Human – computer interaction is a vast area of research that targets two very different groups of subjects: the humans as a biology originated group, and the machines – as human made products. One specific part of this interaction is the human – robot interaction. This paper discusses the challenges that arise in the human – robot interaction. Special accent is put on creating social robots, in particular on creating more human-like empathic robots. Here a concept for creating such robots is proposed, where data from human-human interaction is used for feature selection. As an example of the stated concept a robot has been added the ability to perceive human emotions. In this research the robot uses human sound data as source for perceiving human emotions. Additionally, a biologically driven custom algorithm is used in this research for emotion evaluation classification. The proposed approach is analyzed and evaluated using experimental results done from rea l data. Also, the influence of the emotion aware robots on humans is explained in more detail. At the end, a discussion about the importance of sharing life with these emotion aware robots today and in the future is argued.