This thesis concerns the development of an autonomous mobile robot intended for use in a human environment. From a primary focus on HRI, a novel robotic behaviour algorithm has been developed and validated through simulation as well as real world experiments. Furthermore, the behaviour of the robot is continuously adjusted to that of encountered people, by incorporating CBR based artificial intelligence, implemented by use of a MySQL database. Besides interacting with people, the robot is capable of navigating and localizing itself in a given environment while avoiding unknown obstacles. All robot software, including the implementation of the above algorithm, has been developed for use with the Player robot software framework, and implemented on a FESTO Robotino®. Consequently, due to the Player framework offering a wide range of robot features, the developed system can form the basis of future robotic research. Experiments performed on the robot show promising results, in relation to both the behavioural capabilities and the developed system in general.