mlsanityData to 14 Oct 2026

CALOS: Control-Affine Lyapunov On-manifold Safety Layer for Safe Deep Reinforcement Learning for Quadrotors

arXiv 2609.17758 ↗First posted 15 Sep 2026cs.RO cs.AIcited by 0

On the map: Deep reinforcement learning methods → RL for robot navigation and UAV control

Abstract (arXiv)

Deep Reinforcement Learning has demonstrated remarkable capability in quadrotor control, yet learned policies offer no guarantee of respecting safety constraints during training or deployment. We present CALOS (Control-Affine Lyapunov On-manifold Safety), a runtime safety layer that enforces attitude constraints on a quadrotor without modifying the underlying learning algorithm. CALOS formulates four tilt-angle inequalities and a Lyapunov descent condition as a single quadratic program whose solution is the minimum-norm correction to the nominal torque output of the policy. The quadratic program is solved exactly via active-set enumeration over the three-dimensional torque space, with a computational cost low enough to enforce constraints in real time across thousands of parallel simulation environments, as required by modern massively parallel Deep Reinforcement Learning training. Evaluated on trajectory-tracking tasks in NVIDIA Isaac Lab, CALOS reduces lateral tracking error by 55-60% relative to an unconstrained Proximal Policy Optimization baseline while achieving zero attitude-constraint violations on the training trajectory. By restricting exploration to safe regions of the state space, the safety layer also accelerates training convergence and improves data efficiency without producing suboptimal policies.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in RL for robot navigation and UAV control, one of 15 directions in Deep reinforcement learning methods. Builds on 10 of its 17 references that are in the map. First posted on arXiv 15 Sep 2026.

Builds on

10 of 17 references.

References · 10

  1. 2017Proximal Policy Optimization Algorithmscited by 15,215
  2. 2017Safe Reinforcement Learning via Shieldingcited by 449
  3. 2018Reward Constrained Policy Optimizationcited by 417
  4. 2018A Lyapunov-based Approach to Safe Reinforcement Learningcited by 297
  5. 2019End-to-End Safe Reinforcement Learning through Barrier Functions for Safety-Critical Continuous Control Taskscited by 289
  6. 2019Lyapunov-based Safe Policy Optimization for Continuous Controlcited by 145
  7. 2021Lyapunov-stable neural-network controlcited by 66
  8. 2023Neural Lyapunov Control for Discrete-Time Systemscited by 23
  9. 2024Safe Reinforcement Learning on the Constraint Manifold: Theory and Applicationscited by 15
  10. 2023Stable and Safe Reinforcement Learning via a Barrier-Lyapunov Actor-Critic Approachcited by 13

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026SafePG: Safe and Globally Optimal Reinforcement Learning with Hard Constraints
  2. 2026Online Reinforcement Learning for Safe Gain Scheduling in Nonlinear Quadrotor Control
  3. 2026Learning over Forward-Invariant Policy Classes: Reinforcement Learning without Safety Concerns
  4. 2026Anticipatory Risk-Guided Reinforcement Learning for Safe Flight Through Dynamic Clutter
  5. 2024Sampling-based Safe Reinforcement Learning for Nonlinear Dynamical Systems
  6. 2026Safe Execution of RL Policies Via Acceleration-Based CBF-QP Constraint Enforcement for Real-World Robotic Deployments
  7. 2026High-Speed Vision-Based Flight in Clutter with Safety-Shielded Reinforcement Learning
  8. 2026LEAP-CBF: A Safety Filter for Uncertain Systems with Least-Effort Adversarial Potentials