Searcharxiv⌕ Search

arXiv subjects

Ricardo Garcia

Publications and source records attributed to Ricardo Garcia.

17 recordsLinked to original sources
arXiv2026

Guardian: Detecting Robotic Planning and Execution Errors with Vision-Language Models

Robust robotic manipulation requires reliable failure detection and recovery. Although recent Vision-Language Models (VLMs) show promise in robot failure detection, their generalization is severely limited by the scarcity and narrow coverage of failure data. To address this bottleneck, we propose an automatic framework for generating diverse robotic planning and execution failures across both simulated and real-world environments. Our approach perturbs successful manipulation trajectories to synthesize failures that reflect realistic failure distributions, and leverages VLMs to produce structured step-by-step reasoning traces. This yields GuardianFail-36k, a large-scale failure reasoning dataset built upon the RLBench simulator and the BridgeDataV2 real-robot dataset. Using GuardianFail-36k, we train Guardian, a multi-view reasoning VLM for unified planning and execution verification. Guardian achieves state-of-the-art performance on three unseen real-world benchmarks: RoboFail, RoboVQA, and our newly introduced UR5-Fail. When integrated with a state-of-the-art LLM-based manipulation policy, it consistently boosts task success rates in both simulation and real-world deployment. These results demonstrate that scaling high-quality failure reasoning data is critical for improving generalization in robotic failure detection. Code, Data, and Models are available at https://www.di.ens.fr/willow/research/guardian/.

cs.RO↗
arXiv2025

Linearization Scheme of Shallow Water Equations for Quantum Algorithms

Computational fluid dynamics lies at the heart of many issues in science and engineering, but solving the associated partial differential equations remains computationally demanding. With the rise of quantum computing, new approaches have emerged to address these challenges. In this work, we investigate the potential of quantum algorithms for solving the shallow water equations, which are, for example, used to model tsunami dynamics. By extending a linearization scheme previously developed in [Phys. Rev. Research 7, 013036 (2025)] for the Navier-Stokes equations, we create a mapping from the nonlinear shallow water equation to a linear system of equations, which, in principle, can be solved exponentially faster on a quantum device than on a classical computer. To validate our approach, we compare its results to an analytical solution and benchmark its dependence on key parameters. Additionally, we implement a quantum linear system solver based on quantum singular value transformation and study its performance in connection to our mapping. Our results demonstrate the potential of applying quantum algorithms to fluid dynamics problems and highlight necessary considerations for future developments.

quant-ph↗
arXiv2025

Gondola: Grounded Vision Language Planning for Generalizable Robotic Manipulation

Robotic manipulation faces a significant challenge in generalizing across unseen objects, environments and tasks specified by diverse language instructions. To improve generalization capabilities, recent research has incorporated large language models (LLMs) for planning and action execution. While promising, these methods often fall short in generating grounded plans in visual environments. Although efforts have been made to perform visual instructional tuning on LLMs for robotic manipulation, existing methods are typically constrained by single-view image input and struggle with precise object grounding. In this work, we introduce Gondola, a novel grounded vision-language planning model based on LLMs for generalizable robotic manipulation. Gondola takes multi-view images and history plans to produce the next action plan with interleaved texts and segmentation masks of target objects and locations. To support the training of Gondola, we construct three types of datasets using the RLBench simulator, namely robot grounded planning, multi-view referring expression and pseudo long-horizon task datasets. Gondola outperforms the state-of-the-art LLM-based method across all four generalization levels of the GemBench dataset, including novel placements, rigid objects, articulated objects and long-horizon tasks.

cs.RO↗
arXiv2025

Correlative angstrom-scale microscopy and spectroscopy of graphite-water interfaces

Water at solid surfaces is key for many processes ranging from biological signal transduction to membrane separation and renewable energy conversion. However, under realistic conditions, which often include environmental and surface charge variations, the interfacial water structure remains elusive. Here we overcome this limit by combining three-dimensional atomic force microscopy and interface-sensitive Raman spectroscopy to characterize the graphite-water interfacial structure in situ. Through correlative analysis of the spatial liquid density maps and vibrational peaks within ~2 nm of the graphite surface, we find the existence of two interfacial configurations at open circuit potential, a transient state where pristine water exhibits strong hydrogen bond (HB) breaking effects, and a steady state with hydrocarbons dominating the interface and weak HB breaking in the surrounding water. At sufficiently negative potentials, both states transition into a stable structure featuring pristine water with a broader distribution of HB configurations. Our three-state model resolves many long-standing controversies on interfacial water structure.

physics.chem-ph↗
arXiv2025

Towards Generalizable Vision-Language Robotic Manipulation: A Benchmark and LLM-guided 3D Policy

Generalizing language-conditioned robotic policies to new tasks remains a significant challenge, hampered by the lack of suitable simulation benchmarks. In this paper, we address this gap by introducing GemBench, a novel benchmark to assess generalization capabilities of vision-language robotic manipulation policies. GemBench incorporates seven general action primitives and four levels of generalization, spanning novel placements, rigid and articulated objects, and complex long-horizon tasks. We evaluate state-of-the-art approaches on GemBench and also introduce a new method. Our approach 3D-LOTUS leverages rich 3D information for action prediction conditioned on language. While 3D-LOTUS excels in both efficiency and performance on seen tasks, it struggles with novel tasks. To address this, we present 3D-LOTUS++, a framework that integrates 3D-LOTUS's motion planning capabilities with the task planning capabilities of LLMs and the object grounding accuracy of VLMs. 3D-LOTUS++ achieves state-of-the-art performance on novel tasks of GemBench, setting a new standard for generalization in robotic manipulation. The benchmark, codes and trained models are available at https://www.di.ens.fr/willow/research/gembench/.

cs.RO↗
arXiv2024

SUGAR: Pre-training 3D Visual Representations for Robotics

Learning generalizable visual representations from Internet data has yielded promising results for robotics. Yet, prevailing approaches focus on pre-training 2D representations, being sub-optimal to deal with occlusions and accurately localize objects in complex 3D scenes. Meanwhile, 3D representation learning has been limited to single-object understanding. To address these limitations, we introduce a novel 3D pre-training framework for robotics named SUGAR that captures semantic, geometric and affordance properties of objects through 3D point clouds. We underscore the importance of cluttered scenes in 3D representation learning, and automatically construct a multi-object dataset benefiting from cost-free supervision in simulation. SUGAR employs a versatile transformer-based model to jointly address five pre-training tasks, namely cross-modal knowledge distillation for semantic learning, masked point modeling to understand geometry structures, grasping pose synthesis for object affordance, 3D instance segmentation and referring expression grounding to analyze cluttered scenes. We evaluate our learned representation on three robotic-related tasks, namely, zero-shot 3D object recognition, referring expression grounding, and language-driven robotic manipulation. Experimental results show that SUGAR's 3D representation outperforms state-of-the-art 2D and 3D representations.

cs.CV↗
arXiv2023

PolarNet: 3D Point Clouds for Language-Guided Robotic Manipulation

The ability for robots to comprehend and execute manipulation tasks based on natural language instructions is a long-term goal in robotics. The dominant approaches for language-guided manipulation use 2D image representations, which face difficulties in combining multi-view cameras and inferring precise 3D positions and relationships. To address these limitations, we propose a 3D point cloud based policy called PolarNet for language-guided manipulation. It leverages carefully designed point cloud inputs, efficient point cloud encoders, and multimodal transformers to learn 3D point cloud representations and integrate them with language instructions for action prediction. PolarNet is shown to be effective and data efficient in a variety of experiments conducted on the RLBench benchmark. It outperforms state-of-the-art 2D and 3D approaches in both single-task and multi-task learning. It also achieves promising results on a real robot.

cs.RO↗
arXiv2023

Robust Visual Sim-to-Real Transfer for Robotic Manipulation

Learning visuomotor policies in simulation is much safer and cheaper than in the real world. However, due to discrepancies between the simulated and real data, simulator-trained policies often fail when transferred to real robots. One common approach to bridge the visual sim-to-real domain gap is domain randomization (DR). While previous work mainly evaluates DR for disembodied tasks, such as pose estimation and object detection, here we systematically explore visual domain randomization methods and benchmark them on a rich set of challenging robotic manipulation tasks. In particular, we propose an off-line proxy task of cube localization to select DR parameters for texture randomization, lighting randomization, variations of object colors and camera parameters. Notably, we demonstrate that DR parameters have similar impact on our off-line proxy task and on-line policies. We, hence, use off-line optimized DR parameters to train visuomotor policies in simulation and directly apply such policies to a real robot. Our approach achieves 93% success rate on average when tested on a diverse set of challenging manipulation tasks. Moreover, we evaluate the robustness of policies to visual variations in real scenes and show that our simulator-trained policies outperform policies learned using real but limited data. Code, simulation environment, real robot datasets and trained models are available at https://www.di.ens.fr/willow/research/robust_s2r/.

cs.RO↗
arXiv2022

Instruction-driven history-aware policies for robotic manipulations

In human environments, robots are expected to accomplish a variety of manipulation tasks given simple natural language instructions. Yet, robotic manipulation is extremely challenging as it requires fine-grained motor control, long-term memory as well as generalization to previously unseen tasks and environments. To address these challenges, we propose a unified transformer-based approach that takes into account multiple inputs. In particular, our transformer architecture integrates (i) natural language instructions and (ii) multi-view scene observations while (iii) keeping track of the full history of observations and actions. Such an approach enables learning dependencies between history and instructions and improves manipulation precision using multiple views. We evaluate our method on the challenging RLBench benchmark and on a real-world robot. Notably, our approach scales to 74 diverse RLBench tasks and outperforms the state of the art. We also address instruction-conditioned tasks and demonstrate excellent generalization to previously unseen variations.

cs.RO↗
arXiv2021

Segmenter: Transformer for Semantic Segmentation

Image segmentation is often ambiguous at the level of individual image patches and requires contextual information to reach label consensus. In this paper we introduce Segmenter, a transformer model for semantic segmentation. In contrast to convolution-based methods, our approach allows to model global context already at the first layer and throughout the network. We build on the recent Vision Transformer (ViT) and extend it to semantic segmentation. To do so, we rely on the output embeddings corresponding to image patches and obtain class labels from these embeddings with a point-wise linear decoder or a mask transformer decoder. We leverage models pre-trained for image classification and show that we can fine-tune them on moderate sized datasets available for semantic segmentation. The linear decoder allows to obtain excellent results already, but the performance can be further improved by a mask transformer generating class masks. We conduct an extensive ablation study to show the impact of the different parameters, in particular the performance is better for large models and small patch sizes. Segmenter attains excellent results for semantic segmentation. It outperforms the state of the art on both ADE20K and Pascal Context datasets and is competitive on Cityscapes.

cs.CV↗
arXiv2020

Learning Obstacle Representations for Neural Motion Planning

Motion planning and obstacle avoidance is a key challenge in robotics applications. While previous work succeeds to provide excellent solutions for known environments, sensor-based motion planning in new and dynamic environments remains difficult. In this work we address sensor-based motion planning from a learning perspective. Motivated by recent advances in visual recognition, we argue the importance of learning appropriate representations for motion planning. We propose a new obstacle representation based on the PointNet architecture and train it jointly with policies for obstacle avoidance. We experimentally evaluate our approach for rigid body motion planning in challenging environments and demonstrate significant improvements of the state of the art in terms of accuracy and efficiency.

cs.RO↗
arXiv2020

Flexural Fatigue Life of Woven Carbon/Vinyl Ester Composites under Sea Water Saturation

The adverse effects of sea water environment on the fatigue life of woven carbon fiber/vinyl ester composites are established at room temperature in view of long-term survivability of offshore structures. It is observed that the influence of sea water saturation on the fatigue life is more pronounced when the maximum cyclic displacement approaches maximum quasi-static deflection, that is, the reduction in the number of cycles to failure are comparable between dry and sea water saturated samples at lower strain ranges (~37% at 0.46% strain), but are drastically different at higher strain ranges (~90% at 0.62% strain). Key damage modes that manifest during the fatigue loading is also identified, and a non-linear model is established for predicting low cycle fatigue life of these composites in dry and sea water saturated conditions.

physics.app-ph↗
arXiv2017

Equally Spaced Levels with T=1 two-body Matrix elements

If one examines two-body matrix elements from experiment one notices that not only J=0 T=1 lies low but also J=1T=0 and J=J_{max} =2j T=0. It is sometimes thought that one needs both T=1 and T=0 two-body matrix elements to get equally spaced spectra of even I states i.e. vibraitonal spectra. We here attempt to get equally spaced levels with only those that have T=1 (even J).As an example we perform single-j calculations (f_{7/2}) in ^{44} T and ^{46} Ti. We then shift gears and decide to play around with the input two particle matrix, elements (not worrying about experiment) to generate interesting spectra e.g. rotaional spectra with and then without T=0 two-body matrix element. Of special interest is a new partial dynamical symmetry found when the "123", the "1234" etc.,interactions are used.

nucl-th↗
arXiv2016

Partial Conservation Law in a Schematic Single j Shell Model

We report the discovery of a partial conservation law obeyed by a schematic Hamiltonian of two protons and two neutrons in a j shell. In our Hamiltonian the interaction matrix element of two nucleons with combined angular momentum J is linear in J for even J and constant for odd J. It turns out that in some stationary states the sum J_p + J_n of the angular momenta J_p and J_n of the proton and neutron pairs is conserved. The energies of these states are given by a linear function of J_p + J_n. The systematics of their occurrence is described and explained.

nucl-th↗
arXiv2015

Advanced scanning probe lithography

The nanoscale control afforded by scanning probe microscopes has prompted the development of a wide variety of scanning probe-based patterning methods. Some of these methods have demonstrated a high degree of robustness and patterning capabilities that are unmatched by other lithographic techniques. However, the limited throughput of scanning probe lithography has prevented their exploitation in technological applications. Here, we review the fundamentals of scanning probe lithography and its use in materials science and nanotechnology. We focus on the methods and processes that offer genuinely lithography capabilities such as those based on thermal effects, chemical reactions and voltage-induced processes.

cond-mat.mtrl-sci↗
arXiv1997

Hodge numbers attached to a polynomial map

Given a polynomial map $f:\Bbb C^{n+1}\to\Bbb C$, one can attach to it a geometrical variation of mixed Hodge structures (MHS) which gives rise to a limit MHS. The equivariant Hodge numbers of this MHS are analytical invariants of the polynomial map and reflect its asymptotic behaviour. In this paper we compute them for a class of generic polynomials, in terms of Hodge numbers attached to isolated hypersurface singularities and Hodge numbers of cyclic coverings of projective space branched along a hypersurface.

alg-geom↗