Researchers Program Simon To Bust Its Moves More Like A Human
Humans derive 50 percent to 90 percent of all meaning from nonverbal communication. Researchers at Georgia Tech discovered that people better recognize what robots are doing when the automatons act in a more human-like fashion, with one movement leading into the next.
The researchers studied how easily people could tell what a robot doing by watching it move. The team had programmed the robot, Simon, based on human movements taken in a motion capture lab. They allowed more joints to move at the same time and made movements flow into each other. Optimizing Simon’s movements made him appear more human.
Ph.D. student Micheal Gielniak said that robot movements are unlike human, robot motion is typically characterized by jerky movements with a lot of stops and starts. When the motion was more human-like, human being will more easily able to perceive what the robot was doing. In the future, they will look at getting Simon to perform the same movements in various ways.
Published at :