Model-Based, Coverage-Driven Verification and Validation of Code for Robots in Human-Robot Interactions.

Abstract

Many industrial and social sectors, such as manufacturing and home care, could be transformed by the presence of robotic assistants. Rigorous processes of verification and validation (V&V) are required to demonstrate these robots are safe, fit for purpose, and working as intended. We propose Coverage-Driven Verification (CDV) as a systematic, goal-directed simulation-based V&V method that is capable of exploring systems of realistic detail under a broad range of environment conditions, such as those found in human-robot interaction (HRI). A CDV testbench was developed for robotic code as deployed in our HRI case study. Based on this application of CDV, we compare the merits of different automatic test generation techniques, including constrained pseudorandom and model-based approaches, in achieving thorough exploration of the system. Our results show that CDV provides an efficient and effective V&V strategy for the HRI domain.