IROS 2024 论文列表

Indoor robot localization is important for the realization of autonomous service robots. Various studies have been conducted on "indoor GPS" measurements using wireless distance sensors such as ultrasonic beacons. However, when these beacons encounter non-line-of-sight (NLoS) conditions due to obstacles, accurate distance measurements become challenging because of multipath and other effects. In this study, we propose a method for simultaneously estimating a robot’s position and distance to reflective surfaces in an environment using wireless distance sensors. The proposed method can estimate not only the robot’s position but also the reflection of the beacon signal. First, the wheel odometry of the robot is assumed to be the initial value, and the measured distance from the beacon to the robot is used as a factor to construct the factor graph. Second, the distance to the reflective surface of the beacon signal, which is parallel to the robot’s movement plane, was estimated from the robot position sequence using the GMM and used as a noise model in the factor graph. Finally, the method is evaluated by acquiring data in a real environment with obstacles. Compared with a method that does not consider reflection paths, this method demonstrated improved accuracy and effectiveness.

We propose an incremental, modular, and extensible method for learning robotic manipulation tasks using a limited number of demonstrations provided in Virtual Reality, while assuming minimal prior information about the objects to be manipulated. The developed framework enables an incremental training process in which the operator first demonstrates specialized tasks to the robotic system, and subsequently more complex tasks, exploiting the skills learned during the previous phases. We illustrate and discuss the method at work considering picking tasks performed by manipulators equipped with multi-fingered sensorized hands. The experimental evaluation highlights the feasibility and advantage of the proposed method, particularly in terms of modularity, low number of demonstrations, and reliability of the trained system.

Manipulation tasks often require a high degree of dexterity, typically necessitating grippers with multiple degrees of freedom (DoF). While a robotic hand equipped with multiple fingers can execute precise and intricate manipulation tasks, the inherent redundancy stemming from its extensive DoF often adds unnecessary complexity. In this paper, we introduce the design of a tactile sensor-equipped gripper with two fingers and five DoF. We present a novel design integrating a GelSight tactile sensor, enhancing sensing capabilities and enabling finer control during specific manipulation tasks. To evaluate the gripper’s performance, we conduct experiments involving two challenging tasks: 1) retrieving, singularizing, and classification of various objects embedded in granular media, and 2) executing scooping manipulations of credit cards in confined environments to achieve precise insertion. Our results demonstrate the efficiency of the proposed approach, with a high success rate for singulation and classification tasks, particularly for spherical objects at high as 94.3%, and a 100% success rate for scooping and inserting credit cards.

Multiple approaches have already been proposed to mimic real driver behaviors in simulation. This article proposes a new one, based solely on the exploration of undisturbed observation of intersections. From them, the behavior profiles for each macro-maneuver will be discovered. Using the macro-maneuvers already identified in previous works, a comparison method between trajectories with different lengths using an Extended Kalman Filter (EKF) is proposed, which combined with an Expectation-Maximization (EM) inspired method, defines the different clusters that represent the behaviors observed. This is also paired with a Kullback-Liebler divergent (KL) criteria to define when the clusters need to be split or merged. Finally, the behaviors for each macro-maneuver are determined by each cluster discovered, without using any map information about the environment and being dynamically consistent with vehicle motion. By observation it becomes clear that the two main factors for driver’s behavior are their assertiveness and interaction with other road users.

Robustness against Out-of-Distribution (OoD) samples is a key performance indicator of a trajectory prediction model. However, the development and ranking of state-of-the-art (SotA) models are driven by their In-Distribution (ID) performance on individual competition datasets. We present an OoD testing protocol that homogenizes datasets and prediction tasks across two large-scale motion datasets. We introduce a novel prediction algorithm based on polynomial representations for agent trajectory and road geometry on both the input and output sides of the model. With a much smaller model size, training effort, and inference time, we reach near SotA performance for ID testing and significantly improve robustness in OoD testing. Within our OoD testing protocol, we further study two augmentation strategies of SotA models and their effects on model generalization. Highlighting the contrast between ID and OoD performance, we suggest adding OoD testing to the evaluation criteria of trajectory prediction models.

Many robotic and biological systems exploit their morphology’s interaction with the environment to become more adaptable, more energy efficient, and to simplify their control. The principles of morphological computation (MC) have been increasingly studied in recent years and researchers have investigated theoretical approaches to quantify the contribution of MC for a variety of robotic systems using only simulated models. In this work, we quantify MC in a physical robotic system, utilizing position-controlled legs with two degrees of freedom in two designs of different elastic compliance, on a bespoke test rig to execute a walking gait. The contribution of morphology was estimated by applying a theoretical model at various stages within the gait cycle to quantify the MC. The relationships between MC and the ground reaction forces and actuator energy consumption are analyzed. The results indicating that increasing the compliance in the leg morphology increases the mean MC value (7.70±1.49) relative to a traditional leg design (5.03±2.27). Periods of high MC were found to occur during the swing phase of the walking gait and reduced during the stance phase with ground reaction forces, which correlates with the findings of prior theoretical studies of MC in hopping gaits. The benefits of refining the leg morphology for higher MC is demonstrated by the measurements of cost of transport (COT), where the leg with the higher mean MC of 7.70 has a lower mean COT of 102.8 compared to the other leg’s mean COT of 153.8. The results demonstrate how real-world measurements of MC may help design the morphology of improved robotics systems.

Industrial robots have increased payload, repeatability, and reach compared to collaborative robots, however, they have a fixed position controller and low intrinsic admittance. This makes realizing safe contact challenging due to large contact force overshoots in contact transitions and contact instability when the environment and robot dynamics are coupled. To improve safe contact on industrial robots, we propose an admittance controller with inverse model compensation, designed and implemented outside the position controller. By including both the inner loop and outer loop dynamics in its design, the proposed method achieves expanded admittance in terms of increasing both gain and cutoff frequency of the desired admittance. Results from theoretical analyses and experiments on a commercial industrial robot show that the proposed method improves rendering of the desired admittance while maintaining contact stability. We further validate this by conducting actual assembly tasks of plug insertion with fine positioning, switch insertion onto the rail, and colliding the robot end effector with random objects and surfaces, as seen at https://youtu.be/8XfkdHEdWDs.

Optical sensors, such as the Far Infrared (FIR) sensor, have demonstrated advantages over traditional imaging. For example, 3D reconstruction in the FIR field captures the heat distribution of a scene that is invisible to RGB, aiding various applications like gas leak detection. However, less texture information and challenges in acquiring FIR frames hinder the reconstruction process. Given that implicit neural representations (INRs) can integrate geometric information across different sensors, we propose Implicit Neural Fusion (INF) of RGB and FIR for 3D reconstruction of invisible scenes in the FIR field. Our method first obtains a neural density field of objects from RGB frames. Then, with the trained object density field, a separate neural density field of gases is optimized using limited view inputs of FIR frames. Our method not only demonstrates outstanding reconstruction quality in the FIR field through extensive experiments but also can isolate the geometric information of the invisible, offering a new dimension of scene understanding.

This paper introduces the TactiMesh Teleoperator Interface (TTI), a novel predictive visual and haptic system designed explicitly for human-in-the-loop robot control using a head-mounted display (HMD). By employing simultaneous localization and mapping (SLAM) in tandem with a space carving method (CARV), TTI creates a real-time 3D surface mesh of remote environments from an RGB camera mounted on a Barrett WAM arm. The generated mesh is integrated into a physics simulator, featuring a digital twin of the WAM robot arm to create a virtual environment. In this virtual environment, TTI provides haptic feedback directly in response to the operator’s movements, eliminating the problem with delayed response from the haptic follower robot. Furthermore, texturing the 3D mesh with keyframes from SLAM allows the operator to control the viewpoint of their Head Mounted Display (HMD) independently of the arm-mounted robot camera, giving a better visual immersion and improving manipulation speed. Incorporating predictive visual and haptic feedback significantly improves tele-operation in applications such as search and rescue, inspection, and remote maintenance.

Robotic cutting, a crucial task in applications such as disassembly and decommissioning, faces challenges due to uncertainties in real-world environments. This paper presents a novel approach to enhance sim-to-real transfer of robotic cutting policies, leveraging a hybrid method integrating Gaussian process (GP) regression to model disturbance forces encountered during cutting tasks. By learning from a limited number of real-world trials, our method captures residual process dynamics, enabling effective adaptation to diverse materials without the need for fine-tuning on physical robots. Key to our approach is the utilisation of imitation learning, where expert actions in the uncorrected simulation are paired with GP-corrected observations. This pairing aligns action distributions between simulated and real-world domains, facilitating robust policy transfer. We illustrate the efficacy of our method through real world cutting trials in autonomously adapting to diverse material properties; our method surpasses re-training, while providing similar benefits to fine-tuning in real-world cutting scenarios. Notably, policies transferred using our approach exhibit enhanced resilience to noise and disturbances, while maintaining fidelity to expert behaviours from the source domain.

Reinforcement Learning (RL) has been widely used to solve tasks where the environment consistently provides a dense reward value. However, in real-world scenarios, rewards can often be poorly defined or sparse. Auxiliary signals are indispensable for discovering efficient exploration strategies and aiding the learning process. In this work, inspired by intrinsic motivation theory, we postulate that the intrinsic stimuli of novelty and surprise can assist in improving exploration in complex, sparsely rewarded environments. We introduce a novel sample-efficient method able to learn directly from pixels, an image-based extension of TD3 with an autoencoder called NaSA-TD3. The experiments demonstrate that NaSA-TD3 is easy to train and an efficient method for tackling complex continuous-control robotic tasks, both in simulated environments and real-world settings. NaSA-TD3 outperforms existing state-of-the-art RL image-based methods in terms of final performance without requiring pre-trained models or human demonstrations.

Micro-structured surfaces possess useful properties such as friction modification, anti-fouling, and hydrophobicity. However, manufacturing these surfaces in an affordable, scalable, and efficient manner remains challenging. Standard coverage methods for surface patterning require precise placement of micro-scale features over meter-scale surfaces with expensive tooling for support. In this work, we address the scalability challenge in surface patterning by designing a mobile robot with a credit-card-sized footprint to generate micro-scale divots using a modulated tool tip. We provide a control architecture with a target feature density to specify surface coverage, eliminating the dependence on individual indentation locations. Our robot produces high-fidelity surface patterns and achieves automatic coverage of a surface from sophisticated target images. We validate an exemplary application of such micro-structured surfaces by controlling the friction coefficients at different locations according to the density of indentations. These results show the potential for compact robots to perform scalable manufacturing of functional surfaces, switching the focus from precision machines to small-footprint devices tasked with matching only the density of features.

This paper presents a novel, nonlinearity robust identification method for deriving the inertia matrix of multi-DOF Flexible Joint Robots (FJR), utilizing resonance and anti-resonance frequencies in the Frequency Response Functions (FRF). Our proposed method overcomes the limitations of conventional approaches, which are susceptible to mechanical nonlinearities, leading to inaccurate models. By leveraging frequency domain techniques, our approach effectively mitigates the influence of nonlinear characteristics, providing a more accurate and reliable means of robot control. Furturmore, the paper highlights the benefits of frequency domain system identification, including nonlinear robustness and the ability to decompose the flexible joint into motor and load components. Finally, a novel sequential excitation algorithm is proposed to obtain the inertia matrix of a multi-DOF robot manipulator without relying on complex theories or optimizations. The effectiveness of the proposed algorithm is verified through simulation and experiment.

This study addresses the absence of an identification framework to quantify a comprehensive dynamic model of human and anthropomorphic tendon-driven fingers, which is necessary to investigate the physiological properties of human fingers and improve the control of robotic hands. First, a generalized dynamic model was formulated, which takes into account the inherent properties of such a mechanical system. This includes rigid-body dynamics, coupling matrix, joint viscoelasticity, and tendon friction. Then, we propose a methodology comprising a series of experiments, for step-wise identification and validation of this dynamic model. Moreover, an experimental setup was designed and constructed that features actuation modules and peripheral sensors to facilitate the identification process. To verify the proposed methodology, a 3D-printed robotic finger based on the index finger design of the Dexmart hand was developed, and the proposed experiments were executed to identify and validate its dynamic model. This study could be extended to explore the identification of cadaver hands, aiming for a consistent dataset from a single cadaver specimen to improve the development of musculoskeletal hand models.

Current visual navigation systems often treat the environment as static, lacking the ability to adaptively interact with obstacles. This limitation leads to navigation failure when encountering unavoidable obstructions. In response, we introduce IN-Sight, a novel approach to self-supervised path planning, enabling more effective navigation strategies through interaction with obstacles. Utilizing RGB-D observations, IN-Sight calculates traversability scores and incorporates them into a semantic map, facilitating long-range path planning in complex, maze-like environments. To precisely navigate around obstacles, IN-Sight employs a local planner, trained imperatively on a differentiable costmap using representation learning techniques. The entire framework undergoes end-to-end training within the state-of-the-art photorealistic Intel SPEAR Simulator. We validate the effectiveness of IN-Sight through extensive benchmarking in a variety of simulated scenarios and ablation studies. Moreover, we demonstrate the system’s real-world applicability with zero-shot sim-to-real transfer, deploying our planner on the legged robot platform ANYmal, showcasing its practical potential for interactive navigation in real environments.

This study is twofold. First, a comprehensive simulation framework of cable dynamics is introduced. This framework considers variable length cables and allows to incorporate elements such as buoys, ballast or Inertial Measurement Unit (IMU) sensors. The accuracy of this framework is assessed through experimental data. Second, a novel and improved solution for the instrumentation of a V-shaped buoy-ballast system using IMU sensors is investigated. This latter, designed for a neutrally buoyant tether between a Remotely Operated Vehicle (ROV) and an Unmanned Surface Vehicle (USV), is meant to improve operation safety. The discussed IMU-based solution provides wider information on the interaction between the ROV and the cable, including its 3D orientation and curvature amplitude, which could be used for both the control of the USV trajectory and its onboard winch.

Soft continuum manipulators are celebrated for their versatility and physical robustness to external forces and perturbations. However, this feature comes at a cost. The many degrees of freedom and compliance pose challenges for accurate pose reconstruction, both in terms of distributed sensing and pose reconstruction algorithms. Moreover, soft arms are inherently susceptible to deformation from external forces or loads, meaning that closed-loop control is essential for robust task performance. In this article, we propose the integration of multiple Inertial Measurement Units (IMUs) of a soft robot arm, Helix, for reconstruction of pose under internal and external forces. Furthermore, we integrate this dynamic pose reconstruction for kinematic-based closed-loop control strategies. By serially integrating sensing in the body of the Helix soft manipulator, we provide the system with high-frequency pose reconstruction and demonstrate improvements in end effector position with comparison to open-loop performance.

In recent years, infrastructure-based localization methods have achieved significant progress thanks to their reliable and drift-free localization capability. However, the preinstalled infrastructures suffer from inflexibilities and high maintenance costs. This poses an interesting problem of how to develop a drift-free localization system without using the preinstalled infrastructures. In this paper, an infrastructure-free and drift-free localization system is proposed using the ambient magnetic field (MF) information, namely IDF-MFL. IDF-MFL is infrastructure-free thanks to the high distinctiveness of the ambient MF information produced by inherent ferromagnetic objects in the environment, such as steel and reinforced concrete structures of buildings, and underground pipelines. The MF-based localization problem is defined as a stochastic optimization problem with the consideration of the non-Gaussian heavy-tailed noise introduced by MF measurement outliers (caused by dynamic ferromagnetic objects), and an outlier-robust state estimation algorithm is derived to find the optimal distribution of robot state that makes the expectation of MF matching cost achieves its lower bound. The proposed method is evaluated in multiple scenarios 1, including experiments on high-fidelity simulation, and real-world environments. The results demonstrate that the proposed method can achieve high-accuracy, reliable, and real-time localization without any pre-installed infrastructures.

This paper proposes a new kinematic model for wheeled skid-steer vehicles moving with low inertia on firm slopes that is based on the variation of the Instantaneous Center of Rotation (ICR) of its lateral treads. To this end, the ICR-based model for horizontal surfaces, where constant tread ICRs were considered, has been extended with two additional parameters to account for the sliding down phenomenon on inclined terrains that the former is unable to predict. The current pitch and roll angles of the vehicle together with the speeds of the treads are employed to calculate the changing positions of tread ICRs during turnings. This kinematic approach shows promising results when applied to the heavy robotic rover J8 on a slanted surface of smooth concrete.