A conceptual model of personal space for human-aware robot activity placement

This work proposes a conceptual model of personal space that can be used within symbolic knowledge-based robot architectures. The robot represents personal spaces in terms of affordance spaces attached to humans, i.e., portions of physical space that can be used for interactions. The model does not rely on numeric functions nor does it impose any restrictions upon the geometry of personal space. The approach is applied to human-aware robot activity placement planning: Given an activity type and a robot that is to perform an activity of this type, principles of personal space use determine the reasons pros and cons to place that particular activity at a particular affordance space. The symbolic representations of personal spaces enable robots to make choices and to communicate the reasons for their choices to humans.