Efficient Dynamic Shielding for Parametric Safety Specifications
Davide Corsi, Kaushik Mallik, Andoni Rodriguez, Cesar Sanchez
TL;DR
The paper addresses runtime safety for ML-based autonomous systems by introducing dynamic shielding for parametric safety specifications $P{-}\mathit{Safety}(R)$ built from atomic objectives $\{\Box G_i\}$. It proposes a hybrid offline-online approach: offline synthesis of atomic shields $\mathcal{S}_{G_i}$ for each $\Box G_i$, online composition via $\mathcal{C}^*_{G_1}\bigotimes \cdots \bigotimes \mathcal{C}^*_{G_k}$ and extraction of the largest nonblocking sub-controller to realize $\mathcal{C}^*_R(G)$. The method leverages abstraction-based control (ABC), discretization with feedback refinement relations (FRR), and symbolic data structures (BDDs) implemented in Mascot-SDS to enable scalable shield synthesis. Experiments on robot navigation in unknown terrains show offline design in minutes and online adaptation per step in fractions of a second to seconds, outperforming pure online recomputation by up to fivefold and achieving safety by construction.
Abstract
Shielding has emerged as a promising approach for ensuring safety of AI-controlled autonomous systems. The algorithmic goal is to compute a shield, which is a runtime safety enforcement tool that needs to monitor and intervene the AI controller's actions if safety could be compromised otherwise. Traditional shields are designed statically for a specific safety requirement. Therefore, if the safety requirement changes at runtime due to changing operating conditions, the shield needs to be recomputed from scratch, causing delays that could be fatal. We introduce dynamic shields for parametric safety specifications, which are succinctly represented sets of all possible safety specifications that may be encountered at runtime. Our dynamic shields are statically designed for a given safety parameter set, and are able to dynamically adapt as the true safety specification (permissible by the parameters) is revealed at runtime. The main algorithmic novelty lies in the dynamic adaptation procedure, which is a simple and fast algorithm that utilizes known features of standard safety shields, like maximal permissiveness. We report experimental results for a robot navigation problem in unknown territories, where the safety specification evolves as new obstacles are discovered at runtime. In our experiments, the dynamic shields took a few minutes for their offline design, and took between a fraction of a second and a few seconds for online adaptation at each step, whereas the brute-force online recomputation approach was up to 5 times slower.