Knowledge Management for Robot Activities in a Real World Context. A Case for Task Pattern Analysis (TAPAS)

Knowledge management is used for handling knowledge and activities in real world context. However, knowledge management can also be used for the modelling of robot activities in a real world context. Although robot technology is still under development, Intelligent Service Robots are slowly becoming a reality. The programming of these robots to support a closer interaction with the users is a complex problem including the difficulty of correct modelling of the tasks of the robots. The need for close cooperation between user and robot adds extra complexity, compared to standard models for task modelling. In this paper we outline a representation scheme, Task Pattern Analysis (TAPAS), that is directed towards a notion of activities as rule-based behaviours. In task pattern analysis, the tasks are described in terms of Task Patterns, consisting of a frame rule, and Task Fragments, specifying and constraining the applicability of the rules.