Editor’s note: Find the latest COVID-19 news as well as instructions in Medscape’s Coronavirus Resource Center. Coronavirus Resource Center.
A robot that is able to spot up to six people within 6 feet of each other could be a crucial tool to maintain social distance during a upcoming or current pandemic.
The robot spotted and walked towards groups of two to six people, advising them that they were too close to one another to be tested in the beginning.
The COVID Surveillance robot (CS-robot) discretely displays an on-screen message to alert people who are too close. Researchers from the University of Maryland developed the technology. It also comes with thermal cameras that detect people who are hotter than others and alert them to fever. The robot is able to notify security or health personnel.
The study was published Dec. 1 in PLOS One.
A Sense of Surprise?
It is unknown what the public’s reaction will be to a robot referee. “We have tested mostly in our lab and buildings and some public events,” said the lead study author Dinesh Manocha PhD.
“Many times, humans are shocked when they see a robotic vehicle moving around them or displaying these messages. Robots like this aren’t widely used and it’s difficult to guess public perception to their use,” said Manocha, an associate professor of computer science and electrical and computer engineering at the University of Maryland in College Park.
I’m interested to know what the psychology of this would be, what the sociology would be -What is it to be reprimanded by a machine?
Bruce Hirsch, MD, says there are a variety of things you can think about when it comes to the way people react.
“I’m interested in seeing what psychology will look like, and what the sociology might look like — what is it that would cause a robot to reprimand you?” He said.
“There is already a depersonalization of the health care system, particularly with the COVID epidemic. Hirsh, an infectious diseases doctor at Northwell Health in Manhasset (NY), stated that a machine wouldn’t be perceived as supportive and helpful.
Manocha The first author, Adarsh Jagan Sathyamoorthy and his colleagues tested the autonomous robot in five scenarios within an indoor area measuring 4 meters by 4 metres (about 12 feet by 12 feet). They evaluated the accuracy of the robot in identifying people, detected social breaches, and tracked people standing or walking.
Researchers found that the robot was able to perform in non-laboratory indoor settings, like lobbies or narrow corridors that have different levels of lighting.
The study also compared results of the robot by itself to the robot plus input from wall-mounted closed circuit television (CCTV) cameras. The robot was able to detect and alert those who were not complying with its instructions, however the use of CCTV data significantly improved its performance in all situations.
On the Move
A “lawnmower strategy” allows the robot to cover any gaps in CCTV coverage. It moves to fixed positions and examines its surroundings.
The robot makes use of its cameras and sensors to “lock” to a person who is not socially distancing itself to guide its movement. The CS-robot could locate the position of the person within 0.3 meters during a series of tests. This was regardless of whether the person was standing or walking.
Interestingly, the robot also considers time in addition to proximity. The engineers were trying to avoid flagging less risky situations, such as when two people walk within 6 feet in an aisle.
Engineers also created the robot to stay clear of obstacles and other people. Their Frozone collision avoidance strategy, which was described in an prior study, allows robot’s “red green, red and depth” (RGB–D) camera to track people and predict their movements.
The robot is able to prioritize its actions in order to attend to larger groups or follow people in motion based on their proximity to it.
People Remain Anonymous
Privacy concerns can be raised by robots following you or approaching you with cameras.
The researchers have explained this by using standard de-identification techniques like visual redaction of images for gestures, faces, and gait data. The robot’s visual camera assigns a random number to each person in a group.
The robot decreases risk for humans. The authors pointed out that remotely monitoring temperatures of people reduces the risk of security and health personnel who contract coronavirus.
Still a Role for Humans
The advantages of the robot Hirsch said, included freeing employees from having to watch people. The technology could also make monitoring more reliable.
But, he said, “I’m older and I like people.”
Humans are still more likely to assist in the field of education, which is why social distancing is essential — as well as “reading certain situations,” Hirsch said.
He gave an example of a person who had lost a loved-one and was in tears in a hospital.
“Is a robot going to be able to walk up to that person in a potentially sensitive moment and ask them to move 6 feet away from another person or to turn them in for additional health care monitoring?”
Manocha noted that, although robots were tested in indoors only, “we expect robots to be able to be used outdoors or indoors when humans are in close proximity,”
Researchers could also think about alternative methods that aren’t screen message, to alert people who aren’t distancing. This could include sounds or other signals to warn.
Manocha noted the robot doesn’t differentiate between people who are not from the same household. The authors also stated that “we need to improve humanrobot interaction approaches.”
The investigators also want to study the social impacts of robots.
“We also want to develop methods for detecting if the people in the robot’s surroundings are wearing masks,” they noted.
PLOS One: “COVID surveillance robot”: Monitoring social distancing restrictions in indoor environments.”
Dinesh Manocha PhD, professor of computer science, University of Maryland College Park.
Bruce Hirsch, MD, infectious disease doctor, Northwell Health, Manhasset, NY.
Content Source: https://www.medscape.com/viewarticle/964719?src=rss