IROS 2020 论文列表

Current techniques in robot design and fabrication are time consuming and costly. Robot designs are needed that facilitate low-cost fabrication techniques and reduce the design to production timeline. Here we present an axial-rotational coupled metastructure that can serve as the functional core of a low-cost 3D printed walking robot. Using an origami-inspired assembly technique, the axial-rotational coupled metastructure robot can be 3D printed flat and then folded into a final configuration. This print-then-fold approach allows for the facile integration of critical subcomponents during the printing process. The axial-rotational metastructures eliminate the need for joints and linkages by enabling locomotion through a single compliant structure. Finite element models of the axialrotational metastructures were developed and validated against experimental deformation of 3D printed units under tensile loading. As a proof-of-concept, an ultra low-cost 3D-printed metabot was designed and fabricated using the proposed axial-rotational coupled metastructure and its walking performance was characterized. A top speed of 4.30 mm/s was achieved with an alternating stepping gait at a frequency of 0.8 Hz.

Small unmanned aerial vehicles (UAV) have penetrated multiple domains over the past years. In GNSS-denied or indoor environments, aerial robots require a robust and stable localization system, often with external feedback, in order to fly safely. Motion capture systems are typically utilized indoors when accurate localization is needed. However, these systems are expensive and most require a fixed setup. In this paper, we study and characterize an ultra-wideband (UWB) system for navigation and localization of aerial robots indoors based on Decawave's DWM1001 UWB node. The system is portable, inexpensive and can be battery powered in its totality. We show the viability of this system for autonomous flight of UAVs, and provide open-source methods and data that enable its widespread application even with movable anchor systems. We characterize the accuracy based on the position of the UAV with respect to the anchors, its altitude and speed, and the distribution of the anchors in space. Finally, we analyze the accuracy of the self-calibration of the anchors' positions.

Trajectory prediction has always been a challenging problem for autonomous driving, since it needs to infer the latent intention from the behaviors and interactions from traffic participants. This problem is intrinsically hard, because each participant may behave differently under different environments and interactions. This key is to effectively model the interlaced influence from both spatial context and temporal context. Existing work usually encodes these two types of context separately, which would lead to inferior modeling of the scenarios. In this paper, we first propose a unified approach to treat time and space dimensions equally for modeling spatio-temporal context. The proposed module is simple and easy to implement within several lines of codes. In contrast to existing methods which heavily rely on recurrent neural network for temporal context and hand-crafted structure for spatial context, our method could automatically partition the spatio-temporal space to adapt to the data. Lastly, we test our proposed framework on two recently proposed trajectory prediction dataset ApolloScape and Argoverse. We show that the proposed method substantially outperforms the previous state-of-the-art methods while maintaining its simplicity. These encouraging results further validate the superiority of our approach.

Anomaly detection is a key goal of autonomous surveillance systems that should be able to alert unusual observations. In this paper, we propose a holistic anomaly detection system using deep neural networks for surveillance of critical infrastructures (e.g., airports, harbors, warehouses) using an unmanned aerial vehicle (UAV). First, we present a heuristic method for the explicit representation of spatial layouts of objects in bird-view images. Then, we propose a deep neural network architecture for unsupervised anomaly detection (UAV-AdNet), which is trained on environment representations and GPS labels of bird-view images jointly. Unlike studies in the literature, we combine GPS and image data to predict abnormal observations. We evaluate our model against several baselines on our aerial surveillance dataset and show that it performs better in scene reconstruction and several anomaly detection tasks. The codes, trained models, dataset, and video will be available at https://bozcani.github.io/uavadnet.

Coverage path planning (CPP) is the task of designing a trajectory that enables a mobile agent to travel over every point of an area of interest. We propose a new method to control an unmanned aerial vehicle (UAV) carrying a camera on a CPP mission with random start positions and multiple options for landing positions in an environment containing no-fly zones. While numerous approaches have been proposed to solve similar CPP problems, we leverage end-to-end reinforcement learning (RL) to learn a control policy that generalizes over varying power constraints for the UAV. Despite recent improvements in battery technology, the maximum flying range of small UAVs is still a severe constraint, which is exacerbated by variations in the UAV's power consumption that are hard to predict. By using map-like input channels to feed spatial information through convolutional network layers to the agent, we are able to train a double deep Q-network (DDQN) to make control decisions for the UAV, balancing limited power budget and coverage goal. The proposed method can be applied to a wide variety of environments and harmonizes complex goal structures with system constraints.

This paper presents the modeling and analysis of a novel moving mechanism "tumbling" for asteroid exploration. The system actuation is provided by an internal motor and torque wheel; elastic spring-mounted spikes are attached to the perimeter of a circular-shaped robot, protruding normal to the surface and distributed uniformly. Compared with the conventional motion mechanisms, this simple layout enhances the capability of the robot to traverse a diverse microgravity environment. Technical challenges involved in conventional moving mechanisms, such as uncertainty of moving direction and inability to traverse uneven asteroid surfaces, can now be solved. A tumbling locomotion approach demonstrates two beneficial characteristics in this environment. First, tumbling locomotion maintains contact between the rover spikes and the ground. This enables the robot to continually apply control adjustments to realize precise and controlled motion. Second, owing to the nature of the mechanical interaction of the spikes and potential uneven surface protrusions, the robot can traverse uneven surfaces. In this paper, we present the dynamics modeling of the robot and analyze the motion of the robot experimentally and via numerical simulations. The results of this study help establish a moving strategy to approach the desired locations on asteroid surfaces.

We present TrueÆdapt, a model-free method to learn online adaptations of robot trajectories based on their effects on the environment. Given sensory feedback and future waypoints of the original trajectory, a neural network is trained to predict joint accelerations at regular intervals. The adapted trajectory is generated by linear interpolation of the predicted accelerations, leading to continuously differentiable joint velocities and positions. Bounded jerks, accelerations and velocities are guaranteed by calculating the range of valid accelerations at each decision step and clipping the network's output accordingly. A deviation penalty during the training process causes the adapted trajectory to follow the original one. Smooth movements are encouraged by penalizing high accelerations and jerks. We evaluate our approach by training a simulated KUKA iiwa robot to balance a ball on a plate while moving and demonstrate that the balancing policy can be directly transferred to a real robot.

In this paper, we introduce a novel learning-based approach for grasping known rigid objects in highly cluttered scenes and precisely placing them based on depth images. Our Placement Quality Network (PQ-Net) estimates the object pose and the quality for each automatically generated grasp pose for multiple objects simultaneously at 92 fps in a single forward pass of a neural network. All grasping and placement trials are executed in a physics simulation and the gained experience is transferred to the real world using domain randomization. We demonstrate that our policy successfully transfers to the real world. PQ-Net outperforms other model-free approaches in terms of grasping success rate and automatically scales to new objects of arbitrary symmetry without any human intervention.

Parallel robots with configurable platforms (PRCPs) combine the benefits of parallel robots with additional functionalities such as grasping and cutting. However, some of the theoretical tools used to study classical parallel robots do not apply to parallel robots with configurable platforms. This paper uses screw theory to study the transferable wrenches from the robot's limbs to the configurable platform of an 8-DoF parallel robot. Deriving the transferable wrenches allows one to construct the screw system that is applied to each part of the configurable platform. Based on the analytical expressions of the limb and platform wrenches that have been derived and numerically validated, the mathematical tools that are used to study parallel kinematic structures, such as Grassmann line geometry, can thus be applied to the presented parallel robot with a configurable platform.

Trajectory tracking of flexible link robots is a classical control problem. Historically, the link elasticity was considered as something to be removed. Hence, the control performance was guaranteed by adopting high-gain feedback loops and, possibly, a dynamic compensation with the result to stiffen up the dynamic behavior of the robot. Nowadays, robots are pushed more and more towards a safe physical interaction with a less and less structured environment. Hence, the design and control of the robots moved to an on-purpose introduction of highly compliant elements in the robot bodies, the so-called soft robotics, and towards control approaches that aim to provide the tracking performance without a substantial change in the robot dynamic behavior. Following this approach, we present an iterative learning control that relies mainly on a feedforward component, hence preserves the robot dynamics, for trajectory tracking of a one-link flexible arm. We provide a condition, based on the system dynamics and similar to the Strong Inertially Coupled property, that ensures the applicability of the proposed control method. Finally, we report simulation and experimental tests to validate the theoretical results.

A car driver knows how to react on the gestures of the traffic officers. Clearly, this is not the case for the autonomous vehicle, unless it has road traffic control gesture recognition functionalities. In this work, we address the limitation of the existing autonomous driving datasets to provide learning data for traffic control gesture recognition. We introduce a dataset that is based on 3D body skeleton input to perform traffic control gesture classification on every time step. Our dataset consists of 250 sequences from several actors, ranging from 16 to 90 seconds per sequence. To evaluate our dataset, we propose eight sequential processing models based on deep neural networks such as recurrent networks, attention mechanism, temporal convolutional networks and graph convolutional networks. We present an extensive evaluation and analysis of all approaches for our dataset, as well as real-world quantitative evaluation. The code and dataset is publicly available4.

One approach to control drug delivery in the cochlea is to use a magnetic microrobot powered by externally applied magnetic fields. However, it is necessary to integrate a localization system to ensure the precise navigation of the microrobot in the cochlear canal. To avoid integrating a clinical imaging modality for the navigation of microrobots in the cochlea, we propose in this work the application of magnetic sensors to localize the magnetic microrobot. In our method, we propose a real-time localization system based only on two sensors to keep a precise localization of the spherical magnetic microrobot. The first sensor measures both the magnetic field of the environment and the magnetic field generated by the microrobot (localization sensor). The second sensor (surrounding sensor) is placed away from the localization sensor, this sensor measures the magnetic field of the environment, which will be subtracted from the signal of the localization sensor to determine the value of the magnetic field of the microrobot. We have proposed a new magnetic sensor calibration method and a robust localization algorithm for precise localization of the microrobot. The experiments demonstrate the effectiveness of the designed system and show the precision of the proposed localization strategy.

Mitral regurgitation is one of the most common heart diseases caused by ventricular dysfunction or anatomic abnormality of the mitral valve. The fundamental treatment for mitral regurgitation is to repair/replace the mitral valve through open-heart surgery which is risky and requires more time to recover or through minimally invasive approaches, which have significant challenges and limitations. Through the transcatheter approach, the mitral valve implant is minimally invasively delivered directly to the mitral valve and is clamped onto the leaflet to mitigate or prevent regurgitation. However, this procedure requires delicate manipulation of the catheter in a constrained space and remains a challenging problem. In this work, we present a robotically steerable cathether design for the transcatheter procedure to address mitral regurgitation. The proposed catheter consists of two bending joints, one torsion joint, and implant delivery module at the distal end of the robot. Kinematic models for each joint design are derived and compared with experimental results. Finally, we experimentally demonstrate the feasibility of the proposed catheter to navigate in a phantom heart model. In this demonstration, the bending joint was actuated by 75°, the torsion joint was actuated by 90° and the implant was pushed out by 1.8 mm to deliver the implant.

The integration of computer vision techniques for the accomplishment of autonomous interaction tasks represents a challenging research direction in the context of aerial robotics. In this paper, we consider the problem of contact-based inspection of a textured target of unknown geometry and pose. Exploiting state of the art techniques in computer graphics, tuned and improved for the task at hand, we designed a framework for the projection of a desired trajectory for the robot end-effector on a generically-shaped surface to be inspected. Combining these results with previous work on energy-based interaction control, we are laying the basis of what we call vision-based impedance control paradigm. To demonstrate the feasibility and the effectiveness of our methodology, we present the results of both realistic ROS/Gazebo simulations and preliminary experiments with a fully-actuated hexarotor interacting with heterogeneous curved surfaces whose geometric description is not available a priori, provided that enough visual features on the target are naturally or artificially available to allow the integration of localization and mapping algorithms.

Surgical tool segmentation in endoscopic images is an important problem: it is a crucial step towards full instrument pose estimation and it is used for integration of pre- and intra-operative images into the endoscopic view. While many recent approaches based on convolutional neural networks have shown great results, a key barrier to progress lies in the acquisition of a large number of manually-annotated images which is necessary for an algorithm to generalize and work well in diverse surgical scenarios. Unlike the surgical image data itself, annotations are difficult to acquire and may be of variable quality. On the other hand, synthetic annotations can be automatically generated by using forward kinematic model of the robot and CAD models of tools by projecting them onto an image plane. Unfortunately, this model is very inaccurate and cannot be used for supervised learning of image segmentation models. Since generated annotations will not directly correspond to endoscopic images due to errors, we formulate the problem as an unpaired image-to-image translation where the goal is to learn the mapping between an input endoscopic image and a corresponding annotation using an adversarial model. Our approach allows to train image segmentation models without the need to acquire expensive annotations and can potentially exploit large unlabeled endoscopic image collection outside the annotated distributions of image/annotation data. We test our proposed method on Endovis 2017 challenge dataset and show that it is competitive with supervised segmentation methods.

Multi-human tracking in the crowded environment is a challenging problem due to occlusions, pose change, viewpoint variation and cluttered background. In this work, we propose a robust feature learning for tracking-by-detection methods based on second-order attention network that can capture higher-order relationships between salient features at the early stages of Convolutional Neural Network (CNN). Guided Second-Order Attention Network (GSAN) that, unlike the existing attention learning methods which are weakly-supervised, uses a supervisory signal based on the quality of the self-learned attention maps. More specifically, GSAN looks into the attended maps of a person having the highest confidence and supervise itself to look into the correct regions in the images of the person. Attention maps learned this way are spatially aligned and thus robust to camera-view changes and body pose variations. We verify the effectiveness of our approach by comparing with the state-of-the-art methods on challenging person re-identification and multi object tracking (MOT) datasets.

Providing explanations of chosen robotic actions can help to increase the transparency of robotic planning and improve users' trust. Social sciences suggest that the best explanations are contrastive, explaining not just why one action is taken, but why one action is taken instead of another. We formalize the notion of contrastive explanations for robotic planning policies based on Markov decision processes, drawing on insights from the social sciences. We present methods for the automated generation of contrastive explanations with three key factors: selectiveness, constrictiveness and responsibility. The results of a user study with 100 participants on the Amazon Mechanical Turk platform show that our generated contrastive explanations can help to increase users' understanding and trust of robotic planning policies, while reducing users' cognitive burden.

Object tracking has been broadly applied in unmanned aerial vehicle (UAV) tasks in recent years. However, existing algorithms still face difficulties such as partial occlusion, clutter background, and other challenging visual factors. Inspired by the cutting-edge attention mechanisms, a novel object tracking framework is proposed to leverage multi-level visual attention. Three primary attention, i.e., contextual attention, dimensional attention, and spatiotemporal attention, are integrated into the training and detection stages of correlation filter-based tracking pipeline. Therefore, the proposed tracker is equipped with robust discriminative power against challenging factors while maintaining high operational efficiency in UAV scenarios. Quantitative and qualitative experiments on two well-known benchmarks with 173 challenging UAV video sequences demonstrate the effectiveness of the proposed framework. The proposed tracking algorithm favorably outperforms 12 state-of-the-art methods, yielding 4.8% relative gain in UAVDT and 8.2% relative gain in UAV123@10fps against the baseline tracker while operating at the speed of ~28 frames per second.

This paper presents a method for processing sparse, non-Gaussian multimodal data in a simultaneous localization and mapping (SLAM) framework using factor graphs. Our approach demonstrates the feasibility of using a sum-product inference strategy to recover functional belief marginals from highly non-Gaussian situations, relaxing the prolific unimodal Gaussian assumption. The method is more focused than conventional multi-hypothesis approaches, but still captures dominant modes via multi-modality. The proposed algorithm exists in a trade space that spans the anticipated uncertainty of measurement data, task-specific performance, sensor quality, and computational cost. This work leverages several major algorithm design constructs, including clique recycling, to put an upper bound on the allowable computational expense – a major challenge in non-parametric methods. To better demonstrate robustness, experimental results show the feasibility of the method on at least two of four major sources of non-Gaussian behavior: i) the first introduces a canonical range-only problem which is always underdetermined although composed exclusively from Gaussian measurements; ii) a real-world AUV dataset, demonstrating how ambiguous acoustic correlator measurements are directly incorporated into a non-Gaussian SLAM solution, while using dead reckon tethering to overcome short term computational requirements.