Experts from the University of Jaén applied predictive techniques to relations between robots and people
March 17, 2010
The project explores one of the lines of the current robotic research: adapting the machines to more complex environments and requiring a greater relationship with people. This need requires robots equipped with sophisticated sensors that scan the context to adapt to it. In particular, the Group of the UJA will focus on the sensors associated with the force, i.e. those in which the machine touch an object and applies a force on it. So far, this parameter is controlled with a single sensor. However, the researchers aim to use multiple devices to collect more information. In particular, they will serve of instruments to measure the strength, acceleration, the position and vision. The conjugation of these data will be that the robot has more information in controlling their movements.
The so-called predictive control in addition to this novelty, which includes multitude of sensors, experts. These techniques are based on applying a mathematical model to predict what will happen if a certain decision. The idea is to recreate the future scenario that results after that election. "The robots apply this predictive model to know what will be the behavior of a person and thus take a decision according to such behaviour." "Thus improving the man-machine relationship", explains the Coordinator of the study, Juan Gómez Ortega.