ICRA 2018 论文列表

In this paper, we present a new approach to transfer in Reinforcement Learning (RL) for cross-domain tasks. Unlike, available transfer approaches, where target task learning is accelerated through initialized learning from source, we propose to adapt and reuse the optimal source policy directly in the related domains. We show the optimal policy from a related source task can be near optimal in target domain provided an adaptive policy accounts for the model error between target and the projected source. A significant advantage of the proposed policy augmentation is in generalizing the policies across related domains without having to re-Iearn the new tasks. We demonstrate that, this architecture leads to better sample efficiency in the transfer, reducing sample complexity of target task learning to target apprentice learning.

In the coverage path planning problem, a common assumption is that the robot can fully cover the environment without recharging. However, in reality most mobile robot systems operate under battery limitations. To incorporate this constraint, we consider the problem when the working environment is large and the robot needs to recharge multiple times to fully cover the environment. We focus on a geometric version where the environment is represented as a polygonal grid with a single charging station. Energy consumption throughout the environment is assumed to be uniform and proportional to the distance traveled. We first present a constant-factor approximation algorithm for contour-connected environments. We then extend the algorithm for general environments. We also validate the results in experiments performed with an aerial robot.

In this paper we consider the coverage control problem for a team of wire-traversing robots. The two-dimensional motion of robots moving in a planar environment has to be projected to one-dimensional manifolds representing the wires. Starting from Lloyd's descent algorithm for coverage control, a solution that generates continuous motion of the robots on the wires is proposed. This is realized by means of a Continuous Onto Wires (COW) map: the robots' workspace is mapped onto the wires on which the motion of the robots is constrained to be. A final projection step is introduced to ensure that the configuration of the robots on the wires is a local minimizer of the constrained locational cost. An algorithm for the continuous constrained coverage control problem is proposed and it is tested both in simulation and on a team of mobile robots.

Automated techniques such as model checking have been used to verify models of robotic mission plans based on Markov decision processes (MDPs) and generate counterexamples that may help diagnose requirement violations. However, such artifacts may be too complex for humans to understand, because existing representations of counterexamples typically include a large number of paths or a complex automaton. To help improve the interpretability of counterexamples, we define a notion of explainable counterexample, which includes a set of structured natural language sentences to describe the robotic behavior that lead to a requirement violation in an MDP model of robotic mission plan. We propose an approach based on mixed-integer linear programming for generating explainable counterexamples that are minimal, sound and complete. We demonstrate the usefulness of the proposed approach via a case study of warehouse robots planning.

While the problem of inverse kinematics on serial kinematic chains is well researched, solving motion tasks quickly on more complex robots remains an open problem. Examples include dual-arm manipulation, grasping with multi-finger hands, and full-body motion generation for humanoids. In this paper, we introduce an open-source software package for ROS and MoveIt! that solves inverse kinematics and motion tasks on robots with arbitrary kinematic trees. The underlying memetic algorithm integrates evolutionary optimization, particle swarm optimization, and gradient methods. The optimization respects joint limits, effectively avoids local minima, and achieves fast convergence to accurate solutions. More importantly, the overall motion goal is specified using a set of weighted sub-goals, providing great flexibility and control of secondary objectives. Several application examples demonstrate how to combine the predefined sub-goals to achieve complex motion tasks.

Robust velocity and position estimation is crucial for autonomous robot navigation. The optical flow based methods for autonomous navigation have been receiving increasing attentions in tandem with the development of micro unmanned aerial vehicles. This paper proposes a kernel cross-correlator (KCC) based algorithm to determine optical flow using a monocular camera, which is named as correlation flow (CF). Correlation flow is able to provide reliable and accurate velocity estimation and is robust to motion blur. In addition, it can also estimate the altitude velocity and yaw rate, which are not available by traditional methods. Autonomous flight tests on a quadcopter show that correlation flow can provide robust trajectory estimation with very low processing power. The source codes are released based on the ROS framework.

We demonstrate that moving the wrist while the fingers perform a grasp increases performance. The coordination shapes the interactions between the fingers, the object and its environment to extend the hand capabilities (e.g. higher payload and precision). We evaluated our hypothesis with a human grasping study where the volunteers grasped objects by moving the soft RBO Hand 2 while its fingers closed in a predefined motion. We limited their ability to coordinate their motion with the finger movements using a compliant robot attached to the hand, and observed that their grasp success decreases with increased constraints. We also successfully transferred one of the observed movement patterns to the robot, indicating that adaptive intrinsic/extrinsic motion increases robotic grasp performance as well.

In this paper we present a general, decentralized air traffic control solution using autonomous drones. We challenge some of the most difficult dense traffic situations, namely, crosswalk and package-delivery scenarios, where intelligent collective collision avoidance and motion planning is essential for a jam-free optimal traffic flow. We build up a force-based distributed multi-robot control model using a tunable selection of interaction terms: anisotropic repulsion, behaviour-driven velocity alignment, self-organized queueing and conflict-avoiding self-driving. We optimize the model with evolution in a realistic simulation framework and demonstrate its applicability with 30 autonomous drones in a coordinated outdoor flight within a densely packed virtual arena.

In this paper the modeling and planning problems of a system composed of multiple ground and aerial robots involved in a transportation task are considered. The ground robots rigidly grasp a load, while the aerial vehicles are attached to the object through non-rigid inextensible cables. The idea behind such a heterogeneous multi-robot system is to benefit of the advantages of both types of robots that might be the precision of ground robots, the increased payload of multiple aerial vehicles and their larger workspace. The overall model of the system is derived and its expression and redundancy are exploited by setting a general constrained optimal planning problem. The problem is herein solved by dynamic programming and simulation results validated the proposed scheme.

We present a framework to enable a fleet of rigidly attached quadrotor aerial robots to transport heavy objects along a known reference trajectory without inter-robot communication or centralized coordination. Leveraging a distributed wrench controller, we provide exponential stability guarantees for the entire assembly, under a mild geometric condition. This is achieved by each quadrotor independently solving a local optimization problem to counteract the biased torque effects from each robot in the assembly. We rigorously analyze the controllability of the object, design a distributed compensation scheme to address these challenges, and show that the resulting strategy collectively guarantees full group control authority. To ensure feasibility for online implementation, we derive bounds on the net desired control wrench, characterize the output wrench space of each quadrotor, and perform subsequent trajectory optimization under these input constraints. We thoroughly validate our method in simulation with eight quadrotors transporting a heavy object in a cluttered environment subject to various sources of uncertainty, and demonstrate the algorithm's resilience.

In this work, we address the dual-arm manipulation of a two degrees-of-freedom articulated object that consists of two rigid links. This can include a linkage constrained along two motion directions, or two objects in contact, where the contact imposes motion constraints. We formulate the problem as a cooperative task, which allows the employment of coordinated task space frameworks, thus enabling redundancy exploitation by adjusting how the task is shared by the robot arms. In addition, we propose a method that can estimate the joint location and the direction of the degrees-of-freedom, based on the contact forces and the motion constraints imposed by the object. Experimental results demonstrate the performance of the system in its ability to estimate the two degrees of freedom independently or simultaneously.

This paper studies collaboration through the cloud in the context of cooperative adaptive control for robot manipulators. We first consider the case of multiple robots manipulating a common object through synchronous centralized update laws to identify unknown inertial parameters. Through this development, we introduce a notion of Collective Sufficient Richness, wherein parameter convergence can be enabled through teamwork in the group. The introduction of this property and the analysis of stable adaptive controllers that benefit from it constitute the main new contributions of this work. Building on this original example, we then consider decentralized update laws, time-varying network topologies, and the influence of communication delays on this process. Perhaps surprisingly, these nonidealized networked conditions inherit the same benefits of convergence being determined through collective effects for the group. Simple simulations of a planar manipulator identifying an unknown load are provided to illustrate the central idea and benefits of Collective Sufficient Richness.

In this paper we consider a multi-robot system that shares the environment with external entities, and we propose a methodology for controlling the interaction with them. In particular, we consider the problem of achieving a desired dynamic interaction model, in such a way that the multi-robot system exchanges desired forces with external entities. This is obtained by introducing local deformations of the coupling actions among the robots. The proposed method ensures preservation of the passivity property, which provides safety guarantees in the interaction with the (possibly poorly known) external entities.

Robotic manipulators that are attached to mobile platforms are often used in workspaces that require the end-effector to mobilize beyond the manipulator's limited reach, such as in warehouse shelf stacking and similar applications. However, such assistive robots fall short of completing tasks that require the end-effector to be situated in a specific configuration at a critical time during the task. Traditionally, users control the mobile base to situate the arm such that the task can be completed through continuous operation. This requires an experienced operator who can predict the needed end-effector workspace, and can operate the base accordingly to maximize the likelihood of a successful task while avoiding any floor obstacles. In this work, we propose a straightforward control method that provides sufficient freedom to the end-effector to complete a task that is bound by time-dependent constraints. This is achieved by relaxing the time constraints on the mobile base trajectory in a floor space obstructed by obstacles. The trajectory of the platform is determined by sensor-assisted obstacle avoidance algorithm such that a single degree of freedom mobility can be represented through a safe obstacle-free time-independent path. The proposed control method is implemented in simulation and on physical hardware built in our labs. The simulation included a 5-DoF redundant Planar Mobile Manipulator (PMM). The hardware implementation and testing utilized a 9-DoF redundant mobile manipulator. The implementation results demonstrate the effectiveness of the control method in adjusting the mobile platform motion along its allowed obstacle-free path to enable the end-effector to follow its trajectory for task completion that would otherwise fail to complete when conventional control methods are used.

Ahstract- In this study, we propose a development and formation control of a passive mobile robot equipped only with servo brakes and no motors. The developed passive mobile robot can move forward by utilizing an external pulling force, and steers itself by controlling the servo brakes attached to each wheel. Systems composed of multiple mobile robots are very effective at exploring vast areas. However the coordination and simultaneous control of several robots utilizing simple information is a difficult task. Therefore we propose a leader follower architecture composed of multiple passive follower robots tethered to one active leader. In this paper, we first introduce the developed passive mobile robot. Then, we analyze the fundamental control method of this passive mobile robot from the perspective of the feasible braking control region, which considers the slip of the passive robot and the limitation of the external pulling force from the active leader. Lastly, a control law for the servo brakes attached to each wheel is proposed, followed by a feasibility study to determine whether the passive followers can stay in formation by controlling the servo brakes with the proposed control law.

This paper investigates the energy efficiency potential of continuously controllable clutches between the motors and links of robot joints. Inspired by biological muscles, the clutch enables free, purely gravity driven robot link motion phases gradually disengaged from gear friction, not requiring motor effort. The concept combines the energetic benefits of direct drives during unforced motion phases with the high torque density of conventional and mature geared robotic drive technology during forced motion phases. The paper specifies the general functional principle of the clutch for energy saving independent of any particular clutch implementation. The feasible energy saving of up to 60 % is investigated for harmonic link motions with varying frequencies and with respect to different ratios of link weight to friction torque. The outcomes of the theoretical investigations are supported by first experimental results.

Continuous, easy-to-implement, accurate inference of intended joint angles is important for effectively controlling powered prosthetic devices that can improve the lives and capabilities of upper-limb amputees. Estimation of intended joint angles is difficult because conventional biosignals are not directly related to the intended angle motion. In previous work, we began to address this issue by confirming that both transra-dial amputees and intact subjects, the measured deformation of the muscle bulge on the skin surface change according to the intended wrist joint angle. This paper presents a continuous prosthesis wrist joint control method using this deformation signal. We here verify the effectiveness of the distribution of the muscle bulge for accurate and stable wrist joint angle control in real time. The wrist joint angles were calculated in real time from a muscle viscoelastic model using the previously determined algorithm. We compared the error between measured and estimated angles with a conventional method, the Voigt model, and the KelvinVoigt model. Experimental results obtained for three intact people over three trials of wrist movement tasks gave the accuracy and stability of 7.96±6.16° when using the Voigt model; this is a similar performance compared to related work using a surface electromyogram. A method for continuously controlling the wrist joint angle for a prosthesis using the distribution of the muscle bulge was thus successfully established.

This paper presents a new material deposition strategy for developing a growing robot capable of building its own body. The growing robot is inspired by plant growth and is based on a 3D printer-like mechanism. The plotting of a filament near the tip allows the forward movement of the robot and results in building a tubular body. A material deposition process is introduced to perform a straight continuous growth as well as a turning behavior in order to permit the navigation of the robot in the environment. Bending is achieved by controlling the filament height in each position of the plotting, lowering or increasing plotting velocity with a position PID control algorithm. We demonstrate that the continuous deposition of the filament allows to obtain homogeneous and robust structures, with a significant improvement of the robot's performance compared to our previous version of the system (i.e., more than 100 N pulling force and 200 N shear force). The current version of the robot can sustain its weight, move efficiently by growing in the environment - both air and soil - and penetrate hard medium (up to 60kPa).

Convolutional Neural Networks (CNN) have successfully been utilized for localization using a single monocular image [1]. Most of the work to date has either focused on reducing the dimensionality of data for better learning of parameters during training or on developing different variations of CNN models to improve pose estimation. Many of the best performing works solely consider the content in a single image, while the context from historical images is ignored. In this paper, we propose a combined CNN-LSTM which is capable of incorporating contextual information from historical images to better estimate the current pose. Experimental results achieved using a dataset collected in an indoor office space improved the overall system results to 0.8 m & 2.5° at the third quartile of the cumulative distribution as compared with 1.5 m & 3.0° achieved by PoseNet [1]. Furthermore, we demonstrate how the temporal information exploited by the CNN-LSTM model assists in localizing the robot in situations where image content does not have sufficient features.