IROS 2015 论文列表

We present a localization aware efficient sampling strategy for sampling-based motion planning under uncertainty that uses a new notion of localization ability of a sample. It puts more samples in regions where sensor data is able to achieve higher uncertainty reduction while maintaining adequate samples in regions where uncertainty reduction is poor. This leads to a less dense roadmap and hence results in significant time savings in the path search phase. We provide simulation results that show stochastic planners with our sampling strategy place less samples and find a well-localized path in shorter time with little compromise on the quality of path as compared to existing sampling techniques. We also show that a stochastic planner that uses our sampling strategy is probabilistically complete under some reasonable conditions on parameters.

This paper proposes a novel tongue-operated joy- stick device for severely disabled persons. The human tongue has superior movement and tactile sensations. By providing force information during the device's operation, user can skillfully control life supporting equipment, such as electric powered wheelchairs, robotic manipulators for meals through interaction with this equipment. Therefore, we designed a joystick mechanism that enables not only the generation of control commands for the equipment but also the adjustment of the reaction force during its operation. The adjustment of the reaction force enables the user to know if there is any interference between the equipment and the environment and to improve task performance. In our proposed mechanism, the elastic pins surround the tip of the joystick and restrict its movement by this contact. The reaction force is controlled easily by adjusting the support length of the elastic pins. This paper describes the detailed mechanism and its basic performance through several experiments.

In this paper, a hybrid algorithm for multiple-robot SLAM is proposed that combines the advantages of particle filtering and map merging. The proposed algorithm does not rely on rendezvous and calculates the unknown relative poses from the local maps of the robots. As another contribution, the uncertainty of the relative poses is taken into account by propagating the uncertainty to the past and future information using a novel algorithm. Moreover, once the relative poses are known, the integration of the information from all robots is performed using a novel batch-mode algorithm, which is a fast and efficient approach to deal with the time complexity problem. The experimental results show the effectiveness of the proposed algorithms.

Aiming at the physical coupling feature between the finger and the hand exoskeleton, a human-robot interaction modeling approach is proposed. The muscle motion formulas are established based on the finger physiological structure and Hill model. The equilibrium equations between exoskeleton and finger are connected by static analysis. In order to solve the redundancy problem of the system, a method based on the physiological cross-sectional area (PCSA) is adopted to get the optimized solution of muscle force, and an optimization method based on the total minimum error (TME) is presented to obtain the parameters of Hill model. The experimental setup is established to receive the finger data of motion and force for optimization. The approach proposed can get the quantifiable muscle parameters to study the statistical analysis of muscle motion and rehabilitation state. And it will be possible for the exoskeleton and the finger to be combined as a controlled plant so as to introduce muscle parameters into the controller design.

A human-inspired hand-over control strategy is proposed for the haptic interaction of two dual-fingered hands for the planar case. It is based on a grasp controller for an unknown object which achieves, via fingertip rolling, a stable grasp and a real object mass estimation. Object load transfer is receiver initiated, follows human evidence and involves awareness of the other hand's state based solely on local proprioceptive measurements. Simulation results illustrate the proposed approach.

It has long been assumed that for effective human-robot teaming, it is desirable for assistive robots to infer the goals and intents of the humans, and take proactive actions to help them achieve their goals. However, there has not been any systematic evaluation of the accuracy of this claim. On the face of it, there are several ways a proactive robot assistant can in fact reduce the effectiveness of teaming. For example, it can increase the cognitive load of the human teammate by performing actions that are unanticipated by the human. In such cases, even though the teaming performance could be improved, it is unclear whether humans are willing to adapt to robot actions or are able to adapt in a timely manner. Furthermore, misinterpretations and delays in goal and intent recognition due to partial observations and limited communication can also reduce the performance. In this paper, our aim is to perform an analysis of human factors on the effectiveness of such proactive support in human-robot teaming. We perform our evaluation in a simulated Urban Search and Rescue (USAR) task, in which the efficacy of teaming is not only dependent on individual performance but also on teammates' interactions with each other. In this task, the human teammate is remotely controlling a robot while working with an intelligent robot teammate `Mary'. Our main result shows that the subjects generally preferred Mary with the ability to provide proactive support (compared to Mary without this ability). Our results also show that human cognitive load was increased with a proactive assistant (albeit not significantly) even though the subjects appeared to interact with it less.

This study proposes a high-speed locomotion mechanism for a pipe-inspection robot. As a result of the narrow and complex structures of pipeline networks, it is difficult for robots to move quickly within the pipes. The new pneumatic mechanism proposed here realizes high-speed locomotion along with advantageous features for pipe inspection including a small diameter, flexibility, and low weight. First, we present the design concept of the novel locomotion mechanism using pneumatic flexible hollow-shaft actuators, which was previously developed by the authors. The prototype constructed to realize this concept and the associated mathematical model are then introduced. Second, the basic characteristics of the proposed mechanism are evaluated in terms of the holding force (generated by an expansion mechanism against the pipe wall) and the impellent force that induces forward motion in the robot. Finally, the in-pipe movement performance is confirmed. The experimental results show that the designed robot can be propelled inside a 53-mm-diameter pipe at a maximum speed of 250 mm/s, which is exceedingly faster than conventional designs.

Human activity recognition is an essential task for robots to effectively and efficiently interact with the end users. Many machine learning approaches for activity recognition systems have been proposed recently. Most of these methods are built upon a strong assumption that the labels in the training data are noise-free, which is often not realistic. In this paper, we incorporate the uncertainty of labels into a max-margin learning algorithm, and the algorithm allows the labels to deviate over iterations in order to find a better solution. This is incorporated with a hierarchical approach where we jointly estimate activities at two different levels of granularity. The model is tested on two datasets, i.e., the CAD-120 dataset and the Accompany dataset, and the proposed model shows outperforming results over the state-of-the-art methods.

The use of social robots for elderly care is becoming ever more relevant, thus introducing new challenges which need to be solved to achieve acceptable performance. One fundamental task for a social robot is to move to the person of interest in order to start interacting or perform a service. In this paper we address the task of a robot having to navigate to a possibly occluded person, which needs assistance, based only on audio and range information. Our approach is based on forming a heuristic cost function which is based on combining two audio features, and then moving to the optimum position indicated by this cost function after every interaction with the person. The method is compared to a greedy approach in 20 different tasks using a loud speaker playing at approximately 60dB sound pressure level (SPL) to mimic a human speaker, and the proposed method shows superior performance. A second experiment with an increase of 13dB SPL of the loud speaker is conducted and the proposed method is able to handle this.

Autonomous as well as teleoperated robots find wide applications in various environments. Their capability to accomplish complex and dynamic operations can be significantly improved by fusing human intelligence with autonomous algorithms. In this paper, we propose a haptic shared control algorithm to provide flexible human assistance for semi-autonomous mobile robots. Through the admittance and impedance models, the haptic shared controller smoothly puts together human operator inputs with robot autonomy. Further, the level of autonomy is fully determined by the operator with the grasp motion. A decomposed design has been taken for the autonomous controller of the mobile robot. The algorithm was implemented on the haptic interface omega.7 together with a QBot mobile robot, and its feasibility and efficacy have been validated by experiments.

The surgical robotics community have developed many different flexible robot designs to address the access problems of minimally invasive surgery. In this paper, we present a hand-held mechatronic tool with a miniaturized distal flexible manipulator incorporating a microscopy probe, camera and a light source for diagnostic arthroscopy. Extensive characterization of the flexible manipulator is provided, including an optimization of the manipulator workspace, hysteresis characteristics and repeatability of the instrument. The lateral stiffness of the flexible manipulator for different bending conditions is assessed along with the overall robustness of the platform. A cadaveric study was performed to demonstrate the potential clinical value of the device.

In this paper, we investigate the self-healing problem of mobile robot formation after some robots have been damaged, and present a gradient-based algorithm which enables mobile robots to restore the topology of the formation through local interactions among neighboring robots. Firstly, in order to optimize the repair path in a distributed manner, a gradient generation and diffusion mechanism is proposed to generate a specific gradient distribution in the formation. Then, utilizing several predefined path selection rules, a path selection algorithm is presented to guarantee the optimality of the selected repair path. Furthermore, several optimization indices are presented to quantitatively characterize the performance of self-healing algorithms. Finally, the effectiveness of the proposed algorithm is validated by numerical simulations and the simulation results show that the proposed algorithm can restore the topology of the formation with the fewer repair robots and lower energy consumptions.

Consider a mobile robot exploring an initially unknown school building and assume that it has already discovered some classrooms, offices, and bathrooms. What can the robot infer about the presence and the locations of other classrooms and offices in the school building? This paper makes a step toward providing an answer to the above question by proposing a system based on a generative model that is able to represent the topological structures and the semantic labeling schemas of buildings and to predict the structure and the schema for unexplored portions of these environments. We represent the buildings as undirected graphs, whose nodes are rooms and edges are physical connections between them. Given an initial knowledge base of graphs, our approach, relying on a spectral analysis of these graphs, segments each graph for finding significant subgraphs and clusters them according to their similarity. A graph representing a new building or an unvisited part of a building is eventually generated by sampling subgraphs from clusters and connecting them.

In this paper we present a novel whole-body compliance framework. It is based on the Multi Spring Model, a set of virtual springs interconnecting the limb extremities and the trunk as well. More specifically, six virtual springs connect the trunk of the robot to the feet and the hands and other six springs interconnect the limb extremities. By selecting the stiffness values of each individual spring, complex behaviors can be generated, opening the way for more challenging tasks. The framework is implemented in a multi-layer architecture that distributes the computation on the different control units available on the robot. The proposed method was tested in two different tasks using the COmpliant huMANoid (COMAN), a full-body torque-controlled humanoid robot. For the first task, the robot was able to grasp a box, lift it up and firmly hold it. In the second task, the robot balanced itself while having the hands in contact with two walls. COMAN was able to successfully perform both tasks while perturbed by external disturbances and environmental uncertainties.

The identification of the dynamic model of a robotic manipulator represents a fundamental step for designing high performance model-based controllers. Despite the huge number of works presented on this topic, the symbolic dynamic model reduction (i.e., the identification of the set of parameters observable through the measure of joint torques and positions) still remain a challenging task, characterized from tailored solutions, adapted from time to time to specific families of mechanisms. The work here presented, introduces an automatic and analytical reduction of the dynamic model, based on a multi-dimensional Fourier series decomposition of the dynamic equations. The procedure enables to obtain symbolically the base dynamic parameters (BP) starting from a given kinematic structure. The Fourier based model reduction can be applied indifferently both to open- and closed-chain kinematics. A simulated example shows the effectiveness of the proposed algorithm.

This paper focuses on recognition and prediction of human reaching motion in industrial manipulation tasks. Several supervised learning methods have been proposed for this purpose, but we seek a method that can build models on-the-fly and adapt to new people and new motion styles as they emerge. Thus, unlike previous work, we propose an unsupervised online learning approach to the problem, which requires no offline training or manual categorization of trajectories. Our approach consists of a two-layer library of Gaussian Mixture Models that can be used both for recognition and prediction. We do not assume that the number of motion classes is known a priori, and thus the library grows if it cannot explain a new observed trajectory. Given an observed portion of a trajectory, the framework can predict the remainder of the trajectory by first determining what GMM it belongs to, and then using Gaussian Mixture Regression to predict the remainder of the trajectory. We tested our method on motion-capture data recorded during assembly tasks. Our results suggest that the proposed framework outperforms supervised methods in terms of both recognition and prediction. We also show the benefit of using our two-layer framework over simpler approaches.

Silt accumulation and sedimentation in canal beds leads to deterioration of watercourses over time. Every year a forced closure of the canals in the Indus basin is inevitable for canal cleaning, entailing a very large scale and costly operation. Silt removal precision is prone to inefficiencies due to subjective decision making in the cleaning process. In this paper, we lay out a theoretical framework to map the semi-structured (emptied) canal bed terrains with an Unmanned Aerial Vehicle (UAV) system for quantitative inspection of deposited silt. The study employs Gaussian process regression on sampled points to determine a continuous distribution of silt surface, thereby, predicting the volume of silt on canal bed. Our theoretical analysis builds upon certain mathematical bounds on the variance of estimated volume, while explicitly considering localization error and sensor noise. Essentially, we setup a framework for studying how tolerable are the process and measurement uncertainties, while achieving a desired accuracy in silt profile and corresponding volume. We demonstrate the regression results in simulations as well as lab-scaled-model (LMS151 laser scanner) with different sets of parameters. Volume estimation is verified practically and mathematical performance limits are proposed in established aerial canal inspection system.

This paper presents a new flexible fixtureless assembly (FFA) approach for building T-joint structures automatically. The method is based on a Fixturing Assistive Gripper (FAG) mounted on a robot arm for forming T-joint sections. It eliminates possible uncertainties in terms of positioning and size errors during fit-up process. The method is verified by a robotic assembly workcell which includes dual robots and a simple machine vision unit. The basic algorithms for image processing, task planning, dual robot cooperative motion and path planning are developed for assembling a simple structure which predominantly formed by T-joint sections. Experiments show that the workcell can effectively eliminate the alignment error and successfully complete T-joint assembly task without the complicated sensing system typically required in the existing approaches.

Research on wearable robotic devices is fostered by the need to assist and restore human locomotion, which is essential for most daily life activities. Despite a continuous technological advancement, many challenges are to be faced before establishing the use of these devices as part of the rehabilitation process. A main concern is how to actively engage the users and monitor how they are affected by the exoskeleton. Basic analyses, such as measuring walking speed, must be extended to include analysis on balance, brain activity, and evaluation of neuromechanical effects. This scenario requires the use of different distributed sensor devices that must be carefully synchronized. This work describes the architecture we implemented to extend the capabilities of the H2 exoskeleton. As a proof of the architecture validity, we show the online estimation of users' muscle forces through the acquisition of electromyographic signals, that drive neuromusculoskeletal models. The proposed framework, built upon the Robot Operating System (ROS), aims to be reusable for a wide range of setups, including different exoskeletons and sensors.