Reactive Robot Navigation Using Quasi-conformal Mappings and Control Barrier Functions

Gary P. T. Choi; Gennaro Notomista; Matteo Saveriano

REVIEW 3 major objections 4 minor 43 references

Reactive Robot Navigation Using Quasi-conformal Mappings and Control Barrier Functions

T0 review · 3 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read A robot navigating among non-convex polyhedral obstacles can be kept collision-free and deadlock-free by mapping the workspace to a ball world and reacting through a control-barrier-function quadratic program.

desk verdict A fast quasi-conformal mapping pipeline for reactive navigation with real experiments, but the paper's safety and deadlock guarantees are not actually derived under its time-varying mapping update. read the letter →

arxiv 2411.14908 v1 pith:YLBL422H submitted 2024-11-22 cs.RO math.DSmath.OC

Gennaro Notomista , Gary P. T. Choi , Matteo Saveriano This is my paper · ORCID

classification cs.ROmath.DSmath.OC

keywords quasi-conformalmappingcontrolbarrierfunctionsreactivenavigationballworlddeadlockavoidancecollisionconvexoptimizationrobot

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 aims to make reactive robot navigation safe even when the workspace is cluttered with non-convex, polyhedral obstacles that are only known from sensor readings. The proposed algorithm maps the real workspace to a synthetic ball world in which every obstacle and the workspace boundary are circles, using quasi-conformal mappings that can be recomputed quickly. Safety is then enforced entirely in the ball world: a control-barrier-function quadratic program continuously moves and shrinks the ball obstacles so that the robot's image stays in free space, and the corresponding motion is pulled back to the real world through the mapping's inverse. If this transfer works as claimed, the real robot is collision-free without a planning stage, and the spurious equilibria that standard CBF-QP controllers exhibit in concave environments are avoided. Simulations and experiments with manipulators, pan-tilt cameras, and mobile robots are presented as evidence.

What carries the argument

The two load-bearing objects are quasi-conformal mappings and control barrier functions. A quasi-conformal mapping is an orientation-preserving homeomorphism with bounded distortion satisfying the Beltrami equation; the full map is computed by solving two sparse linear systems, the Laplace equation for a disk harmonic map and then a generalized Laplace equation with the Beltrami coefficient, so it can be refreshed every control iteration. The partial map is a composition of analytic maps, including square-root maps, linear-fractional transforms, and reflection, that sends a polygonal hole to the unit disk and everything outside to the exterior. In the ball world, control barrier functions are defined for each ball obstacle on its center $q_j$ and radius $\rho_j$ with single-integrator dynamics $\dot q_j = u_{q_j}$, $\dot \rho_j = u_{\rho_j}$; the resulting QP enforces conditions (C1)--(C4) and is argued to be always feasible. The diffeomorphism $\varphi$ and its inverse and Jacobian transfer the ball-world obstacle motion into a desired real-world velocity, which a tracking controller executes.

What would settle it

Run the algorithm in a planar polyhedral world with one narrow concave obstacle, using a coarse triangulation for the full quasi-conformal mapping, and log at every control step both the ball-world barrier value $h(q_j,\rho_j)$ and the minimum distance from the real robot to the obstacle boundary. A single sample where $h \ge 0$ while the real distance is negative would falsify the claimed transfer of safety through the discrete mapping update.

Watch

Extended reading notes

Core claim

The central claim is that safety and deadlock-freedom can be separated from the geometry of the real world: navigate in a ball world, then map back. The paper constructs a diffeomorphism from a polyhedral world, a bounded workspace with multiple disjoint polygonal holes, to a ball world using a fast full quasi-conformal map, namely a disk harmonic map followed by a linear Beltrami solve, or, for local updates, a composition of elementary conformal maps. In the ball world each obstacle is parametrized by its center and radius, and a single quadratic program built from control barrier functions chooses velocity and radius-rate inputs so that the robot's mapped state keeps a positive distance from every obstacle, obstacles keep positive distances from each other and from the workspace boundary, and obstacles do not overlap the goal point. Because the mapping is a diffeomorphism updated each control step, forward invariance of the safe set in the ball world is claimed to imply collision avoidance in the real world, and the reactive motion of the ball obstacles removes the undesired stable equilibria that arise when a CBF-QP is run directly on non-convex obstacles.

Load-bearing premise

The load-bearing premise is that keeping the robot's image in free space inside the moving-and-shrinking ball world guarantees the real robot never hits an obstacle, even though the mapping is recomputed discretely each control step and the real robot only tracks the image velocity rather than realizing it exactly.

Editorial extensions

If this is right

A robot can navigate among non-convex polyhedral obstacles using only a convex quadratic program at each step, with no separate motion planner or potential-field construction.
Deadlock equilibria that appear when a CBF-QP acts directly on non-convex obstacles are avoided by reacting through the ball world, as demonstrated in physical mobile-robot experiments.
The full quasi-conformal mapping can be recomputed at control rates, roughly one order of magnitude faster than the harmonic-map alternative, so the approach is usable in changing scenes.
The same safety layer transfers to any robot whose dynamics are near-identity diffeomorphic to a single integrator, feedback linearizable, or differentially flat, covering differential-drive robots, manipulators, and pan-tilt cameras.
Mapping states rather than inputs through the diffeomorphism keeps the inverse mapping smooth and avoids the near-singular Jacobian problems that arise when inputs are mapped numerically.

Reading between the lines

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

We infer that the ball-world obstacle motion could be deliberately biased to absorb tracking error: shrinking obstacles faster when the real robot lags its image would convert the unproved discrete-update coupling into an explicit safety margin, a robustness extension the paper does not develop.
We infer that the partial conformal map's dependence on the tuning parameter $\lambda$ could be automated by choosing $\lambda$ from the current worst-case distortion, which would make the fast local map as drop-in as the full QC map.
We infer that the equilibrium-avoidance argument should transfer to any diffeomorphic image of a ball world, so the same reactive obstacle-motion strategy could be combined with other shape-preserving maps without re-deriving the safety proof.
We infer that the method implicitly assumes the polyhedral world changes slowly relative to the mapping update; extending it to fast-moving obstacles would require a predictor for how the polygonal holes evolve, which the paper names as future work.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

The genuinely new thing here is the mapping pipeline: a Linear Beltrami Solver-based full quasi-conformal map and a zipper-based partial conformal map that turn a polyhedral workspace into a ball world efficiently enough for real-time control. That is a real contribution, and the paper earns credit for it. The computational comparison against harmonic maps shows a clear speed advantage, and the Robotarium experiment plus the manipulator and pan-tilt camera demos give a useful sense of practical behavior. The authors also openly discuss the partial-mapping tuning problem and the near-singular Jacobian issue, which is more honest than most.

The soft spot is the one the stress-test note flags, and it is real. The paper claims guarantees on safety and absence of deadlocks, but Algorithm 1 never enforces q = phi_t(x) at all times. The CBF constraints (24)-(30) are written for a fixed mapping, with the robot's ball-world position q treated as an exogenous signal. The mapping is then updated every iteration, and the robot input is computed by tracking a mapped velocity through the inverse Jacobian of the new map. No equation couples the discrete chart change to the CBF forward-invariance argument. So the transfer from safe-in-the-ball-world to safe-in-the-polyhedral-world is asserted, not derived, and the same applies to the deadlock-free claim. This is a load-bearing gap because the abstract states the guarantee without qualification. A revision should either prove the coupling or soften the claims to match the empirical evidence.

Two smaller points. First, the problem statement is in R^n but the QC and conformal machinery is planar; the paper is really about 2D workspaces, and the general framing overreaches. Second, there is a minor notational inconsistency with the ball-world dimension m versus n. Neither is fatal, but they add to the impression that the theory is being stretched past what is actually shown. Overall, this is a useful and well-illustrated engineering contribution, and the mapping construction deserves attention. But the theoretical guarantees need fixing before the paper can stand as published. I would send it to peer review with a request for major revision.

Referee Report

3 major / 4 minor

Summary. The paper proposes a reactive robot navigation algorithm that maps a polyhedral workspace to a 'ball world' via quasi-conformal (QC) mappings, then uses control barrier functions (CBFs) to move and shrink the ball-shaped obstacles so that the robot's image in the ball world remains safe, and finally maps the resulting velocity back to the physical robot through an inverse Jacobian. The authors claim guarantees both on collision safety and on the absence of deadlocks, and support the approach with simulation studies and real-robot experiments on a differential-drive platform, a manipulator, and a pan-tilt camera. Two mapping variants are presented: a full QC mapping updated at every control iteration and a partial conformal mapping for local obstacle updates.

Significance. If the claimed guarantees were established, this would be a practically valuable extension of the prior CBF-based ball-world approach in [30] to environments whose obstacle shapes are only known through sensor data and have no analytic representation. The computational efficiency argument is a genuine strength: the full QC mapping requires only two sparse linear solves per update, and the paper's timing comparison against harmonic maps (Fig. 5) supports the real-time feasibility claim. The empirical work is extensive, covering three classes of robot dynamics and including Robotarium experiments. However, the central theoretical claim is not actually derived for the implemented algorithm: the link between ball-world CBF constraints and physical safety is not maintained in the discrete, time-varying mapping update of Algorithm 1. The paper, as written, therefore does not yet substantiate the safety and deadlock-avoidance theorems it promises.

major comments (3)

[Algorithm 1, Steps 4-11; Eq. (24)] The safety guarantee for the implemented algorithm is not derived. Condition (C1) in Section III-C states that if the robot is kept in the safe set in the ball world, then the real robot is safe in the polyhedral world. This transfer requires the identity q(t) = phi^(k(t))(x(t)) to hold at all times. Algorithm 1 never enforces or updates this identity: Step 4 computes qdot using phi^(k) and the current x, Steps 7-8 update obstacle centers and radii, Step 9 replaces phi^(k) by phi^(k+1), and Step 10 computes xdot by applying the inverse Jacobian of the new mapping to the old qdot. The coordinate q is not transformed when the chart changes, so after the update q is generally not equal to phi^(k+1)(x). The CBF constraints (24)-(30) treat q and qdot as exogenous signals and certify forward invariance with respect to the obstacle dynamics (22) for a given q trajectory; they do not account for the discrete change of the mapping phi. Consequently, forward invariance of the ball-world safe set is not shown to imply forward invariance of the physical safe set S in (8). This gap directly affects the paper's central claim of guaranteed collision-free motion.
[Abstract and Section I (iv)] The abstract claims 'guarantees both on safety and on the absence of deadlocks', but no theorem or proof is provided that characterizes equilibria or certifies convergence to the goal for the closed-loop system under the time-varying mapping. The only cited support, [30], is invoked for feasibility of QP (31), not for an absence-of-deadlock result in the setting of this paper, where q is exogenous and phi changes at each iteration. The paper should either state and prove a precise no-deadlock property, or weaken the claim to an empirical observation, as the introduction's phrase 'practically preventing the existence of deadlocks' suggests.
[Section III-B vs Section IV-A] There are two different 'partial' mappings in the paper that are not reconciled. Section III-B constructs an analytic map Psi via the geodesic algorithm and a Mobius transformation, with no parameter lambda. Section IV-A and Figs. 3-4 evaluate a 'Partial Conformal Mapping' whose behavior depends on a parameter lambda = 10000, which is defined only in the context of the navigation-function-based map (13), not in Section III-B. The paper does not explain how these two constructions are related, nor why the lambda-parameterized composition is a proper QC diffeomorphism for a polyhedral world. Since the simulation study in Section IV-A is used to compare mapping variants, this conflation of two distinct partial mappings undermines the validation of the partial-mapping claims.

minor comments (4)

[Eq. (9)] The ball-world state is written as q in R^m in (9), while throughout the rest of the paper the state space is R^n (e.g., Eq. (1) and Eq. (10)). The dimension symbol should be made consistent.
[Algorithm 1, Step 10] The notation in Step 10, 'partial phi^(k+1)-1 / partial q', is ambiguous: it should be clarified whether this is the inverse of the Jacobian of phi^(k+1) evaluated at q, or the Jacobian of the inverse mapping, and at which point it is evaluated.
[Fig. 5] The axes of Fig. 5(a) are not labeled in the text or caption beyond tick numbers; the reader must infer the vertical axis is computation time. Please add explicit axis labels and units.
[Eq. (11)] The navigation-function-based mapping (11) is written for planar star-shaped obstacles. For the polyhedral-world setting of the full algorithm, the paper should state the domain dimension and any assumptions needed to apply this construction, or clarify that (11) is used only for the planar examples.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the derivation is self-contained and the cited prior work provides independent, externally checkable support.

full rationale

The paper's derivation chain is: (1) construct a QC mapping from the polyhedral world to the ball world via the linear systems (14)-(15) and the closed-form composition (17)-(21); (2) define the ball-world safe set (10) and enforce it with CBF constraints (23)-(30); (3) formulate the QP (31) that moves and shrinks ball obstacles; and (4) map the resulting velocity back to the physical world through the Jacobian in Algorithm 1. None of these equations define their outputs in terms of the paper's claimed conclusions. The CBF constraints are standard forward-invariance conditions whose theory is external ([15], [31]). The feasibility claim imported from the authors' prior work [30] is a published, parameter-free result that does not depend on any quantity fitted in this paper, so it qualifies as independent support rather than circular reasoning. The QC mapping construction builds on published algorithms ([33], [36], [37], [39]) and is benchmarked against the Harmonic Map approach; no empirical result is renamed as a derivation. The potential gap identified by a careful reader—that Algorithm 1 updates the mapping without maintaining q = phi_t(x), so the ball-world CBF constraints may not certify physical safety—is a correctness or proof-completeness concern, not circularity, because it is not an equation or fitted parameter reducing to its own input. Overall, the paper does not exhibit self-definitional, fitted-input, or self-citation-load-bearing circularity.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The central claim rests primarily on standard CBF results, a feasibility citation to the authors' own [30], the assumption that the QC map is a diffeomorphism, and an unproved coupling between obstacle-motion CBFs and the time-varying mapping. The user-tuned parameters affect behavior but not the core mathematical structure.

free parameters (5)

lambda (partial mapping locality parameter) = manually tuned, e.g., 10000 in Figs. 3-4
Controls the locality of obstacle influence in the partial mapping; Section IV-A admits it requires manual tuning at run time and that the mapping is only a proper QC map for certain lambda values.
Kp (obstacle nominal controller gain) = not specified numerically
Gain in Eq. (33) that determines the nominal obstacle velocity and radius attraction to initial values; chosen by the user.
kappa (position vs radius weight) = not specified
Weight in QP (31) trading off obstacle translation versus radius shrinking; chosen by the user.
alpha (class K comparison function) = not specified
CBF comparison function in constraints (24)-(30); affects feasibility and conservativeness.
hmax (mesh maximum element size) = sampled from [0.05, ..., 0.6] in Section IV-B
Controls discretization of the polyhedral domain and hence the accuracy and computation trade-off in the full QC mapping.

assumptions (6)

standard math Standard CBF forward-invariance theorem
Used to turn constraints (24)-(30) into safety guarantees for ball-world obstacle motion; cited [15].
domain assumption The full QC map is an orientation-preserving homeomorphism or diffeomorphism
Section III-A claims bounded distortion as guaranteed by quasi-conformal theory, but no explicit proof is given that the LBS solution with hole centroids and radii is a diffeomorphism of the whole domain; safety transfer depends on this.
domain assumption Polyhedral obstacles are pairwise disjoint and contained in the workspace
Conditions (7) are assumed to build the functions beta_i and the mapping; overlapping obstacles are not handled.
domain assumption The robot state space is planar or the dynamics fit one of the three listed classes
The QC construction is planar; the Appendix handles near-integrator, feedback-linearizable, and flat systems. The general R^n claim in Section II is not supported by the planar construction.
domain assumption QP (31) is always feasible
Feasibility is cited from the authors' prior work [30]; it is not reproved here and is needed for the algorithm to execute at every step.
domain assumption The robot tracking controller achieves exact tracking of the desired velocity
Algorithm 1 computes u(k+1) to track xdot(k) in Step 11; safety in the ball world transfers to the real world only if tracking is accurate, but no robustness margin is analyzed.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Reactive Robot Navigation Using Quasi-conformal Mappings and Control Barrier Functions." pith.science (2026). https://pith.science/paper/YLBL422H

@misc{pith2026241114908,
  author       = {Pith},
  title        = {Pith review of: Reactive Robot Navigation Using Quasi-conformal Mappings and Control Barrier Functions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YLBL422H}},
  note         = {Machine review of arXiv:2411.14908}
}

read the original abstract

This paper presents a robot control algorithm suitable for safe reactive navigation tasks in cluttered environments. The proposed approach consists of transforming the robot workspace into the \emph{ball world}, an artificial representation where all obstacle regions are closed balls. Starting from a polyhedral representation of obstacles in the environment, obtained using exteroceptive sensor readings, a computationally efficient mapping to ball-shaped obstacles is constructed using quasi-conformal mappings and M\"obius transformations. The geometry of the ball world is amenable to provably safe navigation tasks achieved via control barrier functions employed to ensure collision-free robot motions with guarantees both on safety and on the absence of deadlocks. The performance of the proposed navigation algorithm is showcased and analyzed via extensive simulations and experiments performed using different types of robotic systems, including manipulators and mobile robots.

Figures

Figures reproduced from arXiv: 2411.14908 by the authors.

**Figure 2.** Full QC mapping (using the inverse mapping to transform the states). [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

**Figure 3.** Partial conformal mapping (λ = 10000). In each panel, the left box is the polyhedral (real) world, and the right circle is the ball world. Notice the motion of the obstacles in the ball world to prevent the state of the robot mapped into the ball world from colliding with them at each point in time [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Figures from the paper (7 more)

**Figure 4.** Figure 4: Partial conformal mapping for different values of [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

**Figure 5.** Figure 5: (a) Computational complexity of Harmonic Map [26] and the proposed [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

**Figure 6.** Figure 6: Full QC input mapping (using the mapping Jacobian to transform [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

**Figure 7.** Figure 7: Results obtained with the Full QC mapping in a cluttered office [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

**Figure 9.** Figure 9: Avoiding low manipulability areas with the full QC mapping. In each [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]

**Figure 10.** Figure 10: Avoiding forbidden areas with a pan-tilt camera performing a [PITH_FULL_IMAGE:figures/full_fig_p009_10.png]

**Figure 11.** Figure 11: Snapshots of the robotarium experiments. [PITH_FULL_IMAGE:figures/full_fig_p010_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 36 canonical work pages

[30]

Safety of dynamical systems with multiple non-convex unsafe sets using control barrier functions,

G. Notomista and M. Saveriano, “Safety of dynamical systems with multiple non-convex unsafe sets using control barrier functions,” IEEE Control Syst. Lett. , vol. 6, pp. 1136–1141, 2021

work page 2021
[1]

Safety barrier certificates for collisions-free multirobot systems,

L. Wang, A. D. Ames, and M. Egerstedt, “Safety barrier certificates for collisions-free multirobot systems,” IEEE Trans. Robot. , vol. 33, no. 3, pp. 661–674, 2017

work page 2017
[2]

Persistification of robotic tasks,

G. Notomista and M. Egerstedt, “Persistification of robotic tasks,” IEEE Trans. Control Syst. Technol., vol. 29, no. 2, pp. 756–767, 2020

work page 2020
[3]

Provably safe and robust learning-based model predictive control,

A. Aswani, H. Gonzalez, S. S. Sastry, and C. Tomlin, “Provably safe and robust learning-based model predictive control,”Automatica, vol. 49, no. 5, pp. 1216–1226, 2013

2013
[4]

Learning barrier functions for constrained motion planning with dynamical systems,

M. Saveriano and D. Lee, “Learning barrier functions for constrained motion planning with dynamical systems,” in IEEE/RSJ International Conference on Intelligent Robots and Systems , 2019, pp. 112–119

work page 2019
[5]

Safe motion planning in dynamic environ- ments,

S. Petti and T. Fraichard, “Safe motion planning in dynamic environ- ments,” in IEEE/RSJ International Conference on Intelligent Robots and Systems, 2005, pp. 2210–2215

work page 2005
[6]

The navigation transformation,

S. G. Loizou, “The navigation transformation,” IEEE Trans. Robot. , vol. 33, no. 6, pp. 1516–1523, 2017

work page 2017
[7]

Reactive collision avoidance for asvs based on control barrier functions,

E. H. Thyri, E. A. Basso, M. Breivik, K. Y . Pettersen, R. Skjetne, and A. M. Lekkas, “Reactive collision avoidance for asvs based on control barrier functions,” in IEEE Conference on Control Technology and Applications, 2020, pp. 380–387

work page 2020

Show all 43 references

[8]

Safe autonomy under perception uncertainty using chance-constrained temporal logic,

S. Jha, V . Raman, D. Sadigh, and S. A. Seshia, “Safe autonomy under perception uncertainty using chance-constrained temporal logic,” J. Autom. Reason. , vol. 60, no. 1, pp. 43–62, 2018

2018
[9]

Long-duration robot autonomy: From control algorithms to robot design,

G. Notomista, “Long-duration robot autonomy: From control algorithms to robot design,” Ph.D. dissertation, Georgia Institute of Technology, 2020

2020
[10]

End-to-end safe reinforcement learning through barrier functions for safety-critical continuous control tasks,

R. Cheng, G. Orosz, R. M. Murray, and J. W. Burdick, “End-to-end safe reinforcement learning through barrier functions for safety-critical continuous control tasks,” in AAAI Conference on Artificial Intelligence, vol. 33, no. 1, 2019, pp. 3387–3395

2019
[11]

Safe reinforcement learning using robust control barrier functions,

Y . Emam, G. Notomista, P. Glotfelter, Z. Kira, and M. Egerstedt, “Safe reinforcement learning using robust control barrier functions,” IEEE Robot. Autom. Lett. , no. 99, pp. 1–8, 2022

2022
[12]

Barrier- certified adaptive reinforcement learning with applications to brushbot navigation,

M. Ohnishi, L. Wang, G. Notomista, and M. Egerstedt, “Barrier- certified adaptive reinforcement learning with applications to brushbot navigation,” IEEE Transactions on robotics , vol. 35, no. 5, pp. 1186– 1205, 2019

2019
[13]

A constrained-optimization approach to the execution of prioritized stacks of learned multi-robot tasks,

G. Notomista, “A constrained-optimization approach to the execution of prioritized stacks of learned multi-robot tasks,” arXiv preprint arXiv:2301.05346, 2023

2023 arXiv
[14]

The safety filter: A unified view of safety-critical control in autonomous systems,

K.-C. Hsu, H. Hu, and J. F. Fisac, “The safety filter: A unified view of safety-critical control in autonomous systems,” Annual Review of Control, Robotics, and Autonomous Systems , vol. 7, 2023

2023
[15]

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,” in2019 18th European Control Conference (ECC). IEEE, 2019, pp. 3420–3431

2019
[16]

Predictive control barrier functions for online safety critical control,

J. Breeden and D. Panagou, “Predictive control barrier functions for online safety critical control,” in2022 IEEE 61st Conference on Decision and Control (CDC) . IEEE, 2022, pp. 924–931

2022
[17]

Control barrier function- based quadratic programs introduce undesirable asymptotically stable equilibria,

M. F. Reis, A. P. Aguiar, and P. Tabuada, “Control barrier function- based quadratic programs introduce undesirable asymptotically stable equilibria,” IEEE Control Systems Letters , vol. 5, no. 2, pp. 731–736, 2021

2021
[18]

Exact robot navigation in geomet- rically complicated but topologically simple spaces,

E. Rimon and D. E. Koditschek, “Exact robot navigation in geomet- rically complicated but topologically simple spaces,” in Proceedings., IEEE International Conference on Robotics and Automation . IEEE, 1990, pp. 1937–1942

1990
[19]

The construction of analytic diffeomorphisms for exact robot navigation on star worlds,

——, “The construction of analytic diffeomorphisms for exact robot navigation on star worlds,” Trans. Am. Math. Soc. , vol. 327, no. 1, pp. 71–116, 1991

1991
[20]

Exact robot navigation using artificial potential functions,

——, “Exact robot navigation using artificial potential functions,” IEEE Trans. Robot. Autom., vol. 8, no. 5, pp. 501–518, 1992

1992
[21]

Discrete abstractions for robot motion planning and control in polygonal environments,

C. Belta, V . Isler, and G. J. Pappas, “Discrete abstractions for robot motion planning and control in polygonal environments,” IEEE Trans. Robot., vol. 21, no. 5, pp. 864–874, 2005

2005
[22]

Harmonic-based optimal motion planning in constrained workspaces using reinforcement learning,

P. Rousseas, C. Bechlioulis, and K. J. Kyriakopoulos, “Harmonic-based optimal motion planning in constrained workspaces using reinforcement learning,” IEEE Robot. Autom. Lett., vol. 6, no. 2, pp. 2005–2011, 2021

2005
[23]

A survey of motion planning algorithms from the perspective of autonomous uav guidance,

C. Goerzen, Z. Kong, and B. Mettler, “A survey of motion planning algorithms from the perspective of autonomous uav guidance,” J. Intell. Robot. Syst., vol. 57, pp. 65–100, 2010

2010
[24]

Greedy routing with guaranteed delivery using Ricci flows,

R. Sarkar, X. Yin, J. Gao, F. Luo, and X. D. Gu, “Greedy routing with guaranteed delivery using Ricci flows,” in2009 International Conference on Information Processing in Sensor Networks . IEEE, 2009, pp. 121– 132

2009
[25]

Robot navigation on star worlds using a single-step navigation transformation,

N. Constantinou and S. G. Loizou, “Robot navigation on star worlds using a single-step navigation transformation,” in 2020 59th IEEE Conference on Decision and Control (CDC) . IEEE, 2020, pp. 1537– 1542

2020
[26]

Robot navigation in complex workspaces using harmonic maps,

P. Vlantis, C. Vrohidis, C. P. Bechlioulis, and K. J. Kyriakopoulos, “Robot navigation in complex workspaces using harmonic maps,” in 2018 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2018, pp. 1726–1731

2018
[27]

A conformal mapping-based framework for robot-to-robot and sim-to-real transfer learning,

S. Gao and N. Bezzo, “A conformal mapping-based framework for robot-to-robot and sim-to-real transfer learning,” in 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2021, pp. 1289–1295

2021
[28]

Coverage control for wire-traversing robots,

G. Notomista and M. Egerstedt, “Coverage control for wire-traversing robots,” in 2018 IEEE International Conference on Robotics and Au- tomation (ICRA). IEEE, 2018, pp. 5042–5047

2018
[29]

Robot navigation in complex workspaces using conformal navigation transformations,

L. Fan, J. Liu, W. Zhang, and P. Xu, “Robot navigation in complex workspaces using conformal navigation transformations,” IEEE Robot. Autom. Lett., vol. 8, no. 1, pp. 192–199, 2022

2022
[31]

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 Trans. Autom. Control., vol. 62, no. 8, pp. 3861–3876, 2016

2016
[32]

Lehto, Quasiconformal mappings in the plane

O. Lehto, Quasiconformal mappings in the plane . Springer-Verlag Berlin Heidelberg, 1973, vol. 126

1973
[33]

Efficient conformal parameterization of multiply- connected surfaces using quasi-conformal theory,

G. P. T. Choi, “Efficient conformal parameterization of multiply- connected surfaces using quasi-conformal theory,” J. Sci. Comput. , vol. 87, no. 3, pp. 1–19, 2021

2021
[34]

Computing discrete minimal surfaces and their conjugates,

U. Pinkall and K. Polthier, “Computing discrete minimal surfaces and their conjugates,” Exp. Math., vol. 2, no. 1, pp. 15–36, 1993

1993
[35]

Fast disk conformal parameterization of simply-connected open surfaces,

P. T. Choi and L. M. Lui, “Fast disk conformal parameterization of simply-connected open surfaces,” J. Sci. Comput. , vol. 65, no. 3, pp. 1065–1090, 2015

2015
[36]

Texture map and video compression using Beltrami representation,

L. M. Lui, K. C. Lam, T. W. Wong, and X. Gu, “Texture map and video compression using Beltrami representation,” SIAM J. Imaging Sci., vol. 6, no. 4, pp. 1880–1902, 2013

1902
[37]

FLASH: Fast landmark aligned spherical harmonic parameterization for genus-0 closed brain surfaces,

P. T. Choi, K. C. Lam, and L. M. Lui, “FLASH: Fast landmark aligned spherical harmonic parameterization for genus-0 closed brain surfaces,” SIAM J. Imaging Sci. , vol. 8, no. 1, pp. 67–94, 2015

2015
[38]

Parallelizable global conformal parameterization of simply-connected surfaces via partial welding,

G. P. T. Choi, Y . Leung-Liu, X. Gu, and L. M. Lui, “Parallelizable global conformal parameterization of simply-connected surfaces via partial welding,” SIAM J. Imaging Sci. , vol. 13, no. 3, pp. 1049–1083, 2020

2020
[39]

Convergence of a variant of the zipper algorithm for conformal mapping,

D. E. Marshall and S. Rohde, “Convergence of a variant of the zipper algorithm for conformal mapping,” SIAM J. Numer. Anal., vol. 45, no. 6, pp. 2577–2609, 2007

2007
[40]

The robotarium: Globally impactful opportunities, challenges, and lessons learned in remote-access, distributed control of multirobot systems,

S. Wilson, P. Glotfelter, L. Wang, S. Mayya, G. Notomista, M. Mote, and M. Egerstedt, “The robotarium: Globally impactful opportunities, challenges, and lessons learned in remote-access, distributed control of multirobot systems,” IEEE Control Syst. Mag., vol. 40, no. 1, pp. 2...

2020
[41]

Near-identity diffeomorphisms and exponential/spl epsi/-tracking and/spl epsi/-stabilization of first-order nonholonomic se (2) vehicles,

R. Olfati-Saber, “Near-identity diffeomorphisms and exponential/spl epsi/-tracking and/spl epsi/-stabilization of first-order nonholonomic se (2) vehicles,” in Proceedings of the 2002 American Control Conference (IEEE cat. no. ch37301) , vol. 6. IEEE, 2002, pp. 4690–4695

2002
[42]

H. K. Khalil, Nonlinear control. Pearson New York, 2015

2015
[43]

Flatness and defect of non-linear systems: introductory theory and examples,

M. Fliess, J. L ´evine, P. Martin, and P. Rouchon, “Flatness and defect of non-linear systems: introductory theory and examples,” Int. J. Control , vol. 61, no. 6, pp. 1327–1361, 1995

1995

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [30]

Safety of dynamical systems with multiple non-convex unsafe sets using control barrier functions,

G. Notomista and M. Saveriano, “Safety of dynamical systems with multiple non-convex unsafe sets using control barrier functions,” IEEE Control Syst. Lett. , vol. 6, pp. 1136–1141, 2021

work page 2021

[2] [1]

Safety barrier certificates for collisions-free multirobot systems,

L. Wang, A. D. Ames, and M. Egerstedt, “Safety barrier certificates for collisions-free multirobot systems,” IEEE Trans. Robot. , vol. 33, no. 3, pp. 661–674, 2017

work page 2017

[3] [2]

Persistification of robotic tasks,

G. Notomista and M. Egerstedt, “Persistification of robotic tasks,” IEEE Trans. Control Syst. Technol., vol. 29, no. 2, pp. 756–767, 2020

work page 2020

[4] [3]

Provably safe and robust learning-based model predictive control,

A. Aswani, H. Gonzalez, S. S. Sastry, and C. Tomlin, “Provably safe and robust learning-based model predictive control,”Automatica, vol. 49, no. 5, pp. 1216–1226, 2013

2013

[5] [4]

Learning barrier functions for constrained motion planning with dynamical systems,

M. Saveriano and D. Lee, “Learning barrier functions for constrained motion planning with dynamical systems,” in IEEE/RSJ International Conference on Intelligent Robots and Systems , 2019, pp. 112–119

work page 2019

[6] [5]

Safe motion planning in dynamic environ- ments,

S. Petti and T. Fraichard, “Safe motion planning in dynamic environ- ments,” in IEEE/RSJ International Conference on Intelligent Robots and Systems, 2005, pp. 2210–2215

work page 2005

[7] [6]

The navigation transformation,

S. G. Loizou, “The navigation transformation,” IEEE Trans. Robot. , vol. 33, no. 6, pp. 1516–1523, 2017

work page 2017

[8] [7]

Reactive collision avoidance for asvs based on control barrier functions,

E. H. Thyri, E. A. Basso, M. Breivik, K. Y . Pettersen, R. Skjetne, and A. M. Lekkas, “Reactive collision avoidance for asvs based on control barrier functions,” in IEEE Conference on Control Technology and Applications, 2020, pp. 380–387

work page 2020