University Archives

Autonomous navigation systems require reliable integration of multiple perception, planning, and control subsystems. This project focuses on the system integration architecture for an autonomous ground vehicle competing in the Intelligent Ground Vehicle Competition (IGVC). Our approach uses the ROS2 middleware framework to organize subsystems into modular nodes that communicate through a clearly defined data flow. A needs–metrics analysis was first conducted to identify key system requirements, including reliable communication, emergency stop capability, and the ability to navigate between waypoints. Based on these requirements, a structured communication architecture was developed to define how sensor data, perception outputs, localization estimates, and navigation commands are exchanged between subsystems. Embedded motor control hardware using a microcontroller was integrated with the ROS2 system to provide low-level motor control and encoder feedback. A communication interface was implemented to exchange commands and sensor data between the embedded controller and the ROS2 navigation pipeline. A unified launch-file structure continues to be used to simplify system startup and ensure consistent initialization of all modules. In addition, hardware-in-the-loop testing was used to validate portions of the integration workflow before deployment on the full competition robot. The resulting architecture improves modularity, debugging capability, and system reliability while allowing the system to adapt as additional subsystems are integrated into the Megatron platform.