Robotics, speech recognition and computer vision are relevant problems of computing, signal processing and analysis. Using algorithms for audio and visual information processing allows to control a real mechanism in order to create an autonomous robot-helper. The paper describes the working principles of hardware and software of the robot-guide, their interaction and the implementation of the robot-server interconnection through existing World Wide Web infrastructure. The authors used the Raspberry Pi as a robot control plate, the system of speech recognition module based on PocketSphinx library and the sensor RPLIDAR A1 as the navigation module for computer vision implementation.