The number of robots operating in human environments is increasing every day. In order to operate in such environments, the robot must be able to navigate, interact with human, pick and place different objects. This chapter presents a mobile humanoid robot that is able to localize itself, navigate to the target location, and generates the arm motion based on the specific task. The robot utilizes the Laser Range Finder, camera and compass sensor for localization and navigation. In addition, the robot generates the arm motion satisfying multiple motion criteria, simultaneously. This chapter evolves neural controllers that generate the humanoid robot arm motion in dynamic environment optimizing three different objective functions: minimum time, distance and acceleration. In a single ran of Multi-Objective Genetic Algorithm, multiple neural controllers are generate and the same neural controller can be employed to generate the robot motion for a wide range of initial and goal positions.