During the coronavirus pandemic, social distancing proved very effective in slowing the spread of the disease. But while millions of people remain at home to work together to defeat a dangerous infection, many employees in the food, mining, pharmaceutical and other industries still have to go to work every day. They depend on them to make sure that a food delivery person comes to you tomorrow, or that the nearest pharmacy is always open in case you suddenly have a toothache. So how to ensure safety in these cases? In addition to wearing masks, there is another way to control the distance between people. And artificial intelligence is a big part of it.
To help control the distance between people in the workplace, the startup Landing AI has developed an artificial intelligence-enabled tool to track people’s whereabouts. It can determine whether people are at a safe distance from each other by analyzing video streams from surveillance cameras in real time.
For example, a factory that makes the masks and other protective equipment so needed during a pandemic could integrate this software into its surveillance camera systems. As shown in the demo below, the detector will highlight in red people whose distance between them is less than the minimum allowable distance and draw a line between them to call attention to it. The system will also be able to loudly alert and remind people to keep a safe distance.
The software is installed on the computers that control the security system with video surveillance, and the artificial intelligence itself integrates into the elements it needs. It is a kind of “mini big brother” who monitors every movement of the employee. But not to control whether he’s working or not, but to keep employees at a safe distance from each other. I remember seeing something like that in the TV series “Suspect” – there was an artificial intelligence machine that recognized the movements of every person in the city. It wasn’t really to control the distance between people, though.
Nothing prevents the algorithm from being refined in the future and using such a neural network to monitor a person’s movements on city surveillance cameras. The only thing that this would require is high-quality optics that would be able to recognize a face and characteristic features (gait, physique) without errors. And if a person commits an offence and the camera notices him, later on the neural network will follow him home. We think that such a system will be implemented in the future.
According to Input, Amazon has already implemented similar technology in its warehouses. Employees are also monitored during their shifts and warned that violating social distancing rules could get them fired.
