The DOI matches its Crossref record; 32 of the 34 references it cites were found in a registry or an index.
End-to-End Safe Reinforcement Learning through Barrier Functions for Safety-Critical Continuous Control Tasks
Open accessAbstract
Reinforcement Learning (RL) algorithms have found limited success beyond simulated applications, and one main reason is the absence of safety guarantees during the learning process. Real world systems would realistically fail or break before an optimal controller can be learned. To address this issue, we propose a controller architecture that combines (1) a model-free RL-based controller with (2) model-based controllers utilizing control barrier functions (CBFs) and (3) online learning of the unknown system dynamics, in order to ensure safety during learning. Our general framework leverages the success of RL algorithms to learn high-performance controllers, while the CBF-based controllers both guarantee safety and guide the learning process by constraining the set of explorable polices. We utilize Gaussian Processes (GPs) to model the system dynamics and its uncertainties. Our novel controller synthesis algorithm, RL-CBF, guarantees safety with high probability during the learning process, regardless of the RL algorithm used, and demonstrates greater policy exploration efficiency. We test our algorithm on (1) control of an inverted pendulum and (2) autonomous carfollowing with wireless vehicle-to-vehicle communication, and show that our algorithm attains much greater sample efficiency in learning than other state-of-the-art algorithms and maintains safety during the entire learning process.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Reinforcement learning (RL) algorithms have been increasingly adopted by the robotics community over the past years to control complex robots or multi-robot systems [1], [2], or provide end-to-end policies from perception to control [3].”
From Sim-to-Real Transfer in Deep Reinforcement Learning for Robotics: a Survey (2020) · cites this paper as [3]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Gaussian Processes for Machine Learning
Rasmussen, C. E., & Williams, C. K. I. (2005) · doi:10.7551/mitpress/3206.001.0001 · Cited by 11,705
Continuous control with deep reinforcement learning
Lillicrap, T. P., Hunt, J. J., Pritzel, A., Heess, N., Erez, T., Tassa, Y., Silver, D., & Wierstra, D. (2015) · arXiv · doi:10.48550/arxiv.1509.02971 · Cited by 10,782
(1995) · Cited by 10,558
Trust region policy optimization
Schulman, J., Levine, S., Moritz, P., Jordan, M. I., & Abbeel, P. (2015) · arXiv · doi:10.48550/arxiv.1502.05477 · Cited by 4,578 · cited 13×
“Furthermore, if we use TRPO for the RL algorithm, then the control law Proof.”
MethodInfluentialDeterministic Policy Gradient Algorithms
Silver, D., Lever, G., Heess, N., Degris, T., Wierstra, D., & Riedmiller, M. (2014) · HAL · Cited by 3,275
Control Barrier Function Based Quadratic Programs for Safety Critical Systems
Ames, A. D., Xu, X., Grizzle, J. W., & Paulo Tabuada, B. (2016) · IEEE Transactions on Automatic Control · doi:10.1109/tac.2016.2638961 · Cited by 2,920 · cited 1×
“The existence of a CBF implies that there exists a deterministic controller u CBF : S → A such that the set C is forward invariant for system (1) (Agrawal and Sreenath 2017; Ames et al. 2017).”
BackgroundApprenticeship learning via inverse reinforcement learning
Abbeel, P., & Ng, A. Y. (2004) · International Conference on Machine Learning · doi:10.1145/1015330.1015430 · Cited by 2,855 · cited 1×
“Previous approaches to safe reinforcement learning include reward-shaping , policy optimization with constraints (Gaskett 2003; Moldovan and Abbeel 2012; Achiam et al. 2017; Wachi et al. 2018), or teacher advice (Abbeel and Ng 2004; Abbeel, Coates, and Ng 2010; Tang et al. 2010).”
Background(2015) · Journal of Machine Learning Research · Cited by 1,119
Benchmarking Deep Reinforcement Learning for Continuous Control
Duan, Y., Chen, X., Houthooft, R., Schulman, J., & Abbeel, P. (2016) · arXiv · doi:10.48550/arxiv.1604.06778 · Cited by 1,114 · cited 1×
“This method has been successfully applied to continuous control tasks (Duan et al. 2016; Lillicrap et al. 2015) where controllers have learned to stabilize complex robots (after many policy iterations).”
MethodReinforcement learning of motor skills with policy gradients
Peters, J., & Schaal, S. (2008) · Neural Networks · doi:10.1016/j.neunet.2008.02.003 · Cited by 831
Cited by
Sorted by citations
Items 601–603 of 603
Feng, R., Xu, Z., Ding, X., et al. (2026) · Journal of Electrical Engineering and Technology · doi:10.1007/s42835-026-02995-3
Safe Data-Driven Optimal Output Regulation for Unknown Continuous-Time Linear MIMO SystemsPreprint
Wang, X., Shen, S., & Li, Z. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7629590
Oleynikova, O. O., Potapov, B. V., & Олегович, К. С. (2026) · International Ural Conference on Electrical Power Engineering (UralCon) · doi:10.1109/uralcon71028.2026.11734710
“Safety-constrained reinforcement learning can use barrier-function mechanisms to keep the learned policy within defined safety conditions [13].”
Metadata
- Authors
- Cheng, R., Orosz, G., Murray, R. M., & Burdick, J. W.
- Affiliations
- California Institute of Technology (US)
- University of Michigan (US)
- Venue
- Proceedings of the AAAI Conference on Artificial Intelligence
- Published
- 2019-07-17
- Volume
- 33
- Issue
- 01
- Pages
- 3387-3395
- ISSN
- 2159-5399 (ISSN-L)
- 2374-3468
- OpenAlex
- W2966735560
- Semantic Scholar
- adcd4bfd88213da0c33d3cb7057411dd15b72c7a
- Cited by
- 611
- References
- 34
- Citations by year
- 2026: 79 · 2025: 84 · 2024: 79 · 2023: 102 · 2022: 84 · 2021: 112 · 2020: 60 · 2019: 10
- FWCI
- 25.44
- Type
- conference paper
- Language
- en
- Open access
- diamond
- doi.org
- Licence
- www.aaai.org/ (unspecified)
- Locations
- Proceedings of the AAAI Conference on Artificial Intelligence (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- AAAI Publications (The Association for the Advancement of Artificial Intelligence (AAAI)) (submittedVersion)
- CaltechAUTHORS (California Institute of Technology) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Reinforcement Learning in Robotics
- Physical Sciences › Engineering › Control and Systems Engineering › Adaptive Dynamic Programming Control
- Physical Sciences › Computer Science › Computational Theory and Mathematics › Formal Methods in Verification
- Keywords
- safe reinforcement learning, control barrier functions, safety-critical control, continuous control, model-based control, policy exploration, sample efficiency, inverted pendulum, vehicle-to-vehicle communication
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
