In recent decades, mobile robots have become crucial for the automation and optimization of internal logistics in industry. Automated guided vehicles (AGVs) are mobile robots that follow trajectories marked on the ground. These robots can be perceptually limited, for example, having only sensors for line following. This leads to several limitations: predefined and suboptimal paths, the need for manual code changes when paths are altered, and a lack of central guidance and communication, which can lead to collisions. The goal of this thesis is to develop a centralized system for the reliable navigation of multiple robots on a grid of lines – trajectories, based on the concept of a digital twin. A system was developed, implemented on a Node.js server, that simulates events in real-time, plans optimal paths and prevents collisions. The micro:Maqueen V2 robots, equipped with an ESP32 microcontroller, communicate with the server via a WebSocket connection and follow lines based on received commands. A user interface for visualization and management, as well as an algorithm for simulation calibration were also developed. The developed system enables reliable navigation and coordination between the physical and simulated robots, with occasional errors arising mainly from hardware limitations. The system provides a modular foundation for future upgrades and integration into a demonstration platform.
|
