ICRA 2021 论文列表

Many previous works have facilitated muscle cell (C2C12) alignment to form fiber-like cell structures. However, there still remains a challenge how to induce C2C12 myoblasts in the cell structures to differentiate into matured myocytes to form a functional muscle tissue, while external mechanical stimulation has been proved to have good effects on proliferation and differentiation of myoblasts. In this paper, we proposed a vision-based micro robotic manipulation system to achieve automatic mechanical stimulation for one single muscle fiber-like cell structures (MFCS). A tube, which is attached to a three degree-of-freedom (DOF) manipulator, and a probe are employed to apply the uniaxial mechanical stimulation to train the MFCS. To measure the force applied on MFCS, a vision-based measuring and correction method is utilized, which decrease the error by 74%. Moreover, based on the viscoelastic property of the MFCS, a feedback control algorithm has been applied to compensate for the force loss to realize the force stimulation. And the final value of force remains 699 ± 1μN after 110s experiment.

Learning from demonstrations has made great progress over the past few years. However, it is generally data hungry and task specific. In other words, it requires a large amount of data to train a decent model on a particular task, and the model often fails to generalize to new tasks that have a different distribution. In practice, demonstrations from new tasks will be continuously observed and the data might be unlabeled or only partially labeled. Therefore, it is desirable for the trained model to adapt to new tasks that have limited data samples available. In this work, we build an adaptable imitation learning model based on the integration of Meta-learning and Adversarial Inverse Reinforcement Learning (Meta-AIRL). We exploit the adversarial learning and inverse reinforcement learning mechanisms to learn policies and reward functions simultaneously from available training tasks and then adapt them to new tasks with the meta-learning framework. Simulation results show that the adapted policy trained with Meta-AIRL can effectively learn from limited number of demonstrations, and quickly reach the performance comparable to that of the experts on unseen tasks.

This paper addresses outdoor terrain mapping using overhead images obtained from an unmanned aerial vehicle. Dense depth estimation from aerial images during flight is challenging. While feature-based localization and mapping techniques can deliver real-time odometry and sparse points reconstruction, a dense environment model is generally recovered offline with significant computation and storage. This paper develops a joint 2D-3D learning approach to reconstruct local meshes at each camera keyframe, which can be assembled into a global environment model. Each local mesh is initialized from sparse depth measurements. We associate image features with the mesh vertices through camera projection and apply graph convolution to refine the mesh vertices based on joint 2-D reprojected depth and 3-D mesh supervision. Quantitative and qualitative evaluations using real aerial images show the potential of our method to support environmental monitoring and surveillance applications.

In previous work, using a process we call meshing, the reachable state spaces for various continuous and hybrid systems were approximated as a discrete set of states which can then be synthesized into a Markov chain. One of the applications for this approach has been to analyze locomotion policies obtained by reinforcement learning, in a step towards making empirical guarantees about the stability properties of the resulting system. In a separate line of research, we introduced a modified reward function for on-policy reinforcement learning algorithms that utilizes a "fractal dimension" of rollout trajectories. This reward was shown to encourage policies that induce individual trajectories which can be more compactly represented as a discrete mesh. In this work, we combine these two threads of research by building meshes of the reachable state space of a system subject to disturbances and controlled by policies obtained with the modified reward. Our analysis shows that the modified policies do produce much smaller reachable meshes. This shows that agents trained with the fractal dimension reward transfer their desirable quality of having a more compact state space to a setting with external disturbances. The results also suggest that the previous work using mesh based tools to analyze RL policies may be extended to higher dimensional systems or to higher resolution meshes than would have otherwise been possible.

In this paper, we present a low-damping active knee prosthesis (LDKP) with low noise and high backdrivability. The proposed prosthesis is driven by a motor and then decelerated by a four-stage synchronous belt. High backdrivability given by this structure accelerates the prosthetic response. A control system containing several sensors are embedded in the proposed prosthesis to recognize different modes and provide corresponding strategies. Preliminary experiments were carried out on a transfemoral amputee subject, demonstrating the features of low noise, high backdrivability and ability to reproduce natural walking gaits.

Mechanical intelligence is the use of mechanical and other physical properties to create robotic systems adaptable to new external situations using simple control schemes. Designs of robot hands have successfully been developed and optimised following this principle to produce self-adaptive and versatile power grasps via implementations based on underactuated fingers, elastic components, and open-loop motor control. However, these characteristics, and mechanical-intelligent strategies in general, have been seldom leveraged for precision grasping. This paper proposes a mechanical-intelligent technique to facilitate not only spiral caging power grasp, but also self-adaptive precision grasp with error tolerance. This approach is exemplified by the rigorous analysis, development, and testing of a novel three-fingered, two-actuator, underactuated robot hand, called the helical hand, which is capable of self-adaptive precision grasping, and of generating spiral helical power grasps of unknown objects by simply setting two actuators at a constant speed.

A vehicle is expected to exhibit omni-directional locomotion capability to provide improved rough terrain vehicle functionality. Generally, rough terrain vehicles are not holonomic and cannot travel in the lateral direction, whereas typical omni-directional vehicles have difficulty in traveling on rough terrain. This paper proposes installing a crank leg for a Mecanum-wheeled vehicle ("Mecanum crank") to enhance its rough-terrain locomotion. Compared with other holonomic vehicles designed for rough terrain locomotion, the proposed design exhibited superior capability in the longitudinal wheel direction. To avoid the conflict between the crank leg and Mecanum wheel, this paper proposes the use of a differential gear system. The simple structure of the crank leg enables the implementation of a swing equalizer and roller grousers, which further enhance its rough-terrain locomotion. In longitudinal locomotion experiments, the proposed mechanism could climb a 95 mm-high step, which is 95% of the Mecanum wheel diameter. Furthermore, Mecanum crank could climb miniature scale stairs and execute lateral locomotion on them.

In this paper, we propose a meaningful definition of rotational centroidal orientation which is somewhat missed in the state-of-the-art centroidal momentum and dynamics theory for locomotion robots with one floating base. This centroidal instantaneous orientation rotates as the robot runs, and it is extracted from the total system angular inertia. The new centroidal frame is proposed to be parallel with the principal axes of the centroidal angular inertia, which can describe the whole-robot rotational motion. To avoid high fluctuations of centroidal frame orientation parameters between adjacent control loops, we develop one algorithm to enable the centroidal instantaneous frame to be smooth. The relationship between the centroidal angle rate and the centroidal angular velocity is derived, as well as the relationship in the acceleration level, which can be used for whole-body torque control. The new centroidal orientation or Euler angle is verified by two-scenario simulations, and another scenario is used to track and control the centroidal angular motion in the first-order kinematics level. The idea has considerable potential for system design, motion generation, and torque control in robotics communities with different research topics and theoretical backgrounds.

In this paper, we present a model based approach for predicting the forward kinematics of a tool for picking prepreg fiber plies from a flat table and draping them onto a double curved mold. The tool consists of suction cups interlinked with springs. The tool has 60 actuated and 240 passive degrees of freedom. The prediction is based on establishing a kinematic model of the total potential energy of the tool and minimizing this energy, under the assumptions that dynamic artifacts can be neglected as the movements are relatively slow, and that the springs interpolate the orientation of the suction cups. Various assumptions and simplifications are presented to increase computational speed.Finally, the model predictions are compared to real configurations measured by a camera. Four configurations are simulated and measured and the largest RMS positional deviation measured is 1.99mm, the largest residual measured is 3.68mm. The tolerance of the layup process is 2.5mm so more work needs to be done to reduce the maximum deviations.

In this paper, we study the problem of dynamic interaction between a robot and an unknown object (e.g., catching a ball, or handing off an object during locomotion). In particular, we propose a method for estimating the inertial parameters of an object during dynamic interaction, while minimally altering the trajectory of the object – a minimal interaction approach. Our method combines trajectory estimation (e.g., using standard methods from computer vision) with a model-based estimator that exploits the robot’s known dynamic model. We first develop the method for a generalized three-dimensional problem, and then evaluate the method for the case of an object moving along a linear trajectory. We present experimental results obtained using a KUKA iiwa 7 interacting with rolling balls of varying mass. Our experiments demonstrate that the mass of the objects can be accurately estimated at the moment of impact when accurate object trajectory estimates are available, and that significant improvement can be obtained by incorporating force measurements at the contact point while following the object.

Parallel tracking and mapping (PTAM) as a time-efficient framework for simultaneous localization and mapping (SLAM) has been becoming popular in recent years. However, in this paper, we vigilantly point out that the favorite parallel-pipeline design realized by recent proposed SLAM algorithms may lead to inaccurate state estimates which, as a consequence, cannot always guarantee the performance of the estimators in real application. This is mainly due to the imperfect design for processing loop-closure measurements which accidentally violates the Markov assumption for probabilistic SLAM problem. To address this issue, a novel estimator design is proposed that holds the advantage of parallel processing, while striving to be consistent with the Markov property of the batch probabilistic SLAM estimator, therefore, termed Markov parallel tracking and mapping (MPTAM). Especially, the experiments on challenging visual-inertial datasets are employed to further demonstrate the improvements of proposed estimator in terms of accuracy and efficiency, as compared with the state-of-the-art SLAM system.

In the context of self-driving vehicles there is strong competition between approaches based on visual localisation and Light Detection And Ranging (LiDAR). While LiDAR provides important depth information, it is sparse in resolution and expensive. On the other hand, cameras are low-cost and recent developments in deep learning mean they can provide high localisation performance. However, several fundamental problems remain, particularly in the domain of uncertainty, where learning based approaches can be notoriously over-confident.Markov, or grid-based, localisation was an early solution to the localisation problem but fell out of favour due to its computational complexity. Representing the likelihood field as a grid (or volume) means there is a trade off between accuracy and memory size. Furthermore, it is necessary to perform expensive convolutions across the entire likelihood volume. Despite the benefit of simultaneously maintaining a likelihood for all possible locations, grid based approaches were superseded by more efficient particle filters and Monte Carlo sampling (MCL). However, MCL introduces its own problems e.g. particle deprivation.Recent advances in deep learning hardware allow large likelihood volumes to be stored directly on the GPU, along with the hardware necessary to efficiently perform GPU-bound 3D convolutions and this obviates many of the disadvantages of grid based methods. In this work, we present a novel CNN-based localisation approach that can leverage modern deep learning hardware. By implementing a grid-based Markov localisation approach directly on the GPU, we create a hybrid Convolutional Neural Network (CNN) that can perform image-based localisation and odometry-based likelihood propagation within a single neural network. The resulting approach is capable of outperforming direct pose regression methods as well as state-of-the-art localisation systems.

Maintaining a map online is resource-consuming while a robust navigation system usually needs environment abstraction via a well-fused map. In this paper, we propose a mapless local planner which directly conducts such abstraction on the unfused sensor data. A limited-memory data structure with a reliable proximity query algorithm is proposed for maintaining raw historical information. A sampling-based scheme is designed to extract the free-space skeleton. A smart waypoint selection strategy enables to generate high-quality trajectories within the resultant flight corridors. Our planner differs from other mapless ones in that it can abstract and exploit the environment information more efficiently. The online replan consistency and success rate are both significantly improved against conventional mapless methods.

The paper develops a control scheme for task-space trajectory tracking of robot manipulators in the face of unknown kinematic and dynamic parameters of both the manipulator and picked object/tool, and in the presence of time varying disturbances/noise and certain unmodelled dynamics. The proposed scheme consists of an outer task-space control loop and an inner joint space control loop. The two control loops in combination ensure that the errors tend to zero asymptotically and thus achieve precise task-space tracking. Other features of the proposed method are simple control structure that do not require decomposition of the dynamics into the regressor matrix-parameter vector. Simulation results are provided for a full 6-DOF manipulator performing complex tasks while carrying a tool with unknown kinematics and dynamics parameters.

Robot manipulation in cluttered scenes often requires contact-rich interactions with objects. It can be more economical to interact via non-prehensile actions, for example, push through other objects to get to the desired grasp pose, instead of deliberate prehensile rearrangement of the scene. For each object in a scene, depending on its properties, the robot may or may not be allowed to make contact with, tilt, or topple it. To ensure that these constraints are satisfied during non-prehensile interactions, a planner can query a physics-based simulator to evaluate the complex multi-body interactions caused by robot actions. Unfortunately, it is infeasible to query the simulator for thousands of actions that need to be evaluated in a typical planning problem as each simulation is time-consuming. In this work, we show that (i) manipulation tasks (specifically pick-and-place style tasks from a tabletop or a refrigerator) can often be solved by restricting robot-object interactions to adaptive motion primitives in a plan, (ii) these actions can be incorporated as subgoals within a multi-heuristic search framework, and (iii) limiting interactions to these actions can help reduce the time spent querying the simulator during planning by up to 40× in comparison to baseline algorithms. Our algorithm is evaluated in simulation and in the real-world on a PR2 robot using PyBullet as our physics-based simulator. Supplementary video: https://youtu.be/ABQc7JbeJPM.

Teleoperation provides a way for human operators to guide robots in situations where full autonomy is challenging or where direct human intervention is required. It can also be an important tool to teach robots in order to achieve autonomous behaviour later on. The increased availability of collaborative robot arms and Virtual Reality (VR) devices, provides ample opportunity for development of novel teleoperation methods. Since robot arms are often kinematically different from human arms, mapping human motions to a robot in real-time is not trivial. Additionally, a human operator might steer the robot arm toward singularities or its workspace limits, which can lead to undesirable behaviour. This is further accentuated for the orchestration of multiple robots. In this paper, we present a VR interface targeted to multi-arm payload manipulation, which can closely match real-time input motion. Allowing the user to manipulate the payload rather than mapping their motions to individual arms we are able to simultaneously guide multiple collaborative arms. By releasing a single rotational degree of freedom, and by using a local optimization method, we can improve each arm’s manipulability index, which in turn lets us avoid kinematic singularities and workspace limitations. We apply our approach to predefined trajectories as well as real-time teleoperation on different robot arms and compare performance in terms of end-effector position error and relevant joint motion metrics.

In this paper, a robust RGB-D SLAM system is proposed to utilize the structural information in indoor scenes, allowing for accurate tracking and efficient dense mapping on a CPU. Prior works have used the Manhattan World (MW) assumption to estimate low-drift camera pose, in turn limiting the applications of such systems. This paper, in contrast, proposes a novel approach delivering robust tracking in MW and non-MW environments. We check orthogonal relations between planes to directly detect Manhattan Frames, modeling the scene as a Mixture of Manhattan Frames. For MW scenes, we decouple pose estimation and provide a novel drift-free rotation estimation based on Manhattan Frame observations. For translation estimation in MW scenes and full camera pose estimation in non-MW scenes, we make use of point, line and plane features for robust tracking in challenging scenes. Additionally, by exploiting plane features detected in each frame, we also propose an efficient surfel-based dense mapping strategy, which divides each image into planar and non-planar regions. Planar surfels are initialized directly from sparse planes in our map while non-planar surfels are built by extracting superpixels. We evaluate our method on public benchmarks for pose estimation, drift and reconstruction accuracy, achieving superior performance compared to other state-of-the-art methods. We will open-source our code in the future.

Reliable perception is essential for robots that interact with the world. But sensors alone are often insufficient to provide this capability, and they are prone to errors due to various conditions in the environment. Furthermore, there is a need for robots to maintain a model of its surroundings even when objects go out of view and are no longer visible. This requires anchoring perceptual information onto symbols that represent the objects in the environment. In this paper, we present a model for action-aware perceptual anchoring that enables robots to track objects in a persistent manner. Our rule-based approach considers inductive biases to perform high-level reasoning over the results from low-level object detection, and it improves the robot’s perceptual capability for complex tasks. We evaluate our model against existing baseline models for object permanence and show that it outperforms these on a snitch localisation task using a dataset of 1,371 videos. We also integrate our action-aware perceptual anchoring in the context of a cognitive architecture and demonstrate its benefits in a realistic gearbox assembly task on a Universal Robot.

This paper presents a modular and reconfigurable mini-robotic system with 5 degrees of freedom (DoFs) towards minimally invasive surgery (MIS). The mini-robotic system consists of two modules, a 2-DoFs rotational end-effector, and a 3-DoFs positioning platform. The 2-DoFs rotational end-effector is based on a spring-spherical joint mechanism, whose rotation is controlled by Bowden-cable. The 3-DoFs positioning platform is based on the linear Delta parallel mechanism. Magnetic spherical joints are adopted to replace the traditional spherical joint. The magnetic joint connections enable fast assembling and disassemble of the end platform and kinematic chains. Different surgical instruments can be installed without changing the driver and control system. A flexible shaft actuates the 3-DoFs positioning platform to arrange the motors away from the manipulator side. Based on these structure characteristics, the 3-DoFs positioning platform’s size is dramatically reduced. The outer diameter of the current prototype is 32.5 mm. The single-axis positioning accuracy of the 3-DoFs positioning platform is within -1 mm to 0.85 mm. Three axes tracking experiments are also carried out, with the positioning errors of ± 1.2 mm for cylindrical curves and -1.5 mm to 2 mm for spherical helix curves. Static and dynamic load capabilities are also tested. Finally, the feasibility of the proposed system is demonstrated.