Pith. sign in

REVIEW 3 major objections 6 minor 40 references

Learning Safe Control via On-the-Fly Bandit Exploration

T0 review · 3 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper proves that a robot can keep itself safe while learning unknown dynamics, with no backup controller, by exploring on the fly.

desk verdict The algorithm is novel and the simulations look good, but the central safety theorem has a sign error: condition (16) is a lower bound on Δt where the proof needs an upper bound. read the letter →

arxiv 2506.10279 v1 pith:LBAH5HMC submitted 2025-06-12 cs.RO cs.SYeess.SY

Alexandre Capone , Ryan Cosner , Aaaron Ames , Sandra Hirche This is my paper · ORCID
classification cs.ROcs.SYeess.SY MSC 93C8593D3093C10
keywords controlbarrierfunctionsafeexplorationGaussianprocessbanditmodeluncertaintysafetyfilterlearning-basedquantification
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper addresses the problem of keeping a control system inside a safe set when the dynamics model is unknown and highly uncertain. Standard safety filters become infeasible under large model error, and most existing methods require a safe backup controller to fall back on. The authors propose instead to detect infeasibility, switch to a bandit-style exploration controller that collects data most useful for restoring feasibility, and then return to the safety filter. Their main theorem states that if the sampling time is chosen fast enough, this always restores feasibility before the system leaves the safe set, guaranteeing safety with high probability. This removes the need for a backup controller and works even with a zero-mean prior model and no prior measurements.

What carries the argument

The load-bearing object is the pair of confidence bounds on the time derivative of the control barrier function, $\text{LCB}_N(x,u)$ and $\text{UCB}_N(x,u)$, computed from a Gaussian process posterior over the unknown dynamics. The robust safety filter (12) enforces $\text{LCB}_N(x,u) \geq -\alpha(h(x)) + \epsilon/2$, and when this becomes infeasible the algorithm collects data at the input maximizing $\text{UCB}_N(x,u)$. The proof combines an information-gain bound on the GP posterior variance (Lemmas B.7-B.8) with a bound on how fast the CBF value can decay between samples (Lemma B.10) to show that enough informative data is always gathered before the boundary is reached.

What would settle it

Construct a system satisfying all other assumptions but whose true robust feasibility margin is exactly $\epsilon$ at one state, initialize the system there, and run Algorithm 1 with that chosen $\epsilon$; if the algorithm exits the safe set, the theorem's premise is violated and the guarantee fails. A more practical test is to run the cruise-control or quadrotor simulation with $\epsilon$ deliberately chosen larger than the actual margin and observe a boundary crossing.

Watch

Extended reading notes

Core claim

The central claim is Theorem 3.2: under assumptions bounding the time-derivative of the state, the Lipschitz constants of the safe-set function and the dynamics, a robust feasibility margin, and an RKHS norm bound on the unknown dynamics, there exists a finite data-collection budget $\Delta N_{\max}$ such that, if the sampling time satisfies $\Delta t > \epsilon / (L_\alpha L_h L_{\dot{x}} \Delta N_{\max})$, the closed-loop system under Algorithm 1 is safe with probability at least $1 - \delta$ and stops collecting data after at most $\Delta N_{\max}$ observations. The key mechanism is that whenever the robust safety filter, a second-order cone program built from the lower confidence bound of the CBF time derivative, becomes infeasible, the algorithm applies a control input maximizing the upper confidence bound. This input is aimed at reducing exactly the uncertainty that blocks feasibility, so that the filter becomes feasible again strictly inside the safe set.

Load-bearing premise

The proof requires an a priori known positive margin $\epsilon$ such that the true, unknown dynamics always admit a control input that exceeds the CBF condition by at least $\epsilon$, and this assumption cannot be checked or verified online.

Editorial extensions

If this is right

  • Safety can be guaranteed during learning even when the prior model is zero-mean and no safe backup controller is available.
  • The algorithm provably terminates its exploration phase: after at most $\Delta N_{\max}$ collected observations the safety filter is feasible for all states in the safe set.
  • Higher sampling frequency directly improves the safety guarantee, since the sufficient condition on $\Delta t$ becomes easier to satisfy as $\Delta N_{\max}$ grows.
  • Heuristic exploration based on random inputs is outperformed by UCB-guided exploration in the simulations, supporting the claim that targeted data collection is the reason safety is recovered.
  • The result connects safety guarantees to the regret-like quantity $\beta^2_N \sum_i \gamma_{i,N}$, which grows sublinearly for common kernels, making the condition on $\Delta N_{\max}$ satisfiable for those kernels.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same machinery could be extended to performance objectives: once feasibility is restored, exploration could be re-activated to reduce model uncertainty in regions relevant to tracking, not just safety.
  • The robust feasibility margin $\epsilon$ is an assumption on the true unknown dynamics; a practical extension could estimate it online or adaptively shrink the safe set until a conservative value is identified.
  • The approach suggests a bridge between safe Bayesian optimization and CBF-based control, where the 'reward' being optimized during exploration is the restoration of the safety certificate rather than a task reward.
  • A testable extension is to replace the fixed sampling interval with an event-triggered scheme that explores only when the predicted time-to-infeasibility drops below a threshold, potentially reducing data collection further.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper proposes an online safe-control scheme for control-affine systems with unknown dynamics modeled by Gaussian processes. A CBF-based robust SOCP safety filter (12) is applied whenever it is strictly feasible; when it becomes infeasible, the algorithm switches to an exploration phase in which control inputs maximize the UCB of the CBF derivative (14), collects a measurement after a sampling interval Δt, and updates the GP model. The main theoretical result, Theorem 3.2, claims that under Assumptions 2.1, 2.3, 2.4, 2.6–2.8, a lower-bound condition (16) on Δt, and a finite-data condition (17) on ΔNmax, Algorithm 1 is safe with probability at least 1−δ and stops collecting data after at most ΔNmax observations. Numerical experiments on a cruise-control model and a quadrotor model compare UCB-guided exploration with random exploration at different sampling frequencies.

Significance. If Theorem 3.2 were correct, the paper would represent a useful step: it offers a principled way to handle infeasible robust safety filters without assuming a backup controller, gives an explicit finite-sample data-collection bound, and demonstrates the approach in settings where a zero-mean prior and no prior measurements are used. The GP/RKHS bounding framework and information-gain arguments are standard, and the empirical comparison with random exploration in Section 5.3 is informative. However, the central theorem as stated is not supported by its proof, so the theoretical significance cannot be assessed until the sampling-time condition and the proof of Lemma B.11 are corrected. The paper also does not currently establish a theoretical role for the specific UCB exploration rule, despite claiming that bandit exploration is what makes the data safety-relevant.

major comments (3)
  1. [§3, Theorem 3.2 and Eq. (16); Appendix B, Lemma B.11] The sampling-time condition (16) has the wrong inequality direction for the proof that follows. Lemma B.11's final chain gives Nexit−Nε/2+1 ≥ ε/(2LαLhLxdot Δt) and then claims this is '> ΔNmax from the sampling frequency specification (16)'. Substituting the stated condition Δt > ε/(LαLhLxdot ΔNmax) yields ε/(2LαLhLxdot Δt) < ΔNmax/2, which is the opposite of what is needed. The proof requires an upper bound of the form Δt < ε/(2LαLhLxdot ΔNmax), up to constants, not a lower bound. This is load-bearing: Theorem 3.2 is the paper's central safety guarantee, and the condition as written selects the wrong regime. The experiments in Section 5.3, where failure rates decrease as the sampling frequency (Δt)^{-1} increases, support the corrected direction and further indicate that the stated lower bound does not match the intended claim.
  2. [Appendix B, Lemma B.9] The proof that the safety filter becomes feasible after at most ΔNmax−1 observations never uses the UCB exploration rule (14). It only uses that each collected point is sampled at an infeasible state, i.e. LCB_N(x,u) ≤ max_u LCB_N ≤ −α(h(x))+ε/2, together with Assumption 2.6 and the variance-sum bound. As written, the proof would apply equally to arbitrary exploratory inputs, so the paper's central claim that the bandit/UCB mechanism is what makes the collected data contribute to safety is not established. The authors should either prove a property specific to the UCB rule or revise the contribution statement to separate the broader safety guarantee from the empirical benefit of the UCB choice.
  3. [Assumption 2.6 and Theorem 3.2] The robust-CBF margin ε appears as an input in (12), (16), and (17), but Assumption 2.6 only asserts the existence of some ε for the true unknown dynamics. The paper gives no procedure for selecting or verifying ε in practice. If ε is chosen too large, Assumption 2.6 may fail and the proof of Lemma B.9 collapses; if ε is chosen too small, the safety filter is weaker and the guarantee in Theorem 3.2 is less useful. Since the main theorem is conditioned on this unknown constant, the practical applicability of the central result is not yet established.
minor comments (6)
  1. [Appendix B, Lemma B.3] The proof of Lemma B.3 inserts a spurious '+ε/2' before applying the Lipschitz bound; the inequality ˙h ≥ ∂h/∂x µ − Lh β_N sqrt(tr(Σ_N)) + ε/2 does not follow from Lemma B.1. The lemma itself is still true if the spurious term is removed, so this appears to be a proof typo, but it should be fixed.
  2. [Theorem 3.2] The theorem states 'Choose δ ∈ [0,1]', but β_N in (11) contains log(nδ^{-1}), which is undefined for δ=0 and δ=1; the valid range should be δ∈(0,1).
  3. [Algorithm 1, line 20] Line 20 writes 'Set D_N = D_{N−1} ∩ {z(N), y(N)}'; this should presumably be a set union, not an intersection.
  4. [Appendix B, Lemmas B.7 and B.8] The index notation is inconsistent: Lemma B.7 sums σ^2_{i,q}(z^{(q−1)}), while Lemma B.8 refers to Σ_N(x^{(N)},u^{(N)}) in the summation. The indexing should be aligned so that the posterior variance used in the sum is clearly the one available before the corresponding measurement is incorporated.
  5. [Appendix B, Lemma B.11] The final off-by-one accounting is not consistent: the displayed bound uses (Nexit−Nε/2−1) terms in the sum but concludes Nexit−Nε/2+1. Even after the direction of the sampling condition is corrected, the constants and integer arithmetic should be rechecked.
  6. [Sections 1 and 2.1] The function α is called a 'class-KL function' in the abstract and Algorithm 1, but Definition 2.5 defines a strictly increasing function with α(0)=0, which is a class-K-type function; the terminology should be aligned.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the safety guarantee follows from explicit assumptions and a standard GP confidence bound, not from fitting or from defining the conclusion into the assumptions. A separate algebraic gap in Lemma B.11 is a correctness issue, not circularity.

full rationale

The derivation chain is not circular. The only load-bearing result cited from overlapping-author prior work is Lemma B.1, which is attributed to Capone & Hirche (2019, Lemma 1); it is a standard RKHS/GP uniform confidence bound whose stated assumptions (Assumptions 2.7 and 2.8) do not include the paper's target claim of safety under on-the-fly bandit exploration, so under the review rules it counts as independent support and does not raise the circularity score. Theorem 3.2 then proceeds by showing that strict feasibility of the robust filter (12) implies CBF safety (Lemmas B.4, B.5, Corollary B.6); that infeasibility triggers exploration (13)-(14) and the information-gain sum bound (Lemma B.8) forces strict feasibility after at most DeltaNmax samples (Lemma B.9); and that at least DeltaNmax samples are collected before exit (Lemma B.11). No parameter is fitted to the safety outcome, and Assumption 2.6 is an explicit feasibility margin assumption rather than a disguised version of the conclusion. The paper's own Section 4 flags Assumption 2.6 as potentially strict, which is a practical limitation but not circularity. Separately, there is a serious non-circular correctness gap: the final inequality of Lemma B.11 uses the sampling-frequency specification (16), but (16) is a lower bound on Delta t while the proof requires an upper bound; substituting (16) gives epsilon/(2 Lalpha Lh Lxdot Delta t) < DeltaNmax/2, so the claimed '> DeltaNmax' does not follow. This is an algebraic error invalidating Theorem 3.2 as stated, but it is not a reduction of the theorem to its own inputs.

Assumptions & free parameters 4 free parameters · 7 assumptions · 0 invented entities

The paper introduces no new physical or mathematical entities; it builds on GPs, CBFs, and standard regret bounds. The load-bearing inputs are assumptions about the true dynamics (notably Assumption 2.6 and the known RKHS bounds) plus user-chosen parameters such as epsilon, alpha, and Delta t, with kernel hyperparameters fitted to data in the experiments. The sampling-time theorem condition is internally inconsistent.

free parameters (4)
  • Kernel hyperparameters (squared-exponential lengthscales and amplitude) = Estimated from 10 initial data points via maximum posterior likelihood in both simulations
    The GP posterior mean and variance depend on these fitted values. The theorem treats the kernel and the RKHS bound B_i as known inputs, but the experiments fit the kernel to data, which is not covered by Assumption 2.8 as stated.
  • CBF margin epsilon = Chosen by user; exact value not specified in the experiments
    Appears in Assumption 2.6 and in the safety filter (12) as the margin epsilon/2. The theorem requires Assumption 2.6 to hold for this epsilon, but the paper does not say how to select or validate it.
  • Class-KL function alpha = Chosen by user; explicit form not given
    Shapes the CBF condition. Its Lipschitz constant L_alpha appears in the sampling bound (16), so the user-chosen alpha directly affects the theoretical condition.
  • Sampling time Delta t = 1e-5 s in the cruise control and quadrotor simulations
    Controls the exploration duration and data collection rate. The theorem's condition on Delta t is stated incorrectly (reversed direction), making the selection rule inconsistent with the proof.
assumptions (7)
  • domain assumption Assumption 2.1: there exists a known positive constant L_xdot such that ||xdot||_2 <= L_xdot for all x in X and u in U.
    Used throughout to bound how fast the state and CBF can change during exploration intervals, e.g., in Lemma B.11.
  • domain assumption Assumption 2.3: the safe set C is compact and h admits a global Lipschitz constant L_h.
    Needed for the LCB/UCB definitions and for bounding the CBF decrease rate in the proof of Theorem 3.2.
  • domain assumption Assumption 2.4: the nominal controller is locally Lipschitz continuous.
    Used to guarantee well-defined closed-loop trajectories under the safety filter, via Lemma B.2 and existence arguments.
  • domain assumption Assumption 2.6: there exists a positive margin epsilon such that sup_u dh/dx (f+gu) >= -alpha(h(x)) + epsilon for all x.
    This is the key feasibility assumption on the true unknown dynamics. It is used in Lemma B.9 to show that infeasibility of the filter implies a bounded variance-to-margin gap. It cannot be verified online and is load-bearing.
  • domain assumption Assumption 2.7: measurement noise is i.i.d. sub-Gaussian with parameter sigma_i,ns.
    Required for the GP concentration bound in Lemma B.1 and for the information-gain-based bounds in Lemma B.7.
  • domain assumption Assumption 2.8: each entry of f(x)+g(x)u belongs to the RKHS of the composite kernel with known norm bound B_i.
    This is the standard RKHS assumption that makes uniform GP error bounds possible. The known bound B_i is required in (11) and is hard to obtain in practice when the kernel is estimated from data.
  • standard math Lemma B.7 uses the standard information-gain bounds of Srinivas et al. (2012, Lemmas 5.3 and 5.4).
    The paper cites and applies these bounds without reproving them. They are well-established results for GP posterior variance sums and are used to bound the cumulative uncertainty during exploration.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Learning Safe Control via On-the-Fly Bandit Exploration." pith.science (2026). https://pith.science/paper/LBAH5HMC

@misc{pith2026250610279,
  author       = {Pith},
  title        = {Pith review of: Learning Safe Control via On-the-Fly Bandit Exploration},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LBAH5HMC}},
  note         = {Machine review of arXiv:2506.10279}
}
read the original abstract

Control tasks with safety requirements under high levels of model uncertainty are increasingly common. Machine learning techniques are frequently used to address such tasks, typically by leveraging model error bounds to specify robust constraint-based safety filters. However, if the learned model uncertainty is very high, the corresponding filters are potentially invalid, meaning no control input satisfies the constraints imposed by the safety filter. While most works address this issue by assuming some form of safe backup controller, ours tackles it by collecting additional data on the fly using a Gaussian process bandit-type algorithm. We combine a control barrier function with a learned model to specify a robust certificate that ensures safety if feasible. Whenever infeasibility occurs, we leverage the control barrier function to guide exploration, ensuring the collected data contributes toward the closed-loop system safety. By combining a safety filter with exploration in this manner, our method provably achieves safety in a setting that allows for a zero-mean prior dynamics model, without requiring a backup controller. To the best of our knowledge, it is the first safe learning-based control method that achieves this.

Figures

Figures reproduced from arXiv: 2506.10279 by the authors.

Figure 1
Figure 1. Illustration of our method (Algorithm 1) and theoretical result (Theorem 3.2). 2. We rigorously analyze our approach and discuss our main theorem, which shows the interdependence be￾tween model uncertainty, sampling time, and safety. 3. We showcase our approach using two numerical simu￾lations where we achieve safety using a control barrier function, a dynamics model with a zero-mean prior, and no prior measurements… view at source ↗
Figure 2
Figure 2. CBF and LCB for a single cruise control simulation using a sampling time of ∆t = 10−5 s. 4. Discussion and Limitations Although we assume that the system structure (1) is affine in the control input, this is not strictly a requirement for theoretical guarantees and can be relaxed, provided that the optimization problems (12), (13) and (14) can be solved adequately. Theorem 3.2 points at an interaction between variou… view at source ↗
Figure 3
Figure 3. Visualization of a single trajectory of the quadrotor experiment using a sampling time of ∆t = 10−5 s. 0 10 20 30 40 50 Time 0 5 10 15 20 25 30 C B F l e v e l s e t CBF for height Di,. .atness-based CBF (a) Value of control barrier functions for quadrotor example. 0 10 20 30 40 50 Time -200 0 200 400 600 800 1000 E s t. C B F d e c r e a s e c o n d. CBF for height Di,. .atness-based CBF 0 0.5 1 1.5 2 0 200 400 600… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: CBF and LCB for a single quadrotor run using a sampling time of ∆t = 10−5 s. target vehicle in front, and v denotes the ego vehicle speed. Our setup is fully described in Appendix A.1. As a control barrier function, we employ h(x) = z − Thv, where Th = 1.8, which aims …
Figure 5
Figure 5. Figure 5: Rate of failure using our approach and random inputs for the cruise control (left) and quadrotor (right) settings at varying sampling frequencies (∆t) −1 . keeping altitude higher than zero. The second CBF utilizes the differential flatness of the quadrotor to restrict…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 35 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

  2. [2]

    and Sreenath, K

    Agrawal, A. and Sreenath, K. Discrete control barrier functions for safety-critical control of discrete systems with application to bipedal robot navigation. In Robotics: Science and Systems, volume 13, pp.\ 1--10. Cambridge, MA, USA, 2017

    work page

  3. [3]

    K., Fisac, J

    Akametalu, A. K., Fisac, J. F., Gillula, J. H., Kaynama, S., Zeilinger, M. N., and Tomlin, C. J. Reachability-based safe learning with G aussian processes. In 53rd IEEE Conference on Decision and Control, pp.\ 1424--1431, 2014. doi:10.1109/CDC.2014.7039601

    work page arXiv 2014

  4. [4]

    D., Xu, X., Grizzle, J

    Ames, A. D., Xu, X., Grizzle, J. W., and Tabuada, P. Control Barrier Function Based Quadratic Programs for Safety Critical Systems . IEEE Transactions on Automatic Control, 62 0 (8): 0 3861--3876, August 2017. ISSN 0018-9286, 1558-2523. doi:10.1109/TAC.2016.2638961. URL http://arxiv.org/abs/1609.06408. arXiv: 1609.06408

    arXiv 2017

  5. [5]

    J., As, Y., Li, R., Van Assche , K., Davoodi, A., Cavalcanti, N

    Ao, Y., Esfandiari, H., Carrillo, F., Laux, C. J., As, Y., Li, R., Van Assche , K., Davoodi, A., Cavalcanti, N. A., Farshad, M., Grewe, B. F., Vander Poorten , E., Krause, A., and Fürnstahl, P. Saferplan: Safe deep reinforcement learning for intraoperative planning of pedicle screw placement. Medical Image Analysis, 99: 0 103345, 2025. ISSN 1361-8415

    work page 2025

  6. [6]

    Bansal, S., Chen, M., Herbert, S., and Tomlin, C. J. Hamilton-jacobi reachability: A brief overview and recent advances. In 2017 IEEE 56th Annual Conference on Decision and Control (CDC), pp.\ 2242--2253. IEEE, 2017

    work page 2017

  7. [7]

    Safe model-based reinforcement learning with stability guarantees

    Berkenkamp, F., Turchetta, M., Schoellig, A., and Krause, A. Safe model-based reinforcement learning with stability guarantees. In Advances in Neural Information Processing Systems, pp.\ 908--919, 2017

    work page 2017

  8. [8]

    P., and Krause, A

    Berkenkamp, F., Schoellig, A. P., and Krause, A. No-regret B ayesian optimization with unknown hyperparameters. Journal of Machine Learning Research, 20 0 (50): 0 1--24, 2019

    work page 2019

Show all 40 references
  1. [9]

    and Hirche, S

    Capone, A. and Hirche, S. Backstepping for partially unknown nonlinear systems using G aussian processes. IEEE Control Systems Letters, 3: 0 416--421, 2019

    2019

  2. [10]

    G aussian process uniform error bounds with unknown hyperparameters for safety-critical applications

    Capone, A., Lederer, A., and Hirche, S. G aussian process uniform error bounds with unknown hyperparameters for safety-critical applications. In International Conference on Machine Learning, pp.\ 2609--2624. PMLR, 2022

    2022

  3. [11]

    J., Zhang, B., Tomlin, C

    Castañeda, F., Choi, J. J., Zhang, B., Tomlin, C. J., and Sreenath, K. Pointwise feasibility of G aussian process-based safety-critical control under model uncertainty. In 2021 60th IEEE Conference on Decision and Control (CDC), pp.\ 6762--6769, 2021. doi:10.1109/CDC45484.2021.9683743

    2021

  4. [12]

    J., Jung, W., Zhang, B., Tomlin, C

    Castañeda, F., Choi, J. J., Jung, W., Zhang, B., Tomlin, C. J., and Sreenath, K. Probabilistic safe online learning with control barrier functions. arXiv preprint arXiv:2208.10733, 2022

    2022 arXiv

  5. [13]

    M., and Burdick, J

    Cheng, R., Orosz, G., Murray, R. M., and Burdick, J. W. End-to-end safe reinforcement learning through barrier functions for safety-critical continuous control tasks. In Proceedings of the AAAI conference on artificial intelligence, volume 33, pp.\ 3387--3395, 2019

    2019

  6. [14]

    J., and Sreenath, K

    Choi, J., Castañeda, F., Tomlin, C. J., and Sreenath, K. Reinforcement learning for safety-critical control under model uncertainty, using control lyapunov functions and control barrier functions. Robotics: Science and Systems (RSS), 2020

    2020

  7. [15]

    Deep reinforcement learning in a handful of trials using probabilistic dynamics models

    Chua, K., Calandra, R., McAllister, R., and Levine, S. Deep reinforcement learning in a handful of trials using probabilistic dynamics models. Advances in neural information processing systems, 31, 2018

    2018

  8. [16]

    Efficient model-based reinforcement learning through optimistic policy search and planning

    Curi, S., Berkenkamp, F., and Krause, A. Efficient model-based reinforcement learning through optimistic policy search and planning. In Advances in Neural Information Processing Systems, volume 33, pp.\ 14156--14170. Curran Associates, Inc., 2020

    2020

  9. [17]

    Autonomous landing of quadrotor based on ground effect modelling

    Danjun, L., Yan, Z., Zongying, S., and Geng, L. Autonomous landing of quadrotor based on ground effect modelling. In 2015 34th Chinese Control Conference (CCC), pp.\ 5647--5652, 2015. doi:10.1109/ChiCC.2015.7260521

    2015

  10. [18]

    Differential flatness of quadrotor dynamics subject to rotor drag for accurate tracking of high-speed trajectories

    Faessler, M., Franchi, A., and Scaramuzza, D. Differential flatness of quadrotor dynamics subject to rotor drag for accurate tracking of high-speed trajectories. IEEE Robotics and Automation Letters, 3 0 (2): 0 620--626, 2018. doi:10.1109/LRA.2017.2776353

    2018

  11. [19]

    W., and Trimpe, S

    Fiedler, C., Scherer, C. W., and Trimpe, S. Practical and rigorous uncertainty bounds for G aussian process regression. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, pp.\ 7439--7447, 2021

    2021

  12. [20]

    Towards a framework for realizable safety critical control through active set invariance

    Gurriet, T., Singletary, A., Reher, J., Ciarletta, L., Feron, E., and Ames, A. Towards a framework for realizable safety critical control through active set invariance. In 2018 ACM/IEEE 9th International Conference on Cyber-Physical Systems (ICCPS), pp.\ 98--106. IEEE, 2018

    2018

  13. [21]

    Hewing, L., Kabzan, J., and Zeilinger, M. N. Cautious model predictive control using G aussian process regression. IEEE Transactions on Control Systems Technology, 2019

    2019

  14. [22]

    P., Menner, M., and Zeilinger, M

    Hewing, L., Wabersich, K. P., Menner, M., and Zeilinger, M. N. Learning-based model predictive control: Toward safe learning in control. Annual Review of Control, Robotics, and Autonomous Systems, 3: 0 269--296, 2020

    2020

  15. [23]

    J., and Zamani, M

    Jagtap, P., Pappas, G. J., and Zamani, M. Control barrier functions for unknown nonlinear systems using G aussian processes. In 2020 59th IEEE Conference on Decision and Control (CDC), pp.\ 3699--3704. IEEE, 2020

    2020

  16. [24]

    Khalil, H. K. Noninear systems. Prentice-Hall, New Jersey, 2 0 (5): 0 5--1, 1996

    1996

  17. [25]

    Learning-based model predictive control for safe exploration

    Koller, T., Berkenkamp, F., Turchetta, M., and Krause, A. Learning-based model predictive control for safe exploration. In IEEE Conference on Decision and Control, pp.\ 6059--6066, 2018

    2018

  18. [26]

    and Ong, C

    Krause, A. and Ong, C. Contextual G aussian process bandit optimization. In Advances in Neural Information Processing Systems, volume 24. Curran Associates, Inc., 2011

    2011

  19. [27]

    Safe barrier-constrained control of uncertain systems via event-triggered learning

    Lederer, A., Begzadi \'c , A., Hirche, S., Cort \'e s, J., and Herbert, S. Safe barrier-constrained control of uncertain systems via event-triggered learning. arXiv preprint arXiv:2408.16144, 2024

    2024

  20. [28]

    A., Xu, Y., and Zhang, H

    Micchelli, C. A., Xu, Y., and Zhang, H. Universal kernels. Journal of Machine Learning Research, 7 0 (Dec): 0 2651--2667, 2006

    2006

  21. [29]

    and Sreenath, K

    Nguyen, Q. and Sreenath, K. Exponential control barrier functions for enforcing high relative-degree safety-critical constraints. In 2016 American Control Conference (ACC), pp.\ 322--328. IEEE, 2016

    2016

  22. [30]

    Near-optimal multi-agent learning for safe coverage control

    Prajapat, M., Turchetta, M., Zeilinger, M., and Krause, A. Near-optimal multi-agent learning for safe coverage control. Advances in Neural Information Processing Systems, 35: 0 14998--15012, 2022

    2022

  23. [31]

    Rasmussen, C. E. and Williams, C. K. G aussian processes for machine learning. 2006. The MIT Press, Cambridge, MA, USA, 2006

    2006

  24. [32]

    Rodriguez, I. D. J., Rosolia, U., Ames, A. D., and Yue, Y. Learning unstable dynamics with one minute of data: A differentiation-based G aussian process approach. arXiv preprint, 2021

    2021

  25. [33]

    M., and Seeger, M

    Srinivas, N., Krause, A., Kakade, S. M., and Seeger, M. W. Information-theoretic regret bounds for G aussian process optimization in the bandit setting. IEEE Transactions on Information Theory, 58 0 (5): 0 3250--3265, 2012

    2012

  26. [34]

    Safe exploration for optimization with G aussian processes

    Sui, Y., Gotovos, A., Burdick, J., and Krause, A. Safe exploration for optimization with G aussian processes. In Proceedings of the 32nd International Conference on Machine Learning, volume 37 of Proceedings of Machine Learning Research, pp.\ 997--1005, Lille, France, 07--09 J...

    2015

  27. [35]

    Gosafeopt: Scalable safe exploration for global optimization of dynamical systems

    Sukhija, B., Turchetta, M., Lindner, D., Krause, A., Trimpe, S., and Baumann, D. Gosafeopt: Scalable safe exploration for global optimization of dynamical systems. Artificial Intelligence, 320: 0 103922, 2023. ISSN 0004-3702

    2023

  28. [36]

    Learning for safety-critical control with control barrier functions

    Taylor, A., Singletary, A., Yue, Y., and Ames, A. Learning for safety-critical control with control barrier functions. In Proceedings of the 2nd Conference on Learning for Dynamics and Control, volume 120 of Proceedings of Machine Learning Research, pp.\ 708--717. PMLR, 10--11...

    2020

  29. [37]

    Safe exploration in finite markov decision processes with G aussian processes

    Turchetta, M., Berkenkamp, F., and Krause, A. Safe exploration in finite markov decision processes with G aussian processes. In Advances in Neural Information Processing Systems, pp.\ 4312--4320, 2016

    2016

  30. [38]

    Wabersich, K. P. and Zeilinger, M. N. Linear Model Predictive Safety Certification for Learning - Based Control . In 2018 IEEE Conference on Decision and Control ( CDC ) , pp.\ 7130--7135, December 2018. doi:10.1109/CDC.2018.8619829. ISSN: 2576-2370

    2018

  31. [39]

    and Sui, Y

    Wachi, A. and Sui, Y. Safe reinforcement learning in constrained M arkov decision processes. In Proceedings of the 37th International Conference on Machine Learning, volume 119 of Proceedings of Machine Learning Research, pp.\ 9797--9806. PMLR, 13--18 Jul 2020

    2020

  32. [40]

    Safety-critical model predictive control with discrete-time control barrier function

    Zeng, J., Zhang, B., and Sreenath, K. Safety-critical model predictive control with discrete-time control barrier function. In 2021 American Control Conference (ACC), pp.\ 3882--3889. IEEE, 2021

    2021

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.