Design Criteria

1. The cleaning device must clean the lens leaving it free of smudges or dust/ dirt particles.
2. The cleaning device must not scratch the eyeglasses
3. The cleaning device must be easily transported.
4. The cleaning device must weigh no more than 1 lbs.
5. The cleaning device must not take more than 60 seconds to completely clean lenses.
6. The cleaning device must cost no more than $30.
7. The cleaning device must not produce excessive amounts of waste
8. The cleaning device must clean the tight spaces near the hinge and nosepiece
9. The cleaning device must be visually appealing.

Cleaning Test

The following sheet helped us to keep track of the cleanings as outlined in Element H. The cleaning test was performed for two reasons:

• To determine the cleaning ability of our device
• To test for scratches developing on the lens after many cleanings

The chart below helped us keep track of the cleaning reps as we completed them, when pictures were taken, visible smudges in the light (note: in the light to our eyes, not in the pictures below), and as a result, if the cleaning brush needed to be replaced. This chart helped us to call attention to a couple of issues with our testing procedure and our design. Our initial procedure called for use of butter spray to simulate oils that accumulate on the lens, as well as sand to simulate larger particles that are more damaging. During only the second cleaning using the prototype brush, we realized that the brush had become soiled from the first cleaning, and would no longer perform as desired. This could be attributed to the "heaviness" of the butter spray, and more importantly, the small size of the brush. Not only was the brush becoming completely saturated from oils, but it fails to remove the larger particles and actually picked them right up leaving them to further scratch the lenses. In order to continue our testing procedure we used a microfiber cloth instead of our little brush. This was justified as our design was to essentially be a microfiber cloth style brush. We also used clearly visible fingerprints on the lens rather than the butter spray. Fingerprints simulated natural oils building up much better than butter spray. Note that beyond cleaning number two, the brush replacement column becomes void.

Control Lenses

Below are three different lenses fresh out of the package. These control lenses bring attention to the fact that a lense that is looks clean in the daylight may still be covered in smaller particles. The appearance of these particles poses some questions: To what extent must our cleaning device clean a lens? When particles that remain on lenses are only visible under heavy scrutiny, is the lense still considered clean? These questions may need to be answered through further research.

Test Lense

Why Average Values?

RGB Color values range from 0-255, 0 being completely off or black and 255 being completely on or maxed out for that color. The combinations of these three colors at different intensities give us the visible spectrum. Each combination of values for red, green, and blue correspond to exactly one pixel. These values must be averaged across the lenses due to the way that cameras form pictures. As seen below, two pixels side by side, seen from far enough away look just fine, but when selecting a color value can vary quite wildly.

Interpreting the Data

Our overall experiences as well as the data show that our device requires further development and testing. Our data does not show a clear result to criteria one and two. This could mean that more cleaning repetitions are needed, a different brush design, a different way to average color values, or more likely, all three. It is very possible that in order for microscopic scratches to develop on the lens, many more cleaning repetitions are needed, but it is also possible that our device succeeded in criteria two. Our brush design was centered around microfiber due to it's softer cleaning ability, as well as its reputation among optometrists as seen in our research. Our research implies that our design should not scratch lenses, but testing was necessary to confirm in a scientific manner. It is also possible that our method of detecting and measuring light reflected out of the lens needs to be changed or refined. With more time and resources this would be possible. In short, our current design failed to leave the lens smudge free, and may or may not have succeeded in leaving the lens scratch free.

Future Testing Recommendations

More Accurate RGB Value Testing

Averaging the RGB values for nine pixels across the lens may not be the best way to determine the "average" color seen through the lense. Given more time we may be able to research a more accurate way to determine measurable color values. This may be possible through software or hardware that we did not use.

Pinpoint Light Through Lens

Some of the white glare seen in our testing method across the lense may be observed due to light hitting the front or back of the lense rather than only being directed through the edge. We tried to prevent this by directing the light source with electrical tape, but ensuring light is only traveling through the side of the lense will provide more accurate results.

Different Testing Method

Our initial research showed that a different testing method could be used, such as surface observation technology. Technology like that would allow us to take a deeper look at the lenses and examine what was actually causing the scratches and how to prevent them. However, this technology costs multiple thousands of dollars. It is quite possible that the most effective testing method to observe microscopic scratches is entirely different than the method that we used.

More Cleaning Repetitions/ Automation

Cleaning the lenses more than 50 times, even upwards of 1000 cleaning, would be ideal. This test is meant to simulate cleanings over the lifetime of the glasses spanning possibly over multiple years. An automated device to simulate cleanings would be ideal. This would remove any user error/ uncontrolled variables.

Soiled Lens Simulation Improvements

Our initial testing used butter spray and play sand to simulate dirty lenses. Butter spray simulated grease on the lens far too heavily. Our testing results would be more accurate if our method of simulating a soiled lens were more true to life.