Formal Verification of Real-Time Autonomous Robots: An Interdisciplinary Approach
Mohammed Foughali,Alexander Zuepke
DOI: https://doi.org/10.3389/frobt.2022.791757
2022-04-13
Abstract:Due to the severe consequences of their possible failure, robotic systems must be rigorously verified as to guarantee that their behavior is correct and safe. Such verification, carried out on a model, needs to cover various behavioral properties (e.g., safety and liveness), but also, given the timing constraints of robotic missions, real-time properties (e.g., schedulability and bounded response). In addition, in order to obtain valid and useful verification results, the model must faithfully represent the underlying robotic system and should therefore take into account all possible behaviors of the robotic software under the actual hardware and OS constraints (e.g., the scheduling policy and the number of cores). These requirements put the rigorous verification of robotic systems at the intersection of at least three communities: the robotic community, the formal methods community, and the real-time systems community. Verifying robotic systems is thus a complex, interdisciplinary task that involves a number of disciplines/techniques (e.g., model checking, schedulability analysis, component-based design) and faces a number of challenges (e.g., formalization, automation, scalability). For instance, the use of formal verification (formal methods community) is hindered by the state-space explosion problem, whereas schedulability analysis (real-time systems) is not suitable for behavioral properties. Moreover, current real-time implementations of robotic software are limited in terms of predictability and efficiency, leading to, e.g., unnecessary latencies. This is flagrant, in particular, at the level of locking protocols in robotic software. Such situation may benefit from major theoretical and practical findings of the real-time systems community. In this paper, we propose an interdisciplinary approach that, by joining forces of the different communities, provides a scalable and unified means to efficiently implement and rigorously verify real-time robots. First, we propose a scalable two-step verification solution that combines formal methods and schedulability analysis to verify both behavioral and real-time properties. Second, we devise a new multi-resource locking mechanism that is efficient, predictable, and suitable for real-time robots and show how it improves the latter's real-time behavior. In both cases, we show, using a real drone example, how our approach compares favorably to that in the literature. This paper is a major extension of the RTCSA 2020 publication "A Two-Step Hybrid Approach for Verifying Real-Time Robotic Systems."