Search

Statistical tools are crucial for studying and modeling turbulent flows, where chaotic velocity fluctuations span a wide range of spatial and temporal scales. Advances in image velocimetry, especially in tracking-based methods, now allow for high-speed, high-density particle image processing, enabling the collection of detailed 3D flow fields. This lecture provides a set of tutorials on processing such datasets to extract essential quantities like averages, second-order moments (turbulent stresses) and coherent patterns using modal decompositions such as the Proper Orthogonal Decomposition (POD). After a brief review of the fundamentals of statistical and modal analysis, the lecture delves into the challenges of processing scattered data from tracking velocimetry, comparing it to traditional gridded-data approaches. It also covers research topics, including physics-based Radial Basis Function (RBF) regression for meshless computation of turbulent statistics and the definition of an RBF inner product, which enables meshless computation of data-driven decompositions. These include traditional methods like Proper Orthogonal Decomposition (POD) and Dynamic Mode Decomposition (DMD), as well as advanced variants such as Spectral POD (SPOD) and Multiscale POD (mPOD). We refer to this approach as the "Meshless Data-Driven Decomposition" framework. All codes are available at https://github.com/mendezVKI/PIV_LS_2024_Signal, Comment: Book in preparation! Chapter from VKI Lecture Series "Particle Image Velocimetry 2024", held at the von Karman Institute, 2-5 December 2024

Machine learning methods have a groundbreaking impact in many application domains, but their application on real robotic platforms is still limited. Despite the many challenges associated with combining machine learning technology with robotics, robot learning remains one of the most promising directions for enhancing the capabilities of robots. When deploying learning-based approaches on real robots, extra effort is required to address the challenges posed by various real-world factors. To investigate the key factors influencing real-world deployment and to encourage original solutions from different researchers, we organized the Robot Air Hockey Challenge at the NeurIPS 2023 conference. We selected the air hockey task as a benchmark, encompassing low-level robotics problems and high-level tactics. Different from other machine learning-centric benchmarks, participants need to tackle practical challenges in robotics, such as the sim-to-real gap, low-level control issues, safety problems, real-time requirements, and the limited availability of real-world data. Furthermore, we focus on a dynamic environment, removing the typical assumption of quasi-static motions of other real-world benchmarks. The competition's results show that solutions combining learning-based approaches with prior knowledge outperform those relying solely on data when real-world deployment is challenging. Our ablation study reveals which real-world factors may be overlooked when building a learning-based solution. The successful real-world air hockey deployment of best-performing agents sets the foundation for future competitions and follow-up research directions., Comment: Accept at NeurIPS 2024 Dataset and Benchmark Track

Flatback airfoils, characterized by a blunt trailing edge, are used at the root of large wind turbine blades. A low-drag pocket has recently been identified in the flow past these airfoils at high angles of attack, potentially offering opportunities for enhanced energy extraction. This study uses three-dimensional Detached Eddy Simulations (DES) combined with statistical and data-driven modal analysis techniques to explore the aerodynamics and coherent structures of a flatback airfoil in these conditions. Two angles of attack - one inside $\left(12^{\circ}\right)$ and one outside $\left(0^{\circ}\right)$ of the low-drag pocket - are examined more thoroughly. The spanwise correlation length of secondary instability is analyzed in terms of autocorrelation of the $\Gamma_{1}$ vortex identification criterion, while coherent structures were extracted via the multiscale Proper Orthogonal Decomposition (mPOD). The results show increased base pressure, BL thickness, vortex formation length, and more organized wake structures inside the low-drag regime. While the primary instability (B\'enard-von K\'arm\'an vortex street) dominates in both cases, the secondary instability is distinguishable only for the $12^{\circ}$ case and is identified as a Mode S$^{\prime}$ instability., Comment: 27 pages, 25 figures, submitted to Physical Review Fluids

Sloshing of cryogenic liquid propellants can significantly impact a spacecraft's mission safety and performance by unpredictably altering the center of mass and producing large pressure fluctuations due to the increased heat and mass transfer within the tanks. This study, conducted as part of the NT-SPARGE (Non-isoThermal Sloshing PARabolic FliGht Experiment) project, provides a detailed experimental investigation of the thermodynamic evolution of a partially filled upright cylindrical tank undergoing non-isothermal sloshing in microgravity. Sloshing was induced by a step reduction in gravity during the 83rd European Space Agency (ESA) parabolic flight, resulting in a chaotic reorientation of the free surface under inertia-dominated conditions. To investigate the impact of heat and mass transfer on the sloshing dynamics, two identical test cells operating with a representative fluid, HFE-7000, in single-species were considered simultaneously. One cell was maintained in isothermal conditions, while the other started with initially thermally stratified conditions. Flow visualization, pressure, and temperature measurements were acquired for both cells. The results highlight the impact of thermal mixing on liquid dynamics coupled with the significant pressure and temperature fluctuations produced by the destratification. The comprehensive experimental data gathered provide a unique opportunity to validate numerical simulations and simplified models for non-isothermal sloshing in microgravity, thus contributing to improved cryogenic fluid management technologies.

Data-driven RANS modeling is emerging as a promising methodology to exploit the information provided by high-fidelity data. However, its widespread application is limited by challenges in generalization and robustness to inconsistencies between input data of varying fidelity levels. This is especially true for thermal turbulent closures, which inherently depend on momentum statistics provided by low or high fidelity turbulence momentum models. This work investigates the impact of momentum modeling inconsistencies on a data-driven thermal closure trained with a dataset with multiple fidelity (DNS and RANS). The analysis of the model inputs shows that the two fidelity levels correspond to separate regions in the input space. It is here shown that such separation can be exploited by a training with heterogeneous data, allowing the model to detect the level of fidelity in its inputs and adjust its prediction accordingly. In particular, a sensitivity analysis and verification shows that such a model can leverage the data inconsistencies to increase its robustness. Finally, the verification with a CFD simulation shows the potential of this multi-fidelity training approach for flows in which momentum statistics provided by traditional models are affected by model uncertainties.

On-orbit servicing (OOS) activities will power the next big step for sustainable exploration and commercialization of space. Developing robotic capabilities for autonomous OOS operations is a priority for the space industry. Visual Servoing (VS) enables robots to achieve the precise manoeuvres needed for critical OOS missions by utilizing visual information for motion control. This article presents an overview of existing VS approaches for autonomous OOS operations with space manipulator systems (SMS). We divide the approaches according to their contribution to the typical phases of a robotic OOS mission: a) Recognition, b) Approach, and c) Contact. We also present a discussion on the reviewed VS approaches, identifying current trends. Finally, we highlight the challenges and areas for future research on VS techniques for robotic OOS., Comment: Accepted for publication at the 2024 International Conference on Space Robotics (iSpaRo)

Accurate disturbance estimation is crucial for reliable robotic physical interaction. To estimate environmental interference in a low-cost and sensorless way (without force sensor), a variety of tightly-coupled visual inertial external force estimators are proposed in the literature. However, existing solutions may suffer from relatively low-frequency preintegration. In this paper, a novel estimator is designed to overcome this issue via high-frequency instantaneous accelerometer update., Comment: Accepted by the 40th Anniversary of the IEEE Conference on Robotics and Automation (ICRA@40)

The stability of liquid film flows are important in many industrial applications. In the dip-coating process, a liquid film is formed over a substrate extracted at a constant speed from a liquid bath. We studied the linear stability of this film considering different thicknesses $\hat{h}$ for four liquids, spanning a large range of Kapitza numbers ($\rm Ka$). By solving the Orr-Sommerfeld eigenvalue problem with the Chebyshev-Tau spectral method, we calculated the neutral curves, investigated the instability mechanism and computed the absolute/convective threshold. The instability mechanism was studied through the analysis of vorticity distribution and the kinetic energy balance of the perturbations. It was found that liquids with low $\rm Ka$ (e.g. corn oil, $\text{Ka}$ = 4) have a smaller area of stability than a liquid at high $\rm Ka$ (e.g. Liquid Zinc, $\rm Ka$ = 11525). Surface tension has both a stabilizing and a destabilizing effect, especially for large $\rm Ka$. For long waves, it curves the vorticity lines near the substrate, reducing the flow under the crests. For short waves, it fosters vorticity production at the interface and creates a region of intense vorticity near the substrate. In addition, we discovered that the surface tension contributes to both the production and dissipation of perturbation's energy depending on the $\rm Ka$ number. In terms of absolute/convective threshold, we found a window of absolute instability in the $\text{Re}-\hat{h}$ space, showing that the Landau-Levich-Derjaguin solution ($\hat{h}=0.945 \text{Re}^{1/9}\text{Ka}^{-1/6}$) is always convectively unstable. Moreover, we show that for $\text{Ka}<17$, the Derjaguin's solution ($\hat{h}=1$) is always convectively unstable.

Hummingbirds and insects achieve outstanding flight performance by adapting their flapping motion to the flight requirements. Their wing kinematics can change from smooth flapping to highly dynamic waveforms, generating unsteady aerodynamic phenomena such as leading-edge vortices (LEV), rotational circulation, wing wake capture, and added mass. This article uncovers the interactions of these mechanisms in the case of a rigid semi-elliptical wing undergoing aggressive kinematics in the hovering regime at $Re\sim \mathcal{O}(10^3)$. The flapping kinematics were parametrized using smoothed steps and triangular functions and the flow dynamics were simulated by combining the overset method with Large Eddy Simulations (LES). The analysis of the results identifies an initial acceleration phase and a cruising phase. During the former, the flow is mostly irrotational and governed by the added mass effect. The added mass was shown to be responsible for a lift first peak due to the strong flapping acceleration. The dynamic pitching and the wing wake interaction generate a second lift peak due to a downwash flow and a vortex system on the proximal and distal parts of the wing's pressure side. Conversely, aerodynamic forces in the cruising phase are mainly governed by the growth and the establishment of the LEV. Finally, the leading flow structures in each phase and their impact on the aerodynamic forces were isolated using the extended Proper Orthogonal Decomposition (POD).

Complex phenomena can be better understood when broken down into a limited number of simpler "components". Linear statistical methods such as the principal component analysis and its variants are widely used across various fields of applied science to identify and rank these components based on the variance they represent in the data. These methods can be seen as factorisations of the matrix collecting all the data, assuming it consists of time series sampled from fixed points in space. However, when data sampling locations vary over time, as with mobile monitoring stations in meteorology and oceanography or with particle tracking velocimetry in experimental fluid dynamics, advanced interpolation techniques are required to project the data onto a fixed grid before the factorisation. This interpolation is often expensive and inaccurate. This work proposes a method to decompose scattered data without interpolating. The approach employs physics-constrained radial basis function regression to compute inner products in space and time. The method provides an analytical and mesh-independent decomposition in space and time, demonstrating higher accuracy. Our approach allows distilling the most relevant "components" even for measurements whose natural output is a distribution of data scattered in space and time, maintaining high accuracy and mesh independence.

Recent advancements in Multi-Robot Systems (MRS) and mesh network technologies pave the way for innovative approaches to explore extreme environments. The Artemis Accords, a series of international agreements, have further catalyzed this progress by fostering cooperation in space exploration, emphasizing the use of cutting-edge technologies. In parallel, the widespread adoption of the Robot Operating System 2 (ROS 2) by companies across various sectors underscores its robustness and versatility. This paper evaluates the performances of available ROS 2 MiddleWare (RMW), such as FastRTPS, CycloneDDS and Zenoh, over a mesh network with a dynamic topology. The final choice of RMW is determined by the one that would fit the most the scenario: an exploration of the extreme extra-terrestrial environment using a MRS. The conducted study in a real environment highlights Zenoh as a potential solution for future applications, showing a reduced delay, reachability, and CPU usage while being competitive on data overhead and RAM usage over a dynamic mesh topology, Comment: PrePrint

Electromagnetic wiping systems allow to pre-meter the coating thickness of the liquid metal on a moving substrate. These systems have the potential to provide a more uniform coating and significantly higher production rates compared to pneumatic wiping, but they require substantially larger amounts of energy. This work presents a multi-objective optimization accounting for (1) maximal wiping efficiency (2) maximal smoothness of the wiping meniscus, and (3) minimal Joule heating. We present the Pareto front, identifying the best wiping conditions given a set of weights for the three competing objectives. The optimization was based on a 1D steady-state integral model, whose prediction scales according to the Hartmann number (Ha). The optimization uses a multi-gradient approach, with gradients computed with a combination of finite differences and variational methods. The results show that the wiping efficiency depends solely on Ha and not the magnetic field distribution. Moreover, we show that the liquid thickness becomes insensitive to the intensity of the magnetic field above a certain threshold and that the current distribution (hence the Joule heating) is mildly affected by the magnetic field's intensity and shape.

Generalizable perception is one of the pillars of high-level autonomy in space robotics. Estimating the structure and motion of unknown objects in dynamic environments is fundamental for such autonomous systems. Traditionally, the solutions have relied on prior knowledge of target objects, multiple disparate representations, or low-fidelity outputs unsuitable for robotic operations. This work proposes a novel approach to incrementally reconstruct and track a dynamic unknown object using a unified representation -- a set of 3D Gaussian blobs that describe its geometry and appearance. The differentiable 3D Gaussian Splatting framework is adapted to a dynamic object-centric setting. The input to the pipeline is a sequential set of RGB-D images. 3D reconstruction and 6-DoF pose tracking tasks are tackled using first-order gradient-based optimization. The formulation is simple, requires no pre-training, assumes no prior knowledge of the object or its motion, and is suitable for online applications. The proposed approach is validated on a dataset of 10 unknown spacecraft of diverse geometry and texture under arbitrary relative motion. The experiments demonstrate successful 3D reconstruction and accurate 6-DoF tracking of the target object in proximity operations over a short to medium duration. The causes of tracking drift are discussed and potential solutions are outlined., Comment: Accepted at IEEE International Conference on Space Robotics 2024

The ability to autonomously assemble structures is crucial for the development of future space infrastructure. However, the unpredictable conditions of space pose significant challenges for robotic systems, necessitating the development of advanced learning techniques to enable autonomous assembly. In this study, we present a novel approach for learning autonomous peg-in-hole assembly in the context of space robotics. Our focus is on enhancing the generalization and adaptability of autonomous systems through deep reinforcement learning. By integrating procedural generation and domain randomization, we train agents in a highly parallelized simulation environment across a spectrum of diverse scenarios with the aim of acquiring a robust policy. The proposed approach is evaluated using three distinct reinforcement learning algorithms to investigate the trade-offs among various paradigms. We demonstrate the adaptability of our agents to novel scenarios and assembly sequences while emphasizing the potential of leveraging advanced simulation techniques for robot learning in space. Our findings set the stage for future advancements in intelligent robotic systems capable of supporting ambitious space missions and infrastructure development beyond Earth., Comment: Accepted for publication at the 2024 International Conference on Space Robotics (iSpaRo) | The source code is available at https://github.com/AndrejOrsula/drl_omni_peg

Lunar exploration has become a key focus, driving scientific and technological advances. Ongoing missions are deploying rovers to the surface of the Moon, targeting the far side and south pole. However, these terrains pose challenges, emphasizing the need for precise obstacles and resource detection to avoid mission risks. This work proposes a novel system that integrates eXtended Reality (XR) and Artificial Intelligence (AI) to teleoperate lunar rovers. It is capable of autonomously detecting rocks and recreating an immersive 3D virtual environment of the location of the robot. This system has been validated in a lunar laboratory to observe its advantages over traditional 2D-based teleoperation approaches

We propose a method to obtain superresolution of turbulent statistics for three-dimensional ensemble particle tracking velocimetry (EPTV). The method is ''meshless'' because it does not require the definition of a grid for computing derivatives, and it is ''binless'' because it does not require the definition of bins to compute local statistics. The method combines the constrained radial basis function (RBF) formalism introduced Sperotto et al. (Meas Sci Technol, 33:094005, 2022) with a kernel estimate approach for the ensemble averaging of the RBF regressions. The computational cost for the RBF regression is alleviated using the partition of unity method (PUM). Three test cases are considered: (1) a 1D illustrative problem on a Gaussian process, (2) a 3D synthetic test case reproducing a 3D jet-like flow, and (3) an experimental dataset collected for an underwater jet flow at $\text{Re} = 6750$ using a four-camera 3D PTV system. For each test case, the method performances are compared to traditional binning approaches such as Gaussian weighting (Ag\"u\'i and Jim\'enez, JFM, 185:447-468, 1987), local polynomial fitting (Ag\"uera et al, Meas Sci Technol, 27:124011, 2016), as well as a binned version of the RBF statistics., Comment: 21 pages, 13 figures, submitted to: Experiments in Fluids

In robotics, motion capture systems have been widely used to measure the accuracy of localization algorithms. Moreover, this infrastructure can also be used for other computer vision tasks, such as the evaluation of Visual (-Inertial) SLAM dynamic initialization, multi-object tracking, or automatic annotation. Yet, to work optimally, these functionalities require having accurate and reliable spatial-temporal calibration parameters between the camera and the global pose sensor. In this study, we provide two novel solutions to estimate these calibration parameters. Firstly, we design an offline target-based method with high accuracy and consistency. Spatial-temporal parameters, camera intrinsic, and trajectory are optimized simultaneously. Then, we propose an online target-less method, eliminating the need for a calibration target and enabling the estimation of time-varying spatial-temporal parameters. Additionally, we perform detailed observability analysis for the target-less method. Our theoretical findings regarding observability are validated by simulation experiments and provide explainable guidelines for calibration. Finally, the accuracy and consistency of two proposed methods are evaluated with hand-held real-world datasets where traditional hand-eye calibration method do not work., Comment: Accepted by 3DV 2024

In the emerging space economy, autonomous robotic missions with specialized goals such as mapping and mining are gaining traction, with agencies and enterprises increasingly investing resources. Multirobot systems (MRS) research has provided many approaches to establish control and communication layers to facilitate collaboration from a technical perspective, such as granting more autonomy to heterogeneous robotic groups through auction-based interactions in mesh networks. However, stakeholders' competing economic interests often prevent them from cooperating within a proprietary ecosystem. Related work suggests that distributed ledger technology (DLT) might serve as a mechanism for enterprises to coordinate workflows and trade services to explore space resources through a transparent, reliable, non-proprietary digital platform. We challenge this perspective by pointing to the core technical weaknesses of blockchains, in particular, increased energy consumption, low throughput, and full transparency through redundancy. Our objective is to advance the discussion in a direction where the benefits of DLT from an economic perspective are weighted against the drawbacks from a technical perspective. We finally present a possible DLT-driven heterogeneous MRS for map exploration to study the opportunities for economic collaboration and competitiveness., Comment: 5 pages, 1 figure, 1 table, accepted for conference (ICSE24-NIER)

In the new space economy, space agencies, large enterprises, and start-ups aim to launch space multi-robot systems (MRS) for various in-situ resource utilization (ISRU) purposes, such as mapping, soil evaluation, and utility provisioning. However, these stakeholders' competing economic interests may hinder effective collaboration on a centralized digital platform. To address this issue, neutral and transparent infrastructures could facilitate coordination and value exchange among heterogeneous space MRS. While related work has expressed legitimate concerns about the technical challenges associated with blockchain use in space, we argue that weighing its potential economic benefits against its drawbacks is necessary. This paper presents a novel architectural framework and a comprehensive set of requirements for integrating blockchain technology in MRS, aiming to enhance coordination and data integrity in space exploration missions. We explored distributed ledger technology (DLT) to design a non-proprietary architecture for heterogeneous MRS and validated the prototype in a simulated lunar environment. The analyses of our implementation suggest global ISRU efficiency improvements for map exploration, compared to a corresponding group of individually acting robots, and that fostering a coopetitive environment may provide additional revenue opportunities for stakeholders., Comment: 10 pages, 4 figures, accepted for conference (SAC24)

Compound flows consist of two or more parallel compressible streams in a duct and their theoretical treatment has gained attention for the analysis and modelling of ejectors. Recent works have shown that these flows can experience choking upstream of the geometric throat. While it is well known that friction can push the sonic section downstream the throat, no mechanism has been identified yet to explain its displacement in the opposite direction. This study extends the existing compound flow theory and proposes a 1D model, including friction between the streams and the duct walls. The model captures the upstream and downstream displacements of the sonic section. Through an analytical investigation of the singularity at the sonic section, it is demonstrated that friction between the streams is the primary driver of upstream displacement. The 1D formulation is validated against axisymmetric Reynolds Averaged Navier-Stokes (RANS) simulations of a compound nozzle for various inlet pressure and geometries. The effect of friction is investigated using an inviscid simulation for the isentropic case and viscous simulations with both slip and no-slip conditions at the wall. The proposed extension accurately captures the displacement of the sonic section, offering a new tool for in-depth analysis and modeling of internal compound flows., Comment: 36 pages, 13 figures, in submission to Journal of Fluid Mechanics

The drag-out problem for small Reynolds numbers ($\rm Re$) admits the Landau-Levich-Derjaguin (LLD) solution for small capillary numbers ($\rm Ca$), and Derjaguin's solution for large $\rm Ca$. We investigate whether these solutions are absolutely or convectively unstable, solving the Orr-Sommerfeld eigenvalue problem. For Derjaguin's solution, we show that the LLD solution is convectively unstable for $\text{Ka}<17$ and absolutely unstable for $\text{Ka} \gtrsim 0.15 \,\text{Re}^{1.7}$ for $\text{Re} > 10$. For water ($\text{Ka}=3400$), the LLD solution is always convectively unstable. The absolute instability is observed only when the dip-coated film is additionally fed from above.

Vision-based grasping of unknown objects in unstructured environments is a key challenge for autonomous robotic manipulation. A practical grasp synthesis system is required to generate a diverse set of 6-DoF grasps from which a task-relevant grasp can be executed. Although generative models are suitable for learning such complex data distributions, existing models have limitations in grasp quality, long training times, and a lack of flexibility for task-specific generation. In this work, we present GraspLDM, a modular generative framework for 6-DoF grasp synthesis that uses diffusion models as priors in the latent space of a VAE. GraspLDM learns a generative model of object-centric $SE(3)$ grasp poses conditioned on point clouds. GraspLDM architecture enables us to train task-specific models efficiently by only re-training a small denoising network in the low-dimensional latent space, as opposed to existing models that need expensive re-training. Our framework provides robust and scalable models on both full and partial point clouds. GraspLDM models trained with simulation data transfer well to the real world without any further fine-tuning. Our models provide an 80% success rate for 80 grasp attempts of diverse test objects across two real-world robotic setups. We make our implementation available at https://github.com/kuldeepbrd1/graspldm .

The motion of a gas-liquid interface along a solid wall is influenced by the capillary forces resulting from the interface's shape and its interaction with the solid, where it forms a dynamic contact angle. Capillary models play a significant role in the management of cryogenic propellants in space, where surface tension dominates the behaviour of gas-liquid interfaces. Yet, most empirical models have been derived in configurations dominated by viscous forces. In this study, we experimentally investigate the wetting of a low-viscosity, highly wetting fluid in a reduced gravity environment. Our setup consisted of a transparent and diverging U-tube in which capillary forces sustain the liquid motion. Combining Particle Image Velocimetry (PIV) and high-speed backlighting visualization, the experimental campaign allowed for measuring the interface evolution and the velocity field within the liquid under varying gravity levels. This work reports on the preliminary results from the image velocimetry and shows that the velocity profile within the tube is close to parabolic until a short distance from the interface. Nevertheless, classic 1D models for capillary rise face difficulties reproducing the interface dynamics, suggesting that the treatment of the surface tension in these problems must be reviewed.

This work investigates the capillary rise dynamics of highly wetting liquids in a divergent U-tube in the microgravity conditions provided by 78th European Space Agency (ESA) parabolic flight. This configuration produces a capillary-driven channel flow. We use image recording in backlight illumination to characterize the interface dynamics and dynamic contact angle of HFE7200 and Di-Propylene Glycol (DPG). For the case of HF7200, we complement the interface measurements with Particle Tracking Velocimetry (PTV) to characterize the velocity fields underneath the moving meniscus. In the experiments with DPG, the liquid column reaches different heights within various experiments, and the measurements show a sharp reduction of the meniscus curvature when the contact line moves from a pre-wet to a dry substrate. In experiments with HFE7200, the interface always moves on a pre-wet surface. Yet a curvature reduction is observed due to the inertial forces on the underlying accelerating flow. The PTV measurements show that the distance from the interface within which the velocity profile adapts to the meniscus velocity shortens as the interface acceleration increases.

Digital twins promise to revolutionize engineering by offering new avenues for optimization, control, and predictive maintenance. We propose a novel framework for simultaneously training the digital twin of an engineering system and an associated control agent. The twin's training combines adjoint-based data assimilation and system identification methods, while the control agent's training merges model-based optimal control with model-free reinforcement learning. The control agent evolves along two independent paths: one driven by model-based optimal control and the other by reinforcement learning. The digital twin serves as a virtual environment for confrontation and indirect interaction, functioning as an "expert demonstrator." The best policy is selected for real-world interaction and cloned to the other path if training stagnates. We call this framework Reinforcement Twinning (RT). The framework is tested on three diverse engineering systems and control tasks: (1) controlling a wind turbine under varying wind speeds, (2) trajectory control of flapping-wing micro air vehicles (FWMAVs) facing wind gusts, and (3) mitigating thermal loads in managing cryogenic storage tanks. These test cases use simplified models with known ground truth closure laws. Results show that the adjoint-based digital twin training is highly sample-efficient, completing within a few iterations. For the control agent training, both model-based and model-free approaches benefit from their complementary learning experiences. The promising results pave the way for implementing the RT framework on real systems., Comment: submitted Journal of Computational Science

The thermal management of cryogenic storage tanks requires advanced control strategies to minimize the boil-off losses produced by heat leakages and sloshing-enhanced heat and mass transfer. This work presents a data-assimilation approach to calibrate a 0D thermodynamic model for cryogenic fuel tanks from data collected in real time from multiple tanks. The model combines energy and mass balance between three control volumes (the ullage vapor, the liquid, and the solid tank) with an Artificial Neural Network (ANN) for predicting the heat transfer coefficients from the current tank state. The proposed approach combines ideas from traditional data assimilation and multi-environment reinforcement learning, where an agent's training (model assimilation) is carried out simultaneously on multiple environments (systems). The real-time assimilation uses a mini-batch version of the Limited-memory Broyden-Fletcher-Goldfarb-Shanno with bounds (L-BFGS-B) and adjoint-based gradient computation for solving the underlying optimization problem. The approach is tested on synthetic datasets simulating multiple tanks undergoing different operation phases (pressurization, hold, long-term storage, and sloshing). The results show that the assimilation is robust against measurement noise and uses it to explore the parameter space further. Moreover, we show that sampling from multiple environments simultaneously accelerates the assimilation., Comment: 21 pages, 18 figures, preprint submitted to Energy

Nowadays, realistic simulation environments are essential to validate and build reliable robotic solutions. This is particularly true when using Reinforcement Learning (RL) based control policies. To this end, both robotics and RL developers need tools and workflows to create physically accurate simulations and synthetic datasets. Gazebo, MuJoCo, Webots, Pybullets or Isaac Sym are some of the many tools available to simulate robotic systems. Developing learning-based methods for space navigation is, due to the highly complex nature of the problem, an intensive data-driven process that requires highly parallelized simulations. When it comes to the control of spacecrafts, there is no easy to use simulation library designed for RL. We address this gap by harnessing the capabilities of NVIDIA Isaac Gym, where both physics simulation and the policy training reside on GPU. Building on this tool, we provide an open-source library enabling users to simulate thousands of parallel spacecrafts, that learn a set of maneuvering tasks, such as position, attitude, and velocity control. These tasks enable to validate complex space scenarios, such as trajectory optimization for landing, docking, rendezvous and more.

This work investigates the wetting dynamics of cryogenic fluids in inertia-dominated conditions. We experimentally characterized an oscillating gas-liquid interface of liquid nitrogen in a partially filled U-shaped quartz tube. The experiments were carried out in controlled cryogenic conditions, with interface oscillations produced by releasing the liquid column from an unbalanced position and having nitrogen vapor as the only ullage gas. During the experiments, the interface shape was tracked via image processing and used to fit a model from which the contact angle could be accurately determined. The results show that the dynamic contact angle evolution in advancing conditions is linearly linked to the Capillary number, with a slope depending on whether the interface moves over a dry or a pre-wet surface. However, the contact angle remains close to the one at equilibrium in receding conditions. To analyze the relation between contact angle and interface dynamics, we define an equivalent contact angle as the one that would make a spherical interface produce the same capillary pressure drop as the actual interface shape. The evolution of this equivalent contact angle proved to be independent of the evolution of the actual one, suggesting that the interface shape is not influenced by it. Finally, a theoretical analysis of the interface motion using a simplified model shows that viscous forces dominate the damping of the interface for small tube sizes, while gravity and inertial forces dominate the oscillating dynamics of the liquid column for larger tubes., Comment: 13 pages, 26 pdf figure files grouped in 8 figures

This investigation introduces a novel deep reinforcement learning-based suite to control floating platforms in both simulated and real-world environments. Floating platforms serve as versatile test-beds to emulate micro-gravity environments on Earth, useful to test autonomous navigation systems for space applications. Our approach addresses the system and environmental uncertainties in controlling such platforms by training policies capable of precise maneuvers amid dynamic and unpredictable conditions. Leveraging Deep Reinforcement Learning (DRL) techniques, our suite achieves robustness, adaptability, and good transferability from simulation to reality. Our deep reinforcement learning framework provides advantages such as fast training times, large-scale testing capabilities, rich visualization options, and ROS bindings for integration with real-world robotic systems. Being open access, our suite serves as a comprehensive platform for practitioners who want to replicate similar research in their own simulated environments and labs., Comment: Updated to the version accepted at IROS 2024. Minor revisions based on peer review

We investigate the dynamics of a gas jet impinging on a thin liquid film. This configuration is relevant to the jet-wiping process and is unstable. In particular, we complement previous works that focused on the wiping of liquids with low Kapitza numbers (highly viscous liquids) by numerically analyzing the wiping of liquids with much higher Kapitza numbers, more relevant to industrial processes. The simulations are carried out by combining Volume of Fluid (VOF) and Large Eddy Simulation (LES), and the dynamics of the gas-liquid interaction is analyzed using extended multiscale Proper Orthogonal Decomposition (emPOD). The resolution and flow details captured by the simulations are unprecedented. The results show that, despite the vastly different wiping conditions, the dynamics of the gas-liquid interaction is remarkably similar. This opens new avenues to the study and the scaling of the jet-wiping process.

Accurate global localization is crucial for autonomous navigation and planning. To this end, various GPS-aided Visual-Inertial Odometry (GPS-VIO) fusion algorithms are proposed in the literature. This paper presents a novel GPS-VIO system that is able to significantly benefit from the online calibration of the rotational extrinsic parameter between the GPS reference frame and the VIO reference frame. The behind reason is this parameter is observable. This paper provides novel proof through nonlinear observability analysis. We also evaluate the proposed algorithm extensively on diverse platforms, including flying UAV and driving vehicle. The experimental results support the observability analysis and show increased localization accuracy in comparison to state-of-the-art (SOTA) tightly-coupled algorithms., Comment: Accepted by ICRA 2024

Developing algorithms for extra-terrestrial robotic exploration has always been challenging. Along with the complexity associated with these environments, one of the main issues remains the evaluation of said algorithms. With the regained interest in lunar exploration, there is also a demand for quality simulators that will enable the development of lunar robots. % In this paper, we explain how we built a Lunar simulator based on Isaac Sim, Nvidia's robotic simulator. In this paper, we propose Omniverse Lunar Robotic-Sim (OmniLRS) that is a photorealistic Lunar simulator based on Nvidia's robotic simulator. This simulation provides fast procedural environment generation, multi-robot capabilities, along with synthetic data pipeline for machine-learning applications. It comes with ROS1 and ROS2 bindings to control not only the robots, but also the environments. This work also performs sim-to-real rock instance segmentation to show the effectiveness of our simulator for image-based perception. Trained on our synthetic data, a yolov8 model achieves performance close to a model trained on real-world data, with 5% performance gap. When finetuned with real data, the model achieves 14% higher average precision than the model trained on real-world data, demonstrating our simulator's photorealism.% to realize sim-to-real. The code is fully open-source, accessible here: https://github.com/AntoineRichard/LunarSim, and comes with demonstrations., Comment: 7 pages, 4 figures

This paper introduces a novel GPS-aided visual-wheel odometry (GPS-VWO) for ground robots. The state estimation algorithm tightly fuses visual, wheeled encoder and GPS measurements in the way of Multi-State Constraint Kalman Filter (MSCKF). To avoid accumulating calibration errors over time, the proposed algorithm calculates the extrinsic rotation parameter between the GPS global coordinate frame and the VWO reference frame online as part of the estimation process. The convergence of this extrinsic parameter is guaranteed by the observability analysis and verified by using real-world visual and wheel encoder measurements as well as simulated GPS measurements. Moreover, a novel theoretical finding is presented that the variance of unobservable state could converge to zero for specific Kalman filter system. We evaluate the proposed system extensively in large-scale urban driving scenarios. The results demonstrate that better accuracy than GPS is achieved through the fusion of GPS and VWO. The comparison between extrinsic parameter calibration and non-calibration shows significant improvement in localization accuracy thanks to the online calibration., Comment: Accepted by IEEE ITSC 2023

Mobility on asteroids by multi-limbed climbing robots is expected to achieve our exploration goals in such challenging environments. We propose a mobility strategy to improve the locomotion safety of climbing robots in such harsh environments that picture extremely low gravity and highly uneven terrain. Our method plans the gait by decoupling the base and limbs' movements and adjusting the main body pose to avoid ground collisions. The proposed approach includes a motion planning that reduces the reactions generated by the robot's movement by optimizing the swinging trajectory and distributing the momentum. Lower motion reactions decrease the pulling forces on the grippers, avoiding the slippage and flotation of the robot. Dynamic simulations and experiments demonstrate that the proposed method could improve the robot's mobility on the surface of asteroids., Comment: Paper accepted for presentation at the CLAWAR 2023 (26th International Conference on Climbing and Walking Robots and the Support Technologies for Mobile Machines) (Updated references formatting)

Paints and coatings often feature interfacial defects due to disturbances during the deposition process which, if they persist until solidification, worsen the product quality. In this article, we investigate the stability of a thin liquid film dragged by a vertical substrate moving against gravity, a flow configuration found in a variety of coating processes. The receptivity of the liquid film to three-dimensional disturbances is discussed with Direct Numerical Simulations (DNS), an in-house non-linear Integral Boundary Layer (IBL) film model, and Linear Stability Analysis (LSA). The thin film model, successfully validated with the DNS computations, implements a pseudo-spectral approach for the capillary terms that allows for investigating non-periodic surface tension dominated flows. The combination of these numerical tools allows for describing the mechanisms of capillary and non-linear damping, and identifying the instability threshold of the coating processes. The results show that transverse modulations can be beneficial for the damping of two-dimensional waves within the range of operational conditions considered in this study, typical of air-knife and slot-die coating.

The Sand Hypoplastic (SH) constitutive law by von Wolffersdorff (1996) is a interesting hypoplastic model for soil mechanics. This model includes eight parameters, usually calibrated using the oedometric (OE) and the drained isotropically consolidated triaxial tests (CD). However, previous studies show that the SH model calibration in the CD test has conflicting requirements in predicting the evolution of stresses and strains. In this work, we study the SH model calibration over a wide range of testing conditions using 12 OE and 25 CD tests by Wichtmann and Triantafyllidis (2016) on the Karlsruhe sands. The parameter space is extensively explored via Genetic Algorithm Optimization (GA) using the recently developed open-source software GA-cal (available at \url{https://github.com/FraJoMen/GA-cal}). This exploration allowed us to study the SH model's predictive limits and to identify, using a multi-objective analysis, the main parameters governing the compromise between the accurate prediction of stresses versus strain in the CD tests.

Nonisothermal liquid sloshing in partially filled reservoirs can significantly enhance heat and mass transfer between liquid and ullage gasses. This can result in large temperature and pressure fluctuations, producing thrust oscillations in spacecraft and challenging thermal management control systems. This work presents an experimental characterization of the thermodynamic evolution of a cylindrical reservoir undergoing sloshing-induced thermal de-stratification. We use a 0D model-based inverse method to retrieve the heat and mass transfer coefficients in planar and swirl sloshing conditions from the temperature and pressure measurements in the liquid and the ullage gas. The experiments were carried out in the SHAKESPEARE shaking table of the von Karman Institute in a cuboid quartz cell with a cylindrical cut-out of 80 mm diameter in the centre, filled up to 60mm with the cryogenic replacement fluid HFE-7200. A thermal stratification with 25 K difference between the ullage gas and liquid was set as the initial conditions. A pressure drop of 90% in the ullage gas was documented in swirling conditions. Despite its simplicity, the model could predict the system's thermodynamic evolution once the proper transfer coefficients were derived., Comment: 18 pages, 14 figures, 4 tables

Flapping Wing Micro Air Vehicles (FWMAV) are highly manoeuvrable, bio-inspired drones that can assist in surveys and rescue missions. Flapping wings generate various unsteady lift enhancement mechanisms challenging the derivation of reduced models to predict instantaneous aerodynamic performance. In this work, we propose a robust CFD data-driven, quasi-steady (QS) Reduced Order Model (ROM) to predict the lift and drag coefficients within a flapping cycle. The model is derived for a rigid ellipsoid wing with different parameterized kinematics in hovering conditions. The proposed ROM is built via a two-stage regression. The first stage, defined as `in-cycle' (IC), computes the parameters of a regression linking the aerodynamic coefficients to the instantaneous wing state. The second stage, `out-of-cycle' (OOC), links the IC weights to the flapping features that define the flapping motion. The training and test dataset were generated via high-fidelity simulations using the overset method, spanning a wide range of Reynolds numbers and flapping kinematics. The two-stage regressor combines Ridge regression and Gaussian Process (GP) regression to provide estimates of the model uncertainties. The proposed ROM shows accurate aerodynamic predictions for widely varying kinematics. The model performs best for smooth kinematics that generate a stable Leading Edge Vortex (LEV). Remarkably accurate predictions are also observed in dynamic scenarios where the LEV is partially shed, the non-circulatory forces are considerable, and the wing encounters its own wake., Comment: submitted to Physics of Fluids

The calibration of an advanced constitutive law for soil is a challenging task. This work describes GA-cal, a Fortran software for automatically calibrating constitutive laws using Genetic Algorithms (GA) optimization. The proposed approach sets the calibration problem as a regression, and the GA optimization is used to adjust the model parameters so that a numerical model matches experimental data. This document provides a user guide and a simple tutorial. We showcase GA-cal on the calibration of the Sand Hypoplastic law proposed by von Wolffersdorff, with the oedometer and triaxial drained test data. The implemented subroutines can be easily extended to solve other regression or optimization problems, including different tests and constitutive models. The source code and the presented tutorial are freely available at \url{https://github.com/FraJoMen/GA-cal}.

In robotic manipulation, end-effector compliance is an essential precondition for performing contact-rich tasks, such as machining, assembly, and human-robot interaction. Most robotic arms are position-controlled stiff systems at a hardware level. Thus, adding compliance becomes essential. Compliance in those systems has been recently achieved using Forward dynamics compliance control (FDCC), which, owing to its virtual forward dynamics model, can be implemented on both position and velocity-controlled robots. This paper evaluates the choice of control interface (and hence the control domain), which, although considered trivial, is essential due to differences in their characteristics. In some cases, the choice is restricted to the available hardware interface. However, given the option to choose, the velocity-based control interface makes a better candidate for compliance control because of smoother compliant behaviour, reduced interaction forces, and work done. To prove these points, in this paper FDCC is evaluated on the UR10e six-DOF manipulator with velocity and position control modes. The evaluation is based on force-control benchmarking metrics using 3D-printed artefacts. Real experiments favour the choice of velocity control over position control., Comment: Submitted to RA-L on 15th Sept 2022, for associated video see: https://www.youtube.com/watch?v=iIFscA-CHRU