# SHIELD: Safety on Humanoids via CBFs In Expectation on Learned Dynamics

arXiv: https://arxiv.org/abs/2505.11494 · HTML: https://mlsanity.com/paper/2505.11494/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 16 May 2025.

## Built on by (most cited first, 7 of 7)

- [Contextual Safety Reasoning and Grounding for Open-World Robots](https://mlsanity.com/paper/2602.19983/) (2026)
- [CMP: Robust Whole-Body Tracking for Loco-Manipulation via Competence Manifold Projection](https://mlsanity.com/paper/2604.07457/) (2026)
- [Safe Execution of RL Policies Via Acceleration-Based CBF-QP Constraint Enforcement for Real-World Robotic Deployments](https://mlsanity.com/paper/2607.14488/) (2026)
- [Constrained Whole-Body Tracking for Humanoid Robots](https://mlsanity.com/paper/2606.00374/) (2026)
- [ResSafe: Learning Safety Filtering with Residual Reinforcement Learning for Humanoids](https://mlsanity.com/paper/2609.15988/) (2026)
- [DODGER: Safety-Guided Reinforcement Learning for Robot Navigation Among Dynamic Obstacles](https://mlsanity.com/paper/2609.38873/) (2026)
- [Safe Reinforcement Learning with Receding-Horizon Safety Critic and Differentiable Action Shields](https://mlsanity.com/paper/2610.20255/) (2026)

## Abstract (arXiv)

Robot learning has produced remarkably effective ``black-box'' controllers for complex tasks such as dynamic locomotion on humanoids. Yet ensuring dynamic safety, i.e., constraint satisfaction, remains challenging for such policies. Reinforcement learning (RL) embeds constraints heuristically through reward engineering, and adding or modifying constraints requires retraining. Model-based approaches, like control barrier functions (CBFs), enable runtime constraint specification with formal guarantees but require accurate dynamics models. This paper presents SHIELD, a layered safety framework that bridges this gap by: (1) training a generative, stochastic dynamics residual model using real-world data from hardware rollouts of the nominal controller, capturing system behavior and uncertainties; and (2) adding a safety layer on top of the nominal (learned locomotion) controller that leverages this model via a stochastic discrete-time CBF formulation enforcing safety constraints in probability. The result is a minimally-invasive safety layer that can be added to the existing autonomy stack to give probabilistic guarantees of safety that balance risk and performance. In hardware experiments on an Unitree G1 humanoid, SHIELD enables safe navigation (obstacle avoidance) through varied indoor and outdoor environments using a nominal (unknown) RL controller and onboard perception.