Learning Contextual Runtime Monitors for Safe AI-Based Autonomy

• URL: http://arxiv.org/abs/2601.20666v1
• Date: Wed, 28 Jan 2026 14:49:04 GMT
• Title: Learning Contextual Runtime Monitors for Safe AI-Based Autonomy
• Authors: Alejandro Luque-Cerpa, Mengyuan Wang, Emil Carlsson, Sanjit A. Seshia, Devdatt Dubhashi, Hazem Torfah,
• Abstract summary: We introduce a novel framework for learning context-aware runtime monitors for AI-based control ensembles.<n>Our approach comes with two key benefits: (1) theoretical safety guarantees during controller selection, and (2) improved utilization of controller diversity.
• Score: 42.24674899629751
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: We introduce a novel framework for learning context-aware runtime monitors for AI-based control ensembles. Machine-learning (ML) controllers are increasingly deployed in (autonomous) cyber-physical systems because of their ability to solve complex decision-making tasks. However, their accuracy can degrade sharply in unfamiliar environments, creating significant safety concerns. Traditional ensemble methods aim to improve robustness by averaging or voting across multiple controllers, yet this often dilutes the specialized strengths that individual controllers exhibit in different operating contexts. We argue that, rather than blending controller outputs, a monitoring framework should identify and exploit these contextual strengths. In this paper, we reformulate the design of safe AI-based control ensembles as a contextual monitoring problem. A monitor continuously observes the system's context and selects the controller best suited to the current conditions. To achieve this, we cast monitor learning as a contextual learning task and draw on techniques from contextual multi-armed bandits. Our approach comes with two key benefits: (1) theoretical safety guarantees during controller selection, and (2) improved utilization of controller diversity. We validate our framework in two simulated autonomous driving scenarios, demonstrating significant improvements in both safety and performance compared to non-contextual baselines.

Related papers

• Improving the Resilience of Quadrotors in Underground Environments by Combining Learning-based and Safety Controllers [22.566692834880396]
  We train a normalizing flow-based prior over the environment, which provides a measure of how far out-of-distribution the quadrotor is at any given time.<n>We use this measure as a runtime monitor, allowing us to switch between a learning-based controller and a safe controller when we are sufficiently out-of-distribution.
  arXiv  Detail & Related papers  (2025-09-02T20:22:54Z)
• SafeAuto: Knowledge-Enhanced Safe Autonomous Driving with Multimodal Foundation Models [63.71984266104757]
  We propose SafeAuto, a framework that enhances MLLM-based autonomous driving by incorporating both unstructured and structured knowledge.<n>To explicitly integrate safety knowledge, we develop a reasoning component that translates traffic rules into first-order logic.<n>Our Multimodal Retrieval-Augmented Generation model leverages video, control signals, and environmental attributes to learn from past driving experiences.
  arXiv  Detail & Related papers  (2025-02-28T21:53:47Z)
• Dynamic Simplex: Balancing Safety and Performance in Autonomous Cyber Physical Systems [1.3309898919316483]
  We propose a simplex strategy with an online controller switching logic that allows two-way switching. We show that the proposed approach results in fewer collisions and higher performance than state-of-the-art alternatives.
  arXiv  Detail & Related papers  (2023-02-20T04:00:53Z)
• Recursively Feasible Probabilistic Safe Online Learning with Control Barrier Functions [60.26921219698514]
  We introduce a model-uncertainty-aware reformulation of CBF-based safety-critical controllers. We then present the pointwise feasibility conditions of the resulting safety controller. We use these conditions to devise an event-triggered online data collection strategy.
  arXiv  Detail & Related papers  (2022-08-23T05:02:09Z)
• A Safety-Critical Decision Making and Control Framework Combining Machine Learning and Rule-based Algorithms [15.613795936398606]
  Rule-based methods dominate in safety-critical systems. Yet, the latter cannot compete with the first ones in robustness to multiple requirements. This paper proposes a decision making and control framework, which profits from advantages of both the rule- and machine-learning-based techniques.
  arXiv  Detail & Related papers  (2022-01-30T13:58:51Z)
• safe-control-gym: a Unified Benchmark Suite for Safe Learning-based Control and Reinforcement Learning [3.9258421820410225]
  We propose a new open-source benchmark suite, called safe-control-gym. Our starting point is OpenAI's Gym API, which is one of the de facto standard in reinforcement learning research. We show how to use safe-control-gym to quantitatively compare the control performance, data efficiency, and safety of multiple approaches.
  arXiv  Detail & Related papers  (2021-09-13T21:09:28Z)
• Closing the Closed-Loop Distribution Shift in Safe Imitation Learning [80.05727171757454]
  We treat safe optimization-based control strategies as experts in an imitation learning problem. We train a learned policy that can be cheaply evaluated at run-time and that provably satisfies the same safety guarantees as the expert.
  arXiv  Detail & Related papers  (2021-02-18T05:11:41Z)
• Enforcing robust control guarantees within neural network policies [76.00287474159973]
  We propose a generic nonlinear control policy class, parameterized by neural networks, that enforces the same provable robustness criteria as robust control. We demonstrate the power of this approach on several domains, improving in average-case performance over existing robust control methods and in worst-case stability over (non-robust) deep RL methods.
  arXiv  Detail & Related papers  (2020-11-16T17:14:59Z)
• Training Adversarial Agents to Exploit Weaknesses in Deep Control Policies [47.08581439933752]
  We propose an automated black box testing framework based on adversarial reinforcement learning. We show that the proposed framework is able to find weaknesses in both control policies that were not evident during online testing.
  arXiv  Detail & Related papers  (2020-02-27T13:14:53Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.