Safe and Reliable Training of Learning-Based Aerospace Controllers

TL;DR

This paper introduces new methods for training and verifying deep reinforcement learning controllers to ensure safety and reliability in aerospace applications, combining formal verification techniques and reachability analysis.

Contribution

It presents a design-for-verification approach using k-induction, neural Lyapunov Barrier certificates, and reachability-based methods for verifying safety-critical aerospace controllers.

Findings

Verification of liveness properties using k-induction

Summary of neural Lyapunov Barrier certificates capabilities

Introduction of reachability-based approaches for DRL system verification

Abstract

In recent years, deep reinforcement learning (DRL) approaches have generated highly successful controllers for a myriad of complex domains. However, the opaque nature of these models limits their applicability in aerospace systems and safety-critical domains, in which a single mistake can have dire consequences. In this paper, we present novel advancements in both the training and verification of DRL controllers, which can help ensure their safe behavior. We showcase a design-for-verification approach utilizing k-induction and demonstrate its use in verifying liveness properties. In addition, we also give a brief overview of neural Lyapunov Barrier certificates and summarize their capabilities on a case study. Finally, we describe several other novel reachability-based approaches which, despite failing to provide guarantees of interest, could be effective for verification of other DRL systems, and could be of further interest to the community.