In the video / "Cassie" robot ... a robot that climbs stairs in the dark and delivers parcels in a foggy atmosphere.
American engineers have created a robot that can easily climb stairs in the dark, and it can be used safely in guarding homes and delivering parcels even if the weather is foggy or dark.
The robot "Cassie" climbed the stairs at Oregon State University 80% mastery, without eyes or other sensors.
Scientists trained the two-legged robot to use "proprioception," or body awareness - to navigate uneven surfaces.
This is important, the researchers suggest, if fog, dim lighting, or other factors that limit the robot's acumen or visual ability are present.
Cassie was developed thanks to a dynamic robotics tutor at Oregon State University in 2017.
Although Cassie has no head, her legs and thighs have flexibility that allows her to move her legs forward, backward and sideways, as well as rotating them at the same time.
Additionally, Cassie's powered ankles allow her to stand in place without having to constantly move her feet, shift her weight, or fidget from the long wait.
Cassie was classified as a "walking robot" that could walk more smoothly, human-like and closer to it than other robots, giving her the skill to traverse the complex terrain that usually hinders robots, such as climbing and descending stairs, running, overcoming potholes, potholes and obstacles.
And Cassie was already able to trek off-road and into mountains, storming flames and successfully riding sleds without tripping.
The American researchers were interested in training Cassie to climb "blind" stairs, that is, without seeing a computer or other sensors.
Although most robots are able to overcome stairs with cameras and computer vision, certain conditions, such as low light, are not always ideal for visual input.
Scientists wanted Cassie to go up and down the stairs using "proprioception" or body awareness - in the same way that you or I might creep upstairs at night.
Using a simulator, they taught Cassie how to handle a number of situations, including stairs and flat terrain, thus avoiding "many falls and accidents, especially early in training," the team said.
Finally, it was time to learn how to visualize with the real world and to pass the practical test. Researchers escorted Cassie to the campus, and watched her work with barriers, ladders, and other uneven surfaces that she had never encountered before.
Out of 10 experiments, Cassie had an 80% success rate of climbing stairs and a 100% success rate in descending.
"As far as we know, this is the first control device for a bipedal robot, on a human scale, capable of crossing a variety of realistic stairs and other stair-like obstacles using only proprioception," the research team wrote.
"Cassie had some reservations, including the need to stay at a steady speed up and down so you don't get messy or go haphazard," they added. Indeed, this work demonstrated surprising capabilities for blind mobility of robots, leaving the field open for further development.
Watch the Video
https://www.alroeya.com/?utm_source=KwikPlayer&utm_medium=KwikShare&utm_campaign=KwikMotion
Source:1
