A Centralized Autonomous Vehicle Platform for Indoor Navigation Using ArUco Markers

Sammanfattning

This paper presents the design and implementation of a modular and accessible research platform for small-scale autonomous vehicles. The system is built on a 1/18-scale RC car and is designed to avoid the complexity and cost associated with more advanced robotics platforms. It does not rely on onboard cameras, artificial intelligence, or machine learning. Instead, the platform uses an overhead camera and ArUco markers for real-time localization, with pose estimation performed using OpenCV. A central server handles all decision-making and communicates with each vehicle over TCP sockets. Raspberry Pi units on the cars receive these commands and control the steering and drive motor using PWM signals via a PCA9685 servo driver. The system is reproducible, and easy to extend, making it ideal for education and early-stage robotics research. Complete source code and documentation are provided through an open-source GitHub repository.