ICRA 2022 论文列表

This paper presents a novel robot arm that is capable of switching between a rigid robot arm and a continuum robot arm. Therefore, the novel robot arm can perform adaptive physical interaction and manipulation against complex working environments and tasks. The switch-ability of the robot arm is achieved with two types of joints: knee-like flexible joints and continuum flexible joints, with which the continuum segment of the robot arm is capable of locking and losing, hence the degree of freedom of the robot arm is capable to be switched. In this work, kinematics is established for specifying the relationship between joints space and global coordinates in both rigid and continuum configurations. Then, the posture and workspace in rigid and continuum configurations are analyzed and illustrated with numerical simulations, and compared based on the established kinematic model. Finally, a series of preliminary experimental testing toward the joint motion and stiffness has been carried out to validate the design, the kinematic model, and the motion performance of the proposed robot arm. Both the numerical and experimental results show that the knee-like joints can guarantee favorable motion accuracy, and the motion of continuum segment from the testing is well aligned with the motion calculated from the theoretical model. Moreover, the stiffness of rigid configuration is larger than the continuum configuration based on the stiffness experiment results. Therefore, the proposed novel robot arm is capable to handle adaptive interaction and manipulation in a diverse environment through the switching between the rigid and continuum configurations.

Global registration using 3D point clouds is a crucial technology for mobile platforms to achieve localization or manage loop-closing situations. In recent years, numerous researchers have proposed global registration methods to address a large number of outlier correspondences. Unfortunately, the degeneracy problem, which represents the phenomenon in which the number of estimated inliers becomes lower than three, is still potentially inevitable. To tackle the problem, a degeneracy-robust decoupling-based global registration method is proposed, called Quatro. In particular, our method employs quasi-SO(3) estimation by leveraging the Atlanta world assumption in urban environments to avoid degeneracy in rotation estimation. Thus, the minimum degree of freedom (DoF) of our method is reduced from three to one. As verified in indoor and outdoor 3D LiDAR datasets, our proposed method yields robust global registration performance compared with other global registration methods, even for distant point cloud pairs. Furthermore, the experimental results confirm the applicability of our method as a coarse alignment. Our code is available: https://github.com/url-kaist/quatro

Prioritization of tasks is a common approach to resolve conflicts in instantaneous control of redundant robots. However, the idea of prioritization has not yet been satisfactorily extended to model predictive control (MPC) to allow for real-time robot control. The standard sequential approach for prioritization is unsuitable because of the computational burden involved in solving a nonlinear problem (NLP) at every priority level. We introduce an alternate promising approach of using weighted exact penalties for the MPC stage costs, where a correctly tuned set of weights can introduce strict prioritization. We prove the existence of a set of equivalent weights that provides the same solution as the sequential approach for a local convex approximation of the original NLP and use this insight to design an algorithm to adaptively tune the weights. The weighted method is validated on a dual arm robot task in simulations and also implemented on a physical robot. We report computational times that are fast enough for prioritized MPC of robot manipulators for the first time, to the best of our knowledge.

Humanoid lower limbs with tactile cognition are crucial for future bipedal robots developing advanced bionic intelligence, such as owning autonomous reflexes and performing human-like actions. Most existing robotic lower limbs focus on providing physical support and mobility, with little work on more bionic DoFs or tactile sensing abilities that are more than significant for a fully humanoid system. This paper develops a robotic lower limb with whole-foot tactile sensing capability. An eight-DoF mechanism with humanoid joints is designed comprehensively, and a tactile sensor with electric double-layer capacitors principle wraps the special-shaped foot surface. An STM32-based circuit integrating perception and control with a real-time inverse kinematics algorithm is experimentally demonstrated. This work provides novel insight and methodology for humanoid robots and tactile-based bionic intelligence.

Tensegrity robots, composed of rigid rods and flexible cables, are difficult to accurately model and control given the presence of complex dynamics and high number of DoFs. Differentiable physics engines have been recently proposed as a data-driven approach for model identification of such complex robotic systems. These engines are often executed at a high-frequency to achieve accurate simulation. Ground truth trajectories for training differentiable engines, however, are not typically available at such high frequencies due to limitations of real-world sensors. The present work focuses on this frequency mismatch, which impacts the modeling accuracy. We proposed a recurrent structure for a differentiable physics engine of tensegrity robots, which can be trained effectively even with low-frequency trajectories. To train this new recurrent engine in a robust way, this work introduces relative to prior work: (i) a new implicit integration scheme, (ii) a progressive training pipeline, and (iii) a differentiable collision checker. A model of NASA's icosahedron SUPERballBot on MuJoCo is used as the ground truth system to collect training data. Simulated experiments show that once the recurrent differentiable engine has been trained given the low-frequency trajectories from MuJoCo, it is able to match the behavior of MuJoCo's system. The criterion for success is whether a locomotion strategy learned using the differentiable engine can be transferred back to the ground-truth system and result in a similar motion. Notably, the amount of ground truth data needed to train the differentiable engine, such that the policy is transferable to the ground truth system, is 1% of the data needed to train the policy directly on the ground-truth system.

Wheelchair-mounted robotic manipulators have the potential to help the elderly and individuals living with disabilities carry out their activities of daily living (ADLs) independently. Robotics researchers focus on assistive tasks from the perspective of various control schemes and motion types, whereas, health research focuses on clinical assessment and rehabilitation, arguably leaving important differences between the two domains. In particular, there have been many studies on which activities are relevant to functional independence, but little is known quantitatively about the frequencies of ADLs that are typically carried out in everyday life. Understanding what activities are frequently carried out during the day can help guide the development and prioritization of robotic technology for in-home assistive robotic deployment. Robotics and health care communities have differing terms and taxonomies for representing tasks and motions; we aim to ameliorate taxonomic differences by consolidating quantitative task data with prior results from subjective task priority surveys. This study targets lifelogging databases, where we compute (i) daily activity task frequency from long-term low sampling frequency video and Internet of Things sensor data, and (ii) short term arm and hand movement data from video data of domestic tasks. In this work, we aim to provide deeper insights and meaningful guidelines to focus research and future developments in the field of assistive robotic manipulation that support the needs and performance requirements of the target population.

Manipulation involves a broad spectrum of skills, e.g., polishing, peeling, flipping, screwing, etc., requiring complex and delicate control over both force and position. This paper aims at designing an optimal haptic interface for providing a robot with direct demonstrations of human's innate intelligence in performing a wide range of force-based bimanual manipulation tasks. Based on the proprioceptive actuation mechanism, kinodynamic design parameters of the (dual) 7-DOF haptic arm are optimized so as to maximize the force transparency perceived by the operator over the full real-scale workspace of human arm while also ensuring other important constraints including robot-to-operator collision and singularity avoidance, payload, controlled stiffness, etc. 2.65 kg of average reflective mass and 1500 N/m of controlled stiffness is achieved over the entire workspace. We show the efficacy of our haptic interface by demonstrating various force-based manipulation tasks with a light-weight anthropomorphic bimanual manipulator, LIMS2-AMBIDEX [1].

Robots performing close physical interaction with humans would require a continuously variable transmission to operate in the region around the peak efficiency or peak power of the driving system. Conventional continuously variable transmission (CVT) has shown advantages in energy-efficient driving systems. However, these CVT designs are heavy and large for robotic applications. This paper presents a passively adaptable toroidal-CVT (pat-CVT) that is coupled with a twisted string actuator (TSA). The proposed combination of pat-CVT with TSA expands the operation range of TSA and mimics the torque-speed characteristics of artificial muscles. The contributions of the proposed system are as follows: 1) It has a high transmission ratio (4.6:1) compared to the level of existing CVT, and compact design; 2) We propose a structure that increases the power transmission efficiency by reducing slip rate during rotation through the application of soft material on the input/output disk and roller surfaces of the CVT; 3) Relations between the external load and the transmission ratio can be determined by selecting the spring in the system to optimize the motor operating conditions over the entire range of loads. We show theoretical modeling of pat-CVT with TSA and characterization of the transmission ratio, energy efficiency, and force-velocity curve.

Actuated universal joints, or equivalent joint systems, are found in a number of robotic applications, in particular mobile snake robots, continuum robots and robotic tails. These joints have two degrees of freedom on two axes, each perpendicular to a third axis and to themselves. Such joints use a variety of actuation methods, including direct drive motors, linear screw drives, cable based systems, and hydraulics/pneumatics. In this paper the authors design and validate a mechanism that uses the Twisted String Actuator (TSA) in an antagonistic triad to actuate the universal joint, using orientation sensors and load cells to create a robust cascading closed loop control system. This results in a light, compact, high-performance actuation system that avoids the extra mass and hardware complexity that alternative actuation methods present, with the additional challenge of nonlinearity.

Perching onto objects can allow flying robots to stay at a desired height at low or no cost of energy. This paper presents a novel passive mechanism for aerial perching onto smooth surfaces. This mechanism is made from a bistable mechanism and a soft suction cup. Different from existing designs, it can be easily attached onto and detached from a surface, but it can also hold a large weight when attached to a surface. Further, the mechanism can still work when the suction cup is not precisely aligned with the surface, alleviating the requirement for precise motion control of flying robots. The attachment and detachment are facilitated by the bistable mechanism, while the strong holding is enabled by a locking mechanism that can disable the bistable mechanism. We conduct experiments to characterize the required forces for successful attachments and detachments. We also equip the perching mechanism onto a quadcopter to demonstrate it can be successfully used for perching onto smooth surfaces (e.g., glass).

Despite the advances in the field of human-robot interface (HRI) based on biological neural signal, the use of the sole electroencephalography (EEG) signal to help robotic exoskeleton predict the limb movement is currently no mature in rehabilitation training, due to its unreliability. Multimodal HRI represents a very recent solution to enhance the performance of single modal HRI. These HRI normally include the EEG signal with surface electromyography (sEMG) signal. However, their use for the lower limb movement prediction in hemiplegia is still limited, and the deep fusion feature of sEMG and EEG signal is ignored. This paper proposes a Dense co-attention mechanism-based Multimodal Enhance fusion Network (DMEFNet) for the lower limb movement prediction in hemiplegia. The DMEFNet can realize the mapping and deep fusion between the sEMG and EEG signal features and get a high accuracy movement prediction of the lower limbs. A sEMG and EEG data acquisition experiment and an incomplete asynchronous data collection paradigm are designed to verify the effectiveness of DMEFNet. The experimental results show that DMEFNet has a good movement prediction performance in both within-subject and cross-subject situations, reaching an accuracy of 82.96% and 88.44% respectively.

Navigating natural environments with deformable terrain is a difficult challenge in robotics. Understanding the interaction dynamics between robots and such terrain is an important first step in enabling them to explore these environments. Terramechanics models are largely developed and tested on wheeled and tracked platforms, but with the advent of readily available lightweight legged robots, developing an understanding of how robot feet interact with the terrain becomes increasingly important. Works on estimating terramechanical properties of deformable sands and soils use an underlying assumption that translation of the robot foot along the surface of the terrain is due to internal shear deformation of the soil. We show that for lightweight legged robots, this is not the case. Shear forces acting on the foot of a robot during a stride are not accurately predicted by the widely-used Janosi-Hanamoto formula. We propose a new model in which two forces acting on the foot dominate the foot-terrain interaction - gross sliding friction and bulldozing resistance - and propose a model of how these forces act on the foot. We test this model on multiple soil types with different foot materials. Experimental data, collected on a testbench equipped with actuators and sensors identical to those deployed on a robot in the field, is used to validate our proposed model.

In order to meet the mobility and physical activity needs of people with impaired limbs function, a novel limbs-free variable structure wheelchair system controled by face-computer interface (FCI) was developed in this study. FCI used facial electromyography (fEMG) as a human intention recognition method from 6 facial movements, and the accuracy of intent recognition reached 97.6% under a series of offline optimization including channel optimization based on the Hilbert transform to obtain the envelope of fEMG, features optimization, and channel-independent model optimization. A collection of finite state machines (FSM) was used to control the movement and structural changes of the wheelchair. A shared control strategy called “ Keep Action after Take Over (KAaTO) “ that can reduce user fatigue while increasing safety was used in long-distance movement control of wheelchair. To test the performance of the system, in the braking distance test experiment, the result of 0.429m under KAaTO was better than the EMG-based discrete command control and speech command control method. Finally, an outdoor long-distance control pilot experiment proved the superior performance of the developed system.

In this paper a novel full state feedback approach for control of compliant actuated robot with nonlinear spring characteristics is presented. A multi-DOF elastic robot arm is a multi-input multi-output (MIMO) under-actuated system. By the new novel controller, which is based on motor coordinate transformation and motor inertia shaping, the MIMO system can be converted into a set of decoupled single-input single-output (SISO) systems. Using full state feedback controller, we can configurate the poles of each SISO system. The controller is validated by an 3-DOF elastic robot with nonlinear spring characteristics in simulation of MATLAB/Simulink.

Non-invasive brain-computer interfaces (BCIs) provide us with the unique ability to classify the psychological state of a person using only neurophysiological signals, such as those captured with an electroencephalogram (EEG). With this ability, new avenues for innovation in the field of healthcare arise, especially as it is used for robotics. EEGNet is a novel deep learning technique for the classification of EEG data with a limited training set that generalizes well to a variety of BCI paradigms, and the performance thereof can further be improved. We propose the use of Thomson Multitaper Power Spectral Density estimation in the EEG-BCI classification pipeline as well as a novel convolutional neural network (CNN), which extends EEGNet with sparse feature maps produced by efficient regularized separable convolutions. Further, we test the efficacy of interspersed Gaussian noise as a data augmentation technique. To show the improvements found with this new pipeline, we test on a widely used public EEG dataset related to emotion classification, then perform an ablation study to determine the most contributing factors. The accuracy on this public dataset was 77.16%. These results show that our pipeline improved the classification accuracy by 10.86% when compared with the state-of-the-art.

Certain wheeled mobile robots e.g., electric wheelchairs, can operate through indirect joystick controls from users. Correct steering angle becomes essential when the user should determine the vehicle direction and velocity, in particular for differential wheeled vehicles since the vehicle velocity and direction are controlled with only two actuating wheels. This problem gets more challenging when complex curves should be realized by the user. A novel assistive controller with safety constraints is needed to address these problems. Also, the classic control methods mostly require the desired states beforehand which completely contradicts human's spontaneous decisions on the desired location to go. In this work, we develop a novel assistive control strategy based on differential geometry relying on only joystick inputs and vehicle states where the controller does not require any desired states. We begin with explaining the vehicle kinematics and our designed Darboux frame kinematics on a contact point of a virtual wheel and plane. Next, the geometric controller using the Darboux frame kinematics is designed for having smooth trajectories under certain safety constraints. We experiment our approach with different participants and evaluate its performance in various routes.

Bioinspired robots are useful tools to study complex biomechanical processes of animal locomotion. Key movements and kinematic parameters are under the control of experimenters, which is impossible to perform when experimenting with living animals. The primary challenge to test biological hypotheses is designing realistic robots taking inspiration from swimming snakes. Yet, underlying biomechanics of undulatory swimming i.e., anguilliform swimming, remains poorly understood. Many of underwater snakebots are made of rigid segments that form a broken-line system, unlike the skeleto-muscular systems of living snakes include more than 200 vertebrae, conferring an extreme fluidity. This paper introduces a novel design based on hybrid continuum cable driven robot (HCDR) developed through interaction with biologists and roboticists. This Biology-Push design significantly increases the fluidity and freedom of the robot's motion. Thus, improved mimic snake's locomotion is given using cable-driven to represent linkages between muscles and vertebrae providing good fluidity. In addition, the association of rigid and flexible parts allows a homogeneous distribution of actuators and masses to design autonomous swimming snake robot. Combining literature data and kinematic of swimming snakes’ analyses, we implemented a kinematic model to control prototype’ motion in both a plane and a volume. Finally, a comparative study between the device kinematics and the snake's movements is carried out.

This work presents a novel Aspect Ratio-Modular Vertical Take-Off and Landing (ARM-VTOL) aerial robot, which is a meta-aircraft composed of two or more TiltRotor hybrid aircraft systems capable of magnetically being coupled during hovering flight, and of executing VTOL / Fixed-Wing hybrid missions once combined. The proposed meta-aircraft system carries the advantage of improved aerodynamic efficiency due its increased cumulative planform aspect ratio, which can be leveraged to achieve prolonged flight times in collaborative multi-vehicle flight. We propose an extendable methodology for its control which relies on the multi-body equivalent dynamics, and we present the coupling mechanism design that facilitates its experimental demonstration. We accompany these contributions with a field test-driven evaluation study conducted with a bi-vehicle ARM-VTOL prototype. The presented sequence includes vehicle-to-vehicle magnetic coupling during hovering flight, and is followed by a combined-vehicle mission comprising vertical climb, VTOL-forward transition, fixed-wing flight and maneuvering, and reverse-transition to VTOL and landing.

Agile navigation through uncertain and obstacle-rich environments remains a challenging task for autonomous mobile robots (AMR). For most AMR, obstacles are identified using onboard sensors, e.g., lidar or cameras. The effectiveness of these sensors may be severely limited, however, by occlusions introduced from the presence of other obstacles. The occluded area may contain obstacles, static or dynamic, not included into the motion planning of the robot and could cause potential collisions if they suddenly appear in the field of view of the robot. This paper proposes a general Model Predictive Control (MPC)-based framework for handling occlusions in structured or unstructured environments, that contain known or unknown static or dynamic obstacles. Safety is promoted by commanding velocities that consider surrounding obstacle uncertainty, while perception is promoted through a specially designed objective that can reduce the occluded area created by obstacles. The effectiveness of this framework is validated through simulations that show swift and safe motion in a variety of different environments. Similarly, experimental validation is achieved with a Boston Dynamics' Spot quadruped robot operating in an occluding environment.