Educational Robotics through Web Applications: From Visual Programming to Simulation-Driven Learning

Abstract

This paper explores the role of educational web applications in supporting robotics-based learning, with a primary focus on how such tools can enhance early STEM education. Based on two implementations developed within the MetaRoboLearn project, the paper compares and analyzes two different approaches: one utilizing block-based programming through the Blockly library, and the other using Python-based code editing with an integrated 3D simulator. While both applications control physical and simulated robots, their target audiences and pedagogical strategies differ significantly. The block-based system is tailored for primary school students and emphasizes simplicity, modularity, and abstraction of syntax. The Python-based system, on the other hand, introduces students to real code structures and debugging practices in a controlled, interactive environment. Both systems include simulation environments that allow for experimentation without direct access to hardware, increasing accessibility and safety. The paper argues that such hybrid learning environments – combining visual programming, real-time simulation, and physical interaction – provide a scalable and engaging framework for developing algorithmic thinking, problem-solving skills, and interest in computer science among young learners.