ICRA 2015 论文列表

Robotic odor source localization is a promising tool with numerous applications in safety, search and rescue, and environmental science. In this paper, we present an algorithm for odor source localization using multiple cooperating robots equipped with chemical sensors. Laplacian feedback is employed to maintain the robots in a formation, introducing spatial diversity that is used to better establish the position of the flock relative to the plume and its source. Robots primarily move upwind but use odor information to adjust their position and spacing so that they are centered on the plume and trace its structure. Real-world experiments were performed with an ethanol plume inside a wind tunnel, and used to both validate the algorithm and assess the impact of different formation shapes.

Collisions between moving machinery and human workers in construction job sites are one of the main sources of fatalities and accidents during the execution of construction projects. In this paper, we present a methodology to identify and assess construction project plan dangers before their execution. Our methodology has the following steps: 1) Plans are translated from a high-level activity graph to a discrete event simulation model; 2) Trajectories are simulated using sampling based and combinatorial motion planning algorithms; and 3) Safety scores and risk-based heatmaps are calculated based on the trajectories of moving equipment. Finally, we present an illustrative case study to demonstrate the usability of our model.

Whilst current minimally invasive surgical robots offer many advantages to the surgeon, most of them are still controlled using the traditional master-slave approach, without fully exploiting the complementary strengths of both the human user and the robot. This paper proposes a framework that provides a cooperative control approach to human-robot interaction. Typical teleoperation is enhanced by incorporating haptic guidance-based feedback for surgical tasks, which are demonstrated to and learned by the robot. Safety in the surgical scene is maintained during reproduction of the learned tasks by including the surgeon in the guided execution of the learned task at all times. Continuous Hidden Markov Models are used for task learning, real-time learned task recognition and generating setpoint trajectories for haptic guidance. Two different surgical training tasks were demonstrated and encoded by the system, and the framework was evaluated using the Raven II surgical robot research platform. The results indicate an improvement in user task performance with the haptic guidance in comparison to unguided teleoperation.

Modern approaches to simultaneous localization and mapping (SLAM) formulate the inference problem as a high-dimensional but sparse nonconvex M-estimation, and then apply general first- or second-order smooth optimization methods to recover a local minimizer of the objective function. The performance of any such approach depends crucially upon initializing the optimization algorithm near a good solution for the inference problem, a condition that is often difficult or impossible to guarantee in practice. To address this limitation, in this paper we present a formulation of the SLAM M-estimation with the property that, by expanding the feasible set of the estimation program, we obtain a convex relaxation whose solution approximates the globally optimal solution of the SLAM inference problem and can be recovered using a smooth optimization method initialized at any feasible point. Our formulation thus provides a means to obtain a high-quality solution to the SLAM problem without requiring high-quality initialization.

In this paper we use optimal control on a geared electric DC motor to compare the energetic efficiency of a simulation of conceptual monoped hoppers with either parallel elastic actuation (PEA) or series elastic actuation (SEA). The energy is measured using three cost functions: positive actuator work, electrical losses, and positive electrical work. For PEA, the presence of the motor inertia in the collision losses leads to increased collision losses at large transmission ratios, which lead to energetically costly compensatory strategies where the SEA is at its most efficient. At small transmission ratios, the motor force increases for both cases, leading to increased thermal losses. In agreement with those theoretical predictions, our work shows that for positive actuator work and positive electrical work the optimal parameter choice for SEA is significantly more energetically efficient than the optimal choice for PEA. For electrical losses, a suitable choice of the transmission ratio can lead to negligible cost values for both actuator concepts.

It has recently been shown that intrinsically elastic robots are capable of outperforming rigid robots in terms of peak velocity by making systematic use of energy storage and release. Certainly, high link side velocities are beneficial for performance, however, they also increase the probability of self damage or human injury in case of a collision. To ensure the physical integrity of both human and robot, it is therefore crucial to avoid potentially dangerous collisions and react in a compliant manner if unwanted contact has occurred or may occur unforeseeable. In this paper, we consider the most intuitive collision anticipation and pre-reaction scheme, namely stopping an elastic robot, if possible in minimum time. For 1-DOF elastic joints with limited elastic deflection we extend existing model-based and model-free controllers and compare their performance. Furthermore, we analyze the braking trajectory that is achieved with the different strategies. The 1-DOF solution is extended to the double pendulum case, where we show that feasible estimates for maximum and final position can be obtained at the very first instant of braking.

This paper proposes an environmental representation approach based on hybrid metric and topological maps as a key component for mobile robot navigation. Focus is made on an ego-centric pose graph structure by the use of Keyframes to capture the local properties of the scene. With the aim of reducing data redundancy, suppress sensor noise whilst maintaining a dense compact representation of the environment, neighbouring augmented spheres are fused in a single representation. To this end, an uncertainty error model propagation is formulated for outlier rejection and data fusion, enhanced with the notion of landmark stability over time. Finally, our algorithm is tested thoroughly on a newly developed wide angle 360° field of view (FOV) spherical sensor where improvements such as trajectory drift, compactness and reduced tracking error are demonstrated.

Variable stiffness actuators (VSA) are fostered in robotics for their capability to address physical interaction with a physically adjustable compliance, being advantageous in terms of efficiency, safety and adaptability to unknown environments. Here we introduce the concept of a switching VSA (sVSA), in which a single actuator is used to control the position or the stiffness of a robotic joint according to a mechanical switch. Despite not allowing simultaneous control of both quantities, this architecture has the potential to make the design lighter, requiring one continuously powered actuator, controllable in position, and one additional switch, activated only occasionally between two limit stages: the advantages are the separation of the motors power requirements and a simpler control. A first prototype of a 1-DoF revolute variable-stiffness joint has been built, based on the vsaUT-II developed at the University of Twente, with a novel clutch mechanism allowing continuous and efficient switching. The prototype proved functionality and feasibility of the sVSA concept.

Human action representation, recognition and learning is of importance to guarantee a fruitful human-robot cooperation. In this paper, we propose a novel coordinate-free, scale invariant representation of 6D (position and orientation) motion trajectories. The advantages of the proposed invariant representation are twofold. First the performance of gesture recognition can be improved thanks to its invariance to different viewpoints and different body sizes of the actors. Secondly, the proposed representation is bi-directional. Not only the original Cartesian trajectory can be converted into the 6 invariant values, but also the motion in the original space can be retrieved back from the invariants. While the former aspect handles robust human gesture recognition, the latter allows the execution of robot motions without the need to store the Cartesian data. Experimental results illustrate the effectiveness of the proposed invariant representation for gesture recognition and accurate trajectory reconstruction.

There is considerable freedom in choosing the sensors to be equipped on a robot. Currently many sensing technologies are available (radar, lidar, vision sensors, time-of-flight cameras, etc.). For each class, there are additional choices regarding the exact sensor parameters (spatial resolution, frame rate, etc.). Which sensor is best? In general, this question needs to be qualified. It depends on the task. In an estimation task, the answer depends on the prior for the signal. In a control task, the answer depends exactly on which are the sufficient statistics for computing the control signal. This paper shows that an ulterior qualification that needs to be made: the answer depends on the power available for sensing, even when the task is fixed. We define the “power-performance” curve as the performance attainable on a task for a given level of sensing power. We show that this approach is well suited to comparing a traditional CMOS sensor with the recently available “neuromorphic” sensors. We discuss estimation tasks with different priors for the signal. We find priors for which one sensor dominates the other and vice-versa, priors for which they are equivalent, and priors for which the answer depends on the power available. This shows that comparing sensors is a quite delicate problem. It also suggests that the optimal architecture might have more that one sensor, and would switch sensors on and off according to the performance level required instantaneously.

Recognizing the human gait is an edge research in robotics and rehabilitation. It has been popular to recognize the human gait from kinematic data. However, the recognition from muscle activities, the input of the movement, has not been widely approached. In this paper, we propose a framework to recognize the human walking and running movements by investigating muscle activities through electromyography (EMG). The framework is a Hidden Markov Model (HMM) topology utilizing Locally Linear Embedding (LLE) technique to extract feature vectors. We show that: (1) the high-dimensional EMG data can be embedded into a lower-dimensional space by using a manifold learning algorithm (LLE), primitive components which give meaningful representation of the EMG can be extracted, and (2) our proposed HMM topology whose input are the extracted vectors from EMG can recognize the gait movement at an accuracy rate of over 80%.

Computational thinking is an important part of a modern education, and robotics provides a powerful tool for teaching programming logic in an interactive and engaging way. The robot garden presented in this paper is a distributed multi-robot system capable of running autonomously or under user control from a simple graphical interface. Over 100 origami flowers are actuated with LEDs and printed pouch motors, and are deployed in a modular array around additional swimming and crawling folded robots. The garden integrates state-of-the-art rapid design and fabrication technologies with distributed systems software techniques to create a scalable swarm in which robots can be controlled individually or as a group. The garden can be used to teach basic algorithmic concepts through its distributed algorithm demonstration capabilities and can teach programming concepts through its education-oriented user interface.

Observing human activities can reveal a lot about the structure of the environment, the objects contained therein and also their functionality. This knowledge, in turn, can be useful for robots interacting with humans or for robots performing mobile manipulation tasks. In this paper, we present an approach to infer the geometric and functional structure of the environment and the position of certain relevant objects in it from human activity. We observe this activity using a full-body motion capture system consisting of a set of inertial measurement units. This is a hard problem since our data suit provides odometry estimates only, which severely drift over time. Therefore, we regard the objects inferred from the activities as landmarks in a graph-based simultaneous localization and mapping problem, which we optimize to obtain accurate estimates about the poses of the objects and the trajectory of the human. In extensive experiments, we demonstrate the effectiveness of the proposed method for the reconstruction of 3D representations. The resulting models not only contain a geometric but also a functional description of the environment and naturally provide a segmentation into individual objects.

We present a method for the three-dimensional (3D) steering of a flexible needle under 3D ultrasound guidance. The proposed solution is based on a duty-cycling visual servoing strategy we designed in a previous work, and on a new needle tracking algorithm for 3D ultrasound. The flexible needle modeled as a polynomial curve is tracked during automatic insertion using particle filtering. This new tracking algorithm enables real-time closed-loop needle control with 3D ultrasound feedback. Experimental results of a targeting task demonstrate the robustness of the proposed tracking algorithm and the feasibility of 3D ultrasound-guided needle steering.

This paper introduces an automatic method to visualize 3D needle shapes for reliable assessment of needle placement during needle insertion procedures. Based on partial observations of the needle within a small sample of 2D transverse ultrasound images, the 3D shape of the entire needle is reconstructed. An intensity thresholding technique is used to identify points representing possible needle locations within each 2D ultrasound image. Then, a Random Sample and Consensus (RANSAC) algorithm is used to filter out false positives and fit the remaining points to a polynomial model. To test this method, a set of 21 transverse ultrasound images of a brachytherapy needle embedded within a transparent tissue phantom are obtained and used to reconstruct the needle shape. Results are validated using camera images which capture the true needle shape. For this experimental data, obtaining at least three images from an insertion depth of 50 mm or greater allows the entire needle shape to be calculated with an average error of 0.5 mm with respect to the measured needle curve obtained from the camera image. Future work and application to robotics is also discussed.

This paper addresses the challenge of detecting and localizing a poorly textured known object, by initially estimating its complete 3D pose in a video sequence. Our solution relies on the 3D model of the object and synthetic views. The full pose estimation process is then based on foreground/background segmentation and on an efficient probabilistic edge-based matching and alignment procedure with the set of synthetic views, classified through an unsupervised learning phase. Our study focuses on space robotics applications and the method has been tested on both synthetic and real images, showing its efficiency and convenience, with reasonable computational costs.

This paper presents the mechanical design of a modular robot called the 3D M-Block, a 50mm cubic module capable of both independent and lattice-based locomotion. The first M-Blocks described in [1] could pivot about one axis of rotation only. In contrast, the 3D M-blocks can exert on demand both forward and backward torques about three orthogonal axes, for a total of six directions. The 3D M-Blocks transform these torques into pivoting motions which allow the new 3D M-Blocks to move more freely than their predecessors. Individual modules can employ pivoting motions to independently roll across a wide variety of surfaces as well as to join and move relative to other M-Blocks as part of a larger collective structure. The 3D M-Block maintains the same form factor and magnetic bonding system as the one-dimensional M-Blocks [1], but a new fabrication process supports more efficient and precise production. The 3D M-blocks provide a robust and capable modular self-reconfigurable robotic platform able to support swarm robot applications through individual module capabilities and self-reconfiguring robot applications using connected lattices of modules.

We present a system for the detection of small and potentially obscured obstacles in vegetated terrain. The key novelty of this system is the coupling of a volumetric occupancy map with a 3D Convolutional Neural Network (CNN), which to the best of our knowledge has not been previously done. This architecture allows us to train an extremely efficient and highly accurate system for detection tasks from raw occupancy data. We apply this method to the problem of detecting safe landing zones for autonomous helicopters from LiDAR point clouds. Current methods for this problem rely on heuristic rules and use simple geometric features. These heuristics break down in the presence of low vegetation, as they do not distinguish between vegetation that may be landed on and solid objects that should be avoided. We evaluate the system with a combination of real and synthetic range data. We show our system outperforms various benchmarks, including a system integrating various hand-crafted point cloud features from the literature.

This work aims to accurately match two scene images under large viewpoint changes, which is the key issue in appearance-based localization tasks. In this paper, two key ideas are proposed to solve the challenging problem. First, to detect extreme small overlapping regions between two images, a new approach is developed to estimate the camera motion using only two pairs of matched features, while the state-of-art needs at least five. Second, proper prior knowledge to the environmental structures is utilized to strengthen the outlier rejection. The proposed 2-point approach is tested on challenging scenes and shows good robustness to the drastic occlusion and scaling caused by viewpoint changes.