# Conformal Safety Shielding for Imperfect-Perception Agents

arXiv: https://arxiv.org/abs/2506.17275 · HTML: https://mlsanity.com/paper/2506.17275/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 33 references that are in the map.
- First posted on arXiv 12 Jun 2025.

## Builds on (most cited first, 20 of 20)

- [A Gentle Introduction to Conformal Prediction and Distribution-Free Uncertainty Quantification](https://mlsanity.com/paper/2107.07511/) (2021)
- [Safe Reinforcement Learning via Shielding under Partial Observability](https://mlsanity.com/paper/2204.00755/) (2022)
- [Guaranteeing Safety of Learned Perception Modules via Measurement-Robust Control Barrier Functions](https://mlsanity.com/paper/2010.16001/) (2020)
- [Formal Verification and Control with Conformal Prediction](https://mlsanity.com/paper/2409.00536/) (2024)
- [Learning Safe, Generalizable Perception-based Hybrid Control with Certificates](https://mlsanity.com/paper/2201.00932/) (2022)
- [Online Shielding for Reinforcement Learning](https://mlsanity.com/paper/2212.01861/) (2022)
- [Formal Analysis and Redesign of a Neural Network-Based Aircraft Taxiing System with VerifAI](https://mlsanity.com/paper/2005.07173/) (2020)
- [Robust Guarantees for Perception-Based Control](https://mlsanity.com/paper/1907.03680/) (2019)
- [Learning Robust Output Control Barrier Functions from Safe Expert Demonstrations](https://mlsanity.com/paper/2111.09971/) (2021)
- [Verifying Controllers with Convolutional Neural Network-based Perception: A Case for Intelligible, Safe, and Precise Abstractions](https://mlsanity.com/paper/2111.05534/) (2021)
- [Closed-loop Analysis of Vision-based Autonomous Systems: A Case Study](https://mlsanity.com/paper/2302.04634/) (2023)
- [Enforcing safety for vision-based controllers via Control Barrier Functions and Neural Radiance Fields](https://mlsanity.com/paper/2209.12266/) (2022)
- [Safe Perception-Based Control under Stochastic Sensor Uncertainty using Conformal Prediction](https://mlsanity.com/paper/2304.00194/) (2023)
- [TEMPEST -- Synthesis Tool for Reactive Systems and Shields in Probabilistic Environments](https://mlsanity.com/paper/2105.12588/) (2021)
- [Differentiable Control Barrier Functions for Vision-based End-to-End Autonomous Driving](https://mlsanity.com/paper/2203.02401/) (2022)
- [Safe POMDP Online Planning via Shielding](https://mlsanity.com/paper/2309.10216/) (2023)
- [Safe Reinforcement Learning via Probabilistic Shields](https://mlsanity.com/paper/1807.06096/) (2018)
- [Leveraging Approximate Model-based Shielding for Probabilistic Safety Guarantees in Continuous Environments](https://mlsanity.com/paper/2402.00816/) (2024)
- [State-Dependent Conformal Perception Bounds for Neuro-Symbolic Verification of Autonomous Systems](https://mlsanity.com/paper/2502.21308/) (2025)
- [Scenario-based Compositional Verification of Autonomous Systems with Neural Perception](https://mlsanity.com/paper/2504.20942/) (2025)

## Built on by (most cited first, 3 of 3)

- [Safety Generalization Under Distribution Shift in Safe Reinforcement Learning: A Diabetes Testbed](https://mlsanity.com/paper/2601.21094/) (2026)
- [Latent Q-Barrier Shielding for Safe In-Context Reinforcement Learning](https://mlsanity.com/paper/2605.25267/) (2026)
- [Interval POMDP Shielding for Imperfect-Perception Agents](https://mlsanity.com/paper/2604.20728/) (2026)

## Abstract (arXiv)

We consider the problem of safe control in discrete autonomous agents that use learned components for imperfect perception (or more generally, state estimation) from high-dimensional observations. We propose a shield construction that provides run-time safety guarantees under perception errors by restricting the actions available to an agent, modeled as a Markov decision process, as a function of the state estimates. Our construction uses conformal prediction for the perception component, which guarantees that for each observation, the predicted set of estimates includes the actual state with a user-specified probability. The shield allows an action only if it is allowed for all the estimates in the predicted set, resulting in local safety. We also articulate and prove a global safety property of existing shield constructions for perfect-perception agents bounding the probability of reaching unsafe states if the agent always chooses actions prescribed by the shield. We illustrate our approach with a case-study of an experimental autonomous system that guides airplanes on taxiways using high-dimensional perception DNNs.