For unicycle robots and an arbitrary continuous reference tracking controller, wepropose a control augmentation providing obstacle avoidance properties. The avoidance controllaw is activated in an eye-shaped neighborhood around an obstacle and guarantees that aspecified domain around the obstacle is not entered. Switching between different control lawsis orchestrated through a hybrid systems framework and Zeno behavior is avoided throughappropriate hysteresis regions. Since the size of the eye-shaped neighborhood and its orientationdepend continuously on the orientation and velocity of the robot, it can be guaranteed that thevelocity input of the overall control law is continuous. Numerical simulations are provided toillustrate the performance of the tracking controller endowed with the avoidance augmentation.