Are Sunglasses Proper for Driving? Investigation and Prototype for Public Testing


 Background: Sunglasses safety is important regarding driving conditions and is requirement of sunglasses standard ISO 12312-1, for testing traffic light and visible transmittances on lenses. Methods: We have spectroscopically investigated 232 sunglasses regarding compliance with standard, in the 380 nm – 780 nm range, with 5nm steps. Category (0 – 4) were determined as well as Q factors, which evaluate color distortion for the vision on sunglasses. Furthermore, we developed a prototype for public to self-check sunglasses regarding safety for driving. Combination of white light illumination, 4-channels photo sensor detection and calculation, allows testing simultaneously the colors of traffic light, as well as the visible light transmittances (categories). Q factors are provided by the prototype. Results: Spectroscopy shows that 75.43% complies with ISO 12312-1:2013. Prototype was validated by testing 232 sunglasses, with trained user and 60 from this set were double-checked with non-trained users, and compared to spectroscopic measurements. Bland-Altman analysis presented non-significant biases and narrow 95 % limits of agreement within pre-defined tolerances, for all measurements. Only 2 samples differed in category measurement. Conclusions: Sunglasses on the market should adjust for driving conditions. Prototype it is a fine set up for guiding users on checking proper sunglasses for safe driving Significance: Immediate attention on checking sunglasses for driving conditions is needed.


Background
Efforts of our team has been progressively done regarding sunglasses standard (Masili et  Previous research of our team was the development of a kiosk for the public to self-check their own sunglasses regarding the ultraviolet (UV) protection, as required by the standards (Mello et al. 2014). In that system, the test provides a report of the category of the lenses (0 -4) as well as UVA and UVB protection associated to the sunglasses´ category. Brazilian national survey in order to improve nationalization of sunglasses standards; and studies conducted on revisiting requirements of worldwide sunglasses standards, in which this work is inserted. So, this work is one of the many systems that our lab is developing (Mello et  The public is usually interested on sunglasses style, eye comfort and UV protection when buying a pair of shades. Although these are relevant aspects, several others are noteworthy. One is its safety for driving (Nathan, 1964;Phillips, 1975, Palmer 1997Dain, 2009). Lenses excessively dark could hinder object and traffic signal recognitions at safe distances, particularly in shadow.
Moreover, lenses could excessively enhance or attenuate some colors, leading to distortion on 4 color perception. Thus, the choice of improper sunglasses may ultimately lead to dangerous situations (Dain 2003;Hovis 2011).
The international standard ISO 12312-1 establishes requirements for all afocal sunglasses and clip-ons, intended for protection against solar radiation for general use, including road use and driving.
The standard grades sunglasses in five categories (0 -4) according to visible transmittance (380 nm -780 nm range) of their lenses, i.e., depending on the level of sun glare reduction of their lenses. Sunglasses are recommended for specific situation according to their category. The categories are rated between 0 (clear lenses) and 4 (very dark lenses). Category 4 sunglasses are recommended for environments with high solar incidence, but are not proper for driving.
Although not recommended for normal conditions, category 4 filters may be recommended in extreme high luminance conditions, such as desert and snowfields under full sunlight.
If the transmission of the lens in the visible spectrum (380 nm -780nm) is less than 75 %, sunglasses should not be used for road and driving in twilight or at night. Those with transmittance less than 8 % (category 4) are not appropriate for driving at any time. Additionally, regarding road use and safe driving conditions, the spectral transmittance of filters for the 475 nm -650 nm range shall be not less than 0.2 times the visible transmittance. These specifications are for regular tinted sunglasses, photochromic sunglasses lens for road use and driving requires additional steps described in standards (ISO 2013a).
The visible transmittance, , is defined as in which (λ) is the spectral transmittance of the filter, V(λ) is the spectral luminous efficiency function for photopic vision, (λ) is the visible part of the solar spectrum at sea level for air mass 2 (terrestrial solar spectrum occurring when the sun's position vector is 60.11 degrees from the zenith), (λ) is the luminous weighting function (ISO 2013a).

6
Each category has a recommended use. Darker lenses (less luminous transmittance) are recommended for environments with higher solar incidence, while lenses with luminous transmittance inferior to 3 % are not considered appropriate to be used.
Traffic light transmittances, , are defined for red, yellow, green and blue (some countries has a blueish type of traffic light) as: where, (λ) is the spectral energy distribution of the red, yellow, green, or blue traffic signals, which is different for traffic lights lit by incandescent and LED lamps and is available in the standards (ISO 2013a). There is also a relative visual attenuation quotient for the three colors of traffic light detection, denoted by , is and defined as As stated, traffic light transmittances and consequently, relative visual attenuation quotients are defined for incandescent and LED lights; however, standard requirements take into account only the relative visual attenuation quotients (Q factors), which are related to incandescent lights.
According to the standards, for sunglasses of categories 0, 1, 2 and 3, luminous transmittance should be equal or greater than 8%, and Q factors shall not be less than 0.80 for the red signal light, and not less than 0.60 for the yellow, green and blue signal lights. It is important that community should have a way to test their sunglasses to assess their optical

Results
A set of 232 sunglasses was tested, and 222 were unbranded sunglasses.
The relation of unbranded (222) and tagged sunglasses (10)    Analysis of agreement between our system and gold standard spectrophotometer was assessed by using the Bland-Altman method.
For luminous transmittance, the bias adopted as significant should be greater than 0.5%. The 95% interval of agreement was considered wide if the upper limit is greater than 5% or the lower limit is lesser than -5%. For the Q factors, the bias was adopted as significant if it is greater than 11 0.1. The 95% interval of agreement was considered wide if the upper limit is greater than 0.3% or the lower limit is lesser than -0.3%.
Additionally, sunglasses lenses have been tested on the prototype, by non-trained users, which The bias and the 95 %-limit-of-agreement interval are shown in Table 2.
These values are within pre-defined tolerances.
Bland-Altman plots ( Figure 7) indicated consistent variability across the graphs, without trends, for all plots. On average, prototype measures were lower than gold standard ones except for . 13

Discussion
Bland-Altman plots presented non-significant biases and narrow 95 % limits of agreement within pre-defined tolerances, for all plots.
The plots also indicated consistent variability across the graphs, without trends, for all plots.
Therefore, prototype measurements presented good accuracy compared to spectrophotometer gold standard measurements within pre-defined tolerances.
Although one of the lenses diverged in category measurement, for that particular lens, transmission was in the range of category overlap, bearing the limit of the range (2.3 %).

Conclusions
The purpose of this work was to evaluate a set of sunglasses, which represents the Brazilian The results show that the prototype is a worthwhile system for providing to lay public access to information, as well as educating the public, since it is a quite important safety item to be considered when purchasing a pair of sunglasses.
The easiness on testing individual´s own sunglasses and its self-explanation screens might have a potential market for retailers, as well as adding value to their product, since it provides knowledge to salespeople who also do not commonly have access to this type of information. It may also be used in the stores for checking their own sunglasses as well, for the consumer, working as an additional advertising of their product.
Immediate attention of population should be called about checking their sunglasses for driving conditions before purchasing them. Likewise, the good job that the market has done over the years about the UV protection on sunglasses filters, advertisement should be done on driving conditions.

A. Spectroscopy Measurements
Electromagnetic spectroscopytransmittancewas performed on sunglasses with the CARY 5000 (VARIAN) spectrophotometer, which is a double-beam system, in the visible range, from 380 nm -780 nm, with 5 nm steps.
The standard also establishes that spectroscopic measurements should be performed so that the optical path from the light source to the sensor passes through the geometric center of the lens, in a region of 5 mm in diameter. For linear response photo sensors, the ratio of the measured value of the sample and baseline value is equal to the light transmittance, weighted with a spectral function given by the term-toterm product of the spectral emission of the light source and the spectral response of the photo sensor. If a photo sensor has no linear response, a simple mathematical correction should be made.

17
The TCS3472 photo sensor provides four different responses; therefore, it is possible to measure the transmittance of the sample with four different known weighting functions ( Figure 3).
Furthermore, the four known weighting functions are linearly combined to estimate the desired weighting functions for each type of measurement.

Consent for publication
Not applicable.