Enhancing Obstacle Avoidance in Nav2 Using Control Barrier Function

Abstract

Autonomous Robots, also known as Autonomous Mobile Robots (AMRs), is considered as the transformation of industries by making safe, efficient, and robust operations in unstable environment and complex ones. However, a significant challenge is ensuring AMRs can navigate safely in real-time, especially when faced with unexpected obstacles. The ROS2-based Nav2 stack provides key navigation capabilities, including localization, path planning, and control, but its limitations remain in handling dynamic obstacles and uncertain environments effectively. This thesis dealing with studying the cooperation of Control Barrier Function (CBFs) with the Nav2 stack to enhance safety during navigation. CBFs are control techniques that ensure the robot stays within a safe set, avoiding collisions while continuing to perform tasks. The thesis examines how CBFs can be integrated with existing Nav2 components like planners, controllers, and perception systems, improving real-time decision-making and robustness in dynamic environments. The proposed approach demonstrates significant improvements in safety and efficiency, particularly in environments with moving obstacles and unexpected changes. This work contributes to the field by offering a formal, practical method to enhance the reliability and safety of Nav2-based systems, providing a promising solution for deploying autonomous robots in real-world, dynamic settings where safety is critical.