IROS 2017 论文列表

Pocket-drones are inherently safe for flight near humans, and their small size allows maneuvering through narrow indoor environments. However, achieving autonomous flight of pocket-drones is challenging because of strict on-board hardware limitations. Further challenges arise when multiple pocket-drones operate as a team and need to coordinate their movements. This paper presents a set-up that can achieve autonomous flight in an indoor environment with avoidance of both static obstacles and other pocket-drones. The pocket-drones use only on-board sensing and processing implemented on a STM32F4 microprocessor (168MHz). Experiments were conducted with two 40g pocket-drones flying autonomously in a real-world office while avoiding walls, obstacles, and each-other.

In this paper, a biologically inspired 3D slithering gait for a snake-like robot is designed and implemented for the purpose of target tracking. First, by balancing the forward speed and the stability of the robot, a straight slithering gait is modelled, under which the robot can march straight, fast, and stably. Then, for the purpose of steering, the straight slithering gait is modified into a biased slithering gait. The relationship between turning radius and gait parameters is analyzed by the resistive force theory. With the head composition algorithm, we investigate the orientation problem of the snake robot's head module to obtain stable visual information during the locomotion process. Finally, with the guidance of the vision sensor mounted in the head module, target tracking simulations and prototype experiments are conducted to demonstrate the practicality and effectiveness of the slithering gait in autonomous locomotion scenarios.

Autonomous Underwater Vehicles (AUVs) have been widely used in scientific or industrial operations over the past years. AUV missions require the vehicle to perform a set of actions autonomously and return to the operators. These missions are currently mostly planned offline by the human operators. Existing solutions in commercial planning software usually only provide an estimate of the mission execution time. The uncertain and dynamic underwater environment can have an effect on the mission performance. More time and energy may be required, disallowing successful mission execution. This work proposes the usage of the correlated orienteering problem (COP) that maximises the utility of a sensing mission while respecting energy and time constraints. We propose a heuristic-based on genetic algorithms (GA) for the solution of the COP. This heuristic is compared against optimal mixed-integer quadratic programming (MIQP) solutions. Results show that the quality of the heuristic solution is in the worst tested case 5.5% less than the 1% optimal solutions. The heuristic proves to be at least 3 times more time efficient than the optimal MIQP solutions in the worst case. The heuristic is finally tested on an embedded platform showing its ability to be used on real robotic platforms.

Modern robotic systems often consist of a growing set of information-producing components that need to be appropriately connected for the system to function properly. This is commonly done manually or through relatively simple scripts by specifying explicitly which components to connect. However, this process is cumbersome and error-prone, does not scale well as more components are introduced, and lacks flexibility and robustness at run-time. This paper presents an algorithm for setting up and maintaining implicit subscriptions to information through its semantics rather than its source, which we call semantic subscriptions. The proposed algorithm automatically reconfigures the system when necessary in response to changes at run-time, making the semantic subscriptions adaptive to changing circumstances. To illustrate the effectiveness of adaptive semantic subscriptions, we present a case study with two SoftBank Robotics NAO robots for handling the cases when a component stops working and when new components, in this case a second robot, become available. The solution has been implemented as part of a stream reasoning framework integrated with the Robot Operating System (ROS).

When designing a user-friendly Mobility Assistive Device (MAD) for mobility constrained people, it is important to take into account the diverse spectrum of disabilities, which results to completely different needs to be covered by the MAD for each specific user. An intelligent adaptive behavior is necessary. In this work we present experimental results, using an in house developed methodology for assessing the gait of users with different mobility status while interacting with a robotic MAD. We use data from a laser scanner, mounted on the MAD to track the legs using Particle Filters and Probabilistic Data Association (PDA-PF). The legs' states are fed to an HMM-based pathological gait cycle recognition system to compute in real-time the gait parameters that are crucial for the mobility status characterization of the user. We aim to show that a gait assessment system would be an important feedback for an intelligent MAD. Thus, we use this system to compare the gaits of the subjects using two different control settings of the MAD and we experimentally validate the ability of our system to recognize the impact of the control designs on the users' walking performance. The results demonstrate that a generic control scheme does not meet every patient's needs, and therefore, an Adaptive Context-Aware MAD (ACA MAD), that can understand the specific needs of the user, is important for enhancing the human-robot physical interaction.

In this paper we examine joint torque tracking as well as estimation of external torques for the KUKA Lightweight Robot (LBR) IIWA. To support physical human-robot interaction tasks, we need smooth estimation that allows detection of delicate external events and good control to hide inertial forces. Unfortunately a transmission nonlinearity in the motor to joint gearing injects vibrations and limits the performance of the built-in torque controller and observer. We confirm the nonlinearity to be a spatially periodic deflection between the motor and joint. Identification of this behavior allows us to generate more accurate joint position measurements. We also design a matching spatial filter to remove the vibrations from joint torque measurements. Experiments on an LBR IIWA show that compensating for the nonlinearity provides smoother external torque estimates and improves the torque tracking performance. Furthermore, we are able to increase the gain margin more than three fold over the built-in controller.

We present a micro aerial vehicle (MAV) system, built with inexpensive off-the-shelf hardware, for autonomously following trails in unstructured, outdoor environments such as forests. The system introduces a deep neural network (DNN) called TrailNet for estimating the view orientation and lateral offset of the MAV with respect to the trail center. The DNN-based controller achieves stable flight without oscillations by avoiding overconfident behavior through a loss function that includes both label smoothing and entropy reward. In addition to the TrailNet DNN, the system also utilizes vision modules for environmental awareness, including another DNN for object detection and a visual odometry component for estimating depth for the purpose of low-level obstacle detection. All vision systems run in real time on board the MAV via a Jetson TX1. We provide details on the hardware and software used, as well as implementation details. We present experiments showing the ability of our system to navigate forest trails more robustly than previous techniques, including autonomous flights of 1 km.

We consider an autonomous mapping and exploration problem in which a range-sensing mobile robot is guided by an information-based controller through an a priori unknown environment, choosing to collect its next measurement at the location estimated to yield the maximum information gain within its current field of view. We propose a novel and time-efficient approach to predict the most informative sensing action using a deep neural network. After training the deep neural network on a series of thousands of randomly-generated “dungeon maps”, the predicted optimal sensing action can be computed in constant time, with prospects for appealing scalability in the testing phase to higher dimensional systems. We evaluated the performance of deep neural networks on the autonomous exploration of two-dimensional workspaces, comparing several different neural networks that were selected due to their success in recent ImageNet challenges. Our computational results demonstrate that the proposed method provides high efficiency as well as accuracy in selecting informative sensing actions that support autonomous mobile robot exploration.

The twisted string actuator (TSA) shows promise as a good substitution for biological muscle in robotics. In this paper, a new force sensing mechanism is suggested utilizing the separator which exists in some kind of TSAs. We have identified a relation between the force acting on the separator and the load on the actuator and propose a mathematical model based on this original phenomenon. We also propose an estimation method of the load acting on the TSA using the mathematical model. This relationship has been verified in experiment. A contribution of this research is that the actuation load can be estimated by use of internal structure in TSA which in some circumstances can result in a simpler system compared to direct measurement of load in line.

We address the integrated planning and control problem that enables a single follower robot (the “photographer”) to maintain a moving target (the “subject”) in its field of view for as long as possible. We propose a real-time pursuit algorithm that seamlessly handles the often neglected, yet unavoidable, scenario in which the target escapes the follower's field of view; a scenario that simple, reactive controllers are ill-equipped to handle. Our algorithm aims to minimize the expected time until visual contact is re-established, which enables the photographer to track the subject for as long as possible, even in the presence of loss of visibility. At the core of our pursuit algorithm is an efficient method for sampling plausible trajectories from different homotopy classes. We do this by generating topologically distinct shortest paths by using the Voronoi diagram. We use these paths to make informed, model-based predictions of the likely future locations of the target, given a history of observations. Given these predictions, our algorithm produces pursuit trajectories that approximately minimize the expected time to recover visual contact. We show that constraining the predictive pursuit problem to the space of homotopy classes condenses the expanse of possibilities that our algorithm must consider, which enables target tracking in large occupancy grids, as opposed to many POMDP methods that are constrained to small environments. We benchmark the tracking behavior of our algorithm against the baseline of human subjects who performed the same set of pursuit tasks in simulation, as well as against two other pursuit algorithms that only take into account paths from a single homotopy class. We show that considering homotopy alternatives in 2D pursuit improves the tracking performance and that our algorithm does at least as well as humans in most pursuit scenarios.

This paper introduces a topological localization algorithm that uses visual and Wi-Fi data. Its main contribution is a novel way of merging data from these sensors. By making Wi-Fi signature suited to FABMAP algorithm, it develops an early-fusion framework that solves global localization and kidnapped robot problem. The resulting algorithm is tested and compared to FABMAP visual localization, over data acquired by a Pepper robot in an office building. Several constraints were applied during acquisition to make the experiment fitted to real-life scenarios. Without any tuning, early-fusion surpasses the performances of visual localization by a significant margin: 94% of estimated localizations are less than 5m away from ground truth compared to 81% with visual localization.

Real-time tool segmentation from endoscopic videos is an essential part of many computer-assisted robotic surgical systems and of critical importance in robotic surgical data science. We propose two novel deep learning architectures for automatic segmentation of non-rigid surgical instruments. Both methods take advantage of automated deep-learning-based multi-scale feature extraction while trying to maintain an accurate segmentation quality at all resolutions. The two proposed methods encode the multi-scale constraint inside the network architecture. The first proposed architecture enforces it by cascaded aggregation of predictions and the second proposed network does it by means of a holistically-nested architecture where the loss at each scale is taken into account for the optimization process. As the proposed methods are for real-time semantic labeling, both present a reduced number of parameters. We propose the use of parametric rectified linear units for semantic labeling in these small architectures to increase the regularization of the network while maintaining the segmentation accuracy. We compare the proposed architectures against state-of-the-art fully convolutional networks. We validate our methods using existing benchmark datasets, including ex vivo cases with phantom tissue and different robotic surgical instruments present in the scene. Our results show a statistically significant improved Dice Similarity Coefficient over previous instrument segmentation methods. We analyze our design choices and discuss the key drivers for improving accuracy.

This paper proposes a method of tire force estimation designed for use in navigation decision making for Unmanned Ground Vehicles (UGVs) operating on off-road terrains. This method (called the Centroid Method) uses a 3D point cloud representation of the terrain to determine tire-ground interaction, and a linear spring to determine the corresponding tire forces. Simulation results of tire force estimation of the Centroid Method are compared against the experimentally validated MF-Swift model, including on a fractal terrain surface which simulates a rough off-road terrain. The results show that the Centroid Method performs well for longitudinal and vertical tire force estimation especially for lower frequencies up to about 8-21Hz. Although the Centroid Method does not provide the same detail as the advanced tire models such as MF-Swift, it is significantly less complex and simpler to implement for enabling efficient on-board UGV navigation decision making.

This paper presents methodologies for dexterous manipulation by an underactuated robotic soft gripper with a low DOF. While the softness and passive joints provide a stable grasp, the manipulations become challenging owing to the complexities of modeling and state estimation. To investigate the drawbacks, target manipulations using a support surface and gravity are defined in this study. By realizing the target manipulations, methodologies for utilizing both the support surface and gravity in manipulation are validated. For manipulation utilizing the support surface, the key states - at which the type of gripper motion primitive changes - are defined, and gripper motion at the key states is produced with a DNN-based motion generator. For manipulation utilizing gravity, the fall risk of the object is formulated as a constraint for a criterion function, and a parameters-online-exploration based methodology is adopted. The validity of the methodologies is demonstrated through several experiments.

We present a simple and efficient mathematical framework suitable for generating motion in the context of a variety of robotic motion tasks ranging from low-level motor control up to high-level locomotion planning. Our concept is based on a one-dimensional second-order model that allows analytic computation of its inverse dynamics while respecting physical constraints. This makes it a particularly useful tool for tasks that are expressed only as a start and goal state, such as animation key frames or way points in path planning. By means of time synchronization, the model extends easily to an arbitrary number of dimensions in a way that the target is reached in all dimensions at the same time. The framework excels in terms of execution time, which lies in the microsecond range even for high-dimensional trajectory generation tasks. We demonstrate our method in two different settings - full-body trajectory generation and path planning - and show its benefits in comparison with current state-of-the-art algorithms.

An actuated spine appears to be a critical component for maneuverability in quadruped animals. However, robotic systems have yet to capitalize on this mechanism. This research compares three different spine morphologies in the planar case, namely the rigid, revolute and prismatic spine. Using a wide range of robots sampled from the design space (200 robots sampled at random), large-scale trajectory optimization (60 seed points per robot per spine morphology) was used to determine the best spine morphology in terms of stride averaged acceleration. Bootstrapping was performed on the results to achieve a better statistical representation and this revealed that for 75% of the robots, a prismatic spine design is the most effective at rapid acceleration, followed by the revolute spine at 6% and rigid spine at 18%.

In recent years the robotics field has witnessed an interesting new trend. Several companies started the production of service robots whose aim is to cooperate with humans. The robots developed so far are either rather expensive or unsuitable for manipulation tasks. This article presents the result of a project which wishes to demonstrate the feasibility of an affordable humanoid robot. R1 is able to navigate, and interact with the environment (grasping and carrying objects, operating switches, opening doors etc). The robot is also equipped with a speaker, microphones and it mounts a display in the head to support interaction using natural channels like speech or (simulated) eye movements. The final cost of the robot is expected to range around that of a family car, possibly, when produced in large quantities, even significantly lower. This goal was tackled along three synergistic directions: use of polymeric materials, light-weight design and implementation of novel actuation solutions. These lines, as well as the robot with its main features, are described hereafter.

To improve the obstacle surmounting performance of the quad-rotor vehicle, this paper focuses on designing, kinematically and dynamically characterizing a novel deformable quad-rotor that is based on the scissor-like foldable structures. The foldable structure allows that the volume of the quad-rotor can be tuned to dynamically adapt variously sized obstacles and small spaces. To generate the controllable deformation, the actuated angulated elements that are the essential components of the scissor-like foldable structure play an important role. The element design, its actuation mechanism and the corresponding configuration patterns for the new quad-rotor are presented in the paper in detail. The simulations on deformation properties and obstacle surmounting ability are then performed to verify the deformation capability of the structure. In addition, experiments were extensively conducted to test the controlled deformation of the structure as well to investigate the deformation induced effects to the activated quad-rotor airframe and its aerodynamics. All implementation results validate the effectiveness of the proposed deformable quad-rotor design, that is, it enables the new quad-rotor having excellent obstacle surmounting performance, adaptability, flight maneuverability, as well as minimal aerodynamics influences during deforming.