There are some jobs, though, robots struggle to complete, like dressing a human.
There are a lot of things robots are good at. Automating manufacturing processes and helping humans accomplish repetitive tasks are just two examples. The use of robots in factories and other industries for carrying out a number of menial jobs is not rare but researchers have been able to take a huge step in using robotics for far bigger functions. There are some jobs, though, robots struggle to complete, like dressing a human. The nuance, gentleness, and responsive touch required are hard for an algorithm or bot using machine learning — especially without hurting the subject — but that’s exactly what scientists at MIT wanted to conquer with new research on the robotics movement.
Easy Does It
A short field of vision is one of the challenges in creating a robot that can dress humans. In an interview with MIT’s media arm, the researchers said that a robotic arm that helps a person pull on a jacket sleeve can’t see the entire arm and needs to calculate the position. The scientists plan to expand the types of material the robot can use as well as make them personalized and better able to adapt to individual movement patterns.
Researchers in Japan recently unveiled a robot that was able to complete the task of peeling a banana without causing any harm to the fruit. The video released by the University of Tokyo showed that the robot was able to pick up the banana and peel it in three minutes using both hands.
According to Reuters, the robot was able to successfully complete the task after around 13 hours of training but the success rate of the dual-armed robot still remains at 57 percent for the time being.
The post MIT’s Trained Robot is Capable of Dressing a Human (Perfectly) appeared first on Analytics Insight.
