The animal behavior served as the inspiration for the behavior based control applied to autonomous mobile robots. This is based on simple directives, or reactions, separated by layers with their priorities. When combined, these directives can run the most diverse and complex functions in the environment, making the control itself a segmented task, which divides the main objective into small modules called primary behaviors. Those act independently and, when working in parallel, result in complex behaviors that are capable of executing more complex tasks, an evolution of reactive control. The logic from this technique makes it easy to program and organize robot controllers because it has a modular structure that allows the addition of new sensors (new behaviors) without major changes in the existing code. This thesis develops and implements a behavior based programming on autonomous mobile robots with optical and ultrasonic sensors in very common open source simulation software, called Player/Stage, and validates experimentally in 2 autonomous robots. The architecture used in the development of behavioral layers was the motor schema with potential fields, originally created by Ronald C. Arkin in 1998. The simulated and experimental results were confronted with classical methods of programming and proved all the benefits of behavior based control.