Scientists from Israel, the USA and Europe are developing a "robotic rat"

You will be able to assist in the rescue and rescue of collapsed buildings, and in the exploration of planets

A new initiative, bringing together nine research groups, including roboticists and neuroscientists from Europe, the USA and Israel, was recently founded with the aim of imitating nature. The multinational group is working to create advanced touch and tactile technologies - such as a sensing array that will be based on artificial hairs, and will, in fact, be a sort of robotic rat. The mustachioed robot will be able to locate, recognize and catch fast moving objects. "The sense of touch has been neglected so far when it comes to the development of artificial intelligence," says Prof. Ehud Ahisher from the Department of Neurobiology at the Weizmann Institute of Science, whose group is participating in the multinational research. "Nocturnal animals, or those operating in areas where vision is limited, use this sense, rather than vision, as the primary means to learn and receive physical information about their environment." Prof. Ahishar investigates the way in which rats use Schaffmann bristles for the purpose of forming their perception of space, and the processes of processing the information coming from the bristles in the brain. "If we manage to understand what makes the sense of touch of these animals so effective, we will be able to develop machines that will play it, and use it effectively."

What, then, is the secret of the mustache? Why is sensing by the whiskers of a rat so much faster and more efficient than sensing by the fingertips of an average person? Research by Prof. Ahisher and his group members provides clues and preliminary answers to these key questions. One explanation stems from the active nature of the tactile system: the whisker bristles constantly vibrate and gather information about the immediate environment. Moreover, experiments show that the pattern in which the rat moves its whiskers depends on the navigation and recognition tasks it has to solve. To place objects, for example, the rat brain activates a three-dimensional coding program, encoding the location coordinates in each of the three dimensions - in a unique pattern of operation. Thus, for example, the horizontal dimension is encoded by the timing of the nerve cell activity; The vertical dimension, that is, the height of the bone, is encoded by the spatial structure of the active nerve cells; The third dimension, the distance of the bone from the base of the hair, is encoded through the intensity of the response - the closer the bone is, the more signals the nerve cells send.

In addition, the members of Prof. Ahisher's research group were able to identify the main routes that carry the information from the mustache hairs to the brain: the nerve signals travel in three separate and parallel routes to the thalamus - the "gateway" to the cerebral cortex. One pathway transmits signals related to the movement of the whisker bristles themselves; A second route conveys information about the contact times with objects; And a third route conveys complex information about the interrelationship between the movement of the whisker and the contact with the bone. These three pathways are part of feedback loops that create closed control loops: the movement of the sense organ affects the sensory information in the brain, which, in turn, affects the movement of the sense organ accordingly. It is the presence of many closed loops, linked among themselves in complex interrelationships, that enables rich and precise movement control, and optimal use of the sensory organs - but it is also the one that poses the great challenge to engineers, who still do not know how to build an artificial system that imitates many and integrated feedback loops.

To investigate the effect of the loops on the sensing ability of the mustache bristles, Prof. David Golomb from Ben-Gurion University of the Negev joined the research initiative. Prof. Golomb applies theoretical and computational methods taken from theoretical physics and applied mathematics to develop and study models that describe the neural and mechanical systems involved in the sensing processes. The models are based on the experimental observations of Prof. Ahisher and other scientists, and predictions obtained from the models will be tested in the laboratories of the scientists participating in the project.

Prof. Ahishar: "The research may promote a better understanding of the brain on the one hand, and technological applications, on the other hand. Regarding the applications: we hypothesize that the multiple closed feedback loops are the key to the advantage of the biological systems over the robotic systems that exist today. Realization of the biological knowledge will help to build efficient machines for extraction, rescue and exploration in difficult visibility conditions. Regarding our efforts to understand the ways of the brain: robotics allows us to build, step by step, a brain-like system, and to understand the role of each component within it. In this sense, the robot will serve as a kind of 'laboratory' that will allow us to better understand the living brain." In this way, basic animal research may contribute to human well-being, outside of medicine.

Scientists from universities, research institutes and high-tech companies in the UK, Switzerland, Italy, France, Germany and the USA are also participating in the project, which is largely financed by the European Union.