Generating symbolic representation from sensor data: Inferring knowledge in robotics competitions

Transformation of data gathered by sensors into symbolic knowledge is a recurrent problem in control systems, artificial systems, and robotics in particular. The control architecture of the robot reflects this process. This paper proposes an inference component devoted to transform sensor information from the sub-symbolic layer to a symbolic layer. Its operation has been illustrated using a navigation task on simulated and real scenario based on @home competition from the European Robotics League.