pith. sign in

arxiv: 2512.06610 · v2 · submitted 2025-12-07 · 💻 cs.RO

Towards Robust Optimization-Based Autonomous Dynamic Soaring with a Fixed-Wing UAV

Marvin Harms , Jaeyoung Lim , David Rohr , Friedrich Rockenbauer , Nicholas Lawrance , Roland Siegwart This is my paper

Pith reviewed 2026-05-17 01:38 UTC · model grok-4.3

classification 💻 cs.RO
keywords dynamic soaringfixed-wing UAVrobust path planningwind shearautonomous flightpath following controlleroptimization-based guidance
0
0 comments X

The pith

Robust reference paths and a path-following controller let a fixed-wing UAV perform autonomous dynamic soaring in wind shear.

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

The paper presents a framework that lets a fixed-wing UAV harvest energy from wind shear layers for sustained flight without propulsion. It builds an explicit wind field model, then uses optimization to create point-wise robust reference paths that stay effective even when the wind estimate contains errors. A separate robust controller is added to track those paths accurately. Real flight tests confirm good path following in actual wind, while simulations show the UAV can keep performing dynamic soaring maneuvers when winds, estimation errors, and disturbances vary.

Core claim

Constructing point-wise robust reference paths from an explicit wind field model and combining them with a robust path following controller allows a fixed-wing UAV to achieve autonomous dynamic soaring that remains stable under wind estimation errors and external disturbances.

What carries the argument

Point-wise robust reference paths: trajectories planned so that each point on the path tolerates a range of wind field estimation errors while still extracting net energy from the shear layer.

If this is right

  • The UAV can sustain flight indefinitely inside a wind shear layer without using onboard energy stores.
  • Path tracking errors stay bounded even when the wind estimate used for planning differs from the true wind.
  • Dynamic soaring becomes a practical mode for long-endurance autonomous UAV missions in natural atmospheric shear.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same robust-path idea could be applied to other energy-harvesting flight regimes such as ridge soaring or thermal centering.
  • If wind estimation is fused with real-time sensor feedback, the framework might reduce dependence on a single precomputed wind model.
  • Scaling the approach to smaller or larger UAVs would test whether the robustness margins remain sufficient at different mass and inertia scales.

Load-bearing premise

The wind field can be represented explicitly and estimated accurately enough that the robust paths still work when real estimation errors and disturbances appear.

What would settle it

A flight test in which the UAV, using the estimated wind field and the proposed robust paths plus controller, fails to complete repeated dynamic soaring cycles without losing altitude or needing thrust.

Figures

Figures reproduced from arXiv: 2512.06610 by David Rohr, Friedrich Rockenbauer, Jaeyoung Lim, Marvin Harms, Nicholas Lawrance, Roland Siegwart.

Figure 1
Figure 1. Figure 1: Impression of one of the flight tests to validate tracking performance [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: A simplified dynamic soaring cycle can be described by four phases: [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Frame conventions used during this work. [PITH_FULL_IMAGE:figures/full_fig_p002_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Block diagram of the wind estimation. Measurements from the pitot, [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Robust energy gain paths, colored by the initial velocity at the lowest [PITH_FULL_IMAGE:figures/full_fig_p004_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Proposed INDI control architecture for path following control. [PITH_FULL_IMAGE:figures/full_fig_p004_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Proposed guidance and control architecture for dynamic soaring. The approach uses three separate modules for wind field estimation, trajectory [PITH_FULL_IMAGE:figures/full_fig_p006_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Visualization of global wind estimator. a) Estimated wind shear profile. Convergence of wind shear parameters b) wind shear strength c) wind shear [PITH_FULL_IMAGE:figures/full_fig_p007_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Monte-Carlo simulation of dynamic soaring with randomized wind [PITH_FULL_IMAGE:figures/full_fig_p007_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Path following results with (a) arbitrary inclined ellipse paths and [PITH_FULL_IMAGE:figures/full_fig_p007_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Airspeed evolution over height of the unpowered spiral test. The [PITH_FULL_IMAGE:figures/full_fig_p008_11.png] view at source ↗
read the original abstract

Dynamic soaring is a flying technique to exploit the energy available in wind shear layers, enabling potentially unlimited flight without the need for internal energy sources. We propose a framework for autonomous dynamic soaring with a fixed-wing unmanned aerial vehicle (UAV). The framework makes use of an explicit representation of the wind field and a classical approach for guidance and control of the UAV. Robustness to wind field estimation error is achieved by constructing point-wise robust reference paths for dynamic soaring and the development of a robust path following controller for the fixed-wing UAV. Wind estimation and path tracking performance are validated with real flight tests to demonstrate robust path-following in real wind conditions. In simulation, we demonstrate robust dynamic soaring flight subject to varied wind conditions, estimation errors and disturbances. Together, our results strongly indicate the ability of the proposed framework to achieve autonomous dynamic soaring flight in wind shear.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 1 minor

Summary. The manuscript proposes a framework for autonomous dynamic soaring with a fixed-wing UAV using an explicit wind field model, point-wise robust reference paths to handle estimation errors, and a robust path-following controller. Real flight tests validate wind estimation and path tracking under actual conditions, while simulations demonstrate full dynamic soaring cycles under varied winds, estimation errors, and disturbances. The authors conclude that the combined results indicate the framework enables autonomous dynamic soaring in wind shear.

Significance. If the integrated closed-loop performance holds under real wind shear, the work would represent a meaningful step toward practical energy-harvesting UAVs by combining optimization-based planning with robustness to estimation errors. Credit is due for the real-flight validation of wind estimation and path tracking components plus the simulation coverage of multiple wind conditions and injected errors. However, the absence of full closed-loop dynamic soaring demonstrations in actual wind shear limits the immediate significance to component-level advances rather than end-to-end system validation.

major comments (1)
  1. [Abstract] Abstract: the central claim that the results 'strongly indicate the ability of the proposed framework to achieve autonomous dynamic soaring flight in wind shear' is only partially supported, as real-flight experiments are limited to wind estimation and path tracking while the integrated energy-extracting cycles under the robust references are shown exclusively in simulation with injected errors. This is load-bearing for the claim because the weakest assumption (accurate-enough explicit wind-field estimation for the constructed paths to remain effective) is not tested in the full closed-loop real-world setting.
minor comments (1)
  1. [Abstract] The distinction between simulation results (full soaring cycles) and real-flight results (subcomponent validation) could be stated more explicitly in the abstract and conclusion to prevent over-interpretation of the experimental scope.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their constructive and detailed review of our manuscript. We address the major comment below and agree that the abstract claim requires revision to more precisely reflect the scope of our validation.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that the results 'strongly indicate the ability of the proposed framework to achieve autonomous dynamic soaring flight in wind shear' is only partially supported, as real-flight experiments are limited to wind estimation and path tracking while the integrated energy-extracting cycles under the robust references are shown exclusively in simulation with injected errors. This is load-bearing for the claim because the weakest assumption (accurate-enough explicit wind-field estimation for the constructed paths to remain effective) is not tested in the full closed-loop real-world setting.

    Authors: We agree that the real-flight experiments validate wind estimation and path tracking under actual conditions, while the integrated energy-extracting dynamic soaring cycles with robust references are demonstrated in simulation (including injected estimation errors drawn from real-flight observations, varied winds, and disturbances). The simulations therefore test the closed-loop performance under the key uncertainty the referee identifies. To address the concern directly, we will revise the abstract's concluding sentence to read: 'Together, our results demonstrate robust wind estimation and path tracking in real flight conditions as well as robust dynamic soaring cycles in simulation, indicating the framework's potential to achieve autonomous dynamic soaring flight in wind shear.' This change ensures the claim is fully supported by the presented evidence. revision: yes

Circularity Check

0 steps flagged

No significant circularity; standard optimization and control applied to explicit wind model.

full rationale

The paper constructs robust reference paths via optimization over an explicit wind-field representation and applies a classical robust path-following controller. Real-flight results validate only the subcomponents of wind estimation accuracy and path-tracking performance; integrated dynamic-soaring cycles are shown exclusively in simulation. No equation or claim reduces by construction to a fitted parameter, self-definition, or load-bearing self-citation chain. The derivation therefore remains self-contained against external benchmarks and receives the default non-circularity finding.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The framework rests on an explicit parametric wind-field model whose estimation accuracy is treated as sufficient for robustness guarantees; classical UAV dynamics and guidance laws are assumed known from prior literature.

axioms (2)
  • domain assumption The wind field admits an explicit representation that can be estimated from onboard sensors with bounded error.
    Invoked to enable construction of point-wise robust reference paths.
  • standard math Fixed-wing UAV dynamics and actuator limits are known and can be used in the path-following controller design.
    Classical approach for guidance and control referenced in abstract.

pith-pipeline@v0.9.0 · 5458 in / 1255 out tokens · 50725 ms · 2026-05-17T01:38:10.808534+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Learning step-level dynamic soaring in shear flow

    physics.flu-dyn 2026-04 unverdicted novelty 6.0

    Dynamic soaring emerges from local state-feedback control learned by reinforcement learning, enabling robust omnidirectional navigation in shear flows without cycle-level planning.

Reference graph

Works this paper leans on

26 extracted references · 26 canonical work pages · cited by 1 Pith paper

  1. [1]

    Maximum travel speed performance of albatrosses and uavs using dynamic soaring,

    G. P. Sachs, “Maximum travel speed performance of albatrosses and uavs using dynamic soaring,” inAIAA Scitech 2019 Forum, 2019, p. 0568

    work page 2019

  2. [2]

    The flight of petrels and alba- trosses (procellariiformes), observed in South Georgia and its vicinity,

    C. J. Pennycuick and M. J. Lighthill, “The flight of petrels and alba- trosses (procellariiformes), observed in South Georgia and its vicinity,” Philosophical Transactions of the Royal Society of London. B, Biological Sciences, vol. 300, no. 1098, pp. 75–106, 1982

    work page 1982

  3. [3]

    Fast and fuel efficient? optimal use of wind by flying albatrosses,

    H. Weimerskirch, T. Guionnet, J. Martin, S. A. Shaffer, and D. P. Costa, “Fast and fuel efficient? optimal use of wind by flying albatrosses,” Proceedings of the Royal Society of London. Series B: Biological Sciences, vol. 267, no. 1455, pp. 1869–1874, 2000

    work page 2000

  4. [4]

    Guidance and control for an autonomous soaring uav,

    M. J. Allen, “Guidance and control for an autonomous soaring uav,” 2008

    work page 2008

  5. [5]

    Ardusoar: an open-source thermalling controller for resource-constrained autopilots,

    S. Tabor, I. Guilliard, and A. Kolobov, “Ardusoar: an open-source thermalling controller for resource-constrained autopilots,” in2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2018, pp. 6255–6262

    work page 2018

  6. [6]

    The autosoar autonomous soaring aircraft, part 1: Autonomy algorithms,

    N. T. Depenbusch, J. J. Bird, and J. W. Langelaan, “The autosoar autonomous soaring aircraft, part 1: Autonomy algorithms,”Journal of Field Robotics, vol. 35, no. 6, pp. 868–889, 2018

    work page 2018

  7. [7]

    Optimal patterns of glider dynamic soaring,

    Y . J. Zhao, “Optimal patterns of glider dynamic soaring,”Optimal control applications and methods, vol. 25, no. 2, pp. 67–89, 2004

    work page 2004

  8. [8]

    Closing the loop in dynamic soaring,

    J. J. Bird, J. W. Langelaan, C. Montella, J. Spletzer, and J. L. Grenestedt, “Closing the loop in dynamic soaring,” inAIAA Guidance, Navigation, and Control Conference, 2014, p. 0263

    work page 2014

  9. [9]

    Flight testing of dynamic soaring part-2: open-field inclined circle trajectory,

    M. Bronz, N. Gavrilovic, A. Drouin, G. Hattenberger, and J.-M. Moschetta, “Flight testing of dynamic soaring part-2: open-field inclined circle trajectory,” inAIAA Aviation 2021 Forum, 2021, p. 2803

    work page 2021

  10. [10]

    The soaring of birds,

    Rayleigh, “The soaring of birds,”Nature, vol. 27, pp. 534–535, 1883. [Online]. Available: https://doi.org/10.1038/027534a0

    work page doi:10.1038/027534a0

  11. [11]

    Robust trajectory optimization for dynamic soaring,

    T. Flanzer, G. Bower, and I. Kroo, “Robust trajectory optimization for dynamic soaring,” inAIAA guidance, navigation, and control conference, 2012, p. 4603

    work page 2012

  12. [12]

    Optimal dynamic soaring consists of successive shallow arcs,

    G. D. Bousquet, M. S. Triantafyllou, and J.-J. E. Slotine, “Optimal dynamic soaring consists of successive shallow arcs,”Journal of The Royal Society Interface, vol. 14, no. 135, p. 20170496, 2017. [Online]. Available: https://royalsocietypublishing.org/doi/abs/10.1098/ rsif.2017.0496

    work page arXiv 2017

  13. [13]

    Minimum shear wind strength required for dynamic soaring of albatrosses,

    G. Sachs, “Minimum shear wind strength required for dynamic soaring of albatrosses,”Ibis, vol. 147, no. 1, pp. 1–10, 2005

    work page 2005

  14. [14]

    Deep reinforcement learning approach for integrated updraft mapping and exploitation,

    S. Notter, C. Gall, G. M ¨uller, A. Ahmad, and W. Fichter, “Deep reinforcement learning approach for integrated updraft mapping and exploitation,”Journal of Guidance, Control, and Dynamics, vol. 46, no. 10, pp. 1997–2004, 2023

    work page 1997

  15. [15]

    Robotic technologies for solar-powered UA Vs: Fully autonomous updraft-aware aerial sensing for multiday search-and-rescue missions,

    P. Oettershagen, T. Stastny, T. Hinzmann, K. Rudin, T. Mantel, A. Melzer, B. Wawrzacz, G. Hitz, and R. Siegwart, “Robotic technologies for solar-powered UA Vs: Fully autonomous updraft-aware aerial sensing for multiday search-and-rescue missions,”Journal of Field Robotics, vol. 35, no. 4, pp. 612–640, 2018. [Online]. Available: https://onlinelibrary.wiley...

    work page doi:10.1002/rob.21765 2018

  16. [16]

    Path planning for autonomous soaring flight in dynamic wind fields,

    N. R. Lawrance and S. Sukkarieh, “Path planning for autonomous soaring flight in dynamic wind fields,” in2011 IEEE international conference on robotics and automation. IEEE, 2011, pp. 2499–2505

    work page 2011

  17. [17]

    Reinforcement learning for autonomous dynamic soaring in shear winds,

    C. Montella and J. R. Spletzer, “Reinforcement learning for autonomous dynamic soaring in shear winds,” in2014 IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE, 2014, pp. 3423– 3428

    work page 2014

  18. [18]

    Learning to soar: Resource-constrained exploration in reinforcement learning,

    J. J. Chung, N. R. Lawrance, and S. Sukkarieh, “Learning to soar: Resource-constrained exploration in reinforcement learning,”The inter- national journal of robotics research, vol. 34, no. 2, pp. 158–172, 2015

    work page 2015

  19. [19]

    Deep neural network- based feedback control for dynamic soaring of unpowered aircraft,

    S.-h. Kim, J. Lee, S. Jung, H. Lee, and Y . Kim, “Deep neural network- based feedback control for dynamic soaring of unpowered aircraft,” IFAC-PapersOnLine, vol. 52, no. 12, pp. 117–121, 2019

    work page 2019

  20. [20]

    Autonomous control for orographic soaring of fixed-wing uavs,

    T. Suys, S. Hwang, G. C. De Croon, and B. D. Remes, “Autonomous control for orographic soaring of fixed-wing uavs,” in2023 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2023, pp. 5338–5344

    work page 2023

  21. [21]

    Si- multaneous wind field measurements with doppler lidar, quadrotor and fixed-wing uav,

    M. Bronz, N. Gavrilovic, G. Hattenberger, and J.-M. Moschetta, “Si- multaneous wind field measurements with doppler lidar, quadrotor and fixed-wing uav,” inAIAA SCITECH 2023 Forum, 2023, p. 0625

    work page 2023

  22. [22]

    Online parameter estimation within trajectory optimization for dynamic soaring,

    C. A. McKenna and A. Gorodetsky, “Online parameter estimation within trajectory optimization for dynamic soaring,” inAIAA SCITECH 2023 Forum, 2023, p. 1482

    work page 2023

  23. [23]

    Global Trajectory-tracking Control for a Tailsitter Flying Wing in Agile Uncoordinated Flight,

    E. A. Tal and S. Karaman, “Global Trajectory-tracking Control for a Tailsitter Flying Wing in Agile Uncoordinated Flight,” inAIAA Aviation 2021 Forum, 2021. [Online]. Available: https: //arc.aiaa.org/doi/abs/10.2514/6.2021-3214

    work page doi:10.2514/6.2021-3214 2021

  24. [24]

    D. G. Hull.,Atmosphere, Aerodynamics, and Propulsion. Heidelberg: Springer Berlin, 2015, p. 50. [Online]. Available: https://doi.org/10. 1007/978-3-540-46573-7

    work page 2015

  25. [25]

    Lecture Notes in Model Predictive Control,

    M. Zeilinger, “Lecture Notes in Model Predictive Control,” D-MA VT, Eidgen¨ossische Technische Hochschule Z ¨urich, April 2021

    work page 2021

  26. [26]

    Engineless unmanned aerial vehicle propulsion by dynamic soaring,

    M. Deittert, A. Richards, C. A. Toomer, and A. Pipe, “Engineless unmanned aerial vehicle propulsion by dynamic soaring,”Journal of guidance, control, and dynamics, vol. 32, no. 5, pp. 1446–1457, 2009

    work page 2009