IROS 2018 论文列表

Search and rescue operations in the context of emergency response to human or natural disasters have the major goal of finding potential victims in the shortest possible time. Multi-agent teams, which can include specialized human respondents, robots and canine units, complement the strengths and weaknesses of each agent, like all-terrain mobility or capability to locate human beings. However, efficient coordination of heterogeneous agents requires specific means to locate the agents, and to provide them with the information they require to complete their mission. The major contribution of this work is an application of Wireless Sensor Networks (WSN) to gather information from a multi-agent team and to make it available to the rest of the agents while keeping coverage. In particular, a canine agent has been equipped with a mobile node installed on a harness, providing information about the dog's location as well as gas levels. The configuration of the mobile node allows for flexible arrangement of the system, being able to integrate static as well as mobile nodes. The gathered information is available at an external database, so that the rest of the agents and the control center can use it in real time. The proposed scheme has been tested in realistic scenarios during search and rescue exercises.

In this work we propose a method for integrating motion planning and in-hand manipulation. Commonly addressed as a separate step from the final execution, in-hand manipulation allows the robot to reorient an object within the end-effector for the successful outcome of the goal task. A joint achievement of repositioning the object and moving the manipulator towards its desired final pose saves time in the execution and introduces more flexibility in the system. We address this problem using a pivoting strategy (i.e. in-hand rotation)for repositioning the object and we integrate this strategy with a path planner for the execution of a complex task. This method is applied on a Baxter robot and its efficacy is shown by experimental results.

In partially observed environments, it can be useful for a human to provide the robot with declarative information that represents probabilistic relational constraints on properties of objects in the world, augmenting the robot's sensory observations. For instance, a robot tasked with a search-and-rescue mission may be informed by the human that two victims are probably in the same room. An important question arises: how should we represent the robot's internal knowledge so that this information is correctly processed and combined with raw sensory information? In this paper, we provide an efficient belief state representation that dynamically selects an appropriate factoring, combining aspects of the belief when they are correlated through information and separating them when they are not. This strategy works in open domains, in which the set of possible objects is not known in advance, and provides significant improvements in inference time over a static factoring, leading to more efficient planning for complex partially observed tasks. We validate our approach experimentally in two open-domain planning problems: a 2D discrete gridworld task and a 3D continuous cooking task. A supplementary video can be found at http://tinyurl.com/chitnis-iros-18.

We present a robotic system for picking a target from a pile of objects that is capable of finding and grasping the target object by removing obstacles in the appropriate order. The fundamental idea is to segment instances with both visible and occluded masks, which we call `instance occlusion segmentation'. To achieve this, we extend an existing instance segmentation model with a novel `relook' architecture, in which the model explicitly learns the inter-instance relationship. Also, by using image synthesis, we make the system capable of handling new objects without human annotations. The experimental results show the effectiveness of the relook architecture when compared with a conventional model and of the image synthesis when compared to a human-annotated dataset. We also demonstrate the capability of our system to achieve picking a target in a cluttered environment with a real robot.

Belt conveyors are transport systems composed by a plurality of belt lines. When those systems are used to transport fragile materials or liquid containers, it is necessary to minimize the oscillations of the transported objects in order to avoid damage, spillage and particle degradation. Those oscillations are regularly caused by steps (i.e. differences in level) at the joints of the belt lines, and for that reason, it is necessary to inspect those steps during installation and maintenance. Regular inspections involve the visual verification of the steps, which stops production, takes significant time, occasionally requires system disassembly and is subjected to human errors. In this paper, a novel belt conveyor inspection system which is able to detect and measure the steps at the joints of the belt lines is presented. This system consists in the acquirement of data of the belt conveyor with an inertial measurement unit (IMU), and the processing of this data with original algorithms for zero offset filtering, sensor progressing direction detection, step event detection and step height calculation. The presented system had successfully detected and measured the steps of a complex belt conveyor with an accuracy of ±0.3 mm. It is demonstrated that this cost-effective and ready to use system enables an automatic and prompt inspection of the whole belt conveyor system at once, thus reducing the workload, time and errors of the belt conveyor inspection.

In this paper, we present a novel approach to tightly couple visual and inertial measurements in a fixed-lag visual-inertial odometry (VIO) framework using information sparsification. To bound computational complexity, fixed-lag smoothers typically marginalize out variables, but consequently introduce a densely connected linear prior which significantly deteriorates accuracy and efficiency. Current state-of-the-art approaches account for the issue by selectively discarding measurements and marginalizing additional variables. However, such strategies are sub-optimal from an information-theoretic perspective. Instead, our approach performs a dense marginalization step and preserves the information content of the dense prior. Our method sparsifies the dense prior with a nonlinear factor graph by minimizing the information loss. The resulting factor graph maintains information sparsity, structural similarity, and nonlinearity. To validate our approach, we conduct real-time drone tests and perform comparisons to current state-of-the-art fixed-lag VIO methods in the EuRoC visual-inertial dataset. The experimental results show that the proposed method achieves competitive and superior accuracy in almost all trials. We include a detailed run-time analysis to demonstrate that the proposed algorithm is suitable for real-time applications.

Human teleoperation of robots and autonomous operations go hand in hand in today's service robots. While robot teleoperation is typically performed on low to medium levels of abstraction, automated planning has to take place on a higher abstraction level, i.e. by means of semantic reasoning. Accordingly, an abstract state of the world has to be maintained in order to enable an operator to switch seamlessly between both operational modes. We propose a novel approach that combines simulation based geometric tracking and semantic state inference by means of so called State Inference Entities to overcome this issue. We also demonstrate how Evolutionary Strategies can be employed to refine simulation parameters. All experiments are demonstrated in real-world experiments conducted with the humanoid robot Rollin' Justin.

This work addresses the design and implementation of a filter that estimates the orientation of the body-fixed z axis and the velocity of a quadrotor UAV from the inertial measurement unit (IMU) given a known yaw. The velocity and attitude estimation is possible since the filter employs a linear drag model measuring the drag forces on the quadrotor through the IMU. These forces are functions of the robot's velocity and attitude. In addition, the filter estimates the linear drag parameters and thrust coefficient for the propellers. These parameters may be fed back into a controller to improve tracking performance. Experimental results are used to validate the proposed approach.

Indoor localization for pedestrians has gained increasing popularity among the rich body of literature for the last decade. In this paper, a low-cost indoor mapping and localization solution is proposed using the opportunistic signals from ambient indoor environments with a smartphone. It is composed of GraphSLAM-based offline mapping and Bayesian filtering-based online localization using generated signal maps. The GraphSLAM front-end is constructed by motion constraints from pedestrian dead-reckoning (PDR), loop-closure constraints identified by magnetic sequence matching with WiFi signal similarity validation, and observation constraints from opportunistic magnetic headings after error rejection. Globally consistent trajectories are created by graph optimization, after which signal maps (e.g., WiFi, magnetic fields, lights) are generated by Gaussian Processes Regression (GPR) for later localization. We propose to use the pseudo-wall constraints from the GPR variance map of magnetic fields and the lights measurements as observations for particle filtering. The proposed method is evaluated on several datasets collected from both the in-compass office buildings and outside public areas. Real-time localization is demonstrated on a smartphone in an office building covering 2000 square meters with the 50- and 90-percentile accuracies being 2.30 m and 3.41 m, respectively.

Efficient skill acquisition, representation, and online adaptation to different scenarios has become of fundamental importance for assistive robotic applications. In the past decade, dynamical systems (DS) have arisen as a flexible and robust tool to represent learned skills and to generate motion trajectories. This work presents a novel approach to incrementally modify the dynamics of a generic autonomous DS when new demonstrations of a task are provided. A control input is learned from demonstrations to modify the trajectory of the system while preserving the stability properties of the reshaped DS. Learning is performed incrementally through Gaussian process regression, increasing the robot's knowledge of the skill every time a new demonstration is provided. The effectiveness of the proposed approach is demonstrated with experiments on a publicly available dataset of complex motions.

The Label Propagation (LP) algorithm, first introduced by Zhu and Ghahramani [1], is a semi-supervised method used in transductive learning scenarios, where all data are available already in the beginning. In this work, we present a novel extension of the LP algorithm for applications where data samples are observed sequentially - as is the case in autonomous driving. Specifically, our “Incremental Label Propagation” algorithm efficiently approximates the so called harmonic solution on a nearest-neighbor graph that is regularly updated by new labeled and unlabeled nodes. We achieve this by reformulating the original algorithm based on an active set of nodes and by introducing a threshold to decide whether the label of a given node should be updated or not. Our method can also deal with graphs that are not fully connected, and we give a formal convergence proof for this general case. In experiments on the challenging KITTI benchmark data stream, we show superior performance in terms of both test accuracy and number of required training labels compared to state-of-the-art online learning methods.

Collecting 3D object data sets involves a large amount of manual work and is time consuming. Getting complete models of objects either requires a 3D scanner that covers all the surfaces of an object or one needs to rotate it to completely observe it. We present a system that incrementally builds a database of objects as a mobile agent traverses a scene. Our approach requires no prior knowledge of the shapes present in the scene. Object-like segments are extracted from a global segmentation map, which is built online using the input of segmented RGB-D images. These segments are stored in a database, matched among each other, and merged with other previously observed instances. This allows us to create and improve object models on the fly and to use these merged models to reconstruct also unobserved parts of the scene. The database contains each (potentially merged) object model only once, together with a set of poses where it was observed. We evaluate our pipeline with one public dataset, and on a newly created Google Tango dataset containing four indoor scenes with some of the objects appearing multiple times, both within and across scenes.

3D visual understanding of the surrounding environment is vital for successful mobile robotic tasks such as autonomous navigation or general object interaction. However, current systems have limited perceptual capabilities in the sense that they are not very well adaptable to unknown environments. Human operators, on the other hand, are experts in adapting to previously unknown information. Hence, human-robot teaming in which the human helps the robot to adapt to new environments and the robot assists in automated object recognition to efficiently feed the control environment of the operator is advantageous. In this work, we propose an object recognition and localization system for mobile robots, based on deep learning, and we study the adaptation of the resulting robotic perception to a new environment. We propose two methods to teach the robot a new object category: using prior knowledge and using limited operator input. We conducted several experiments to show the feasibility of proposed methods.

The most widely used methods for toolpath planning in fused deposition 3D printing slice the input model into successive 2D layers in order to construct the toolpath. Unfortunately slicing-based methods can incur a substantial amount of wasted motion (i.e., the extruder is moving while not printing), particularly when features of the model are spatially separated. In recent work we have introduced a new paradigm that constructs the toolpath in 3D and prints local features to minimize wasted motion. Our algorithm is based on a local search and we have demonstrated substantial improvements in the efficiency of the resulting toolpaths. Our approach is amenable to incorporating physical constraints of the 3D printing process, and, in this paper we extend our approach to incorporate kinematic properties into toolpath optimization. With an accurate kinematic model of the extruder, our algorithm is able to model the real-world fabrication time of the model with a high degree of accuracy. To our knowledge, this toolpath optimization algorithm is the first to encode real-world fabrication time as the objective function. We demonstrate the real-world improvement in fabrication time that is possible with our algorithm on a benchmark of almost 600 models. We find improvement in nearly every toolpath generated for our benchmark set (with a mean of 3.2%), but substantially larger improvements for some models. To rationalize these results, we introduce a metric for model characterization that we call “oriented compactness” and show that it correlates positively with our observations. We believe this metric can be an important tool in the setup of fabrication (e.g., by guiding an orientation search of the model).

While there has been significant progress in solving the problems of image pixel labeling, object detection and scene classification, existing approaches normally address them separately. In this paper, we propose to tackle these problems from a bottom-up perspective, where we simply need a semantic segmentation of the scene as input. We employ the DeepLab architecture, based on the ResNet deep network, which leverages multi-scale inputs to later fuse their responses to perform a precise pixel labeling of the scene. This semantic segmentation mask is used to localize the objects and to recognize the scene, following two simple yet effective strategies. We evaluate the benefits of our solutions, performing a thorough experimental evaluation on the NYU Depth V2 dataset. Our approach achieves a performance that beats the leading results by a significant margin, defining the new state of the art in this benchmark for the three tasks comprising the scene understanding: semantic segmentation, object detection and scene categorization.

Behavior Trees (BTs) have become a popular framework for designing controllers of autonomous agents in the computer game and in the robotics industry. One of the key advantages of BTs lies in their modularity, where independent modules can be composed to create more complex ones. In the classical formulation of BTs, modules can be composed using one of the three operators: Sequence, Fallback, and Parallel. The Parallel operator is rarely used despite its strong potential against other control architectures such as Finite State Machines. This is due to the fact that concurrent actions may lead to unexpected problems similar to the ones experienced in concurrent programming. In this paper, we outline how to tackle the aforementioned problem by introducing Concurrent BTs (CBTs) as a generalization of BTs in which we include the notions of progress and resource usage. We show how CBTs allow safe concurrent executions of actions and we analyze the approach from a mathematical standpoint. To illustrate the use of CBTs, we provide a set of use cases in realistic robotics scenarios.

Indoor mobile robots navigation must use external sensors to complement odometry. This paper analyses two different external sensors such as a laser LMS-200 and an omnidirectional camera Mobotix C2S. Experiments with only one of these sensors and with both integrated are carried out on a tour guide robot in order to obtain conclusions about their contribution to robot pose estimation, and how to locate external landmarks in the environment.

We present an evaluation of several representative sampling-based and optimization-based motion planners, and then introduce an integrated motion planning system which incorporates recent advances in trajectory optimization into a sparse roadmap framework. Through experiments in 4 common application scenarios with 5000 test cases each, we show that optimization-based or sampling-based planners alone are not effective for realistic problems where fast planning times are required. To the best of our knowledge, this is the first work that presents such a systematic and comprehensive evaluation of state-of-the-art motion planners, which are based on a significant amount of experiments. We then combine different stand-alone planners with trajectory optimization. The results show that the combination of our sparse roadmap and trajectory optimization provides superior performance over other standard sampling-based planners' combinations. By using a multi-query roadmap instead of generating completely new trajectories for each planning problem, our approach allows for extensions such as persistent control policy information associated with a trajectory across planning problems. Also, the sub-optimality resulting from the sparsity of roadmap, as well as the unexpected disturbances from the environment, can both be overcome by the real-time trajectory optimization process.

The development of good experimental methodologies for robotics takes often inspiration from general principles of experimental practice. Repeatability prescribes that experiments should involve several trials in order to guarantee that results are not achieved by chance, but are systematic, and statistically significant trends can be identified. In this paper, we propose an approach to improve the repeatability of experiments performed in robotics. In particular, we focus on the domain of SLAM (Simultaneous Localization And Mapping) and we introduce a system that exploits simulations to generate a large number of test data on which SLAM algorithms are automatically evaluated in order to obtain consistent results, according to the principle of repeatability.