Swiss Researchers Develop A Bionic Arm With Lightning Fast Reflexes
[ Watch the Video: Ultra-Fast Robotic Arm Can Catch Objects On The Fly ]
Lawrence LeBlond for redOrbit.com – Your Universe Online
Researchers from the Learning Algorithms and Systems Laboratory (LASA) at EPFL in Switzerland have developed a bionic arm that can catch a multitude of objects skillfully with little or no need for luck.
The arm, which measures nearly five feet long and keeps an upright position, remains completely motionless with an open palm. Then, when something flies into its field of vision, it grabs it out of the air with split-second accuracy. In an experiment, the robot easily caught a tennis racket, a ball, a bottle, and a hammer. The arm has three joints and the sophisticated hand has four fingers.
The LASA team designed the arm to test robotic solutions for capturing objects. It is unique, as it has the ability to catch projectiles of various shapes and sizes in less than five hundredths of a second. The team unveiled the project in a new paper published today in the journal IEEE Transactions on Robotics.
“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,” said Aude Billard, head of LASA. “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.”
The robotic arm already has real-world potential for space-based applications. EPFL’s Swiss Space Center is already tying this new arm to the Clean- mE project, which aims to develop technologies for the recovery and disposal of space debris orbiting Earth. By fitting this lightning fast arm on a satellite, it would have the ability to catch debris as it passes by.
As it stands now, it is impossible to predict the flight dynamics of space debris. By putting the robot arm in space, the team says it can study such dynamics with precision by observing approaching objects. Once it can determine flight dynamics, it can grab such objects rather easily.
However, it requires the integration of several parameters and the reaction to unforeseen events over milliseconds, perhaps, in order to have the skill to accurately grab objects in space.
“Today’s machines are often pre-programmed and cannot quickly assimilate data changes,” added Billard. “Consequently, their only choice is to recalculate the trajectories, which requires too much time from them in situations in which every fraction of a second can be decisive.”
The team was inspired by the way humans learn – via imitation and trial and error – to obtain the desired speed and adaptability for their robotic arm. The robotic arm, however, doesn’t necessarily rely on human learning to achieve its goals. Instead, it learns by studying possible trajectories, consisting of manually guiding the arm to the projected target and repeating the exercise several times.
The objects used in the experiment – four objects in all (however, the bottle was used either empty or half full) – were selected because they offer a varied range of situations in which the part of the object that the robot has to catch (as in the racket and the hammer) does not correspond to its center of gravity. The half-full bottle offers additional challenges since its center of gravity changes several times during its trajectory. As these objects fly through the air, they make complex movements that involve several axes. As a result, when the moving objects are submitted to the robot’s abilities, the outcomes get rather interesting.
In the learning phase, the robot uses several cameras to monitor the movements and subsequently creates models for the objects’ kinetics based on the trajectories, speeds and rotational movements. The scientists translate the data into an equation which then allows the robotic arm to position itself very quickly in the right direction whenever the object is thrown.
In a few short milliseconds during the approach, the arm refines and corrects the trajectory and then makes a high precision capture of the object. The efficiency is further enhanced by the development of controllers that couple and synchronize the movements of the hand and fingers.