At this point, a review of the components introduced so far is in order as well as a discussion of the integrated system. The flow between perceptual input, the graphical output, the time series processing and the learning system are presented in detail and different modes of operation are enumerated. The types of operation can be conceptually split into three categories: human-human interaction, human-machine interaction and machine-machine interaction. We specifically describe the ARL framework as it is used for behaviour learning, interaction, prediction, filtering and simulation. In addition, we point out some practical issues in the integrated system such as the dimensionality of the data in the flow and the temporal constraints of each module.