Collision-free navigation of a 2-dimentional wheeled mobile robot in the presence of static and quasi-static obstacles is undertaken in the present study. Two soft-computing approaches based on sensory detection namely; Goal-oriented fuzzy strategy and Genetic-fuzzy approach have been developed for this purpose. The main advantage of sensor-based approaches is that the robot can navigate safely in a dynamic environment by reacting to obstacles detected by sensors. A major drawback is that due to the limitation of sensors, the robot may get lost even if a path to the goal exists. In fuzzy approach, some short-range sensors such as infrared proximity light sensors or touch sensors are used to keep the robot free from collisions. In this approach, there is no explicit fuzzy rule for motion planning. In other words, information about obstacles is used simultaneously as local information, which may result in a shortsighted behavior in some situations. The proposed Genetic-fuzzy approach optimizes two desired parameters namely, length of path and difference of robot rotations in every two steps, by finding an optimal fuzzy rule base. Rather than the fuzzy strategy, in genetic-fuzzy approach, the fuzzy rule base is explicit. Fuzzy functions consist of two inputs and one output. The first input is as distance form and second input and output are as angle form. The results of a number of problem scenarios in GUI MATLAB and also comparision between two methods are presented in the results. Key Words Sensor-based, fuzzy approach, genetic-fuzzy, obstacle, mobile robot, navigation.