In harsh and unstructured environments, such as disaster-stricken areas filled with debris, effective robot navigation is critical for tasks like search and rescue, inspection, and material transport. Traditional path planning algorithms like A* often generate paths with numerous sharp turns, significantly increasing traversal time due to the robot’s need to manoeuvre through these abrupt corners. This paper proposes an enhanced multi-resolution path optimization technique that smooths the A*-generated paths, reducing the sharp turns and improving the robot’s movement efficiency. The proposed method applies optimization at multiple resolution levels, balancing path smoothness and obstacle avoidance through the integration of curve fitting and adaptive interpolation techniques. This approach eliminates sharp turns and maintains collision-free navigation in both sparse and cluttered scenarios. Experimental evaluations in the developed simulated environment demonstrated average improvements of 15.10% in traversal time and 56.53% in the number of turns compared to traditional A* paths, validating the method’s efficiency and robustness. The results highlight the potential of this approach for real-world applications where efficient robotic movement in challenging environments is essential.