Pith. sign in

REVIEW 4 major objections 6 minor 58 references

AAM-SEALS: Developing Aerial-Aquatic Manipulators in SEa, Air, and Land Simulator

T0 review · 4 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper introduces a simulator, AAM-SEALS, in which an aerial-aquatic manipulator can fly, dive, and grasp objects, with particle-based water claimed to match real drop tests closely enough for simulation-first robot development.

desk verdict A genuinely new integrated simulator for cross-medium manipulation, but the abstract's quantitative fidelity claim is not backed by the reported validation. read the letter →

arxiv 2412.19744 v4 pith:5KA22WZU submitted 2024-12-27 cs.RO

William Yang , Karthikeya Kona , Yashveer Jain , Tomer Atzili , Abhinav Bhamidipati , Xiaomin Lin , Yantian Zha This is my paper · ORCID
classification cs.RO
keywords aerial-aquaticmanipulationcross-mediumroboticsposition-baseddynamicsparticle-basedhydrodynamicsrobotsimulationreinforcementlearningfromdemonstrationsmovablecenterofgravitycontrolunderwater
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 introduces a new class of robot, the Aerial-Aquatic Manipulator (AAM), and a simulator, AAM-SEALS, in which such a robot can fly, swim, enter the water, and grasp objects in one continuous world. The central claim is that particle-based hydrodynamics, implemented through position-based dynamics, makes the water realistic enough to support controller design and learning without a physical robot: the authors report water-entry damping, wave impacts, buoyancy, and underwater light effects, and they validate the fluid model by dropping a 3D-printed AAM into a tank and comparing its acceleration trace to the simulation. The paper also contributes a control architecture that recomputes the rotor-force allocation matrix as the manipulator shifts the robot's center of gravity, and it demonstrates reinforcement learning, including learning from demonstrations, on tasks such as capturing a moving crab. The authors position AAM-SEALS as the first high-fidelity simulator for cross-medium manipulation and frame it as a precursor to building physical AAMs.

What carries the argument

The central mechanism is position-based dynamics (PBD) for the fluid: water is a discrete set of particles whose positions are iteratively corrected so that the local density satisfies C = ρ_i/ρ_0 − 1 ≤ 0, with boundary particles included in the density estimate, giving stable free-surface waves, splashes, and buoyancy at interactive rates. The second load-bearing piece is the control allocation matrix, recomputed each time step from the instantaneous center of gravity, which maps desired force and torque into rotor speeds and lets a quadrotor keep tracking a command while its manipulator moves. These two pieces are what make the claimed realism and the demonstrated cross-medium trajectories possible.

What would settle it

Run repeated instrumented drops of objects with different masses, shapes, and impact speeds into a real water tank and into the simulator, and require that the acceleration traces agree quantitatively — for example, that the peak deceleration at water entry and the settling time match within a pre-specified margin across all objects. Alternatively, measure closed-loop position-tracking error during a diving trajectory on a physical prototype and compare it to the same trajectory in SEALS; if the errors diverge beyond sensor noise, the paper's claim of quantitative hydrodynamic validation would be refuted.

Watch

Extended reading notes

Core claim

The paper's central result is that a single particle-based fluid model can carry an aerial robot through the full air-to-water transition with enough realism to make simulation-first development plausible. In AAM-SEALS, water is modeled not as a rigid-body force field but as position-based dynamics (PBD) particles that satisfy a density constraint at every step, with boundary particles providing the pressures that produce buoyancy, splashes, and damping. On top of this fluid, the robot is controlled by a velocity PID and a joint PD controller, and the key mechanical idea is an allocation matrix A whose entries are recomputed continuously so the torque balance tracks the changing center of gravity as the arm moves. The evaluation claims that the resulting position-tracking error stays small in hover and on an oval trajectory that crosses the air-water boundary, and that both visual reinforcement learning and reinforcement learning from demonstrations converge in the simulator.

Load-bearing premise

The load-bearing premise is that the particle-based water in the simulator produces the same forces an actual robot feels when hitting and moving through water; in the paper, this is supported by only one drop-test comparison with no error metric, so if that comparison is unrepresentative, the hydrodynamic validation collapses.

Editorial extensions

If this is right

  • Simulation-first development of AAMs becomes feasible: control, perception, and learning can be tested in SEALS before investing in waterproofed physical hardware.
  • The dynamic allocation matrix offers a template for any aerial manipulator whose payload or arm shifts the center of gravity, not only water-crossing robots.
  • The simulator gives robot learning a benchmark where a single policy must handle aerial, aquatic, and transitional phases, with photorealistic cameras and contact sensors for visual RL.
  • The search-and-capture challenge with controllable crabs and sea spiders provides a reproducible new task for evaluating cross-medium manipulation without requiring live animals in training.
  • If the hydrodynamic fidelity holds under broader conditions, the same PBD approach could be applied to other free-surface robotics problems, such as boats, wave energy, or flooded-environment navigation.

Reading between the lines

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

  • The paper's validation rests on a single comparison with no error metric; a careful reader should treat 'quantitatively validated' as a goal rather than a demonstrated fact until repeated trials with error statistics appear.
  • The dynamic center-of-gravity allocation matrix is a general idea that could transfer to any aerial manipulator with a moving payload, an implication the paper only states in the AAM context.
  • A natural next experiment the paper does not run is closed-loop sim-to-real transfer of a trained grasping policy; SEALS would be the right testbed for that.
  • The PBD fluid model may be more credible for qualitative training (splashes, damping, visual appearance) than for exact force prediction; until error metrics are reported, policies trained on it should be treated as prescreened rather than final validation.
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

4 major / 6 minor

Summary. The paper introduces AAM-SEALS, a simulator built on NVIDIA Isaac Sim for Aerial-Aquatic Manipulators (AAMs) that aims to integrate flight, swimming, and manipulation across sea, air, and land. The system uses position-based dynamics (PBD) via PhysX for hydrodynamics, simplified linear drag for aerodynamics, PID/PD controllers for the vehicle and manipulator, and offers cameras, contact sensors, and RL interfaces. The evaluation claims quantitative validation of hydrodynamic fidelity by comparing simulated and real-world drop tests, demonstrates position-tracking for hovering and cross-medium trajectory following, and reports RL and RLfD training results. The paper also describes a new robot class, a photorealistic environment with aquatic animal models, and an open-source commitment.

Significance. If the validation were made rigorous, AAM-SEALS would be a valuable and timely contribution: it is apparently the first integrated simulator for aerial-aquatic manipulation that spans sea, air, and land, with photorealistic rendering, particle-based hydrodynamics, and learning interfaces. The authors provide a real-world drop-test data point, detailed appendices on PBD and RL hyperparameters, and a stated intention to open-source code and data; these are concrete assets. However, the central fidelity claim currently rests on a single qualitative acceleration-curve comparison, and the paper's own appendix admits that a rigid-body hydrodynamics baseline was integrated but never reported. The contribution is original and the direction is sound, but the evidence presented does not yet support the advertised quantitative validation.

major comments (4)
  1. [Abstract and Sec. V-A] The abstract's central validation claim—'quantitatively validate the fidelity of particle-based hydrodynamics by comparing position-tracking errors across real-world and simulated systems'—does not match the experiment reported in Sec. V-A, which compares one simulated z-axis acceleration-over-time curve with one real-world IMU curve (Fig. 10). No position-tracking error, numerical error metric, error bars, or repeat trials are reported, so the claimed quantitative validation is absent; the abstract and experiment need to be reconciled.
  2. [Sec. V-A and Appendix A] Appendix A states that a rigid-body hydrodynamics baseline was integrated into AAM-SEALS alongside position-based hydrodynamics 'to compare the two hydrodynamics models,' yet no baseline comparison appears in the evaluation. Without this baseline, the qualitative similarity in Fig. 10 cannot be attributed to PBD rather than to the linear drag model, buoyancy, or other tuned simulation parameters; the authors should report the baseline comparison or remove the claim that it provides insight.
  3. [Sec. V-A] The real-world and simulated drop tests are underconstrained: the mass, inertia, center of gravity, drop height, and water tank properties of the 3D-printed AAM are not specified as matched simulation parameters. The reported qualitative agreement therefore does not establish predictive fidelity; at minimum, the authors should report these parameters and ideally include a sensitivity analysis to show which parameters drive the observed agreement.
  4. [Sec. V-B and Eqs. (6)-(7)] The control evaluation reports position tracking within ±0.015 m, ±0.003 m, and ±0.2 m for X, Y, and Z during hovering, but the PID gains in Eq. (6) and PD gains in Eq. (7) are not reported, and no repeat trials or disturbance conditions are described. Since the adaptive-allocation mechanism in Eq. (4) is central to the claimed robustness to a changing center of gravity, the tracking results need at least gain values and a statement of how many runs they summarize.
minor comments (6)
  1. [Fig. 3 caption] The caption misspells 'Proportional' as 'Propotional' twice; please correct both instances.
  2. [Fig. 15 caption] The caption ends with an incomplete fragment, 'steps'; it should be completed or removed.
  3. [Sec. IV-B] The sentence 'This system gives SEAL a strong and cutting-edge balance' should read 'SEALS' rather than 'SEAL'.
  4. [Sec. V-A] The sentence 'All of the objects are equipped with an IMU sensor' is not supported by the reported experiments, which describe only the 3D-printed AAM; please clarify which objects were actually tested.
  5. [Appendix C] The reward regions are described as 'outer (distance greater than 1 meter), inner (distance between 1 meter and dt), and success (distance less than dt),' where dt is 10^{-2} m; this leaves the interval (dt, 1 m) ambiguously assigned, so the boundaries should be restated more precisely.
  6. [Sec. VI] The Limitations paragraph already acknowledges that Sim2Real transfer is not fully verified; this is appropriate, but it should be cross-referenced with the fidelity claims made in the abstract and Sec. V-A so that readers are not misled.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: hydrodynamics fidelity is checked against external real-world drop data, not fitted parameters; self-citations are peripheral.

full rationale

Walking the paper's claimed derivation chain, the central fidelity claim rests on Sec. V-A, where simulated and real-world free-fall acceleration curves are compared in Fig. 10. No coefficient in that comparison is fitted to the real IMU data; the particle-based hydrodynamics equations (Eqs. 10-12) are imported from external literature [25,26,31], and the simulator is built on NVIDIA Isaac Sim. The controller equations (Eqs. 1-7) and the CoG-adaptive allocation matrix (Eq. 4) do not define their outputs in terms of the quantities they are later used to evaluate. The self-citations [22,33,47,52] appear in related work or as implementation choices for RLfD; none carries the load of the fidelity or novelty claims, and the Pegasus-based control adaptation [13] is not a self-citation. Appendix A promises a rigid-body hydrodynamics baseline comparison that is never reported, and the abstract overstates Sec. V-A as comparing 'position-tracking errors' when only acceleration curves are shown without error metrics or repeat counts; these are evidentiary and soundness weaknesses, not circularity. The Limitations section explicitly disclaims full Sim2Real verification. The score of 1 reflects only the presence of non-load-bearing self-citations, not any reduction of a central claim to its own inputs.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central claim rests on unvalidated assumptions about PBD water fidelity, exact simulator-known CoG, unmeasured linear drag, and hand-tuned controller gains. No new physical entities such as new forces or particles are introduced; the AAM is assembled from existing robot components and the animal models are simulation assets.

free parameters (4)
  • PID gains of quadrotor controller (Kp, Kd, Ki)
    Controller gains in Eq. 6 are not specified; their values directly determine the reported position-tracking errors and are needed for reproduction.
  • PD gains of manipulator joints (Kmp, Kmd)
    Gains in Eq. 7 are not reported; they determine manipulator tracking behavior and the stability of the arm during flight and diving.
  • Linear drag coefficient c
    Aerodynamic drag in Eq. 8 is a per-axis constant in [0,1), but its value is not given or justified by measurement.
  • PBD solver parameters (rest density, particle radius, smoothing length)
    Water fidelity depends on PhysX PBD parameters; only the simulation timestep dt=0.004 is reported in Appendix C.
assumptions (5)
  • domain assumption Isaac Sim PhysX position-based dynamics accurately represent free-surface hydrodynamics including water-entry damping, buoyancy, and waves.
    Sections IV-B and IV-C rely on this for the central fidelity claim; the only supporting evidence is a qualitative drop-test comparison.
  • ad hoc to paper Linear drag model Fd = c v adequately captures aerial aerodynamic effects on the AAM.
    Equation 8 is inherited from Pegasus-style simulation, and the constant c is unmeasured and unreported.
  • domain assumption The center of gravity of the AAM is available exactly from the simulator at every timestep.
    Section III-B updates the allocation matrix using simulator CoG values, while real-world CoG estimation is acknowledged as an open problem in the footnote.
  • domain assumption Hand-designed RL rewards and termination conditions are sufficient for policy learning in the simulator.
    Appendix C defines dense distance-based rewards and a 0.01 m success threshold, but no sensitivity analysis or comparison to alternative reward designs is provided.
  • standard math Standard rigid-body dynamics and Jacobian kinematics equations apply to the AAM.
    Equations 1, 2, and 3 are standard results used in aerial manipulator modeling; they are not the subject of empirical validation in this paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of AAM-SEALS: Developing Aerial-Aquatic Manipulators in SEa, Air, and Land Simulator." pith.science (2026). https://pith.science/paper/5KA22WZU

@misc{pith2026241219744,
  author       = {Pith},
  title        = {Pith review of: AAM-SEALS: Developing Aerial-Aquatic Manipulators in SEa, Air, and Land Simulator},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5KA22WZU}},
  note         = {Machine review of arXiv:2412.19744}
}
read the original abstract

Current mobile manipulators and high-fidelity simulators lack the ability to seamlessly operate and simulate across integrated environments spanning sea, air, and land. To address this gap, we introduce Aerial-Aquatic Manipulators (AAMs) in SEa, Air, and Land Simulator (SEALS), a comprehensive and photorealistic simulator designed for AAMs to operate and learn in these diverse environments. The development of AAM-SEALS tackles several significant challenges, including the creation of integrated controllers for flying, swimming, and manipulation, and the high-fidelity simulation of aerial dynamics and hydrodynamics leveraging particle-based hydrodynamics. Our evaluation demonstrates smooth operation and photorealistic transitions across air, water, and their interfaces. We quantitatively validate the fidelity of particle-based hydrodynamics by comparing position-tracking errors across real-world and simulated systems. AAM-SEALS benefits a broad range of robotics communities, including robot learning, aerial robotics, underwater robotics, mobile manipulation, and robotic simulators. We will open-source our code and data to foster the advancement of research in these fields. The overview video is available at https://youtu.be/MbqIIrYvR78. Visit our project website at https://aam-seals.umd.edu for more details.

Figures

Figures reproduced from arXiv: 2412.19744 by the authors.

Figure 1
Figure 1. Showing a wide range of critical applications that leverage AAMs’ unique capabilities across sea, air, and land. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Schematic of a representative Aerial Aquatic Manipulator. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Block diagram of controller for AAM To obtain the required rotor angular velocities ω, the inverse of the allocation matrix A−1 is calculated and applied to the vector of desired force and torques [F, τx, τy, τz] T : ω 2 = A−1     F τx τy τz     (5) The squared angular velocities ω 2 are then processed to ensure they are non-negative, followed by normalization if any value exceeds the maximum permissible squ… view at source ↗
Figures from the paper (13 more)
Figure 4
Figure 4. Figure 4: Overview of our SEa, Air, and Lands Simulator (SEALS). [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Video frames of our Aerial-Aquatic Manipulator entering the water while enduring an ocean wave, with water damping and light refraction effects. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 8
Figure 8. Figure 8: Top: Simulated crab walking slowly on the sea floor; Bottom: Simulated sea spider captured by our Aerial-Aquatic Manipulator. This behavioral realism is crucial for practical applications in in-land aquaculture assistance, marine biology sampling, and fish farming, as …
Figure 6
Figure 6. Figure 6: Video frames showing light attenuation when the robot dives deeper. [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 9
Figure 9. Figure 9: Camera positions (red, green, and blue arrows are X, Y, and Z axes, [PITH_FULL_IMAGE:figures/full_fig_p006_9.png]
Figure 7
Figure 7. Figure 7: Visual representations of an underwater environment illustrating the [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 10
Figure 10. Figure 10: Left: A simulated AAM and its acceleration along the z-axis over time as it falls from air into water. Middle: The 3D-printed AAM and the water tank used in real-world experiments. Right: Acceleration along the z-axis over time for the real-world AAM falling into wate…
Figure 11
Figure 11. Figure 11: Plot of acceleration along x, y, and z directions over time while AAM falls freely from the air and gets hit by a wave. [PITH_FULL_IMAGE:figures/full_fig_p008_11.png]
Figure 12
Figure 12. Figure 12: Top: Sample video frames showcasing the AAM hovering with a moving manipulator; Bottom: Position-tracking curves over time for the X, Y, and Z directions [PITH_FULL_IMAGE:figures/full_fig_p009_12.png]
Figure 13
Figure 13. Figure 13: Oval trajectory following results [PITH_FULL_IMAGE:figures/full_fig_p009_13.png]
Figure 14
Figure 14. Figure 14: Reinforcement Learning from Demonstration results [PITH_FULL_IMAGE:figures/full_fig_p009_14.png]
Figure 16
Figure 16. Figure 16: Reinforcement Learning (Soft-Actor Critic) results [PITH_FULL_IMAGE:figures/full_fig_p014_16.png]
Figure 17
Figure 17. Figure 17: Crab Mesh in Blender After creating the URDF and meshes of the crab, we then need to model it and assign controllers to joints. The crab model consists of 18 joints, resulting in a total of 18 degrees of freedom (DoF) for the body. Managing such a high-DoF agent can b…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

58 extracted references · 51 canonical work pages

  1. [1]

    Autonomous Control of a Quadrotor-Manipulator; Application of Extended State Disturbance Observer

    Kabir Abdulmajeed. Autonomous control of a quadrotor- manipulator; application of extended state disturbance observer. arXiv preprint arXiv:1910.09052 , 2019

    work page Pith review arXiv 1910

  2. [2]

    Versatile surface tension and adhesion for sph fluids

    Nadir Akinci, Gizem Akinci, and Matthias Teschner. Versatile surface tension and adhesion for sph fluids. ACM Transactions on Graphics (TOG), 32(6):1–8, 2013

    work page 2013

  3. [3]

    Loon copter: Implementation of a hybrid unmanned aquatic–aerial quadcopter with active buoyancy control

    Hamzeh Alzu’bi, Iyad Mansour, and Osamah Rawashdeh. Loon copter: Implementation of a hybrid unmanned aquatic–aerial quadcopter with active buoyancy control. Journal of field Robotics , 35(5): 764–778, 2018

    work page 2018

  4. [4]

    Unav-sim: A visually realistic under- water robotics simulator and synthetic data-generation framework

    Abdelhakim Amer, Olaya ´Alvarez-Tu˜n´on, Halil ˙Ibrahim U˘gurlu, Jonas Le Fevre Sejersen, Yury Brodskiy, and Erdal Kayacan. Unav-sim: A visually realistic under- water robotics simulator and synthetic data-generation framework. In 2023 21st International Conference on Advanced Robotics (ICAR), pages 570–576. IEEE, 2023

    work page 2023

  5. [5]

    Comparison between smoothed-particle hydrodynamics and position based dynamics for real-time water simulation, 2023

    Rasmus Andersson and Erica Tjernell. Comparison between smoothed-particle hydrodynamics and position based dynamics for real-time water simulation, 2023

    work page 2023

  6. [6]

    Genetically optimized anfis- based pid controller design for posture-stabilization of self-balancing-robots under depleting battery conditions

    Omer Saleem Bhatti, Mohsin Rizwan, Panos S Sh- iokolas, and Babar Ali. Genetically optimized anfis- based pid controller design for posture-stabilization of self-balancing-robots under depleting battery conditions. Journal of Control Engineering and Applied Informatics, 21(4):22–33, 2019

    work page 2019

  7. [7]

    Au- tonomous underwater panel operation by girona500 uvms: A practical approach to autonomous underwater manipulation

    Patryk Cieslak, Pere Ridao, and Mariusz Giergiel. Au- tonomous underwater panel operation by girona500 uvms: A practical approach to autonomous underwater manipulation. In 2015 IEEE International conference on robotics and automation (ICRA) , pages 529–536. IEEE, 2015

    work page 2015

  8. [8]

    Uwsim: A simulator for underwater sensor networks

    Sanjay K Dhurandher, Sudip Misra, Mohammad S Obai- dat, and Sushil Khairwal. Uwsim: A simulator for underwater sensor networks. Simulation, 84(7):327–338, 2008

    work page 2008

Show all 58 references
  1. [9]

    Smoothed Particle Hydrody- namics

    Carlos Alberto Dutra Fraga Filho, Carlos Alberto Dutra Fraga Filho, and Castro. Smoothed Particle Hydrody- namics. Springer, 2019

    2019

  2. [10]

    Waymax: An accelerated, data-driven simulator for large-scale autonomous driving research

    Cole Gulino, Justin Fu, Wenjie Luo, George Tucker, Eli Bronstein, Yiren Lu, Jean Harb, Xinlei Pan, Yan Wang, Xiangyu Chen, et al. Waymax: An accelerated, data-driven simulator for large-scale autonomous driving research. Advances in Neural Information Processing Systems, 36, 2024

    2024

  3. [11]

    Soft actor-critic: Off-policy maximum entropy deep reinforcement learning with a stochastic actor

    Tuomas Haarnoja, Aurick Zhou, Pieter Abbeel, and Sergey Levine. Soft actor-critic: Off-policy maximum entropy deep reinforcement learning with a stochastic actor. In International conference on machine learning , pages 1861–1870. PMLR, 2018

    2018

  4. [12]

    A fast learn- ing control strategy for unmanned aerial manipulators

    Nursultan Imanberdiyev and Erdal Kayacan. A fast learn- ing control strategy for unmanned aerial manipulators. Journal of Intelligent & Robotic Systems , 94:805–824, 2019

    2019

  5. [13]

    Pegasus simulator: An isaac sim framework for multiple aerial vehicles simulation

    Marcelo Jacinto, Jo ˜ao Pinto, Jay Patrikar, John Keller, Rita Cunha, Sebastian Scherer, and Ant ´onio Pas- coal. Pegasus simulator: An isaac sim framework for multiple aerial vehicles simulation. arXiv preprint arXiv:2307.05263, 2023

    2023 arXiv

  6. [14]

    The effects of smoothing length on the onset of wave breaking in smoothed particle hydrodynamics (sph) simulations of highly directionally spread waves

    Taiga Kanehira, Mark L McAllister, Samuel Draycott, Takuji Nakashima, David M Ingram, Ton S van den Bremer, and Hidemi Mutsuda. The effects of smoothing length on the onset of wave breaking in smoothed particle hydrodynamics (sph) simulations of highly directionally spread wav...

    2022

  7. [15]

    Aerial manipulation—a lit- erature survey

    Hossein Bonyan Khamseh, Farrokh Janabi-Sharifi, and Abdelkader Abdessameud. Aerial manipulation—a lit- erature survey. Robotics and Autonomous Systems , 107: 221–235, 2018

    2018

  8. [16]

    Free-surface flow over a step

    AC King and MIG Bloor. Free-surface flow over a step. Journal of Fluid Mechanics , 182:193–208, 1987

    1987

  9. [17]

    Estimation, control, and planning for autonomous aerial transportation

    Hyeonbeom Lee and H Jin Kim. Estimation, control, and planning for autonomous aerial transportation. IEEE Transactions on Industrial Electronics, 64(4):3369–3379, 2016

    2016

  10. [18]

    Control of an aerial manipulator using on-line parameter esti- mator for an unknown payload

    Hyeonbeom Lee, Suseong Kim, and H Jin Kim. Control of an aerial manipulator using on-line parameter esti- mator for an unknown payload. In 2015 IEEE interna- tional conference on automation science and engineering (CASE), pages 316–321. IEEE, 2015

    2015

  11. [19]

    A robust neural con- troller for underwater robot manipulators

    Minho Lee and Hyeung-Sik Choi. A robust neural con- troller for underwater robot manipulators. IEEE Trans- actions on Neural Networks , 11(6):1465–1470, 2000

    2000

  12. [20]

    igibson 2.0: Object-centric simulation for robot learning of everyday household tasks

    Chengshu Li, Fei Xia, Roberto Mart ´ın-Mart´ın, Michael Lingelbach, Sanjana Srivastava, Bokui Shen, Kent Vainio, Cem Gokmen, Gokul Dharan, Tanish Jain, et al. igibson 2.0: Object-centric simulation for robot learning of everyday household tasks. arXiv preprint arXiv:2108.03272, 2021

    2021 arXiv

  13. [21]

    Gpu- accelerated robotic simulation for distributed reinforce- ment learning

    Jacky Liang, Viktor Makoviychuk, Ankur Handa, Nut- tapong Chentanez, Miles Macklin, and Dieter Fox. Gpu- accelerated robotic simulation for distributed reinforce- ment learning. In Conference on Robot Learning , pages 270–282. PMLR, 2018

    2018

  14. [22]

    Oystersim: Underwater simulation for enhancing oyster reef moni- toring

    Xiaomin Lin, Nitesh Jha, Mayank Joshi, Nare Kara- petyan, Yiannis Aloimonos, and Miao Yu. Oystersim: Underwater simulation for enhancing oyster reef moni- toring. In OCEANS 2022, Hampton Roads , pages 1–6. IEEE, 2022

    2022

  15. [23]

    Tj-flyingfish: Design and implementation of an aerial-aquatic quadrotor with tiltable propulsion units

    Xuchen Liu, Minghao Dou, Dongyue Huang, Songqun Gao, Ruixin Yan, Biao Wang, Jinqiang Cui, Qinyuan Ren, Lihua Dou, Zhi Gao, et al. Tj-flyingfish: Design and implementation of an aerial-aquatic quadrotor with tiltable propulsion units. In 2023 IEEE International Conference on Ro...

    2023

  16. [24]

    Marus-a marine robotics simulator

    Ivan Lon ˇcar, Juraj Obradovi ´c, Natko Kra ˇsevac, Luka Mandi´c, Igor Kvasi ´c, Fausto Ferreira, Vladimir Slo ˇsi´c, Dula Nad, and Nikola Miˇskovi´c. Marus-a marine robotics simulator. In OCEANS 2022, Hampton Roads , pages 1–

    2022

  17. [25]

    Position based fluids

    Miles Macklin and Matthias M ¨uller. Position based fluids. ACM Transactions on Graphics (TOG) , 32(4): 1–12, 2013

    2013

  18. [26]

    Unified particle physics for real-time applications

    Miles Macklin, Matthias M ¨uller, Nuttapong Chentanez, and Tae-Yong Kim. Unified particle physics for real-time applications. ACM Transactions on Graphics (TOG) , 33 (4):1–12, 2014

    2014

  19. [27]

    Uuv simulator: A gazebo-based package for underwater intervention and multi-robot simulation

    Musa Morena Marcusso Manh ˜aes, Sebastian A Scherer, Martin V oss, Luiz Ricardo Douat, and Thomas Rauschen- bach. Uuv simulator: A gazebo-based package for underwater intervention and multi-robot simulation. In OCEANS 2016 MTS/IEEE Monterey , pages 1–8. IEEE, 2016

    2016

  20. [28]

    Smoothed particle hydrodynamics implementation of the standard viscous- plastic sea-ice model and validation in simple idealized experiments

    Oreste Marquis, Bruno Tremblay, Jean-Franc ¸ois Lemieux, and Mohammed Islam. Smoothed particle hydrodynamics implementation of the standard viscous- plastic sea-ice model and validation in simple idealized experiments. The Cryosphere Discussions , 2022:1–33, 2022

    2022

  21. [29]

    Minimum snap trajectory generation and control for quadrotors

    Daniel Mellinger and Vijay Kumar. Minimum snap trajectory generation and control for quadrotors. In 2011 IEEE International Conference on Robotics and Automation, pages 2520–2525, 2011. doi: 10.1109/ ICRA.2011.5980409

    2011

  22. [30]

    J. J. Monaghan. Smoothed particle hydrodynamics. Annual review of astronomy and astrophysics , 30:543– 574, 1992

    1992

  23. [31]

    Position based dynamics

    Matthias M ¨uller, Bruno Heidelberger, Marcus Hennix, and John Ratcliff. Position based dynamics. Journal of Visual Communication and Image Representation, 18(2): 109–118, 2007

    2007

  24. [32]

    Overcoming ex- ploration in reinforcement learning with demonstrations

    Ashvin Nair, Bob McGrew, Marcin Andrychowicz, Wo- jciech Zaremba, and Pieter Abbeel. Overcoming ex- ploration in reinforcement learning with demonstrations. In 2018 IEEE international conference on robotics and automation (ICRA), pages 6292–6299. IEEE, 2018

    2018

  25. [33]

    Chatsim: Underwater simulation with natural language prompting

    Aadi Palnitkar, Rashmi Kapu, Xiaomin Lin, Cheng Liu, Nare Karapetyan, and Yiannis Aloimonos. Chatsim: Underwater simulation with natural language prompting. In OCEANS 2023-MTS/IEEE US Gulf Coast , pages 1–7. IEEE, 2023

    2023

  26. [34]

    Holoocean: An underwater robotics simulator

    Easton Potokar, Spencer Ashford, Michael Kaess, and Joshua G Mangelson. Holoocean: An underwater robotics simulator. In 2022 International Conference on Robotics and Automation (ICRA) , pages 3040–3046. IEEE, 2022

    2022

  27. [35]

    Habitat 3.0: A co-habitat for humans, avatars, and robots

    Xavier Puig, Eric Undersander, Andrew Szot, Mikael Dallaire Cote, Tsung-Yen Yang, Ruslan Partsey, Ruta Desai, Alexander Clegg, Michal Hlavac, So Yeon Min, et al. Habitat 3.0: A co-habitat for humans, avatars, and robots. In The Twelfth International Conference on Learning Repr...

  28. [36]

    Nonlinear control of aerial manipulation sys- tems

    Z Samadikhoshkho, S Ghorbani, F Janabi-Sharifi, and K Zareinia. Nonlinear control of aerial manipulation sys- tems. Aerospace Science and Technology , 104:105945, 2020

    2020

  29. [37]

    Habitat: A platform for embodied ai research

    Manolis Savva, Abhishek Kadian, Oleksandr Maksymets, Yili Zhao, Erik Wijmans, Bhavana Jain, Julian Straub, Jia Liu, Vladlen Koltun, Jitendra Malik, et al. Habitat: A platform for embodied ai research. In Proceedings of the IEEE/CVF international conference on computer vision, ...

    2019

  30. [38]

    Control and dynamic motion planning for a hybrid air-underwater quadrotor: Minimizing energy use in a flooded cave environment

    Ilya Semenov, Robert Brown, and Michael Otte. Control and dynamic motion planning for a hybrid air-underwater quadrotor: Minimizing energy use in a flooded cave environment. arXiv preprint arXiv:2301.00936 , 2023

    2023 arXiv

  31. [39]

    Airsim: High-fidelity visual and physical sim- ulation for autonomous vehicles

    Shital Shah, Debadeepta Dey, Chris Lovett, and Ashish Kapoor. Airsim: High-fidelity visual and physical sim- ulation for autonomous vehicles. In Field and Service Robotics: Results of the 11th International Conference , pages 621–635. Springer, 2018

    2018

  32. [40]

    Underwater manipulators: A review

    Satja Siv ˇcev, Joseph Coleman, Edin Omerdi ´c, Gerard Dooly, and Daniel Toal. Underwater manipulators: A review. Ocean engineering, 163:431–450, 2018

    2018

  33. [41]

    Habitat 2.0: Training home assistants to rearrange their habitat

    Andrew Szot, Alexander Clegg, Eric Undersander, Erik Wijmans, Yili Zhao, John Turner, Noah Maestre, Mustafa Mukadam, Devendra Singh Chaplot, Oleksandr Maksymets, et al. Habitat 2.0: Training home assistants to rearrange their habitat. Advances in neural information processing ...

    2021

  34. [42]

    Design of a morphable multirotor aerial-aquatic vehicle

    Yu Herng Tan and Ben M Chen. Design of a morphable multirotor aerial-aquatic vehicle. In Oceans 2019 Mt- s/IEEE Seattle, pages 1–8. IEEE, 2019

    2019

  35. [43]

    A morphable aerial- aquatic quadrotor with coupled symmetric thrust vector- ing

    Tan, Yu Herng and Chen, Ben M. A morphable aerial- aquatic quadrotor with coupled symmetric thrust vector- ing. In 2020 IEEE International Conference on Robotics and Automation (ICRA) , pages 2223–2229. IEEE, 2020

    2020

  36. [44]

    Fluid–structure in- teraction analyses of biological systems using smoothed- particle hydrodynamics

    Milan Toma, Rosalyn Chan-Akeley, Jonathan Arias, Gre- gory D Kurgansky, and Wenbin Mao. Fluid–structure in- teraction analyses of biological systems using smoothed- particle hydrodynamics. Biology, 10(3):185, 2021

    2021

  37. [45]

    Smoothed particle hydrodynamics (sph) for free-surface flows: past, present and future

    Damien Violeau and Benedict D Rogers. Smoothed particle hydrodynamics (sph) for free-surface flows: past, present and future. Journal of Hydraulic Research , 54 (1):1–26, 2016

    2016

  38. [46]

    Phobos: A tool for creating complex robot models

    Kai von Szadkowski and Simon Reichel. Phobos: A tool for creating complex robot models. Journal of Open Source Software, 5(45):1326, 2020

    2020

  39. [47]

    Marvis: Motion & geometry aware real and virtual image segmentation

    Jiayi Wu, Xiaomin Lin, Shahriar Negahdaripour, Cor- nelia Ferm¨uller, and Yiannis Aloimonos. Marvis: Motion & geometry aware real and virtual image segmentation. arXiv preprint arXiv:2403.09850 , 2024

    2024 arXiv

  40. [48]

    Design and demon- stration of a tandem dual-rotor aerial–aquatic vehicle

    Sihuan Wu, Maosen Shao, Sifan Wu, Zhilin He, Hui Wang, Jinxiu Zhang, and Yue You. Design and demon- stration of a tandem dual-rotor aerial–aquatic vehicle. Drones, 8(3):100, 2024

    2024

  41. [49]

    Survey on smoothed particle hydrodynamics and the particle sys- tems

    Runping Xi, Zhangcai Luo, David Dagan Feng, Yanning Zhang, Xiaopeng Zhang, and Tianyi Han. Survey on smoothed particle hydrodynamics and the particle sys- tems. IEEE Access, 8:3087–3105, 2019

    2019

  42. [50]

    Motion planning survey for autonomous mobile manipulators underwater manip- ulator case study

    Dina Youakim and Pere Ridao. Motion planning survey for autonomous mobile manipulators underwater manip- ulator case study. Robotics and Autonomous Systems , 107:20–44, 2018

    2018

  43. [51]

    Simulating complex fluids with smoothed particle hydrodynamics

    Vito Zago, Giuseppe Bilotta, Annalisa Cappello, Robert Dalrymple, Luigi Fortuna, Gaetana Ganci, Alexis H´erault, and Ciro Del Negro. Simulating complex fluids with smoothed particle hydrodynamics. Annals of Geophysics, 2017

    2017

  44. [52]

    Learning from ambiguous demonstrations with self- explanation guided reinforcement learning

    Yantian Zha, Lin Guan, and Subbarao Kambhampati. Learning from ambiguous demonstrations with self- explanation guided reinforcement learning. In Proceed- ings of the AAAI Conference on Artificial Intelligence , volume 38, pages 10395–10403, 2024

    2024

  45. [53]

    Dave aquatic virtual environment: Toward a general underwater robotics simulator

    Mabel M Zhang, Woen-Sug Choi, Jessica Herman, Du- ane Davis, Carson V ogt, Michael McCarrin, Yadunund Vijay, Dharini Dutia, William Lew, Steven Peters, et al. Dave aquatic virtual environment: Toward a general underwater robotics simulator. In 2022 IEEE/OES Au- tonomous Underw...

    2022

  46. [54]

    Design and dynamic analysis of an underwater manipulator

    Wuxiang Zhang, Hongcheng Xu, and Xilun Ding. Design and dynamic analysis of an underwater manipulator. In Proceedings of the 2015 Chinese Intelligent Automation Conference: Intelligent Technology and Systems , pages 399–409. Springer, 2015

    2015

  47. [55]

    Creating a synthetic underwater dataset for egomotion estimation and 3d reconstruction

    Peder Georg Olofsson Zwilgmeyer. Creating a synthetic underwater dataset for egomotion estimation and 3d reconstruction. Master’s thesis, NTNU, 2021. APPENDIX A POTENTIAL QUESTIONS AND OUR ANSWERS In this section, we aim to address potential concerns and questions and hope to ...

    2013

  48. [56]

    Designing the Robot Model: • Begin by designing the robot model in SolidWorks (a 3D CAD Design Software) • Create the individual parts and assemble them, ensuring all joints and kinematic properties are accurately defined

  49. [57]

    Generating the Mesh Files and .urdf File: • Create mesh files to represent the robot’s physical structure visually and geometrically Note: These meshes provide a realistic appearance in the simulation and can be exported alongside the .urdf file • Export these meshes alongside...

  50. [58]

    Importing into Isaac Sim: • Import the .urdf file into the Isaac Sim simulator • Convert the .urdf file into a .usd (Universal Scene Description) file Note: The .usd format is essential because it enables seamless integration and manipulation within the simulator, ensuring tha...

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.