Adaptive GR(1) Specification Repair for Liveness-Preserving Shielding in Reinforcement Learning
This addresses the issue of environment assumption violations in RL shielding for safety-critical applications, offering an incremental improvement over static methods.
The paper tackles the problem of static shielding in reinforcement learning by developing an adaptive shielding framework that repairs GR(1) specifications online using Inductive Logic Programming, resulting in near-optimal reward and perfect logical compliance in case studies like Minepump and Atari Seaquest.
Shielding is widely used to enforce safety in reinforcement learning (RL), ensuring that an agent's actions remain compliant with formal specifications. Classical shielding approaches, however, are often static, in the sense that they assume fixed logical specifications and hand-crafted abstractions. While these static shields provide safety under nominal assumptions, they fail to adapt when environment assumptions are violated. In this paper, we develop the first adaptive shielding framework - to the best of our knowledge - based on Generalized Reactivity of rank 1 (GR(1)) specifications, a tractable and expressive fragment of Linear Temporal Logic (LTL) that captures both safety and liveness properties. Our method detects environment assumption violations at runtime and employs Inductive Logic Programming (ILP) to automatically repair GR(1) specifications online, in a systematic and interpretable way. This ensures that the shield evolves gracefully, ensuring liveness is achievable and weakening goals only when necessary. We consider two case studies: Minepump and Atari Seaquest; showing that (i) static symbolic controllers are often severely suboptimal when optimizing for auxiliary rewards, and (ii) RL agents equipped with our adaptive shield maintain near-optimal reward and perfect logical compliance compared with static shields.