December 15, Public Interest Finding For Prismatic Retroreflective Sheeting (3M DG3 Sheeting) Dear Division Administrator:

Transcription

1 December 5, 2006 Division Administrator Federal Highway Administration Galtier Plaza, Box th St. East, Suite 500 St. Paul, Minnesota Re: Public Interest Finding For Prismatic Retroreflective Sheeting (3M DG3 Sheeting) Dear Division Administrator: We are requesting written concurrence from your agency in our decision to use a specification that results in a sole source for prismatic retroreflective sheeting material (3M Diamond Grade DG 3 sheeting) for all guide signs, chevrons, markers and delineators statewide based on the detailed evaluation included in this letter. Whereas the information given in this letter shows the benefits to the public of using DG 3 sheeting material in the aforementioned applications, we have reservations about the applicability of 23 CFR to this Public Interest Finding (PIF). There is, in our understanding, a difference between a sole source and a proprietary product. Federal law specifically targets proprietary products for competitive bidding requirements. While the manufacturing processes and the optical construction of DG 3 may be proprietary and, therefore, cannot be duplicated by others (similar to many other sheeting materials), the performance level attained by the DG 3 sheeting is not proprietary. Other suppliers are free to provide materials with comparable performance if they choose to develop them. Therefore, a unique product such as DG 3 that meets a non-proprietary performance level should be considered as a sole source product, rather than a proprietary product. Being the only commercially available product in a given ASTM Type classification does not mean a product is proprietary; ASTM types have always been developed without control over how many products may be available within a given ASTM Type classification. Patents on product characteristics and manufacturing processes do not define proprietary products; all sheeting materials have patents behind them one way or another. Last but not least, being the first product in the marketplace to achieve a certain performance level does not make the product proprietary. Our free enterprise system promotes competition for every innovative material that proves to be of value in the market. In our view, DG 3 provides a unique and unprecedented performance which makes it a sole source

2 product. It is not proprietary since there is no restriction preventing other products from being developed that would offer comparable performance. Nonetheless, the specification of a product with this unprecedented performance level has been deemed proprietary. Demonstrating a unique performance level then becomes the pretext for this PIF, which, ironically, is also what this PIF intends to prove. In our opinion, and for the sake of innovation, any manufacturer that can attain higher levels of performance of a product should not be discouraged from doing so. It is in our interest to serve the driving public with the best performing materials that promote public safety. There is no patent in existence that prohibits any sheeting manufacturer from attaining the performance of DG 3, or even producing a better performing sheeting material, although the manufacturing processes and the optics to achieve that performance level have to be patented for the innovation to survive. There are unlimited ways of achieving retroreflectivity levels equal to or exceeding that of DG 3, while any patents pertaining to DG 3 claim only one of the ways to achieve that performance level. Many existing sheeting materials have patents that protect them from duplication in a similar manner. In fact, there are patents issued to competitors of 3M covering other designs for high efficiency retroreflective sheetings that, if followed, could result in sheeting with comparable performance. Manufacturing and marketing these products is completely at their discretion. Over the years many innovative sheetings (ASTM Type I, Type III and Type IX, for examples) were originally offered by a single supplier and erroneously called proprietary. The fact that other suppliers now offer comparable products belies this label. In our best understanding, DG 3 is not a proprietary product but a sole source product and is not well-suited insofar as the proprietary scope of 23 CFR is considered. Nonetheless, this PIF outlines our understanding of the superior performance of the DG 3 sign sheeting material in our justification to use DG 3 sheeting for the benefit of the road user. We have met with industry experts to learn about the unique performance characteristics of the new DG 3 sheeting as well as currently available truncated-cube design prismatic sheeting materials. We have also reviewed and analyzed all published documents comparing these sheeting materials, including the recently distributed FHWA PIF guidance document to obtain supporting information to reinforce our proposal to use DG 3 sheeting. We found that discrepancies exist in some documents in the analysis of the unique performance characteristics of these prismatic sheeting materials and in the conclusions reached as a result of these analyses. Following is the engineering/economic analysis justifying our request to use DG 3 sheeting in lieu of all other available prismatic sheeting materials. 2

3 DESCRIPTION OF NEED According to the USDOT FHWA Older Drivers Design Handbook, the older driver is considered to be the design driver of the 2 st Century. Older drivers, due to their diminished visual, cognitive and motor capabilities, need signs that are larger and brighter, for both conspicuity and legibility. Since increasing the size of signs would likely require modification or replacement of sign structures statewide, and this is not a cost effective measure, we prefer to have brighter signs. We intend to use a sheeting material that best serves our entire driver population day and night, including older drivers and truck drivers that would be underserved by lesser performing sheeting materials. FHWA s Guidelines and Recommendations to Accommodate Older Drivers and Pedestrians specifically recommends high retroreflectance overall, and particularly at the widest available observation angles, be used to provide increased sign conspicuity and legibility for older drivers. Based on the computer model analysis given in the FHWA PIF guidance document and additional computer model analysis of DG 3 vs. various other available sign sheetings (see Figure on page 9, and Figure 2 on page 20), it is clear that DG 3 provides the best observation angle capability, while maintaining long distance performance for higher conspicuity, unlike any other sheeting material. Following the FHWA recommendation, we intend to use DG 3 sheeting to accommodate as many older drivers as possible. The percent accommodation figures are provided in the Engineering / Economic Analysis Supporting the Requested Action section. Table. Percentage of Older Drivers in Various States State MN AZ IL MD VA Percent Drivers above the age of % 26.9% Many of our roadways are heavily used by commercial trucks at night. These drivers rely extensively on traffic signs, as many of them are unfamiliar with the roadways they travel. We would like to use the sheeting material that best serves truck drivers, who view signs at large observation angles. Also as a general trend, trucks are increasingly using newer harmonized VOA type headlight systems, which reduce the available light to most traffic signs. The supplied computer analysis indicates that DG 3 serves the highest segment of this population. Along with the suggestion in the FHWA PIF guidance document, we developed the following table to summarize the percentage of truck drivers at night. Truck traffic is projected to increase in the future, and nearly double the 998 levels nationwide by the year 2020 []. Therefore, their safety is a main concern to us. Although their percentages may be small, a crash involving a truck is much more costly to society than most other crashes. From an economical perspective, the percentage of truck drivers served only by DG 3 should be considered far beyond its face value. 3

4 Table 2. Truck Traffic (VMT) as a Percentage of Total Traffic (VMT) in Various States State MN AZ IL MD VA Percent Truck Vehicle Miles Traveled by Trucks per Day 8,263,000 2,499,000 There are over 00,000 guide signs on the state trunk highway system (including interstate highways) in Minnesota that we plan to upgrade with DG 3 sheeting. On a daily basis, these signs, in aggregate, are viewed millions of times by the hundreds of thousands of drivers traveling statewide. The number of times these signs are viewed by the percentile of drivers served only by DG 3 is expected to be in tens of thousands of viewing during nighttime hours. Although a 5 percent difference in percent drivers served may seem to be a small difference, when the number of looks at all guide signs is considered in aggregate, a 5 percent difference in number of drivers will translate to millions of occurrences of sign viewings. Our intent is to improve the performance of our signs to achieve as many successful and safe sign viewings as possible. The innovative DG 3 prismatic retroreflective sheeting material has a 58 percent active retroreflective surface area (the only sheeting material that achieves a light return efficiency this high), which is almost twice that of the current truncated-cube design prismatic sheetings which have approximately 32 percent active retroreflective surface area. Neither ASTM Types VII, VIII or X sheeting materials (designed to best work at low entrance and observation angles), nor ASTM Type IX sheeting materials (better legibility at closer distances, higher entrance and observation angles) match the luminance performance curves of DG 3 (see Figure through Figure 4). DG 3 makes the best use of available headlight illumination, which is known to be gradually decreasing with the use of VOA headlamps, and will help us achieve our objective of serving as many drivers as possible. ENGINEERING / ECONOMIC ANALYSIS SUPPORTING THE REQUESTED ACTION We support the percent of drivers served concept for individual sheeting materials in establishing the overall cost-benefit analysis of sheeting materials. We also want to emphasize the concept of percent drivers not served, because those are the drivers likely to be involved in a crash, or miss a turn, and increase overall system cost to society. In terms of percent drivers not served, the FHWA PIF guidance document estimates that DG 3 provides substantial gains, i.e. 4 percent not served by Type IX (Avery Dennison Omniview or 3M VIP) vs. 0 percent not served by DG 3. Thus, DG 3 reduces the percent drivers not served by approximately 30 percent. Further analysis of luminance curves actually suggest that the gains are even beyond 30 percent when DG 3 is used. Likewise, we believe that a 30 percent reduction in number of drivers not served would correlate to similar reduction in crash risks. More importantly, the higher the percentile served, the greater the risk mitigation, because the percentile driver not served is more likely to be the truck driver or the 4

5 older driver with clear visual deterioration, and is a higher risk than a driver in lower percentiles. In other words, the topmost 0 percent drivers (90 th to 00 th percentile) are of a greater risk than the average 0 percent drivers (i.e. 45 th to 55 th percentile), and therefore represent a higher potential in risk and crash mitigation, and should be considered accordingly when assessing economic impact. We offer the following evaluation for the supplied luminance and driver demand curves given in the FHWA PIF guidance document: The slanted lines are referred to as the demand curves in the FHWA PIF guidance (Figure and Figure 2). Although we agree with the concept, these are threshold curves rather than demand curves. In general, a true demand is best met when the supply matches the demand. However, a sheeting material that matches these demand curves will not perform well, because these curves are established as a threshold, which suggests unlimited viewing time, static visual acuity, and dark rural conditions with no clutter, which corresponds to a mere minimum for the given percent accommodation level. However, traffic signs need to perform in limited time (i.e. within LOOK3 while in motion) and mostly under suboptimal dynamic driving conditions. The FHWA PIF guidance refers to two milestone studies [2, 3] conducted by Dr. Carlson and his team at Texas Transportation Institute (TTI). However, the figures provided in the FHWA PIF guidance do not agree with the accommodation threshold luminance requirements reported in the cited references. For 40 foot/inch (threshold) legibility at 640feet, the following accommodation threshold luminance values are taken from Table 2 in [3] and from Table 5 in [4]: 50 th percentile accommodation threshold: 2.30 cd/m 2 75 th percentile accommodation threshold: 5.70 cd/m 2 85 th percentile accommodation threshold:.7 cd/m 2 95 th percentile accommodation threshold: 9.2 cd/m 2 98 th percentile accommodation threshold: 3.5 cd/m 2 NOTE: The methodology in attaining the threshold figures for overall percentile drivers served in the FHWA PIF guidance, using the threshold data from only drivers above the age of 55 years with a perfect age-vision correlation assumption is misleading. A 40 foot/inch legibility index may not apply to drivers with deteriorated vision; therefore the 640 foot benchmark may not be valid for 6 inch letter height for the higher threshold driver percentiles. Instead, shorter distances within the LOOK3 range may be more relevant. The aforementioned optimal (static acuity, dark rural, no glare, unlimited viewing time) conditions apply to the above accommodation threshold luminance values. Static acuity is known to be better than dynamic acuity for a given person. The signs we would like to replace with DG 3 will be viewed in real-world situations, where we know the driver will be moving toward the sign (and therefore requiring dynamic acuity). Dynamic acuity is related to static acuity, but it is also a function of age [5]. As age increases, dynamic visual acuity decreases more dramatically than does static visual acuity (see Figure 3). Therefore, static experimentation (as was conducted in the aforementioned studies) only remotely represents the dynamic requirements of drivers, especially in the case of older 5

6 drivers. In real world driving conditions, there will also likely be other light sources (distractive and glaring), oncoming vehicles that contribute to glare, and the driver will have limited viewing time. Consequently, the driver will need more luminance than the threshold levels suggested in these referenced studies. This fact is emphasized in the following paragraph excerpt from [4], referring to the threshold accommodation levels given above: It is important to note that these values were obtained in a dark rural environment with little ambient light. Research has shown that as the background environment becomes more complex and the ambient light level rises (conditions typically found with overhead guide signs), drivers need more luminance to read signs. Therefore, these numbers represent ideal conditions and should be considered absolute minimums. A threshold versus supply luminance curve graph from [4] is shown in Figure 7 for reference, which is different than the values shown in the FHWA PIF guidance. Based on the above percentile accommodation thresholds, and computer analyses of supply luminances for the same type of sheeting materials given in the FHWA PIF guidance, we have reviewed the supply and threshold luminance curves that were redeveloped by industry experts for various sheeting materials (refer to Figure 4 and Figure 5). ASTM Type X (NCI Crystal Grade) sheeting material was excluded from this analysis because its luminance is very similar to ASTM Type VIII. ASTM Type VII sheeting (3M LDP) was also excluded since it will be discontinued in the near future. Figure 4 and Figure 5 show that the threshold curves given in the FHWA PIF guidance are slightly misleading. The 95 th percentile demand is shown to exceed 20cd/m 2 at 640 feet in the FHWA PIF guidance, whereas the actual references indicate that the 95 th percentile demand is 9.2 cd/m 2. Conversely, 85 th percentile demand is less than 0 cd/m 2 in the FHWA PIF guidance, whereas references indicate.7 cd/m 2. Hence the figure in FHWA PIF guidance for 85 th percentile is optimistic, whereas the 95 th percentile threshold seems to be a bit conservative. Therefore, we question the validity of Table in the FHWA PIF guidance. Our analysis shows that the percentage of drivers accommodated only by DG 3 is higher than that suggested by Table in the FHWA PIF guidance: Figure 5 shows that DG 3 accommodates 85 percent of truck drivers for an overhead guide sign with 6 inch high white letters, on green background. The next highest accommodation level is provided by ASTM Type VIII sheeting, which accommodates less than the 80 th percentile. Thus, the percentage of drivers served only by DG 3 is more than 5 percent, not the 2-3 percent suggested by the FHWA PIF guidance. For these conditions, DG 3 is superior to any other retroreflective sign sheeting material. Eye tracking and sign legibility literature [7, 8] indicate that drivers read signs within the theoretical legibility range in a pseudo-random fashion during their approach to a sign before reaching the outof-view distance (OOVD), rather than at a single point such as 640 feet. However, percent accommodation levels in Table of the FHWA PIF guidance are based solely on the luminance provided at 640 feet for the selected 6 inch letter height for 40 foot/inch legibility threshold. Older drivers, which heavily represent the high percentiles in the demand curves, may not have visual acuity to read at the 40 foot/inch legibility index. This single point viewing approach suggests that 6

7 materials are ranked in performance by their luminance strictly at 640 feet under an UMTRI 2004 median headlight beam [9]. The aforementioned eye-tracking studies [7, 8] suggest that final eye fixations (reading and confirming the sign message) occur closer to the signs -- typically within the 50 foot to 450 foot range. Information acquisition (LOOK3) distance distribution also seems to correlate with age, in that older drivers execute their final fixation closer to the signs than do younger drivers [8]. Therefore, the luminance profile of a sheeting material within the entire legibility range has to be considered in determining the percent drivers served. When the luminance profiles within the entire range of LOOK3 (information acquisition range) are considered, DG 3 outperforms all other materials in terms of supplied luminance, especially closer to the signs. Although the threshold (demand) curves in the FHWA PIF guidance are helpful in determining performance and percent driver served this additional luminance supply over threshold is assumed to serve no purpose while calculating the percent drivers served. In the particular case described in Figure and Figure 2, the entire legibility range is 640 feet-320 feet = 320 feet. At 55mph, this legibility range is traveled in 3.82 seconds. Drivers spend less than 3.82 seconds in their final fixations, which further suggests sign reading range will be a stochastic function within this range, possibly of age, speed, weather conditions, visual acuity and contrast sensitivity. Note that older drivers, who read signs closer, are also likely to represent the higher percentile threshold lines in Figure and Figure 2. This strengthens the performance of DG 3 over all other sheeting materials, because the difference in luminance profile between DG 3 and all other materials is even greater for those percentiles at closer distances. In fact, the higher the percentile driver, the higher the performance of DG 3. Thus, we think that the actual percent drivers served in the real world by DG 3 is substantially underrepresented when only the luminance at 640 feet is considered. While threshold demand determines the minimum requirement for a particular percentile, the same percentile driver deserves to read signs easily and with minimal effort to promote a safe driving environment. In the scientific literature, such effortless reading is referred to as the confident legibility, indicating the optimal luminous characteristics and reading distance of a sign that drivers prefer. A set of University of Iowa studies [0,] indicates that confident legibility occurs when the letter size is increased by approximately one third of the original letter height (without changing the luminance or luminance contrast). To model the optimal legibility needs, a computer analysis was conducted to superimpose the confident legibility requirements onto the supplied luminance profiles as shown in Figure 6 and Figure 7. For a 98 th percentile truck driver, no sheeting material provides confident legibility. However, for a 95 th percentile truck driver, DG 3 provides confident legibility at 405 feet to the sign. Type IX sheeting meets the 95 th percentile confident legibility requirement at about 360 feet. Before OOVD, DG 3 allows more than second for the 95 th percentile confident legibility, better than any other sheeting, yet overlooked when only the luminance at 640 feet is considered. We feel that confident legibility holds great value in determining actual sign performance, as signs that provide optimal luminance are read quicker and without stress, capturing the attentional resources of the driver for the shortest time, and allowing the driver to redirect his attention onto the roadway and thereby promoting safety. DG 3 optimally serves more than half the drivers of heavy trucks throughout the entire legibility range. For an SUV, DG 3 serves approximately 70 percent of 7

8 the drivers at 640 feet optimally, and as the distance to the sign decreases, the confident legibility performance of DG 3 is not met by any other sheeting material. At 435 feet, DG 3 serves 98 percent of the SUV drivers optimally, whereas the same driver needs to travel another 25 feet before confidently reading a sign with Type VII or Type VIII sheeting. We believe this is one of the key aspects of DG 3 sheeting that will promote safety and the usage of this sheeting material is in the best interest of the public beyond intuitive sense. There is no question in our minds that better legibility of signs contributes to improved safety, and legibility of typical signs seems to improve up to (and even exceeding) luminance levels of 82cd/m 2 [0]. Confident or optimal legibility studies [0, ] suggest not only comfortable, but also rapid reading of a given sign. Additional luminance over threshold provides quicker information acquisition rates. Ample evidence in the scientific literature [2-4] suggests that increasing luminance on a display helps observers acquire information from the display quickly, efficiently, and accurately. Loftus [2] found that in order to acquire the same amount of information from a display in low luminance conditions, observers required approximately twice the time needed for high luminance conditions. Therefore the luminance provided above the threshold levels should benefit the driver by decreasing the information acquisition times. Signs are forms of information displays, and the same principles typically apply to signs at night. This correlates with the recommended practice for sign lighting by the Illumination Engineering Society of North America (IESNA) [5]. The Recommended Practice for Roadway Sign Lighting suggests that improving the luminance of a sign (in this case by external lighting) can aid the driver in rapid, accurate recognition and understanding of the sign s message, and this serves to improve safety by reducing the possibility that motorists will stop or drastically reduce speed at signs that may otherwise be difficult to read. These studies and recommendations also point to the value of added luminance over threshold as thresholds are established for unlimited viewing times. To acquire the information on a sign, the driver needs luminance over the threshold. This situation is illustrated in Figure 5. When the 95 th percentile accommodation threshold is considered, the threshold line almost coincides with the Type VIII material luminance curve near the end of the LOOK3 information acquisition range. Since the 95 th percentile accommodation threshold is established in dark rural conditions with unlimited viewing times, the Type VIII material cannot serve the 95 th percentile driver in a dynamic viewing situation under limited viewing time at any distance to the sign. Yet, DG 3 provides luminance over the 95 th percentile threshold, although not starting at 640 feet. Yet, the minimum required visibility distance (MRVD) may also not be 640 feet. Most drivers will not look at signs at exactly 640 feet, but most likely closer to the sign. In fact, DG 3 becomes legible at around 560 feet for the 95 th percentile, before any other sheeting material, and stays legible throughout the approach. Although illegible for the 95 th percentile at 640 feet, DG 3 is likely to serve this population within the acquisition range better than any other sheeting. There are two benefits of providing luminance over threshold for a particular accommodation level. First, it can serve a higher percentile driver within the acquisition range. Second, for the particular accommodation level, it improves the speed of information acquisition. The higher the luminance, the faster the acquisition of information from the sign, which therefore leads to shorter minimum required visibility distance (MRVD). It is highly unlikely that the driver reads the sign at theoretical legibility threshold, because he does not have unlimited time. Acquisition is executed 8

9 within a range between theoretical threshold and the last look distance (OOVD), but not throughout. Also, the quality and the speed of information transfer depend on the luminance provided over the threshold. Another reason luminance provided over threshold (suprathreshold) is highly important in realworld sign applications is due to sight distance limitations. Most signs in the real world roadway environment are not viewed from an unlimited sight distance. Only about 20 percent of freeway guide signs are in direct line of sight at 600 feet or more [6]. Thus, the 640 foot legibility benchmark does not apply to most freeway guide signs. Most signs only allow for glance legibility at distances closer than 640 feet. For these blocked signs, a legibility criterion at a distance shorter than 640 feet should apply. When viewed at closer distances and at high speeds blocked signs limit the information acquisition time, thus requiring higher luminance exceeding the threshold levels to serve the driving public. DG 3 seems to be the only material that can currently achieve this requirement especially in sub-optimal viewing conditions. Note that computer models (ERGO, Tarvip, etc) only provide results based on measurements of brand new sheeting materials out of the box. It is our experience that manufacturers typically provide materials that exceed the original product approval specifications by about 0 percent, while the warranties stipulate varying percentages of the original specification retroreflectivity coefficient (R A ) be retained at the end of the warranty period. We have examined the product specifications of various sheeting materials, and found that at the end of the warranty period, the product warranties stipulate the following: Type III sheeting to retain 80 percent of the original reflectivity, while all other prismatic sheetings except Avery s Omniview to retain 70 percent reflectivity, and Avery Omniview (the luminance curves of which are simulated using 3M VIP Type IX sheeting in the figures) to retain 50 percent of the original reflectivity of those levels provided in the computer models (see Attachment ). Since we expect these sheeting materials to last in excess of ten years while deteriorating toward the warranted levels, additional analyses were conducted to represent the end-of-warranty luminances. We assumed that the reduction in standard angle R A applies uniformly to the sheeting reflectivity at all angles. Figure 8 through Figure 5 are based on the research guidance on threshold percentile luminances [3, 4] as well as for the threshold luminances given in the FHWA PIF guidance. Figure 8 illustrates the luminance curves expected at the end of the warranty period for various sheeting materials with superimposed threshold legibility accommodation levels. ASTM sheeting Type I and Type II are excluded from the analysis to minimize figure complexity. At the end of their service lives the best performing sheetings can provide threshold legibility luminance for only about 85 percent of drivers in a typical SUV at a distance of 640 feet from signs. Figure 9 shows the same luminance curves for confident legibility accommodation levels. Only half the SUV drivers can confidently read a 6 inch-high legend for the best performing sheeting materials at 640 feet. For distances closer to the sign, DG 3 increasingly serves higher percentile SUV drivers to easily and confidently read the messages on guide signs. We would like to reiterate that such conditions are not uncommon in the real world. Figure 0 and Figure illustrate the same end-of-warranty luminance curves for the NCHRP 4-29 Heavy Vehicle for threshold and confident legibility accommodation levels, respectively. DG 3 is 9

10 the only material that provides 80 th percentile threshold legibility at 640 feet. These graphs suggest that the next best sheeting material would provide only about 75 th percentile threshold legibility at 640 feet. The difference is in the order of 5 percent by the end of the warranty period between DG 3 and the next best performing sheeting, which is more than that suggested in Table in the FHWA PIF guidance. In fact, when the accommodation levels given in the FHWA PIF guidance are used, the difference reaches 0 percent (as shown in Figure 4). At any higher percentile level, DG 3 still provides a wider range of distances than any other sheeting material. Figure shows that no sheeting material can provide confident legibility at a median level at 640 feet, although DG 3 comes the closest. Likewise, for any given distance within the information acquisition range, DG 3 accommodates the highest percentile of drivers for threshold and for confident legibility conditions. Slightly different versions of the figures for the threshold percentile accommodation curves given in the FHWA PIF guidance have also been developed to represent end-of-warranty conditions. Confident legibility curves are derived from the FHWA PIF guidance threshold data, with the assumption that for confident reading, the visual angle subtended by the letter height would be 33 percent higher than the threshold level []. Figure 2, Figure 3, Figure 4, and Figure 5 illustrate the luminance curves for various sheeting materials at their end-of-warranty predicted reflectivity levels for a typical SUV for threshold, typical SUV for confident, typical heavy truck for threshold, and typical heavy truck for confident levels, respectively (based on the FHWA PIF guidance figures). For the most part, the increase in the percentile driver served is even more accentuated when the threshold legibility levels given in the FHWA PIF guidance document are used for end-ofwarranty sheeting performance. The FHWA PIF guidance refers to human factors studies conducted in TTI to establish the threshold luminance levels for legibility. The established minimum levels seem to assume that visual performance is a linear function of age, and vice versa. This is an incorrect assumption leading to an optimistic estimate of percent drivers served. The study indicates that 89 percent of drivers are under the age of 55, and 96 percent drivers are under the age of 65. With a direct age-visual capability assumption, a median driver above the age of 65 is assumed to represent the 98 th percentile of overall drivers. This simple math assumes that the visual capability of the best 65 year old is worse than the worst 64 year old driver. Many studies, including the Mercier and TTI studies referenced in [2] suggest otherwise as shown in Figure 6. The computer model analysis, which seems to be the very best available, does not consider headlight misaim, weather conditions or lens dirt that may adversely affect sign visibility. Based on the above discussion, DG 3 will be the most forgiving sheeting material available to the public. There are many vehicles on the roadways without new headlights, with headlamp dirt, misaim, lens discoloration, driving under rain, snow, fog, etc. It is in our interest to provide the driving public with the highest luminance where they need it, so when the conditions are suboptimal, they have the most forgiving traffic signs. Conditions in the real world are usually less than optimal, as opposed to most of the studies based on ideal conditions. We would like to emphasize a study conducted in Arizona investigating the luminance provided by different sheeting materials for overhead guide signs [6], and observers ratings of various sign sheeting materials. In a 2005 Arizona Department of Transportation study of overhead guide signs 0

11 (which compared prismatic reflective sheeting materials submitted by the major manufacturers), the DG3 sheeting performed at a higher level, preferentially and photometrically, than any of the other prismatic materials at all five distances evaluated. We think that it would be unfair to disregard this study as being subjective and thus having no scientific value. Although not universally true, subjective ratings do have substantial real-world validity. For example, discomfort glare is a subjective phenomenon, and is hard to quantify, yet it is very real, and it affects driving performance of almost every nighttime driver. In fact, discomfort glare is the real reason behind having low beams. Although it is not quantified, a nominal scale known as the DeBoer scale is established for assessing discomfort glare, and the DeBoer scale is widely used in many scientific studies as a correlative measure with notable validity. Likewise, we believe that the subjective rating of the sheeting materials and indisputable performance of DG 3 over all other sheetings carry scientific weight and should not be prematurely dismissed. In the Arizona study the DG 3 prismatic sheeting material performed better than currently available prismatic reflective sheeting materials. ASTM Type IX was tied with DG3 sheeting for the highest ranking material at the shortest distance but was rated the lowest at the farthest distances. ASTM Types VIII and X were tied with DG3 sheeting for the highest ranked materials at the farthest two distances but were rated the lowest at the closest distance. These results were consistent among older drivers and participants in heavy vehicles (see Attachment 2 for conclusions of that study). This study validates the earlier technical analysis in a field study, confirming the superior performance of DG 3 sheeting material over other available sheeting materials. Another study, a day and night field review, was conducted by Mn/DOT at the 3M Safety Center on June 5, 2005 to evaluate this higher performance prismatic retroreflective sheeting. This review was attended by approximately 00 individuals, consisting of FHWA, Mn/DOT Central Office and district traffic, maintenance and traffic services staff. A majority of viewers that observed ground mounted regulatory and warning signs and ground mounted and overhead mounted guide signs indicated that the DG 3 sheeting performed better than VIP sheeting during the daytime and 90 percent of viewers indicated the DG 3 sheeting performed better during the nighttime observations. These results are consistent with the previous discussion, further strengthening the in-the-field performance of DG 3. (See for full report). White on green DG 3 sheeting materials provide a contrast ratio of 0: at all observation and entrance angles. A contrast level of 0: coupled with higher levels of sign legend and background luminance is likely to increase legibility distance [0,, 7, 8], over those estimated by the threshold curves presented in the FHWA PIF guidance, which were obtained for a lower contrast ratio of 5: [2, 3]. At the luminances achieved by DG 3, a contrast of 0: improves legibility over a sign having 5: contrast. Typically Type IX prismatic sheetings achieve an 8: luminance contrast for white on green signs. The higher contrast for DG 3 is viable and sustained because green DG 3 has a very high retroreflective performance relative to other sheeting types. This high background luminance prevents overglow which affects signs when legend-to-background luminance contrast is high and background luminance is low. We intend to use DG 3 sheeting for both backgrounds and legends to prevent such overglow while maximizing legibility.

12 Recent data suggests that DG 3 shows considerable dew resistance compared to other available sheetings. In most dew-out conditions, DG 3 maintains its retroreflectivity much better than ASTM Type IX microprismatic sign sheeting, and in some cases, comparable to Type IX sheeting covered with dew-resistant film. The use of dew-resistant films on signs is usually impractical due to the films short lifecycle times (approximately 5 years). The luminance of DG 3 alongside standard Type IX, and Type IX with two different dew-resistant films were measured overnight at different nights in May through August of Figure 8 through Figure 20 show the luminance of untreated DG 3 achieved under dew-out conditions in three separate nights. In all cases, standard Type IX luminance approaches nearly to zero in dew-out, whereas DG 3 maintains approximately two to ten times the brightness of untreated Type IX sheeting, and half as much retroreflectivity as those with dew-resistant films. In summary, the dew data suggests the following: Compared to all other materials tested, during the entire testing period, DG³ showed significantly more days without any visible dew formation (42 days vs. 32 days). Even well into the dew-out conditions, DG 3 appears brighter than the other 3 materials for, in most cases, about hour. The dew resistance of DG 3 makes it a very desirable sheeting material, considering the dew-out conditions that render most other sheeting materials almost completely ineffective under dew. DG 3 is a sensible alternative to mitigate the adverse effects of dew, and to serve drivers the best and most cost-effective way. In light of all given data, we believe that there is no suitable alternative to DG 3, and using DG 3 will definitely benefit the driving public and provide a safer roadway environment at no additional cost. It is, in fact, the most economical alternative when all factors are considered. In light of the above discussion, DG 3 seems to be the best sheeting material for guide signs in the interest of the driving public, because: i. DG 3 has an unsurpassed observation angle capability serving the highest percentage of truck and SUV drivers ii. DG 3 has the highest total light return at about 58 percent, almost twice that of truncated-cube prismatic sheetings iii. DG 3 reduces the percent drivers NOT served by 30 percent or more, which has a potential to reduce crashes and inefficiencies proportionately with this reduction iv. DG3 sheeting is the best performing sheeting on signs read closer than threshold conditions in a random fashion. Typically, drivers do not read signs at threshold conditions. The threshold levels given in the FHWA PIF guidance are based on optimal viewing conditions, and may be misleading, because: a. They do not agree with the referenced studies in the PIF guidance b. They represent optimal viewing conditions, i.e. dark, rural, static viewing conditions with unlimited viewing distance c. The drivers that participate in such field studies are usually self-selected, are confident about their nighttime vision and driving, and fail to represent the actual driving public 2

13 d. Participants have unlimited viewing time v. DG 3 provides the highest luminance throughout the information acquisition range, where LOOK3 takes place for both trucks, and smaller vehicles, therefore: a. Serves a higher percentile than those suggested in the FHWA PIF guidance document, in which the performance evaluation is based on the luminance at 640 feet b. Has more potential to be read due to higher luminance over threshold within the limited time when the sign is blocked c. Provides easy, effortless, and quick legibility for those drivers with better vision, and for whom the luminance is over the threshold vi. vii. DG 3 provides the highest percentiles of confident legibility for all drivers, DG 3 serves the highest number of drivers in real-world suboptimal driving conditions (which increase the luminance demand) that involve: a. Limited sign viewing time b. Limited line of sight to the distance to signs c. Misaimed/dirty headlamps d. Rain, fog, snow e. Dynamic visual acuity f. Oncoming vehicle glare g. Visual clutter viii. DG 3 serves the highest percentage of drivers throughout its 2-year service life, while widening the performance gap with other sheeting materials ix. The field study in Arizona and the field review in 3M Cottage Grove Transportation Safety Research Center are valuable field validations of superior DG 3 sheeting performance over other sheeting for guide signs, x. DG 3 has outstanding dew resistance unmatched by any other untreated sheeting xi. material The differences in the percentage of drivers served only by DG 3 is likely to be much larger than those suggested in the PIF guidance There are additional benefits of using a better performing sheeting material, where the cost/benefit analysis is difficult to quantify or goes unnoticed. With better performing signs, there will be fewer drivers missing their exits, or unfamiliar drivers performing erratic maneuvers due to poor legibility of signs. Traffic system efficiency will be best maintained with best performing sign performance, where signs provide the information to the driver as quickly and efficiently as possible, and right on time. The costs associated with non-crash incidents are not reported, but society will reap the benefits. We also would like to use DG 3 for chevrons, delineators, and markers. Many run-off-the-road (ROR) crashes occur in curves, especially at night [9]. ROR crashes are a leading cause of fatal crashes, and are a great safety concern to us. Chevrons, delineators, and markers have to perform at all times and are directly associated with driver safety, which is also identified as the first priority in the FHWA Vital Three initiative. 3

14 One of the strategies identified in most state safety plans is to reduce lane departure crashes, which typically occur on rural highways. We feel we can reduce run-off-the-road crashes by providing information about roadway geometry using the most conspicuous sheeting material available, especially under adverse weather conditions. Chevrons are essential warning signs for driver safety, and their visibility is highly critical. They warn the driver about an existing hazard, for which the driver has to take action. Furthermore, a single chevron provides an incomplete message. Chevrons have to be seen in groups of two or more as specified in the Federal MUTCD. They not only provide valuable information to the driver about an approaching curve, but also its curvature. Markers and delineators work in a similar fashion. Since there is no control over the type of vehicle, type of headlamp, or the weather conditions under which these devices should function, any sheeting used on chevrons, markers, and delineators should accommodate all drivers, in all weather conditions, to the extent possible with the greatest margin of safety. Under normal conditions, chevrons, markers and delineators generally supplement other visual cues and guidance, such as pavement markings. However, the life-critical functionality of these devices is realized under adverse weather conditions, when there is snow on the road surface, under rain and fog, or when pavement markings are wet. During such conditions these devices function as standalone guidance to the drivers. A stop sign should be visible at a minimum of 582 feet [2] for a driver traveling at 70 mph to come to a safe stop by the sign. This distance incorporates the distance traveled during information acquisition, decision making, motor tasks, and vehicle deceleration. Since chevrons only command a speed adjustment for safe curve negotiation and lane keeping, we do not expect chevron visibility distances beyond 700 feet to be highly important. At a minimum, at least two chevrons should be detected by the approaching driver early enough to make the proper speed adjustments. 700 feet is likely to be beyond such minimum required distance according to many eye-tracking and human factors studies [7, 7, 20]. In ideal viewing conditions (dark, rural, no glare, no visual clutter, clear and dry weather, correctly aimed headlights), we expect that the sheeting materials out of the box given in Figure 2 and Figure 22 may all perform adequately, as they will all be visible under unlimited sight distance. For heavy vehicles and SUV type vehicles, we expect that DG 3 will provide the strongest visual signal to the driver at all relevant distances (i.e. within 700 feet), whereas Type IX and Type VIII sheetings will still be visible yet provide more subtle visual signals than DG 3. Although all of these sheetings will be visible in ideal viewing conditions, we have to consider adverse weather conditions, weathered and aged materials, imperfect headlights, oncoming headlamp glare (which is a significant issue in two-way rural roadways), and deteriorated vision, while designing the roadway system for optimal driver safety. Furthermore, DG 3 will provide the strongest visual signal not only for a single chevron, marker or delineator, but also for all chevrons, markers or delineators in a series. This is highly important especially when visibility is compromised due to adverse weather. Type VIII sheeting is slightly brighter for SUV drivers for distances beyond 700 feet. Notwithstanding the relevance of luminance at distances beyond 700 feet to driver safety, the performance of Type VIII sheeting severely suffers at distances closer to the curves. These closer distances are critical [7, 7, 20] where the driver has to make speed adjustments and is acquiring curvature information based on visual cues from all visible chevrons, markers and delineators. That is where DG 3 is the best 4

15 performing sheeting material. Similar concerns previously presented for the guide signs also apply to chevrons, markers and delineators. For instance, when retroreflectivity values are adjusted to end-of-warranty levels, DG 3 will dramatically perform better than other prismatic sheeting materials for practically all vehicle types and drivers. Figure 23 and Figure 24 provide end-of-warranty luminances for Type VIII, two types of Type IX, and DG 3. Omniview retroreflectivity values are simulated using 3M VIP Type IX sheeting, where the retroreflectivity is reduced by the initial-to-end of warranty retroreflectivity ratio given in Omniview warranty specification -- approximately a 50 percent reduction. DG 3 dramatically outperforms the three other sheetings at the end of product warranty. Although the difference between DG 3 and Type IX VIP is relatively maintained, the difference between DG 3 and Type IX-Omniview is further widened at the end of the typical warranty period of 2 years. In light of the luminance analysis, we have concluded that the DG 3 white sheeting material with a significantly higher coefficient of retroreflection (570 candelas/lux/square meter) provides higher luminance at the lower observation angles than Type IX sheeting. Upgrading the sheeting materials on all white markers and delineators to DG 3 sheeting will maximize their impact on road users by improved conspicuity without sacrificing close-in viewing. Although this is not currently quantifiable from a crash report standpoint, we feel that the usage of the brightest material available, at no extra cost, is in the public s best interest and contributes to safety of the road user. The DG 3 fluorescent yellow sheeting material, which we propose to use on all chevrons, yellow markers and delineators, provides higher coefficients of retroreflection at most observation and entrance angles compared to ASTM Type IX yellow or fluorescent yellow sheeting material requirements (see Attachment 3). These higher values equate to brighter materials providing road users with the longest viewing distance. Fluorescent yellow DG 3 sheeting on all chevrons, yellow markers and delineators will make them more visible than standard yellow prismatic reflective sheeting during dusk, dawn and overcast conditions. There is no doubt that DG 3 is the highest performing prismatic sheeting material (as indicated by the initial field studies and computer models) designed to serve the greatest range of drivers both day and night. While situations may arise where it can be shown that another sheeting may perform comparably to DG3 or even in some special circumstances out-perform DG 3, the cost of managing and inventorying more than one type of sheeting for signs based on their environment (rural vs. urban) would far outweigh any potential savings in doing so. We are confident that DG 3 is the best sheeting material that can be used in all guide signs, chevrons, markers and delineators beyond any doubt. DURATION OF APPROVAL We request your concurrence for a three year period. During this three year period, we will evaluate all comparable sheetings materials from other manufacturers that may meet or exceed the level of performance established by the 3M DG 3 sheeting. 5

16 Based on the above analysis we ask that you concur in our proposal to use the DG 3 prismatic retroreflective sheeting material for guide signs, chevrons, markers and delineators. If you have any questions, contact Michael Weiss, State Signing Engineer at Sincerely, State Traffic Engineer Enclosures: Attachments -3 6

17 REFERENCES. Freight Analysis Framework, USDOT, Federal Highway Administration, Office of Freight Management and Operations, Washington, D.C., October Carlson, P.J., H.G. Hawkins, G.F. Schertz, D.J. Mace, and K.S. Opiela. Developing Updated Minimum In-Service Retroreflectivity Levels for Traffic Signs. In Transportation Research Record 824, TRB, National Research Council, Washington, DC, 2003, pp Carlson, P. J., and H. G. Hawkins, Updated Minimum Retroreflectivity Levels for Traffic Signs, FHWA-RD-03-08, USDOT, FHWA, Washington, D.C., Carlson, P. J., Evaluation of Clearview Alphabet with Microprismatic Retroreflective Sheetings, Report No. FHWA/TX-02/4049-, Final Report to Texas DOT, October Burg, Albert, Visual Acuity as Measured by Dynamic and Static Tests: A Comparative Evaluation, Journal of Applied Psychology, December 966, 50(6), pp Joseph E. Hummer, Changseok Baek, Line of Sight Distances to Signs, Proceedings, 83 rd Annual meeting of the Transportation Research Board, January 2004, Washington, D.C. 7. Helmut T Zwahlen, Advisory Speed Signs and Curve Signs and Their Effect on Driver Eye Scanning and Driving Performance, Transportation Research Record No:, p Frank Schieber, David Burns, Jason Myers, Nicholas Willan and Jess Gilland, Driver Eye Fixation and Reading Patterns while Using Highway Signs under Dynamic Nighttime Driving Conditions: Effects of Age, Sign Luminance and Environmental Demand, Paper presented at the Annual TRB meeting, 2004, Washington D.C. 9. Brandon Schoettle, Michael Sivak, Michael J. Flannagan, Walter J. Kosmatka, A Market- Weighted Description of Low-Beam Headlighting Patterns in the U.S.: 2004, UMTRI , September Schnell, T., Aktan, F., Li, C., Traffic Sign Luminance Requirements of Nighttime Drivers for Symbolic Signs. Transportation Research Record No: 862, Journal of the Transportation Research Board, Washington, D.C., pp , Aktan, F., Schnell, T., Defining Threshold and Confident Legibility Contrast and Luminance Needs of Nighttime Drivers for Achromatic and Chromatic Signs. Paper presented in the 7th Biennial Symposium on Visibility and Traffic Control Devices. The Transportation Research Board, Washington, D.C., April Loftus, G. R., Picture Perception: Effects of Luminance on Available Information and Information Extraction Rate. Journal of Experimental Psychology: General, Vol. 4, No. 3, pp , Loftus, G. R., Ruthruff, E., A Theory of Visual Information Acquisition and Visual Memory with Special Application to Intensity-Duration Tradeoffs. Journal of Experimental Psychology: Human Perception and Performance, Vol. 20, pp , Herron, J. M., Bender, T. M., Campbell, W. L., Sumkin, J. H., Rockette, H. E., Gur, D., Effects of Luminance and Resolution on Observer Performance with Chest Radiographs. Radiology, Vol. 25, No., pp , 5. IESNA Recommended Practice for Roadway Sign Lighting. IESNA RP-9-0, Illumination Engineering Society of North America, Driver Ratings of Overhead Guide Sign Legends, Study conducted for the Arizona Department of Transportation, Chalmers Engineering Services Inc. in Association with Paul J. 7

18 Carlson, Ph.D., P.E. 7. Zwahlen, H.T., A. Russ, and S. Vatan, Nighttime Photometric Evaluation of Unlighted Overhead Guide Signs. Paper Number , Transportation Research Board 82nd Annual Meeting, Washington, D.C., January Mace, D.J. Sign Legibility and Conspicuity, Transportation in an Aging Society: Improving Mobility and Safety for Older Persons, Special Report 28, Vol. 2, Transportation Research Board, Washington, D.C., Helmut T. Zwahlen, Jin Young Park, Curve Radius Perception Accuracy as a Function of the Number of Delineation Devices (Chevrons), Transportation Research Record No: 495, Journal of the Transportation Research Board, Washington, D.C., Helmut T. Zwahlen, Traffic Sign Reading Distances and Sign Reading Times When Driving at Night, Paper presented at the 2 th Biennial Symposium on Improving Visibility for the Night Traveler, National Academy of Sciences, Washington, D.C, and May 23-24, NCHRP Project 4-29, Selection of Materials to Optimize Sign Performance, Interim Report, National Cooperative Highway Research Program, National Academy of Sciences, February

19 00.0 Last Look Legibility Threshold Type I Type II 95th % demand 85th % demand Type III - 3M Beaded HI Luminance (cd/m^2) th % demand 50th % demand Type VII - 3M LDP Type VIII - AD 7500 Type IX - 3M VIP AD - Omniview 3M - DG3 50th %ile demand Acquistion Area (LOOK3) 75th %ile demand 85th %ile demand 95th %ile demand Distance to sign (ft) Figure. Computer evaluation of supply and demand curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 SUV. 9

20 00.0 Last Look Legibility Threshold Type I Type II Luminance (cd/m^2) th % demand 85th % demand 75th % demand 50th % demand Type III - 3M Beaded HI Type VII - 3M LDP Type VIII - AD 7500 Type IX - 3M VIP AD - Omniview 3M - DG3 50th %ile demand Acquistion Area (LOOK3) 75th %ile demand 85th %ile demand 95th %ile demand Distance to sign (ft) Figure 2. Computer evaluation of supply and demand curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 heavy vehicle. 20

21 Note: Figure adapted from [5]. Figure 3. Binocular visual acuity as a function of age, sex, and speed of target movement. 2

22 00 Type I Type II Type III Type VIII Type IX DG3 95th Percentile Threshold 85th Percentile Threshold 75th Percentile Threshold Median 98th Percentile Range of Information Acquisition (third and final look at the sign) 98th Percentile Threshold Legibility 95th Percentile Threshold Legibility 0 85th Percentile Threshold Legibility Luminance [cd/m^2 75th Percentile Threshold Legibility Median Threshold Legibility Legibility Threshold 6" letter at 40ft/inch benchmark - 640ft Distance to Sign [ft] Luminance curves are for on overhead sign observed by a driver of a typical SUV (from NCHRP project 4-29). The center of the sign is directly over the center of travel lane, 25 ft above the roadway with no tilt. Both headlamps are the US median beams as per UMTRI Figure 4. Computer evaluation of supply and threshold legibility curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 SUV including 98 th percentile accommodation threshold. 22

23 00 Type I Type II Type III Type VIII Type IX DG3 95th Percentile Threshold 85th Percentile Threshold 75th Percentile Threshold Median 98th Percentile Threshold Range of Information Acquisition (third and final look at the sign) 98th Percentile Threshold 95th Percentile Threshold 0 85th Percentile Threshold Luminance [cd/m^2 75th Percentile Threshold Median Threshold Legibility Threshold - 640ft Luminance curves are for an overhead guide sign, observed by a driver of a typical heavy vehicle (from NCHRP project 4-29). The center of the sign is directly over the center of travel lane, 25 ft above the roadway with no tilt. Both headlamps are the US median beams as per UMTRI Distance to Sign [ft] Figure 5. Computer evaluation of supply and threshold legibility curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 Heavy Vehicle including 98 th percentile accommodation threshold. 23

24 000 Type I Type II Type III Type VIII Type IX DG3 98th Percentile Confident 95th Percentile Confident 85th Percentile Confident 75th Percentile Confident Median Confident Range of Information Acquisition (third and final look at the sign) Luminance [cd/m^ th Percentile Confident Legibility 95th Percentile Confident Legibility 85th Percentile Confident Legibility 75th Percentile Confident Legibility Median Confident Legibility 0. Legibility Threshold for 6" letter height - 640ft Distance to Sign [ft] Luminance curves are for on overhead sign observed by a driver of a typical SUV (from NCHRP project 4-29). The center of the sign is directly over the center of travel lane, 25 ft above the roadway with no tilt. Both headlamps are the US median beams as per UMTRI Figure 6. Computer evaluation of supply and confident (optimal) legibility curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 SUV. 24

25 000 Type I Type II Type III Type VIII Type IX DG3 98th Percentile Confident 95th Percentile Confident 85th Percentile Confident 75th Percentile Confident Median Confident Range of Information Acquisition (third and final look at the sign) 98th Percentile Confident Legibility 00 95th Percentile Confident Legibility 85th Percentile Confident Legibility Luminance [cd/m^2 0 75th Percentile Confident Legibility Median Confident Legibility Typical Legibility Threshold for 6" letter height - 640ft Luminance curves are for an overhead guide sign, observed by a driver of a typical heavy vehicle (from NCHRP project 4-29). The center of the sign is directly over the center of travel lane, 25 ft above the roadway with no tilt. Both headlamps are the US median beams as per UMTRI Distance to Sign [ft] Figure 7. Computer evaluation of supply and confident (optimal) legibility curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 Heavy Vehicle. 25

26 00 Type III Type VIII Type IX-Omniview Type IX - VIP DG3 95th Percentile Threshold 85th Percentile Threshold 75th Percentile Threshold Median 98th Percentile Range of Information Acquisition (third and final look at the sign) 98th Percentile Threshold Legibility 95th Percentile Threshold Legibility 0 85th Percentile Threshold Legibility Luminance [cd/m^2 75th Percentile Threshold Legibility Median Threshold Legibility Legibility Threshold 6" letter at 40ft/inch benchmark - 640ft Predicted end-of-warranty luminance curves for on overhead sign observed by a driver of a typical SUV (from NCHRP project 4-29). The center of the sign is directly over the center of travel lane, 25 ft above the roadway with no tilt. Both headlamps are the US median beams as per UMTRI Distance to Sign [ft] Figure 8. Computer evaluation of end-of-warranty supply and threshold legibility curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 SUV for legibility threshold levels for 40ft/inch legibility index given by Carlson [3, 4]. 26

27 000 Type III Type VIII Type IX-Omniview Type IX - VIP DG3 98th Percentile Confident 95th Percentile Confident 85th Percentile Confident 75th Percentile Confident Median Confident Range of Information Acquisition (third and final look at the sign) Luminance [cd/m^ th Percentile Confident Legibility 95th Percentile Confident Legibility 85th Percentile Confident Legibility 75th Percentile Confident Legibility Median Confident Legibility Legibility Threshold for 6" letter height - 640ft Predicted end-of-warranty luminance curves for on overhead sign observed by a driver of a typical SUV (from NCHRP project 4-29). The center of the sign is directly over the center of travel lane, 25 ft above the roadway with no tilt. Both headlamps are the US median beams as per UMTRI Distance to Sign [ft] Figure 9. Computer evaluation of end-of-warranty supply and confident legibility curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 SUV for legibility threshold levels for 40ft/inch legibility index given by Carlson [3, 4]. 27

28 00 Type III Type VIII Type IX - Omniview Type IX-VIP DG3 95th Percentile Threshold 85th Percentile Threshold 75th Percentile Threshold Median 98th Percentile Threshold Range of Information Acquisition (third and final look at the sign) 98th Percentile Threshold 95th Percentile Threshold Luminance [cd/m^2 0 85th Percentile Threshold Median Threshold Legibility Threshold - 640ft 75th Percentile Threshold Predicted end-of-warranty luminance curves for an overhead guide sign, observed by a driver of a typical heavy vehicle (from NCHRP project 4-29). The center of the sign is directly over the center of travel lane, 25 ft above the roadway with no tilt. Both headlamps are the US median beams as per UMTRI Distance to Sign [ft] Figure 0. Computer evaluation of end-of-warranty supply and threshold legibility curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 Heavy Vehicle for legibility threshold levels for 40ft/inch legibility index given by Carlson [3, 4]. 28

29 000 Type III Type VIII Type IX - Omniview Type IX-VIP DG3 98th Percentile Confident 95th Percentile Confident 85th Percentile Confident 75th Percentile Confident Median Confident 00 Range of Information Acquisition (third and final look at the sign) 98th Percentile Confident Legibility 95th Percentile Confident 85th Percentile Confident Luminance [cd/m^2 0 75th Percentile Confident Median Confident Legibility 0. Typical Legibility Threshold for 6" letter height - 640ft Distance to Sign [ft] Predicted end-of-warranty luminance curves for an overhead guide sign, observed by a driver of a typical heavy vehicle (from NCHRP project 4-29). The center of the sign is directly over the center of travel lane, 25 ft above the roadway with no tilt. Both headlamps are the US median beams as per UMTRI Figure. Computer evaluation of end-of-warranty supply and confident legibility curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 Heavy Vehicle for legibility threshold levels for 40ft/inch legibility index given by Carlson [3,4]. 29

30 00 Type III Type VIII Type IX-Omniview Type IX - VIP DG3 95th Percentile Threshold 85th Percentile Threshold 75th Percentile Threshold Median Threshold Range of Information Acquisition (third and final look at the sign) 95th Percentile Threshold Legibility (PIF guidance) Luminance [cd/m^2 0 85th Percentile Threshold Legibility (PIF guidance) 75th Percentile Threshold Legibility (PIF Guidance) Median Threshold Legibility (PIF Guidance) Legibility Threshold 6" letter at 40ft/inch benchmark - 640ft 0. Predicted end-of-warranty luminance curves for on overhead sign observed by a driver of a typical SUV (from NCHRP project 4-29). The center of the sign is directly over the center of travel lane, 25 ft above the roadway with no tilt. Both headlamps are the US median beams as per UMTRI Distance to Sign [ft] Figure 2. Computer evaluation of end-of-warranty supply and threshold legibility curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 SUV for legibility threshold levels for 40ft/inch legibility index given in the FHWA PIF guidance document. 30

31 000 Type III Type VIII Type IX-Omniview Type IX - VIP DG3 95th Percentile Confident 85th Percentile Confident 75th Percentile Confident Median Confident Range of Information Acquisition (third and final look at the sign) 00 95th Percentile Confident Legibility (PIF guidance) Luminance [cd/m^2 0 85th Percentile Confident Legibility (PIF guidance) 75th Percentile Confident Legibility (PIF Guidance) Median Confident Legibility (PIF Guidance) 0. Legibility Threshold for 6" letter height - 640ft Distance to Sign [ft] Predicted end-of-warranty luminance curves for on overhead sign observed by a driver of a typical SUV (from NCHRP project 4-29). The center of the sign is directly over the center of travel lane, 25 ft above the roadway with no tilt. Both headlamps are the US median beams as per UMTRI Figure 3. Computer evaluation of end-of-warranty supply and confident legibility curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 SUV for legibility threshold levels for 40ft/inch legibility index given in the FHWA PIF guidance document. 3

32 00 Type III Type VIII Type IX - Omniview Type IX-VIP DG3 95th Percentile Threshold 85th Percentile Threshold 75th Percentile Threshold Median Threshold Range of Information Acquisition (third and final look at the sign) 95th Percentile Threshold Legibility (PIF guidance) Luminance [cd/m^ Figure 4. 75th Percentile Threshold Legibility (PIF guidance) Median Threshold Legibility (PIF guidance) 85th Percentile Threshold Legibility (PIF guidance) Legibility Threshold - 640ft Predicted end-of-warranty luminance curves for an overhead guide sign, observed by a driver of a typical heavy vehicle (from NCHRP project 4-29). The center of the sign is directly over the center of travel lane, 25 ft above the roadway with no tilt. Both headlamps are the US median beams as per UMTRI Distance to Sign [ft] Computer evaluation of end-of-warranty supply and threshold legibility curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 Heavy Vehicle for legibility threshold levels for 40ft/inch legibility index given in the FHWA PIF guidance document. 32

33 000 Type III Type VIII Type IX - Omniview Type IX-VIP DG3 95th Percentile Confident 85th Percentile Confident 75th Percentile Confident Median Confident Range of Information Acquisition (third and final look at 00 95th Percentile Confident Legibility (PIF guidance) Luminance [cd/m^2 0 85th Percentile Confident Legibility (PIF guidance) 75th Percentile Confident Legibility (PIF guidance) Median Confident Legibility (PIF guidance) 0. Typical Legibility Threshold for 6" letter Distance to Sign [ft] Predicted end-of-warranty luminance curves for an overhead guide sign, observed by a driver of a typical heavy vehicle (from NCHRP project 4-29). The center of the sign is directly over the center of travel lane, 25 ft above the roadway with no tilt. Both headlamps are the US median beams as per UMTRI Figure 5. Computer evaluation of end-of-warranty supply and confident legibility curves for overhead guide sign mounted 25-feet high, centered on travel lane with no tilt, illuminated by UMTRI 2004 headlamp and viewed from NCHRP 4-29 Heavy Vehicle for legibility threshold levels for 40ft/inch legibility index given in the FHWA PIF guidance. 33

34 Figure 6. Scatterplot of Mercier and TTI required luminance data as a function of driver age. 34

35 Figure 7. Passenger car luminance supply curves versus the threshold accommodation curves reported in [4]. 35

36 600 T start TEnd Dew Out Time DG VIP with 3M DR Film DG VIP with NCI Dewtect Film DG VIP Standard DG^3 Standard :00 22:3 22:26 22:39 22:52 23:05 23:8 23:3 23:44 23:57 0:0 0:23 0:36 0:49 :02 :5 :28 :4 :54 2:07 2:20 2:33 2:46 2:59 3:2 3:25 3:38 3:5 4:04 4:7 4:30 4:43 4:56 5:09 5:22 5:35 5:48 Luminance R max R R min Time Figure 8. The yellow (Type IX Standard) approaches zero luminance while DG 3 maintains reasonable retroreflectivity over the entire dew-out period. 36

37 :00:34 22:5:35 22:30:35 22:45:36 23:00:37 23:5:37 23:30:37 23:45:38 0:00:38 0:5:39 0:30:39 0:45:40 :00:40 :5:4 :30:4 :45:42 2:00:42 2:5:43 2:30:44 2:45:44 3:00:44 3:5:45 3:30:46 Luminance 3:45:46 4:00:47 4:5:47 4:30:48 4:45:48 5:00:48 5:5:49 5:30:49 5:45:50 Time Figure 9. Dew Period: Start of DG³ at :00, severe dew out; recovering at 5: DG VIP with 3M DR Film DG VIP with NCI Dewtect Film DG VIP Standard DG^3 Standard

38 DG VIP with 3M DR Film DG VIP with NCI Dewtect Film DG VIP Standard DG^3 Standard 000 Luminance Figure :00 22:3 22:26 22:39 22:52 23:05 23:8 23:3 23:44 23:57 0:0 0:23 0:36 0:49 :02 :5 :28 :4 :54 2:07 2:20 2:33 2:46 2:59 3:2 3:25 3:38 3:5 4:04 4:7 4:30 4:43 4:56 5:09 5:22 5:35 5:48 Time (a) Dew Period : Light Dew Out: DG³ standard is not affected; dew out for DR materials (light) and DG VIP standard (severe); Recovery of DR materials at :00; DG VIP standard does not recover at all. (b) Dew Period 2: Heavy Dew Out: DG³ starts significant later (3:00) and remains at a higher level than DG VIP standard, but does also not recover until sunrise. 38

39 000 st chevron at the beginning of the curve,.5m high, 6ft offset from right edgeline 00 Luminance [cd/m^ Type VIII Type IX Proposed Type XI Luminance [cd/m^2 0 2nd chevron 40m into the curve,.5m high, 6ft offset from right edgeline Type VIII Type IX Proposed Type XI Distance to Chevron [ft] 3rd chevron 80m into the curve,.5m high, 6ft offset from right edgeline Distance to Chevron [ft] 00 4th chevron 20m into the curve,.5m high, 6ft offset from right edgeline Luminance [cd/m^2 0 Type VIII Type IX Proposed Type XI Luminance [cd/m^2 0 Type VIII Type IX Proposed Type XI Distance to Chevron [ft] Distance to Chevron [ft] Figure 2. Computer evaluation of luminance values for four consecutive chevrons placed on the right shoulder, 6ft offset from the right edgeline of a 200m radius right curve after a 200m straight approach, starting at the beginning of the curve. Chevrons are 5ft high and have no tilt, and are placed according to MUTCD guidelines. Each figure shows the distance to the corresponding chevron. Analysis conducted for an NCHRP 4-29 Heavy Vehicle equipped with the 50 th percentile 2004 UMTRI US low beam headlamp. 39

40 000 st chevron at the beginning of the curve,.5m high, 6ft offset from right edgeline 00 Luminance [cd/m^ Type VIII Type IX Proposed Type XI Luminance [cd/m^2 0 2nd chevron 40m into the curve,.5m high, 6ft offset from right edgeline Type VIII Type IX Proposed Type XI Distance to Chevron [ft] Distance to Chevron [ft] Luminance [cd/m^2 0 3rd chevron 80m into the curve,.5m high, 6ft offset from right edgeline Type VIII Type IX Proposed Type XI Luminance [cd/m^2 0 4th chevron 20m into the curve,.5m high, 6ft offset from right edgeline Type VIII Type IX Proposed Type XI Distance to Chevron [ft] Distance to Chevron [ft] Figure 22. Computer evaluation of luminance values for four consecutive chevrons placed on the right shoulder, 6ft offset from the right edgeline of a 200m radius right curve after a 200m straight approach, starting at the beginning of the curve. Chevrons are 5ft high and have no tilt, and are placed according to MUTCD guidelines. Each figure shows the distance to the corresponding chevron. Analysis conducted for an NCHRP 4-29 SUV equipped with the 50 th percentile 2004 UMTRI US low beam headlamp. 40

41 00 00 Luminance [cd/m^2] 0 0. st chevron at the beginning of the curve,.5m high, 6ft offset from right edgeline, curves reflect expected end-of-warranty retroreflectivities based on the warranty specifications. Type VIII Type IX-VIP Type IX-Omniview Proposed Type XI Luminance [cd/m^2] 0 2nd chevron 40m into the curve,.5m high, 6ft offset from right edgeline, curves reflect expected end-of-warranty Type VIII retroreflectivities based on the warranty Type IX-VIP specifications. Type IX-Omniview Proposed Type XI Distance to Chevron [ft] Distance to Chevron [ft] Luminance [cd/m^2] 0 3rd chevron 80m into the curve,.5m high, 6ft offset from right edgeline, curves reflect expected end-of-warranty retroreflectivities based on the warranty specifications. Luminance [cd/m^2] 0 4th chevron 20m into the curve,.5m high, 6ft offset from right edgeline, curves reflect expected end-of-warranty retroreflectivities based on the warranty specifications. Type VIII Type IX-VIP Type IX-Omniview Proposed Type XI Type VIII Type IX-VIP Type IX-Omniview Proposed Type XI Distance to Chevron [ft] Distance to Chevron [ft] Figure 23. Computer evaluation of end-of-warranty luminance values for four consecutive chevrons placed on the right shoulder, 6ft offset from the right edgeline of a 200m radius right curve after a 200m straight approach, starting at the beginning of the curve. Chevrons are 5ft high and have no tilt, and are placed according to MUTCD guidelines. Each figure shows the distance to the corresponding chevron. Analysis conducted for an NCHRP 4-29 Heavy Vehicle equipped with the 50 th percentile 2004 UMTRI US low beam headlamp. 4

42 000 st chevron at the beginning of the curve,.5m high, 6ft offset from right edgeline, curves reflect expected end-of-warranty retroreflectivities based on the warranty specifications. 00 Luminance [cd/m^2] 00 0 Type VIII Type IX-VIP Luminance [cd/m^2] 0 Type IX-Omniview Proposed Type XI nd chevron 40m into the curve,.5m high, 6ft offset from right Type VIII edgeline, curves reflect expected end-of-warranty Type IX-VIP retroreflectivities based on the warranty specifications. Type IX-Omniview Proposed Type XI Distance to Chevron [ft] Distance to Chevron [ft] Type VIII Type IX Type IX-Omniview Proposed Type XI Luminance [cd/m^2] 0 3rd chevron 80m into the curve,.5m high, 6ft offset from right edgeline, curves reflect expected Type VIII end-of-warranty retroreflectivities based on the Type IX-VIP warranty specifications. Type IX-Omniview Proposed Type XI Distance to Chevron [ft] Luminance [cd/m^2] 0 4th chevron 20m into the curve,.5m high, 6ft offset from right edgeline, curves reflect expected end-of-warranty retroreflectivities based on the warranty specifications Distance to Chevron [ft] Figure 24. Computer evaluation of end-of-warranty luminance values for four consecutive chevrons placed on the right shoulder, 6ft offset from the right edgeline of a 200m radius right curve after a 200m straight approach, starting at the beginning of the curve. Chevrons are 5ft high and have no tilt, and are placed according to MUTCD guidelines. Each figure shows the distance to the corresponding chevron. Analysis conducted for an NCHRP 4-29 SUV equipped with the 50 th percentile 2004 UMTRI US low beam headlamp. 42

43 ATTACHMENT

44

45

46

47 ATTACHMENT 2

48

49

50

51