Weekly email of our best stuff
Hello and welcome to contact us page at eProfits
How can we help you today?
Researchers at the Swiss Federal Institute of Technology (EPFL) have developed a robotic arm that can easily catch ...
Researchers at the Swiss Federal Institute of Technology (EPFL) have developed a robotic arm that can easily catch objects – and the most impressive part is, that it has learned to do so by observing human movements.
Ashwini Shukla, a PhD candidate who was has been working on the algorithms for the robotic arm, explains that the reason why the arm moves with such precision is because the robot can predict the flight trajectory of the object, hence, the actual catching happens at the predicted location.
And how does the robot predict? In order to build the mathematical model, the researchers had to throw objects towards the arm, but they didn't expect the robot to catch them. Instead, the arm simply observed the objects' flight paths with 18 infrared cameras. The robot's “data glove” collects data from the way humans move their hands and fingers to catch things, and by observing these movements, the robot learns how to do it itself. Having collected the data, the robot would predict the motion of the object and where it would land.
With its upright, meter and a half, three joints and four finger arm, this amazing robotic arm manages to tackle a tennis racket, catch a ball, and a bottle in a blink of an eye. See how it's done:
The robotic arm was programmed at the Learning Algorithms and Systems Laboratory, in the French-based EPFL, with the purpose of testing robotic solutions for grasping moving objects. The Head of the Laboratory, Aude Billard, says, “Increasingly present in our daily lives and used to perform various tasks, robots will be able to either catch or dodge complex objects in full-motion.”
For example, utilizing this technology, a self-driving car would be able to react to other traffic or pedestrians who cross the street with the same speed.
Billard also adds, “Not only do we need machines able to react on the spot, but also to predict the moving object’s dynamics and generate a movement in the opposite direction.”