Hand in Hand with Robots
A report from the Institute of Intelligent Systems and Robotics: Assistance with Movements and Therapeutic Applications (Isir - Agathe) in Paris. The objective of the Agathe team is to assist human movements using robotic devices.
-
To guide the prosthesis towards the yellow targets, sensors on the side of the “amputated” arm record the movements of the torso, shoulder and upper arm. The prosthesis is able to interpret this information and determine how to complete the movement naturally. -
Guillaume Morel, lecturer-researcher in robotics, directs the Agathe team and Carnot Interfaces, a structure that federates five complementary research units of excellence, specialized in digital technology and tasked with developing innovative communication solutions between the human and digital worlds. -
To guide the prosthesis towards the yellow targets, sensors on the side of the “amputated” arm record the movements of the torso, shoulder and upper arm. The prosthesis is able to interpret this information and determine how to complete the movement naturally. -
It is difficult to know, without looking at it, whether a prosthesis is properly oriented. Lucas Lavenir, engineering student, is testing two-way communication. The muscle contractions of the healthy arm, captured using electrodes, command the robotic arm: certain movements of the healthy arm, such as bending the wrist, are associated with movements of the robotic arm, such as raising the hand, shown here. -
In return, the prosthesis can "speak": via vibrating devices installed on the healthy arm, the participant receives vibrating signals indicating the direction of the robotic hand. Objective of the experiment: succeed in placing the prosthesis in a specific position, and then ensure that it is well placed by "feeling" the information that it sends.
-
Nathanaël Jarrassé is handling an iBox: This box, equipped with strength sensors and inertial units, measures the pressure exerted by the hand as well as acceleration and speed. Transposable to a hospital setting, it evaluates the motor capacities of the upper limbs for diagnostic purposes. -
The sensor-laden gloves make it possible to know the position of the various segments of the hand and to better understand how the person handles the iBox. Among the tasks requested: lift the object up and put it back down, raise it to the mouth and tilt it like a drinking glass, pass it from one hand to the other… -
For a deeper understanding of the characteristics of walking using a frame, the movements of Éléonore Ferrier-Barbot, cognitive science intern, are meticulously analyzed: infrared optical cameras receive signals sent by the markers affixed to her leg, modeling the assisted walking. -
Alexandre Peudpiece, assistant ingénieur, répare un prototype de déambulateur motorisé : grâce à des caméras surveillant la position des pieds ou à des poignées qui analysent la force de pression des mains, cet équipement peut par exemple aider à prévenir les chutes. Ce type de recherches menées au sein d’Agathe sont à l’origine de la création de la start-up Gema.
In the Agathe (Assistance to Gesture with Applications to THErapy) team, a strange experiment is taking place: a three-armed man is trying to reach a yellow target with one hand! In reality, Étienne de Montalivet is a research engineer in robotics and his third arm is actually a robotic prosthesis intended for use by amputees.
The point of the exercise? Evaluate his ability to control the movements of this prosthesis. But above all: invent the prosthetics of the future – which will be even more efficient. "These days, robots are deployed almost everywhere in our societies. For a long time, they were seen as a simple tool to replace humans when doing such and such a task. Here the aim is for robots to work with humans", explains Guillaume Morel, director of the team located at the Institute of Intelligent Systems and Robotics – which obtained the Inserm labeled in 2014.
For the thirty or so people working there, this approach offers great therapeutic potential: neuromotor rehabilitation in the event of stroke, for example, could be more effective if the patient were equipped with a robotic exoskeleton which does not impose movements that he or she would have trouble performing, but rather assists and corrects gradually. "To do that we need to develop a language between the human and the robot, and also between the robot and the human, so that the information flows in both directions", describes Nathanaël Jarrassé, team member and CNRS researcher. This language is still quite basic at present, but the scientists are working on improving it.
A plunge into this laboratory where infrared sensors rub shoulders with prostheses, exoskeletons and other motorized assistance devices!
