From: https://www.youtube.com/watch?v=L3-XaUzaJHU
Image from: https://www.visualcapitalist.com/the-math-behind-social-distancing/

Social Distance Detector (click picture):

QUESTION 1: Was your social distance detector effective at detecting potential violations?

Yes, I believe that my social distance detector was effective at detecting potential violations. It’s hard without a scale, but there is a correlation between what the model marks as safe and what the model marks as a violation.

QUESTION 2: Do you think this approach would be effective for estimating new infections in real time? How would you implement such an approach in response to the COVID-19 pandemic we are currently experiencing?

I believe that this approch could be effective for estimating new infections in real time. However, at this time I believe that there are too many variables in the spread of COVID-19 that a one size fits all solution would not be effective. For example, just a few weeks ago at the Florida town hall meeting, I was shocked and amused by the people who gathered to talk about how wearing masks would restrict oxygen flow to the body. Not only that, but the homeless people who have no healthcare and no access to masks are at risk of infection and by extension infecting others. Additionally, people who ignore guidelines and unknowing superspreaders who either don’t know they have COVID-19 or don’t care also contribute to the spread. Implementing this could help people understand how much space is too little, for example, not everyone knows how large of a distance 6 feet really is. This could help in that situation, however, this social distance detector does not address the problems mentioned above.

QUESTION 3: What limitations or improvements might you include in order to improve your proposed design?

One limitation is that this social distance detector does not detect masks. Masks significantly reduce the wearer’s chance to infect another person and reduce the wearer’s chance of getting infected. As such, I believe that adding a mask detector could be good in differentiating between the contamination rates. There are some experts who advocate 32 feet, and I believe that it is a good rule for those who do not wear masks, since bodily fluids such as sneezing and coughing are propelled through the air very quickly. The social distance detector could be improved such that it can differentiate between those who wear masks and those who don’t. Therefore, the detector could apply different rules (maybe 32 for non-mask wearers and 6 for mask wearers) to each person to make a better social distance detector. Also, I noticed that the detector also did not differentiate between families and non-families and instead calculated each person as an individual, increasing the distance violations. I believe that it would be better if the detector was trained to recognize families due to their distance. Also, another camera could be put at the parking lot to detect families that leave a car together. They would be classified as a singular individual as opposed to multiple which would falsely increase the distance violations registered by the social distance detector.