The class project of cyber-driver

Abstract

This article describes a class project of cyber-driver, i.e. an android robot controlling a mobile platform. This project can be used both to teach high-school students in programming, hardware design and foundations of robotic science by solving some important and scientifically interesting tasks. Despite the great number of modern android robotics systems, their software controlling systems often have serious issues. Such actions as rough terrain movement or stairs climbing can be performed simply and accurately by both humans and animals but not by biosimilar robots. The traditional systems in these cases demand many complex and precise settings to perform such movements and nevertheless they remain very vulnerable to random factors. In robotic perception models the situation is similar: casual human tasks as image recognition or scene analysis need complex algorithms and huge computer resources if performed by robots. Biologically-inspired models can improve matters in robotic science. Human movement copying is often used for solving this problem in practice but it does not allow copying considerable aspects of movement controlling in nervous system. Both controlling quality and testing quality for the aforementioned tasks can be essentially improved if the robot uses techniques similar to central motor programs of human. This project helps students to study this important domain. Besides, robotic perception can imitate biological systems more closely. Such biologically-inspired perception models give structure to computer vision systems and allow developing these systems quicker. The cyber-driver project is an example of biologically-inspired educational system. The system proposed consists of android robot that can manipulate levers, wheels and buttons, and mobile robotic platform. Educational robotic platform YARP-3 allows easy modification of constructing arms and grippers encouraging students’ creativity. The software of the robot can automatically perform some of the driver’s tasks (indicators recognition, obstacle avoiding and movement without collisions). The project uses some key features of robotic science, which are interdisciplinarity and competitive spirit, to improve students’ experience. The project’s architecture allows modifications in both mathematical and practical aspects of the result system’s description. Besides that, some extra suggestions about project-based inquiry in robotics are made.