Pith. sign in

REVIEW 3 major objections 5 minor 30 references

Embedded Safe Reactive Navigation for Multirotors Systems using Control Barrier Functions

T0 review · 3 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

Pith's one-line read This paper claims that a live-range composite control barrier function can be embedded in the PX4 autopilot to filter acceleration setpoints and prevent collisions in unknown environments without a map.

desk verdict A clean, honest embedded-CBF paper: real PX4 integration, open code, hardware demos, but the limited-FoV safety argument is soft where it matters most. read the letter →

arxiv 2504.15850 v1 pith:DC7WUDWW submitted 2025-04-22 cs.RO

classification cs.RO
keywords controlbarrierfunctionssafetyfiltercollisionavoidancemultirotorPX4autopilotreactivenavigationcompositeCBFembedded
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

This paper claims that a composite control barrier function (CBF) built only from instantaneous onboard range measurements can act as a complete safety filter inside the open-source PX4 autopilot, correcting acceleration setpoints before they reach the attitude controller. The aim is to make reactive collision avoidance for small multirotors practical without any consistent map or external computation. The claim is supported by two hardware flights: one where a human operator deliberately commanded the drone into walls, and one where the drone tracked an aggressive ellipse and a panel suddenly appeared in its path. If the claim holds, a standard flight controller can provide a last-resort safety layer that runs in about 2 ms with field-of-view constraints, and even faster without them.

What carries the argument

The central object is the composite control barrier function, a smooth under-approximation of the minimum over individual obstacle constraints that uses a log-sum-exp soft minimum with a tanh saturation, parameterized by $\kappa$ and $\gamma$. Each obstacle contributes a relative-degree-two exponential CBF: $\nu_{i,0} = \|p_i - p\|^2 - \epsilon^2$ and $\nu_{i,1} = \dot{\nu}_{i,0} - p_0 \nu_{i,0}$. The field of view is imposed by two linear constraints on velocity expressed in the yaw-aligned vehicle frame, which are relaxed with slack variables to keep the quadratic program feasible. When only the composite CBF is active, the safe acceleration is computed analytically by a projection formula; when the FoV constraints are active, the problem is solved with qpOASES, and a first-order low-pass filter on the setpoint reduces chattering.

What would settle it

Command the quadrotor to fly along a velocity vector outside the horizontal field of view of the depth sensor with an obstacle hidden in that direction and check whether the filter prevents a collision; the paper's own Fig. 9 already shows the FoV CBF value crossing below zero during aggressive yaw, indicating the premise that keeps the safety argument valid is not always enforced.

Watch

Extended reading notes

Core claim

The central discovery claim is that the obstacle-avoidance problem for an acceleration-controlled multirotor can be expressed as a single composite CBF $h(x) = -(\gamma/\kappa) \ln \sum_i \exp(-\kappa s(\nu_{i,1}/\gamma))$, where each $\nu_{i,1}$ is an exponential CBF of relative degree two derived from the squared distance to a measured obstacle point. The filter then solves a quadratic program that minimally modifies the nominal acceleration setpoint to satisfy $L_f h(x) + L_g h(x) a \ge -\alpha(h(x))$. In addition, the paper claims that the sensor field of view can be incorporated as two soft acceleration constraints in a yaw-aligned frame, that the unconstrained problem admits a closed-form projection, and that the full implementation runs on the embedded PX4 controller with average solve times of ~2 ms (with FoV constraints) and 50–400 µs (without). The two experiments are presented as evidence that this filter prevents collisions during both adversarial velocity commands and fast trajectory tracking in unknown environments.

Load-bearing premise

The formal guarantee against collisions from unseen obstacles rests on the premise that the vehicle's velocity always points inside the sensor frustum; in the actual implementation this is enforced only by two soft, relaxable constraints, and the yaw-aligned approximation drops the time derivative of the rotation matrix.

Editorial extensions

If this is right

  • A multirotor autopilot can host a reactive safety filter in its standard position/velocity control loop, removing the need for a separate compute board for low-level avoidance.
  • Because the filter uses only current range measurements, collision avoidance is unaffected by map drift, localization failure, or memory of outdated obstacles.
  • Adversarial or erroneous operator commands are overridden only as much as needed: the vehicle is deflected around obstacles and brought to a stop in front of a wall, then resumes normal tracking.
  • Computation scales linearly with the number of obstacle points (about 50 µs for 25 points, 400 µs for 200), so dense point clouds remain feasible on embedded hardware.
  • The FoV constraints intentionally trade tracking performance for safety, reducing lateral accelerations and thus degrading aggressive trajectory tracking.

Reading between the lines

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

  • A convincing stress test of the limited-FoV safety claim would be to command a velocity vector outside the sensor frustum; the paper's theory covers only the configuration where velocity stays inside the frustum, so a collision in that scenario would show the gap between the guarantee and the deployed soft constraints.
  • The same composite-CBF construction should transfer to other acceleration-controlled mobile robots (e.g., ground vehicles or manipulators) wherever relative-degree-two distance constraints appear.
  • Adding a short-term memory of recently seen obstacle points could smooth the jerky behavior caused by obstacles entering and leaving the frustum, a fix the authors mention as future work.
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 / 5 minor

Summary. The paper proposes an embedded safety filter for multirotor UAVs based on composite control barrier functions (CBFs), integrated into the PX4 autopilot stack. The filter modifies the acceleration setpoint produced by a nominal position/velocity controller to enforce a minimum distance to obstacles measured online by an onboard range sensor. A quadratic program (Eq. 17) combines an obstacle-avoidance CBF with two soft field-of-view (FoV) constraints; the implementation includes point-cloud sparsification, body-frame composition, and low-pass filtering of the filter input and output. The approach is evaluated in two hardware experiments: adversarial joystick commands in a hallway and aggressive ellipse tracking with a suddenly inserted obstacle. The paper claims that the architecture ensures collision avoidance and is computationally efficient (50–400 microseconds per solve without FoV constraints, about 2 ms with FoV constraints).

Significance. If the claims are substantiated, the paper offers a practical, open-source, real-time integration of CBF-based safety filtering in a widely used autopilot, with hardware validation in dynamic, unknown environments. The computational benchmarks and the two experiments are valuable, and the extension of the authors' prior composite-CBF work [20] to acceleration control in a cascaded architecture is a useful contribution. However, the formal safety guarantee is weakened by implementation choices: the FoV constraints are soft (Eq. 17c), the yaw-rate term in the FoV derivative is neglected (Eq. 16), low-pass filters are applied to the filter input/output (Eq. 20), and the experimental CBF values go negative (Figs. 7 and 9). The unqualified 'ensures collision avoidance' claim in the abstract and conclusion exceeds what the theory and experiments establish. The contribution is nonetheless a useful engineering advance that could be publishable after a substantial revision of the claims and supporting analysis.

major comments (3)
  1. [Sections III-IV-C (Eqs. 15-17) and Fig. 9] The limited-FoV safety argument in Section III assumes the vehicle's velocity is contained in the sensor frustum, but the implementation does not enforce this condition: Eq. (17c) treats the FoV constraints as soft constraints with slack variables delta_fi, and the yaw-aligned approximation in Eqs. (15)-(16) drops the term V e_j^T dot(R_VI) v, justified only by the claim of small yaw rates. Experiment 2 (Fig. 9) shows one FoV CBF repeatedly negative, which the text attributes to skidding at high yaw rates. Consequently, the theoretical premise for reducing risk from unseen obstacles is not satisfied in the experiments, and the abstract/conclusion statement that the filter 'ensures collision avoidance' is not supported. Please either enforce the FoV constraints as hard constraints (with a feasibility analysis), or restrict the formal safety claim to the currently visible obstacles, or present the FoV behavior as an empirical heuristic rather than as part of the safety certificate.
  2. [Section V-D (Eq. 20)] The safety filter output a* is passed through a first-order low-pass filter before being sent to the attitude controller (Eq. 20). The CBF condition (2) is verified only for the unfiltered QP solution; the filtered command may violate the constraint between samples or during transients. Similarly, pre-filtering the nominal input asp changes the problem data. No argument is given that the filtered commands preserve the forward-invariance guarantee. The authors should either provide an input-to-state safety or filter-aware analysis, or explicitly limit the formal guarantee to the unfiltered command and present the filtered implementation as a heuristic for chattering reduction.
  3. [Sections VI-B and VI-C (Figs. 7 and 9)] The experimental CBF values violate the safe set in both experiments: Fig. 7 shows the composite obstacle CBF h dropping below zero for a few instances in Experiment 1, and Fig. 9 shows one FoV CBF repeatedly negative in Experiment 2. A negative composite CBF means the state is outside the obstacle-avoidance safe set defined by the current point cloud, and a negative FoV CBF means the velocity-in-frustum condition is violated. The paper attributes these dips to sudden appearance of obstacles, measurement noise, and acceleration tracking errors, but it provides no robustness analysis (e.g., input-to-state safety or a disturbance bound) that reconciles these violations with the formal guarantee. Without such analysis, the 'ensures collision avoidance' wording should be tempered to a claim of demonstrated practical collision avoidance in the tested scenarios, or supplemented with a robustness framework.
minor comments (5)
  1. [Title and Abstract] The phrase 'Multirotors Systems' in the title and abstract should be corrected to 'Multirotor Systems.'
  2. [Tables I and II] In Table I the low-pass time constant tau is listed with range [0.01, 0.1], but Table II reports the experimental value tau = 0.5, outside this range; please reconcile the discrepancy.
  3. [Section IV-B] Please define e_j explicitly as the unit vector in the vehicle frame V pointing along the inward normal of the j-th frustum plane; the current text refers to an 'inward-facing normal on the plane' but does not specify its frame or how it relates to the two horizontal constraints used in the QP.
  4. [Section V-C, Algorithm 1] In Algorithm 1, the variable nu1 is used where the text uses nu_{i,1}; also the notation Bp[i]^T is ambiguous and should clearly indicate a row vector expressed in the body frame.
  5. [Section VI-A] The computational benchmarks (50 microseconds, 400 microseconds, 2 ms) are reported for the embedded board, but the number of obstacles used in the FoV-constrained solve is not stated; adding this detail would facilitate reproducibility.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the safety argument rests on external CBF/ECBF theory and independent flight experiments; self-citation to [20] is design lineage, not a load-bearing circular reduction.

full rationale

The derivation of the safety filter is not circular. The obstacle-avoidance CBF in Section IV-A is constructed directly from range measurements, and its forward-invariance argument invokes external results: Definition 2 and Theorem 1 cite [21] and [22], and the composite soft-min CBF composition is attributed to [23]. The paper's own prior work [20] is the source of the composite-CBF formulation, as stated in 'In particular, we adapt the formulation introduced in [20] to acceleration control,' but the current safety argument does not reduce to [20]: the CBF composition and ECBF conditions are cited to [23] and [22], and the hardware validation in Section VI is an independent empirical check. Parameters are tuned rather than fitted to reproduce the safety outcome, so there is no fitted input renamed as a prediction. The limited-FoV premise (velocity contained in the sensor frustum) is implemented with soft constraints and an approximate derivative, and Fig. 9 shows the FoV CBFs going negative; this is a validity gap between the formal assumption and the implementation, not a circular derivation. Therefore no circular step is exhibited.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

The design relies on seven hand-tuned parameters and several modeling approximations from the CBF literature. No new physical entities are introduced.

free parameters (7)
  • epsilon = 0.7 m
    Avoidance radius, chosen by hand to define the minimum distance to each obstacle point.
  • kappa = 70
    Soft-minimum sharpness parameter in the composite CBF; larger values make the approximation less conservative.
  • gamma = 40
    Saturation sensitivity in the CBF composition; larger values make the filter react to farther obstacles.
  • alpha = 2
    Extended class-K gain for the composite CBF; controls filter sensitivity.
  • p0 = -2.5
    ECBF pole for the relative-degree-2 distance constraint; controls damping of the response.
  • alpha_f = 6
    Gain for the field-of-view CBF soft constraints; controls how aggressively the FoV is enforced.
  • tau = 0.5
    Time constant of the first-order low-pass filters applied before and after the safety filter; larger values give stronger acceleration smoothing.
assumptions (6)
  • domain assumption System translation dynamics are modeled as a double integrator with direct acceleration control (Eq. 5), with the inner attitude loop assumed to track commands exactly.
    The safety filter design and QP operate on this model; real actuator delays and attitude dynamics are not included.
  • standard math Composite soft-minimum CBF composition (Eqs. 4, 8) yields a valid CBF for sufficiently large kappa, per Molnar and Ames [23].
    The paper uses this theorem to justify that h in Eq. (8) is a CBF.
  • standard math ECBF theorem of Nguyen and Sreenath [22] ensures that forward invariance of C_r implies forward invariance of C_0 given initial conditions in the intersection.
    Used to convert relative-degree-2 barrier functions into relative-degree-1 conditions via nu_i,1.
  • domain assumption The yaw-aligned vehicle frame V approximates the body frame, and the term V e_j^T dot(R_VI) v in Eq. (16) is small and neglected.
    This approximation holds only for small roll/pitch and low yaw rates, which is assumed during non-agile maneuvers but not enforced.
  • domain assumption Sensor field of view and point density are sufficient to capture the relevant environment, and the current velocity is contained in the sensor frustum (Section III).
    This is the premise that makes limited-FoV mapless safety plausible; it is only weakly enforced by soft constraints.
  • ad hoc to paper Low-pass filtering (Eq. 20) and soft FoV constraints do not destroy the safety property in practice.
    No proof is given that the filtered or slack-violating input preserves the CBF invariance condition; the paper relies on empirical behavior.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Embedded Safe Reactive Navigation for Multirotors Systems using Control Barrier Functions." pith.science (2026). https://pith.science/paper/DC7WUDWW

@misc{pith2026250415850,
  author       = {Pith},
  title        = {Pith review of: Embedded Safe Reactive Navigation for Multirotors Systems using Control Barrier Functions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DC7WUDWW}},
  note         = {Machine review of arXiv:2504.15850}
}
read the original abstract

Aiming to promote the wide adoption of safety filters for autonomous aerial robots, this paper presents a safe control architecture designed for seamless integration into widely used open-source autopilots. Departing from methods that require consistent localization and mapping, we formalize the obstacle avoidance problem as a composite control barrier function constructed only from the online onboard range measurements. The proposed framework acts as a safety filter, modifying the acceleration references derived by the nominal position/velocity control loops, and is integrated into the PX4 autopilot stack. Experimental studies using a small multirotor aerial robot demonstrate the effectiveness and performance of the solution within dynamic maneuvering and unknown environments.

Figures

Figures reproduced from arXiv: 2504.15850 by the authors.

Figure 2
Figure 2. Problem definition and frame conventions used in this work. The [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Simplified planar navigation example with a constrained field of [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Cascade control scheme diagram with the safety filter introduced. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figures from the paper (5 more)
Figure 5
Figure 5. Figure 5: Safety filter response for different parameter values. [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Visualization of the environment and flown trajectory during Experiment 1. The start and end of the mission are marked with the cyan and magenta [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: Values of the composite CBF (top) and FoV CBFs (bottom) during [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 8
Figure 8. Figure 8: A short section from experiment 2. Along the path from the cyan [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]
Figure 9
Figure 9. Figure 9: CBF values during experiment 2. At around [PITH_FULL_IMAGE:figures/full_fig_p007_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 18 canonical work pages

  1. [20]

    Safe Quadrotor Navigation using Composite Control Barrier Functions

    M. Harms, M. Jacquet, and K. Alexis, “Safe quadrotor navigation using composite control barrier functions,” 2025 IEEE International Conference on Robotics and Automation (ICRA) , 2025. [Online]. Available: https://arxiv.org/abs/2502.04101

  2. [1]

    Cooperative large area surveillance with a team of aerial mobile robots for long endurance missions,

    J. J. Acevedo, B. C. Arrue, I. Maza, and A. Ollero, “Cooperative large area surveillance with a team of aerial mobile robots for long endurance missions,” Journal of Intelligent & Robotic Systems, vol. 70, pp. 329–345, 2013

  3. [2]

    An informative path planning framework for uav-based terrain monitoring,

    M. Popovi ´c, T. Vidal-Calleja, G. Hitz, J. J. Chung, I. Sa, R. Siegwart, and J. Nieto, “An informative path planning framework for uav-based terrain monitoring,” Autonomous Robots, vol. 44, no. 6, pp. 889–911, 2020

  4. [3]

    Graph-based path planning for autonomous robotic exploration in subterranean environments,

    T. Dang, F. Mascarich, S. Khattak, C. Papachristos, and K. Alexis, “Graph-based path planning for autonomous robotic exploration in subterranean environments,” in 2019 IEEE/RSJ International Confer- ence on Intelligent Robots and Systems (IROS) . IEEE, 2019, pp. 3105–3112

  5. [4]

    Online informative path planning for active information gathering of a 3d surface,

    H. Zhu, J. J. Chung, N. R. Lawrance, R. Siegwart, and J. Alonso-Mora, “Online informative path planning for active information gathering of a 3d surface,” in 2021 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2021, pp. 1488–1494

  6. [5]

    The open motion planning library,

    I. A. Sucan, M. Moll, and L. E. Kavraki, “The open motion planning library,” vol. 19, no. 4, pp. 72–82, 2012

  7. [6]

    Cerberus in the darpa subterranean challenge,

    M. Tranzatto, T. Miki, M. Dharmadhikari, L. Bernreiter, M. Kulkarni, F. Mascarich, O. Andersson, S. Khattak, M. Hutter, R. Siegwart, and K. Alexis, “Cerberus in the darpa subterranean challenge,” vol. 7, no. 66, 2022

  8. [7]

    Present and future of SLAM in extreme environments: The DARPA subt challenge,

    K. Ebadi, L. Bernreiter, H. Biggie, G. Catt, Y . Chang, A. Chatterjee, C. E. Denniston, S.-P. Deschˆenes, K. Harlow, S. Khattak, L. Nogueira, M. Palieri, P. Petr´aˇcek, M. Petrl´ık, A. Reinke, V . Kr´atk´y, S. Zhao, A.-a. Agha-mohammadi, K. Alexis, C. Heckman, K. Khosoussi, N. Kottege, B. Morrell, M. Hutter, F. Pauling, F. Pomerleau, M. Saska, S. Scherer,...

Show all 30 references
  1. [8]

    Nanomap: Fast, uncertainty-aware proximity queries with lazy search over local 3d data,

    P. Florence, J. Carter, J. Ware, and R. Tedrake, “Nanomap: Fast, uncertainty-aware proximity queries with lazy search over local 3d data,” in 2018 IEEE international conference on robotics and automa- tion (ICRA), 2018, pp. 7631–7638

  2. [9]

    Flying on point clouds: Online trajectory generation and autonomous navigation for quadrotors in cluttered environments,

    F. Gao, W. Wu, W. Gao, and S. Shen, “Flying on point clouds: Online trajectory generation and autonomous navigation for quadrotors in cluttered environments,” JOURNAL of Field Robotics , vol. 36, no. 4, pp. 710–733, 2019

  3. [10]

    Swarm of micro flying robots in the wild,

    X. Zhou, X. Wen, Z. Wang, Y . Gao, H. Li, Q. Wang, T. Yang, H. Lu, Y . Cao, C. Xu, and F. Gao, “Swarm of micro flying robots in the wild,” vol. 7, no. 66, 2022

  4. [11]

    Control barrier function based quadratic programs for safety critical systems,

    A. D. Ames, X. Xu, J. W. Grizzle, and P. Tabuada, “Control barrier function based quadratic programs for safety critical systems,” IEEE Transactions on Automatic Control , vol. 62, no. 8, pp. 3861–3876, 2017

  5. [12]

    A predictive safety filter for learning-based control of constrained nonlinear dynamical systems,

    K. P. Wabersich and M. N. Zeilinger, “A predictive safety filter for learning-based control of constrained nonlinear dynamical systems,” Automatica, vol. 129, p. 109597, 2021

  6. [13]

    Safety-critical control of a 3d quadrotor with range-limited sensing,

    G. Wu and K. Sreenath, “Safety-critical control of a 3d quadrotor with range-limited sensing,” 10 2016, p. V001T05A006

  7. [14]

    Safe control synthesis for multicopter via control barrier function backstepping,

    J. Kim and Y . Kim, “Safe control synthesis for multicopter via control barrier function backstepping,” in 2023 62nd IEEE Conference on Decision and Control (CDC) , 2023, pp. 8720–8725

  8. [15]

    Control barrier functions in dynamic uavs for kinematic obstacle avoidance: A col- lision cone approach,

    M. Tayal, R. Singh, J. Keshavan, and S. Kolathaya, “Control barrier functions in dynamic uavs for kinematic obstacle avoidance: A col- lision cone approach,” in 2024 American Control Conference (ACC) , 2024, pp. 3722–3727

  9. [16]

    Time-varying soft-maximum barrier functions for safety in unmapped and dynamic environments,

    A. Safari and J. B. Hoagg, “Time-varying soft-maximum barrier functions for safety in unmapped and dynamic environments,” 2024. [Online]. Available: https://arxiv.org/abs/2409.01458

  10. [17]

    Learning safe, generalizable perception-based hybrid control with certificates,

    C. Dawson, B. Lowenkamp, D. Goff, and C. Fan, “Learning safe, generalizable perception-based hybrid control with certificates,” IEEE Robotics and Automation Letters , vol. 7, no. 2, pp. 1904–1911, 2022

  11. [18]

    Neural control barrier functions for safe navigation,

    M. Harms, M. Kulkarni, N. Khedekar, M. Jacquet, and K. Alexis, “Neural control barrier functions for safe navigation,” 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pp. 10 415–10 422, 2024. [Online]. Available: https://api.semanticscholar.or...

  12. [19]

    Dynamic control barrier function-based model predictive control to safety-critical obstacle-avoidance of mobile robot,

    Z. Jian, Z. Yan, X. Lei, Z. Lu, B. Lan, X. Wang, and B. Liang, “Dynamic control barrier function-based model predictive control to safety-critical obstacle-avoidance of mobile robot,” in 2023 IEEE International Conference on Robotics and Automation (ICRA) , 2023, pp. 3679–3685

  13. [21]

    Control barrier functions: Theory and applications,

    A. D. Ames, S. Coogan, M. Egerstedt, G. Notomista, K. Sreenath, and P. Tabuada, “Control barrier functions: Theory and applications,” in 2019 18th European Control Conference (ECC) , 2019, pp. 3420– 3431

  14. [22]

    Exponential control barrier functions for enforcing high relative-degree safety-critical constraints,

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

  15. [23]

    Composing control barrier functions for complex safety specifications,

    T. G. Molnar and A. D. Ames, “Composing control barrier functions for complex safety specifications,” IEEE Control Systems Letters , vol. 7, pp. 3615–3620, 2023

  16. [24]

    Px4: A node-based multithreaded open source robotics framework for deeply embedded platforms,

    L. Meier, D. Honegger, and M. Pollefeys, “Px4: A node-based multithreaded open source robotics framework for deeply embedded platforms,” in 2015 IEEE international conference on robotics and automation (ICRA). IEEE, 2015, pp. 6235–6240

  17. [25]

    Multirotor aerial vehicles: Modeling, estimation, and control of quadrotor,

    R. Mahony, V . Kumar, and P. Corke, “Multirotor aerial vehicles: Modeling, estimation, and control of quadrotor,” IEEE robotics & automation magazine, vol. 19, no. 3, pp. 20–32, 2012

  18. [26]

    qpOASES: A parametric active-set algorithm for quadratic program- ming,

    H. Ferreau, C. Kirches, A. Potschka, H. Bock, and M. Diehl, “qpOASES: A parametric active-set algorithm for quadratic program- ming,” Mathematical Programming Computation , vol. 6, no. 4, pp. 327–363, 2014

  19. [27]

    An online active set strategy to overcome the limitations of explicit mpc,

    H. Ferreau, H. Bock, and M. Diehl, “An online active set strategy to overcome the limitations of explicit mpc,” International Journal of Robust and Nonlinear Control , vol. 18, no. 8, pp. 816–830, 2008

  20. [28]

    Control barrier functions and input-to-state safety with application to auto- mated vehicles,

    A. Alan, A. J. Taylor, C. R. He, A. D. Ames, and G. Orosz, “Control barrier functions and input-to-state safety with application to auto- mated vehicles,” IEEE Transactions on Control Systems Technology , vol. 31, no. 6, pp. 2744–2759, 2023

  21. [29]

    Multi-step model pre- dictive safety filters: Reducing chattering by increasing the prediction horizon,

    F. P. Bejarano, L. Brunke, and A. P. Schoellig, “Multi-step model pre- dictive safety filters: Reducing chattering by increasing the prediction horizon,” in 2023 62nd IEEE Conference on Decision and Control (CDC), 2023, pp. 4723–4730

  22. [30]

    Robust visual in- ertial odometry using a direct ekf-based approach,

    M. Bloesch, S. Omari, M. Hutter, and R. Siegwart, “Robust visual in- ertial odometry using a direct ekf-based approach,” in 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2015, pp. 298–304

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.