of the flashing cerned the color and the intensity of the strobe lamps, for which pattern on the measured results. For this analysis, only the only the strobe panel lighting configurations were used. amber panel lighting configurations were used to investigate each of the dependent variables. There were no significant Flash Patterns and Intensity findings for the daytime versus nighttime attention-getting Only the amber lamps were used in this analysis. None of ratings. the four dependent variables (daytime attention-getting rating, nighttime attention-getting rating, peripheral de- Peripheral Detection Angle tection angle, and glare rating) showed significant results There was a significant four-way interaction for abil- for flash pattern or intensity. This outcome suggests that ity to detect the flash patterns in the peripheral vision. the dual-flash versus quad-flash characteristics of the This interaction is difficult to interpret because of its light source are not significant considerations in the complexity. In general, a lower effective intensity re- requirements for lighting. sulted in a worse detection angle; the LED source per- Color and Intensity formed better than the halogen source for older drivers, Only the dual flash patterns were used in this analysis. but was not significantly better for younger drivers; The results for the four dependent variables (daytime the steady condition had worse results than the flash- attention-getting rating, nighttime attention-getting rat- ing conditions; the higher frequency flashing seems to ing, peripheral detection angle, and glare rating) showed be just slightly worse than the lower frequency condi- that the color is significant for peripheral detection, as well tion; and the synchronous flash was slightly better than as for glare, and that intensity is significant for daytime the asynchronous. attention-getting.

OCR for page 17
19 For peripheral detection, both amber and white lighting during the nighttime, but did not rise to the level of the provided better performance than either red or blue. lighted conditions. For glare, amber and white lighting had a greater glare Glare rating than the red and blue systems. As expected, the glare from the retroreflective tape was For daytime attention-getting, a higher effective intensity minimal as compared to the light sources. results in a higher attention-getting rating. Summary The results of the passive condition showed that there must be internally illuminated sources in order to max- Rotating Beacons imize the attention-getting and the peripheral detection Two analyses were performed with the rotating beacons. One factors. These sources will increase the glare experienced analysis concerned the speed and effective intensity of the ro- by the observer but the resulting increase in conspicuity tating beacon, and the other analysis concerned the speed and is likely justified. the color of the beacons, for which only the rotating beacon The passive tape provided an additional impact during the configurations were used. nighttime condition and can be a suitable supplement to provide an additional source of nighttime visibility. Speed and Intensity However, it should not be relied upon as a sole source of Only the amber color results were used for this analysis. warning information. Neither speed nor intensity was significant for any of the four dependent variables (daytime attention-getting rat- Beacon Type and Position ing, nighttime attention-getting rating, peripheral de- One of the issues with a beacon-type source is the daylight tection angle, and glare rating). This outcome suggests infringement behind the light. As mentioned, this issue was that the effective intensity and speed of rotating beacons tested by placing a series of beacons either on the top of the do not influence the human response to them. test platform or on a shelf with the test platform behind the Speed and Color light source. The results of this test were analyzed using lamp These results show that color is significant for peripheral type and location. For this analysis, the sources were all in detection, and amber and white lighting exhibit better beacon format and were all amber in color. performance than red and blue lighting. Analyses also showed that the speed was significant for Attention-Getting Rating peripheral detection. In this case, the slower beacon pro- The daytime attention-getting rating was lower than the vides a larger (better) peripheral detection angle than nighttime, which is consistent with previous results. the faster beacon. This outcome is likely related to the In terms of lamp type, the LED and rotator sources did duration of the flash. not differ in attention-getting, but the strobe performed worse than either of them. It is likely that these results are Passive related to the light characteristics of the sources. The ro- tating beacons provided a higher intensity over time than In this analysis, the passive treatment (retroreflective tape) the strobes and the LEDs. Similarly, the LEDs provided was compared to the panel lighting condition. For this com- higher color purity than the other two sources. parison, the low-intensity steady condition for each color was Peripheral Detection Angle compared to the passive tape condition. The lower shelf location provided a higher performance than the location on the top of the test panel. The location Peripheral Detection Angle result shows that less light is lost to the sky when there is a The performance of the retroreflective tape was signifi- backing behind the light. It should be noted that the test cantly lower than that of the other conditions for pe- panel in this configuration was black and that a different ripheral detection. This outcome was to be expected be- background color may change the impact of the location. cause retroreflective tape relies on the headlamps of an For lamp type, rotating beacons provided the best per- approaching vehicle for its luminance, and the Peripheral formance and the strobe and the LED provided the Detection test was a daytime evaluation of the lighting worse but similar performance. The rotating beacon per- conditions. formance is likely related to the higher time-averaged Attention-Getting Rating intensity provided by this configuration. As with all of the lighting conditions tested, the nighttime Glare condition showed a higher rating than the daytime con- The rotating beacon and the LED beacon performed dition. The passive condition was significantly increased equivalently, while the strobe had a lower glare rating--

OCR for page 17
20 a similar grouping to that found with the attention- Attention-Getting Ratings getting rating. As indicated before, the higher attention- There was a positive correlation between daytime getting sources also have a higher glare rating. attention-getting and effective intensity for all colors Summary and lamp types. Amber lights were typically rated higher The results of this test show that having the beacon appear regardless of effective intensity, with halogen amber against a background is important to the detectability of lights being rated the best. It appears that the relation- the source. The background provides a consistent contrast ship of the rating to the effective intensity of the light for the light source and therefore higher and predictable source is a linearlog relationship, suggesting that the performance. However, this option must be traded off gains in attention-getting diminish with brighter light against the alternative of having more beacons so that sources. This relationship is typical for most human they can be seen from the front of the vehicle as well responses to light sources. as the back. There was also a positive correlation between nighttime attention-getting rating and effective intensity that is similar to, but smaller than, that between daytime Vehicle Recognition Results attention-getting rating and effective intensity. Ratings The vehicle recognition questionnaire was used to identify of low-effective-intensity lights were much higher at any lighting patterns that might resemble a standard pattern night because of the high contrast of the lights and their not used for the maintenance vehicles that were part of this background. Ratings of relatively high-effective-intensity investigation. In other words, what lighting patterns had some lights did not increase as much, likely because of ceiling sort of intrinsic meaning for the viewer, such as "This must effects of the rating scale. In this scenario, the blue LEDs be a law enforcement vehicle"? performed well for conspicuity even though they were among the least intense. The ratings of the white LEDs also increased dramatically during nighttime trials. The Color ratings of the amber halogens, which were among the Blue was predominantly recognized as law enforcement. best in the daytime, remained relatively unchanged at This observation was to be expected as blue is the light- nighttime. ing color used on law enforcement vehicles in the Peripheral Detection Angle Commonwealth of Virginia. Strobes performed much better for the peripheral detec- Amber and white were predominantly associated with tion angle than other lights with similar effective inten- maintenance and construction, including towing. sity, likely because of the strobes' relatively fast flashing Red was associated with medical and fire response. patterns. Overall Lighting Aspects Amber strobes provided the best performance, better than The second aspect of the vehicle recognition question- the amber LEDs that were more than twice as intense. The naire was the identification of which of the lighting as- linearlog relationship of the peripheral detection angle pects are important to the identification. Color was to the photometric effective intensity was more dramatic the predominant response, followed by flash cycle. The in this comparison. light position was not considered to be important by Discomfort Glare participants; however, the lighting configurations were Blue LEDs and white LEDs performed poorly on the dis- predominantly in the same location throughout the in- comfort glare scale, even though they had low effective vestigation. Effective intensity also did not seem to be intensity. The same linearlog relationship exists in this important in the vehicle recognition. comparison. Photometric Limits Photometric Comparison The photometric levels required for the warning lights on the vehicles can be established based on the relation- Comparing the results of the photometric testing with that ship of the rating scales to the photometric measure- of the static testing provides some insight into the required ments. The glare rating would serve as the upper limit limits for lighting on maintenance vehicles. The Form Factor of the specification for the effective intensity because method was shown to be the best method of photometric the relationship shows that the higher the effective inten- testing for effective intensity for the purpose of this study. Each sity, the greater the glare. The attention-getting rating of the individual static testing metrics was compared to the would then be used as the minimum level because it is Form Factor results of the light sources. These comparisons a target that must be surpassed to provide adequate were only made for the panel lights. conspicuity of the light sources.