INVARIANT INFERRING AND MONITORING IN ROBOTIC SYSTEMS

System monitoring can help to detect abnormalities and avoid failures. Crafting monitors for today’s robotic systems, however, can be very difficult due to the systems’ inherent complexity and its rich operating environment. In this work we address this challenge through an approach that automatically infers system invariants and synthesizes those invariants into monitors. This approach is inspired by existing software engineering approaches for automated invariant inference, and it is novel in that it derives invariants by observing the messages passed between system nodes and the invariants types are tailored to match the spatial, time, temporal, and architectural attributes of robotic systems. Further, our approach automatically classifies and synthesizes invariants into a monitor node that can be seamlessly integrated into systems built on top of publish-subscribe architectures. The monitor can be also tailored to trigger actions when an invariant is violated. We have assessed the approach in the context of three UAV systems to better understand its potential. In our case study, we found that invariants can be useful for developers and that the synthesized monitor can reduce system failure rate when facing unexpected faults from 76.2% to 10.6%. Adviser: Sebastian Elbau