A Comparative Study of INDI and NDI with Nonlinear Disturbance Observer for Aerial Robotics
Pith reviewed 2026-05-08 09:20 UTC · model grok-4.3
The pith
INDI exhibits stronger robustness than NDI with nonlinear disturbance observer under model mismatches and stresses in aerial robot simulations.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
Through a systematic simulation campaign across representative operating scenarios, the study shows that INDI demonstrates stronger robustness in several model-mismatch and combined-stress cases, while NDI+NDO primarily matches nominal performance but exhibits greater sensitivity under several non-ideal conditions.
What carries the argument
The side-by-side simulation evaluation of Incremental Nonlinear Dynamic Inversion (INDI) versus Nonlinear Dynamic Inversion with nonlinear disturbance observer (NDI+NDO) on tracking, robustness, and effort metrics under controlled parametric and disturbance variations.
If this is right
- INDI becomes the preferred choice for aerial robot tasks involving significant model uncertainty or simultaneous disturbances.
- NDI+NDO remains viable for systems with accurate models and low variation but requires extra safeguards under non-ideal conditions.
- Control engineers should weigh incremental versus observer-based inversion when designing for reliability in uncertain environments.
- Simulation campaigns of this type can narrow design options before hardware implementation.
- The relative sensitivity of NDI+NDO highlights the value of testing inversion methods under combined stresses rather than isolated variations.
Where Pith is reading between the lines
- These robustness differences imply that INDI may lower the calibration effort needed for aerial robots operating in variable conditions.
- Similar comparative studies could be extended to hybrid controllers that combine incremental and observer elements.
- Real-world validation on physical platforms under actual wind and payload changes would strengthen or qualify the simulation trends.
- The findings connect to broader questions of how inversion methods scale when model knowledge is partial or time-varying.
Load-bearing premise
The chosen simulation scenarios and parametric variations are representative of the disturbances and uncertainties encountered by real fully actuated aerial robots.
What would settle it
A hardware flight test on a physical fully actuated aerial robot under comparable model mismatches, combined disturbances, and noise that shows NDI+NDO matching or exceeding INDI in robustness metrics would falsify the comparative claim.
Figures
read the original abstract
This work presents a simulation-based comparative robustness analysis of Incremental Nonlinear Dynamic Inversion (INDI) and Nonlinear Dynamic Inversion augmented with a nonlinear disturbance observer (NDI+NDO) for fully actuated aerial robots. A systematic simulation campaign across representative operating scenarios is conducted, where we compare tracking performance, robustness, control effort, under parametric variations, external disturbances, and measurement noise. Results show that INDI demonstrates stronger robustness in several model-mismatch and combined-stress cases, while NDI+NDO primarily matches nominal performance but exhibits greater sensitivity under several non-ideal conditions. These findings provide practical guidance on the relative strengths and limitations of incremental and observer-based inversion strategies for aerial robotic applications.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. This manuscript presents a simulation-based comparative robustness analysis of Incremental Nonlinear Dynamic Inversion (INDI) versus Nonlinear Dynamic Inversion augmented with a nonlinear disturbance observer (NDI+NDO) for fully actuated aerial robots. It evaluates tracking performance, robustness, and control effort across operating scenarios involving parametric variations, external disturbances, and measurement noise, concluding that INDI exhibits stronger robustness under model-mismatch and combined-stress conditions while NDI+NDO matches nominal performance but shows greater sensitivity to non-ideal conditions, thereby offering practical guidance for controller selection in aerial robotics.
Significance. If the simulation results are representative, the work provides empirical evidence on the relative robustness trade-offs between incremental and observer-based inversion strategies, which could inform practical controller design choices for aerial robots. The study is purely comparative and empirical with no new theoretical derivations, proofs, or parameter-free results; its value is limited by the absence of hardware validation or cross-checks against real flight data.
major comments (2)
- [Abstract] Abstract: the claim that INDI 'demonstrates stronger robustness in several model-mismatch and combined-stress cases' is not supported by any quantitative metrics, error norms, statistical tests, or tabulated performance values, making it impossible to assess the practical magnitude of the reported gaps.
- [Simulation Results] Simulation campaign (assumed §4–5): the extension to 'practical guidance for aerial robotic applications' rests on the unverified assumption that the chosen parametric variations, disturbance models, and noise levels adequately proxy real-world conditions such as unmodeled actuator lags, aerodynamic effects, or sensor biases; no hardware experiments or flight-data cross-validation are reported to support this.
minor comments (2)
- [Results] Clarify the exact performance indices (e.g., RMSE, control effort integrals) and statistical methods used to declare 'stronger robustness' or 'greater sensitivity' in the results section.
- Ensure consistent notation for controller parameters and disturbance models across figures and text.
Simulated Author's Rebuttal
We thank the referee for the constructive comments on our simulation-based comparative study. We address each major comment below and outline revisions to improve clarity and qualification of our results.
read point-by-point responses
-
Referee: [Abstract] Abstract: the claim that INDI 'demonstrates stronger robustness in several model-mismatch and combined-stress cases' is not supported by any quantitative metrics, error norms, statistical tests, or tabulated performance values, making it impossible to assess the practical magnitude of the reported gaps.
Authors: We agree that the abstract would benefit from quantitative backing. In the revised manuscript, we will incorporate specific metrics from our simulation results, including RMS tracking errors, peak control efforts, and relative robustness margins under the model-mismatch and combined-stress scenarios, to substantiate the claims and allow assessment of the observed differences. revision: yes
-
Referee: [Simulation Results] Simulation campaign (assumed §4–5): the extension to 'practical guidance for aerial robotic applications' rests on the unverified assumption that the chosen parametric variations, disturbance models, and noise levels adequately proxy real-world conditions such as unmodeled actuator lags, aerodynamic effects, or sensor biases; no hardware experiments or flight-data cross-validation are reported to support this.
Authors: We acknowledge that the study is purely simulation-based and does not include hardware validation or flight-data cross-checks, which limits direct extrapolation to physical systems. The chosen variations and noise models were selected as representative of typical aerial robot challenges drawn from prior literature. We will revise the manuscript to explicitly state these limitations, qualify the simulation assumptions (e.g., ideal actuators and absence of unmodeled aerodynamics), and rephrase the 'practical guidance' as simulation-informed insights for controller selection rather than validated real-world recommendations. revision: partial
- Absence of hardware experiments or real flight-data validation, as the manuscript is limited to simulation studies.
Circularity Check
No circularity: purely empirical controller comparison
full rationale
The paper performs a simulation-based robustness comparison of two established controllers (INDI and NDI+NDO) across parametric variations, disturbances, and noise. No derivations, first-principles predictions, fitted parameters renamed as outputs, or uniqueness theorems appear. Claims rest on direct simulation outcomes rather than any self-referential reduction, self-citation chain, or ansatz smuggling. The work is self-contained as an empirical study with no load-bearing steps that collapse to inputs by construction.
Axiom & Free-Parameter Ledger
Reference graph
Works this paper leans on
- [1]
-
[2]
A. Alharbat and C. Gabellieri and A. Mersha and A. Franchi , title =. 2025 , volume =
work page 2025
-
[3]
M. Mizzoni and A. Afifi and A. Franchi , title =. 2025 , volume =. doi:10.1109/LCSYS.2025.3601657 , keywords =
-
[4]
A. Ollero and A. Suarez and C. Papaioannidis and I. Pitas and J. M. Marredo and V. Duong and E. Ebeid and V. Kr\'atk\'y and M. Saska and C. Hanoune and A. Afifi and A. Franchi and C. Vourtsis and D. Floreano and G. Vasiljevic and S. Bogdan and A. Caballero and F. Ruggiero and V. Lippiello and C. Matilla and G. Cioffi and D. Scaramuzza and J. R. Martinez-d...
work page 2025
-
[5]
C. Gabellieri and L. Teeuwen and Y. Shen and A. Franchi , title =. 2025. 2025 , month =
work page 2025
- [6]
-
[7]
B. Gerges and B. Bazzana and N. Botteghi and Y. Aboudorra and A. Franchi , title =. 2025. 2025 , month =
work page 2025
-
[8]
G. Indukumar and A. Saccon and A. Franchi and C. Gabellieri , title =. 2025. 2025 , month =
work page 2025
-
[9]
M. Mizzoni and P. Unified Feedback Linearization for Nonlinear Systems with Dexterous and Energy-Saving Modes , journal = lcss, year =
-
[10]
F. F. Nyboe and A. Afifi and P. Embedded Robust Model Predictive Path Integral Control using Sensitivity Tubes and. 2025. 2025 , address =
work page 2025
-
[11]
A. Srour and A. Franchi and P. Experimental Validation of Sensitivity-Aware Trajectory Planning for a Redundant Robotic Manipulator Under Payload Uncertainty , journal = ral, year =. doi:10.1109/LRA.2024.3519857 , keywords =
- [12]
-
[13]
A. Afifi and T. Belvedere and A. Pupa and P. Safe and Robust Planning for Uncertain Robots: A Closed-Loop State Sensitivity Approach , journal = ral, year =. doi:10.1109/LRA.2024.3468088 , keywords =
-
[14]
G. Lai and C. Shi and K. Wang and Y. Yu and Y. Dong and A. Franchi , title =. 2024 , volume =. doi:10.1109/LRA.2024.3407579 , keywords =
-
[15]
C. Gabellieri and A. Franchi , title =. 2024. 2024 , address =. doi:10.1109/ICUAS60882.2024.10556930 , keywords =
-
[16]
B. Bazzana and R. Brantjes and C. Gabellieri and A. Franchi , title =. 2024. 2024 , address =. doi:, keywords =
work page 2024
-
[17]
A. Srour and S. Marcellini and T. Belvedere, M. Cognetti and A. Franchi and P. Experimental Validation of Sensitivity-Aware Trajectory Planning for a Quadrotor. 2024. 2024 , address =. doi:, keywords =
work page 2024
-
[18]
F. Zoric and A. Franchi and M. Orsag and Z. Kovacic and C. Gabellieri , title =. 2024. 2024 , address =. doi:, keywords =
work page 2024
-
[19]
Y. Aboudorra and A. Saini and A. Franchi , title =. 2024. 2024 , address =. doi:, keywords =
work page 2024
-
[20]
Y. Aboudorra and C. Gabellieri and R. Brantjes and Q. Sabl\'e and A. Franchi , title =. 2024 , volume =. doi:10.1007/s10846-024-02054-x , keywords =
-
[21]
E. Ferrentino and H. Savino and A. Franchi and P. Chiacchio , title =. 2023 , volume =. doi:10.1109/TASE.2023.3330371 , keywords =
-
[22]
M. Hamandi and I. Al-Ali and L. Seneviratne and A. Franchi and Y. Zweiri , title =. 2023 , volume =. doi:10.1109/LRA.2023.3290422 , keywords =
-
[23]
A. Srour and A. Franchi and P. Controller and Trajectory Optimization for a Quadrotor. 2023. 2023 , address =
work page 2023
-
[24]
C. Gabellieri and M. Tognon and D. Sanalitro and A. Franchi , title =. 2023. 2023 , address =
work page 2023
-
[25]
C. Gabellieri and M. Tognon and D. Sanalitro and A. Franchi , title =. 2023 , volume =. doi:10.1109/TRO.2023.3279033 , keywords =
-
[26]
A. Afifi and G. Corsini and Q. Sabl\'e and Y. Aboudorra and D. Sidobre and A. Franchi , title =. 2023. 2023 , address =
work page 2023
- [27]
-
[28]
G. Silano and A. Afifi and M. Saska and A. Franchi , title =. 2023. 2023 , address =
work page 2023
- [29]
-
[30]
E. Rossi and M. Tognon and L. Ballotta and R. Carli and J. Cort\'es and A. Franchi and L. Schenato , title =. 2023 , volume =. doi:10.1016/j.automatica.2023.110941 , keywords =
-
[31]
A. Coelho and A. Albu-Schaeffer and A. Sachtler and H. Mishra and D. Bicego and C. Ott and A. Franchi , title =. 61th. 2022 , address =
work page 2022
- [32]
-
[33]
G. Corsini and M. Jacquet and H. Das and A Afifi and D. Sidobre and A. Franchi , title =. 2022. 2022 , address =
work page 2022
-
[34]
M. Rothammer and A. Coelho and H. Mishra and C. Ott and A. Franchi and A. Albu-Schaeffer , title =. 2022 , volume =. doi:10.1109/LCSYS.2022.3187349 , keywords =
-
[35]
R. Rashad and D. Bicego and J. Zult and S. Energy Aware Impedance Control of a Flying End-effector in the Port-. 2022 , volume =. doi:10.1109/TRO.2022.3183532 , keywords =
-
[36]
A. Jimenez-Cano and D. Sanalitro and M. Tognon and A. Franchi and J. Cort\'es , title =. 2022 , volume =. doi:10.1007/s10846-022-01668-3 , keywords =
-
[37]
D. Sanalitro and M. Tognon and A. Jimenez-Cano and J. Cort\'es and A. Franchi , title =. 2022 , volume =. doi:10.1109/LRA.2022.3176457 , keywords =
-
[38]
A. Alharbat and H. Esmaeeli and D. Bicego and A. Mersha and A. Franchi , title =. 2022. 2022 , address =
work page 2022
-
[39]
A. Afifi and M. Toward Physical Human-Robot Interaction Control with Aerial Manipulators: Compliance, Redundancy Resolution, and Input Limits , booktitle = "2022 "#icra, year =
work page 2022
-
[40]
M. Jacquet and M. Kivits and H. Das and A. Franchi , title =. 2022 , volume =. doi:10.1109/LRA.2022.3143218 , keywords =
-
[41]
F. Pierri and A. Franchi , title =. Encyclopedia of Robotics , year =. doi:10.1007/978-3-642-41610-1_177-1 , keywords =
-
[42]
M. Mohammadi and D. Bicego and A. Franchi and D. Barcelli and D. Prattichizzo , title =. Applied Sciences, Special issue on Aerial Robotics for Inspection and Maintenance. 2021 , volume =. doi:10.3390/app11198955 , keywords =
-
[43]
M. Hamandi and Q. Sable and M. Tognon and A. Franchi , title =. 1st. 2021 , address =
work page 2021
-
[44]
G. Corsini and M. Jacquet and A. E. A General Control Architecture for Visual Servoing and Physical Interaction Tasks for Fully-actuated Aerial Vehicles , booktitle = "1st "#airpharo, year =
-
[45]
Towards Safe Human-Quadrotor Interaction: Mixed-Initiative Control via Real-Time
B. Towards Safe Human-Quadrotor Interaction: Mixed-Initiative Control via Real-Time. 2021 , volume =. doi:10.1109/LRA.2021.3096502 , keywords =
-
[46]
A. Coelho and Y. Sarkisov and J. Lee and R. Balachandran and A. Franchi and K. Kondak and C. Ott , title =. 2021. 2021 , address =
work page 2021
-
[47]
A. Ollero and M. Tognon and A. Suarez and D. J. Lee and A. Franchi , title =. 2021 , volume =. doi:10.1109/TRO.2021.3084395 , keywords =
-
[48]
The International Journal of Robotics Research40(8-9), 1015–1044 (2021)
M. Hamandi and F. Usai and Q. Sable and N. Staub and M. Tognon and A. Franchi , title =. 2021 , volume =. doi:10.1177/02783649211025998 , keywords =
-
[49]
M. Ryll and D. Bicego and M. Giurato and M. Lovera and A. Franchi , title =. 2021 , volume =. doi:10.1109/TMECH.2021.3099197 , keywords =
-
[50]
D. Horla and M. Hamandi and W. Giernacki and A. Franchi , title =. 2021 , volume =. doi:10.1109/LRA.2021.3067229 , keywords =
-
[51]
E. Baskaya and M. Hamandi and M. Bronz and A. Franchi , title =. 2021 , volume =. doi:10.1109/LRA.2021.3067182 , keywords =
-
[52]
A. Coelho and Y. Sarkisov and X. Wu and H. Mishra and H. Singh and A. Dietrich and A. Franchi and K. Kondak and C. Ott , title =. 2021 , volume =. doi:10.1007/s10846-021-01365-7 , keywords =
-
[53]
M. Jacquet and A. Franchi , title =. 2021 , volume =. doi:10.1109/LRA.2020.3045654 , keywords =
-
[55]
E. Umili and M. Tognon and D. Sanalitro and G. Oriolo and A. Franchi , title =. 2020. 2020 , address =
work page 2020
-
[56]
M. Hamandi and K. Sawant and M. Tognon and A. Franchi , title =. 2020. 2020 , address =
work page 2020
-
[57]
A. Petitti and D. Sanalitro and M. Tognon and A. Milella and J. Cort\'es and A. Franchi , title =. 2020. 2020 , address =
work page 2020
-
[58]
M. Tognon and A. Franchi , isbn =. Theory and Applications for Control of Aerial Robots in Physical Interaction Through Tethers , publisher = sv, year =. doi:10.1007/978-3-030-48659-4 , series = star, keywords =
- [59]
-
[60]
G. Michieletto and A. Cenedese and L. Zaccarian and A. Franchi , title =. 2020 , volume =. doi:10.1016/j.automatica.2020.108991 , keywords =
-
[61]
M. Hamandi and M.Tognon and A. Franchi , title =. 2020. 2020 , address =
work page 2020
-
[62]
M. Jacquet and G. Corsini and D. Bicego and A. Franchi , title =. 2020. 2020 , address =
work page 2020
-
[63]
D. Sanalitro and H. J. Savino and M. Tognon and J. Cort\'es and A. Franchi , title =. 2020 , volume =. doi:10.1109/LRA.2020.2969930 , keywords =
-
[64]
G. Nava and Q. Sabl\'e and M. Tognon and D. Pucci and A. Franchi , title =. 2020 , volume =. doi:10.1109/LRA.2019.2958473 , keywords =
-
[65]
C. Gabellieri and M. Tognon and D. Sanalitro and L. Palottino and A. Franchi , title =. 2020 , volume =. doi:10.1007/s11721-019-00178-7 , keywords =
-
[66]
M. Furci and C. Nainer and L. Zaccarian and A. Franchi , title =. 2019 , volume =. doi:10.1109/TCST.2019.2944341 , keywords =
-
[67]
G. Michieletto and A. Cenedese and A. Franchi , title =. 58th. 2019 , address =
work page 2019
-
[68]
E. Rossi and M. Tognon and R. Carli and L. Schenato and J. Cort\'. Cooperative Aerial Load Transportation via Sampled Communication , journal = lcss, year =. doi:10.1109/LCSYS.2019.2924413 , keywords =
-
[69]
The International Journal of Robotics Research 38(9): 1045--1062
M. Ryll and G. Muscio and F. Pierri and E. Cataldi and G. Antonelli and F. Caccavale and D. Bicego and A. Franchi , title =. 2019 , volume =. doi:10.1177/0278364919856694 , keywords =
-
[70]
V. Walter and N. Staub and A. Franchi and M. Saska , title =. 2019 , volume =. doi:10.1109/LRA.2019.2901683 , keywords =
-
[71]
Y. Sarkisov and M. J. Kim and D. Bicego and D. Tsetserukou and C. Ott and A. Franchi and K. Kondak , title =. 2019. 2019 , address =
work page 2019
-
[72]
M. Tognon and H. A. A Truly Redundant Aerial Manipulator System with Application to Push-and-Slide Inspection in Industrial Plants , journal = ral, year =. doi:10.1109/LRA.2019.2895880 , keywords =
-
[73]
A. Franchi , title =. Aerial Robotic Manipulation , year =. doi:10.1007/978-3-030-12945-3 , keywords =
-
[74]
uksel and C. Secchi and H. H. B\
B. Y\"uksel and C. Secchi and H. H. B\"ulthoff and A. Franchi , title =. 2019 , volume =
work page 2019
-
[75]
A. Franchi and P. Online Leader Selection for Improved Collective Tracking and Formation Control: the Second Order Case , journal = tcns, year =. doi:10.1109/TCNS.2019.2891011 , keywords =
-
[76]
V. Arellano-Quintana and E. Merchan-Cruz and A. Franchi , title =. 2018 , volume =. doi:10.1109/ACCESS.2018.2879636 , keywords =
-
[77]
N. Staub and M. Mohammadi and D. Bicego and Q. Delamare and H. Yang and D. Prattichizzo and P. 2018 , volume =. doi:10.1109/MRA.2018.2871344 , keywords =
-
[78]
C. Masone and M. Mohammadi and P. Shared Planning and Control for Mobile Robots with Integral Haptic Feedback , journal = ijrr, year =
-
[79]
H. Yang and N. Staub and A. Franchi and D. Lee , title =. 2018. 2018 , address =. doi:10.1109/IROS.2018.8593834 , keywords =
-
[80]
F. Morbidi and D. Bicego and M. Ryll and A. Franchi , title =. 2018. 2018 , address =
work page 2018
-
[81]
A. Franchi and A. Petitti and A. Rizzo , title =. 2019 , volume =. doi:10.1109/TCNS.2018.2873153 , keywords =
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.