Scientists at the European research project XPERO have invented a robot that has the ability to learn basic physical and mathematical concepts and use them to move, announced today the University of Bonn-Rhein-Sieg in Sankt Augustin (West Germany).
concepts such as location or orientation through a system of coordinates.
Initially, the robot moves aimlessly around and fixed simultaneously recorded sensory data, which then used to produce a pattern or model. This pattern makes it possible, later, the android can anticipate the location of objects and how they vary their position when it moves.
The authors of the algorithm, and Ivan Jure Zabkar Bratke, University of Ljubljana (Slovenia) stresses that "what a person is something trivial, a robot can lead to an enormous difficulty." He further explained that the android designed "has less knowledge than a baby," since it does not distinguish objects, but only spots of color and the edges of them.
Distinguish MANCHAS
"You do not have notions of the concept of an object or its position in a coordinate system, nor is it aware of how the system varies according to their own movements," he adds.
Therefore, the robot must not learn to tell a coordinate system, or how.
Scientists have developed a mechanism that allows the PLC to establish a routine that helps to remove the sensory data and to translate them into a model to help explain and anticipate what surrounds it. Using the same algorithm we have taught some physical concepts such as "flexibility" of an object or the "degree of freedom of movement."
IMPORTANT DISCOVERY
What originally was considered a problem was academic in nature has a great technical importance, says project coordinator Erwin Prassler of the German University of Bonn-Rhein-Sieg (the west).
The project XPERO sets the groundwork for the first time in the future that will develop the key technology to give birth to the next generation of service robots that can clean the house, mow the lawn or cleaning the shoes. The androids that had been invented to date lack intelligence because they are pre-programmed devices, which are unable to export data that does not know or use operating systems that have not previously studied. However, service robots of the future may treat a large number of new concepts and models on the basis of knowledge and learning and sensory-alerts, and perform new tasks.
Automata XPERO project will provide a demonstration at the conference technological FET (Future and Emerging Technologies) this year, taking place in Prague between now and Thursday, April 23.
European scientists have invented a robot that learns math concepts
Robotics
Robotics is a new discipline with its own problems, its foundations and its laws. Is twofold: the theory and practice. In the theoretical Gathering input from the auto, computer science and artificial intelligence. On a practical or technical aspects of construction are (mechanical, electronics) and management (control, programming). The robot then presents a decidedly transdisciplinary.
The first industrial robot was installed in 1961 at a General Motors plant in United States. After United States, the first to robotic industry were Japan and Sweden, while the remaining countries of the OECD robots was first introduced during the decade of the seventies.
However, the growing market of robotics did not take place until the mid-eighties when the cumulative annual rates of growth in the number of robots made more than 20 per 100. The strong growth in demand was mainly due to improved technology, the price of robots in industrial processes and added a number of factors directly related to the competitiveness of business.
The impact of robotic systems is bound to other variables (such as the introduction of new systems of organization, new technologies, globalization, offshoring of business, etc.) Influencing the trend of declining labor conflicts that characterized, according to experts interviewed, the development of advanced technological economies in the coming years.
Flexible automation technologies such as robots and flexible manufacturing systems are one of the elements that supports the "light output".
This production is characterized by the increase in organizational flexibility that allows the manufacture of a greater variety of products in short series. The slight shift to the production requires a series of multi-organizational changes of workers, new relationships with suppliers, etc., but also requires the adoption of technologies that integrate information flows, and flexible manufacturing technologies increase the degree of variability in production.
The robots we make the procedures seem very simple and, with a foolproof security, again and again the same operations. But appearances are deceptive: after a simple motion of lifting and placing a lot of underlying technology.