The CogRob Workshop edition is already an instance in the AI community. As an bi-annual event, in 2016 it took place for the 10th time, bringing together researchers and practitioners interested in AI and robotics. Research in cognitive robotics is interested in the cognitive aspects of robots such as perception processing, attention allocation, anticipation, planning, complex motor coordination, reasoning about other agents and perhaps even about their own mental states. A central aspect is the behaviour modelling of intelligent agents. Here, a number of AI techniques such as agent architectures or high-level planning are traditionally used. Besides the theoretical interest in organizing complex decision making processes of intelligent robots the real implementation of such systems is in the focus. The proposed 2018 edition of the CogRob workshop aims at connecting the CogRob community with recent advances in the field of machine learning and agent technologies in order to allow new insights how such techniques could be used beneficially for the behaviour modelling of intelligent agents. In particular we like to discuss the limitations of quite complementary approaches such as the successful and popular area of machine learning and classical AI and to investigate how a combination of approaches from different areas may overcome that limitations.