Toward a More Dependable Software Architecture for Autonomous Robots

Abstract

Autonomous robots are complex systems that require the interaction/cooperation of numerous heterogeneous software components. Nowadays, robots are getting closer to humans and as such are becoming critical systems which must meet safety properties including in particular logical, temporal and real-time constraints. We present an evolution of the LAAS Architecture for Autonomous System and its tool GenoM. This evolution is based on the BIP (Behaviors Interactions Priorities) component based design framework which has been successfully used in other domains (e.g. embedded systems). In this study, we show how we seamlessly integrate BIP in the preexisting methodology. We present the componentization of the functional level of a robot, the synthesis of an execution controller as well as validation techniques for checking essential safety properties. This approach has been integrated in the LAAS architecture and we have performed a number of experiment in simulation but also on a real robot (DALA).