A Modelling and Formalisation Tool for Use Case Design in Social Autonomous Robotics.

The modelling of use cases in Social Robotics and its formalisation in a control architecture is a complex and time consuming task, which can be seen as a knowledge engineering process. This manuscript focuses on the description of a novel tool created to simplify the process of modelling these interactions through the definition of such use cases as nominal behaviours with exogenous events. The control architecture of the robotic platform used is based on Automated Planning, allowing robots to deal with stochastic environments. The tool comprises two main objectives: to ease the knowledge engineering process to model social interactions and to generate Planning Domain Description Language code (PDDL) as a formalisation that can be injected into a control architecture. We model different use cases and test the PDDL formalisation by using the control architecture. The test carried out shows evidence of good results, encouraging further research of this type of tools to provide autonomy and facility to experts and non-experts in the field.