They create an ultra-fast bionic arm capable of catching objects on the fly

They create an ultra-fast bionic arm capable of catching objects on the fly

Developed by Swiss researchers, it is capable of reacting in less than five hundredths of a second

    A robot developed by researchers of the EPFL (Federal Polytechnic School of Lausanne), in Switzerland, is able to react at the moment and capture objects with complex shapes and trajectories in less than five hundredths of a second. With his palm open, the robot is completely immobile. A fraction of a second later, he suddenly unrolls and catches all kinds of flying objects thrown in his direction: a tennis racket, a ball, a bottle.

   This arm measures 1.5 meters in length and maintains a vertical position. It has three joints and a sophisticated hand with four fingers. It was programmed and designed in the Learning Algorithms and Systems Laboratory (LASA) of this organization to test robotic solutions for the capture of objects in movement. It is unique in that it has the ability to trap projectiles of different irregular shapes in less than five hundredths of a second.

This invention is described in an article published in the leading magazine in this field, IEEE Transactions on Robotics. "Increasingly present in our daily lives, robots can also catch or dodge complex objects in complete motion," said Aude Billard, head of LASA. "We not only need machines capable of reacting on the spot, but also to predict the dynamics of the moving object and generate a movement in the opposite direction."

Remove space junk

This robotic arm already has a very real potential application in space. It has partnered with the Clean-mE project carried out by the Space Center of the EPFL in Switzerland, which aims to develop technologies for the recovery and disposal of space debris in orbit around the Earth. Conditioned in a satellite, the arm would have the task of capturing the debris in orbit, whose dynamics is only partially known. Therefore, the robot will not be able to work on that dynamics accurately until it is in space, by observing the movement of the approaching objects.

The ability to capture things that fly requires the integration of various parameters and react to unforeseen events in record time. "Today's machines are often pre-programmed and can not quickly assimilate data changes," Aude Billard added, "so your only option is to recalculate the trajectories, which requires a lot of time in situations where each fraction of a second can be decisive. "

To obtain the speed and adaptability desired, LASA researchers were inspired by the way humans themselves learn: by imitation and trial and error. This technique, called programming by demonstration, does not give specific instructions to the robot. Instead, it shows examples of possible trajectories to it. It consists of guiding the arm manually towards the projected objective and repeating this exercise several times.

The robot also learns

The investigation was carried out with a ball, an empty bottle, a half-full bottle, a hammer and a tennis racket. These five common objects were selected, since they offer a varied range of situations on the part of the object that the robot has to grasp: the handle of the racquet, for example, does not correspond to its center of gravity. The case of the bottle even offers an additional challenge, since from its center of gravity it moves several times during its trajectory.

When it is projected in the air, all these elements will make movements even more complex movements, often with several axes. As a result, when objects in motion are assimilated by the capabilities of the robot, the results are quite interesting.

In the first phase of learning, objects are thrown several times in the direction of the robot. Through a series of cameras located around it, the robot creates a model for the kinetics of objects based on their trajectories, speeds and the movement of rotation. Scientists translate them into an equation that then allows the robot to quickly position itself in the right direction each time an object is launched. During the few milliseconds of the approach, it refines the machine and corrects the trajectory for a high-precision, real-time capture. This efficiency is further improved with the development of controllers that associate and synchronize the movements of the hand and fingers.