You have heard Artificial Intelligence’s growing application in tracking the progress of Covid-19. You might have also noticed how AI tools have been helpful in ascertaining and maintaining the report on its spread. But, do you think Ai could determine which face masks work best for Covid-19? Possibly.
No other object would better symbolize the pandemic than the face mask. They have become the world’s most coveted commodity. Even Prada and Gucci have started manufacturing them. There is no official statistic as to how many face masks have been created and sold, but the count could go well above 7 billion – equivalent to the population of the world – since everyone owns a mask now.
However, it has been observed that the mask reveals far more than what it is expected to cover. Not all masks popping out on the store fronts in different colors and designs prove their effectiveness against Covid-19.
Ever since countries have stated mandatory face mask orders, almost everyone capable has started their own small-scale industry. Clearly, not every mask out there is N95 standard, but it helps curb the spread of the virus to some extent. However, the recent cases of people attracting the virus even with a mask have started creating concerns about its robustness.
Mask makers barely have any recognized standard or governmental instructions to look up to when manufacturing them. And, it’s not their fault. The pandemic never left much time for any authority to set up the required institution.
Face masks and AI
The problem, however, existed and someone at the Duke University was watching.
And they though of merging face masks and AI. Together, face masks and AI can work better than both working solo.
About the device
A new technique developed at the university uses a new technique to figure out which masks work best. The measurement of effectiveness is based on counting the respiratory droplets the masks emit during speech.
The equipment is a small black box that comes with an in-built laser and a cell-phone camera.
The box has a big hole in one end and a camera perfectly placed at the centre of the other end. A laser passes between ends. A person with any mask places her/his mouth at the big hole of the box and speaks, “Stay healthy, people.”
As s/he repeats the sentence, the airborne droplets coming out of the mouth of the speaker hits the laser beam. Next, the camera records the water particles emitted. The final stage has an AI-based feature-detection algorithm that calculates the number of droplets emerging by measuring the amount of lights getting scattered during the speech.
This was done with 14 most common types of masks. A patch of mask material was separately tested. Finally, professionally fit-tested valved and non-valved N95 masks were tested.
Cotton coverings came third in the test. They were led by 3-layer surgical masks. Fitted N95 masks led both of them.
The only difference between valved and non-valved N95 masks was, the latter was only beneficial to the user. This was further explained in a study by the researchers, “While the valve does not compromise the protection of the wearer, it can decrease protection of people surrounding the wearer. In comparison, the performance of the fitted, non-valved N95 mask was far superior.”
Advice for mask makers
Researchers have claimed that the cost of the equipment would barely cross $200. Hence, they suggest mask-making organizations to use the method and test out their designs before going for large-scale production.
Until a vaccine is here, masks are your only respite. It is the responsibility of everyone to ensure that faulty masks do not get sold or distributed.
We are sure many of you reading this blog would be using some kind of mask right now. But now that you know which mask stands where, go ahead and find the right mask for you and your family. Share this blog and let others know this vivid truth about masks.
Stay healthy, people.