IROS 2017 Papers

This paper presents an anisotropic shadow-based operation assistant method for a multilink-articulated wheeled pipeline-inspection robot by using a single illuminator and camera. By displacing the position of the illuminator relative to that of the head camera, a crescent-shaped shadow appears in the images captured in a bent pipe. The size, position, and orientation of the shadow depend on the robot's orientation around the pipe axis, and the shadow disappears in a certain robot's orientation (anisotropic shadow). Generally, as for shadow based navigation systems, disappearances of the shadow should be avoided because the robot loses its way. However, our previously developed robot (AIRo-2) adapts to a bent pipe without any control when the robot's orientation and the pathway direction of the bent pipe are aligned. By aligning those two specific orientations, we propose operation assistant system to pass through winding pipes. In this paper, the shadow region is extracted using two types of image binarization. The proposed system was experimentally verified in pipelines including seven bent pipes by applying the pathway direction of the bent pipe (obtained from the shadow) to the rolling movement of the robot.

3D laser scanners are frequently used sensors for mobile robots or autonomous cars and they are often used to perceive the static as well as dynamic aspects in the scene. In this context, matching 3D point clouds of objects is a crucial capability. Most matching methods such as numerous flavors of ICP provide little information about the quality of the match, i.e. how well do the matched objects correspond to each other, which goes beyond point-to-point or point-to-plane distances. In this paper, we propose a projective method that yields a probabilistic measure for the quality of matched scans. It not only considers the differences in the point locations but can also take free-space information into account. Our approach provides a probabilistic measure that is meaningful enough to evaluate scans and to cluster real-world data such as scans taken with Velodyne scanner in urban scenes in an unsupervised manner.

This study analyzed the precision grip force for the underactuated gripper for power and precision grasp (uGRIPP). uGRIPP is a robot hand developed for dual arm manipulation. It can grasp objects through power and precision grips. For a more detailed analysis of the performance on the grip force and control of the grip force, a relationship was derived between the grip and actuation forces by using an analytical method based on the principle of virtual power. Finally, theoretical values of the grip force were calculated according to the analysis, and the actual grip force was measured experimentally. The experimental results show the validity of the analysis.

This paper presents a model of navigational habituation for human passengers riding an autonomous robotic wheelchair. We present an example on how robot technology is utilized to understand human behavior which is useful to build human-like navigation models. The work consists of three parts, data collection from manual driving participants, habituation modeling and autonomous navigation experimentation. Manual driving data collection consisted of recording human driving control behavior using a robotic wheelchair in a complex labyrinth-like environment, while measuring physiological activity and emotional state via subjective reports (questionnaires). The purpose of this first part was to collect a sufficient representation of the evolution of human driving behavior in a complex environment (23 participants), such that a numerical model could be found to approximate the trend in velocity usage; velocity usage characterizes navigational habituation. The complex environment habituation model was tested and evaluated, on a different set of 16 participants, in a fully autonomous navigation setting where the robotic wheelchair followed a pre-recorded path. Self-reports and physiological measures confirmed that the proposed habitation velocity regulation model provides a more comfortable ride compared to slow and fast pre-defined static velocity profiles.

Underwater electrosense is bio-inspired by weakly electric fishes that use an electric field to see the objects in the water. Current studies on engineering electrosense focus on designing sophisticated sensors and algorithms for emulating biological functions including localization and identification. This work aimed to develop a planar sensor equipped with a dense electrode array that is capable of providing accurate and dense data for identifying objects of similar volume and aspect ratio, which has been a challenge in underwater sensing. After sensor design and implementation were presented, convolutional neural networks (CNN), which are widely used in digital image recognition, was trained using both simulation and experimental data. In the simulation, the overall success rate on identifying the sphere, cube, and rod is 92.6% by a 28 × 28 electrode array. In the preliminary experimental tests, a sensor with 16 × 16 electrode array achieved an overall success rate of 90.4% on identifying a sphere and a rod.

Multicopters stand to revolutionize parcel delivery because of their capability to operate in areas with unsuitable road infrastructure and precisely maneuver in cluttered environments. However, current multicopters for delivery can be dangerous for people, and are difficult to store and transport. Safety issues arise because users are exposed to unshielded spinning propellers. Transportation to the place of deployment and storage is often impaired by the large size that is required for heavy lifting. This paper addresses these limitations by proposing the integration of a quadcopter into a foldable protective cage. The cage provides an all-round protective structure that physically separates the propellers from the environment, ensuring the safety of people. The drone and the cage can be easily folded with a single movement, significantly reducing its size for ease of storage and transportation. This design has been validated with a quadcopter that can lift parcels up to 500 g and reduce its storage volume by 92% when folded.

We present a novel approach to trajectory generation that enables an autonomous vehicle to accurately follow a lead vehicle tracked by on-board sensors. In contrast to other approaches, we ignore the structure of the environment (e.g., lane markings), focusing purely on following the path driven by the vehicle ahead. Based on the leader's velocity, its distance to the ego vehicle and its recorded path, a continuous-curvature trajectory is generated using Sequential Quadratic Programming. As the optimization process takes the ego vehicle's kinematic and dynamic constraints into account, the resulting trajectory is guaranteed to be feasible and safe. The algorithm was tested extensively during practical experiments with an actual autonomous car. Our tests were conducted in both on- and off-road environments, with speeds ranging from 1 m/s to 15 m/s. Ground truth data shows that the system achieves a high degree of accuracy even in difficult scenarios.

Optical tracking is widely used in robotics applications such as unmanned aerial vehicle (UAV) localization. Unfortunately, such systems require many cameras and are, consequently, expensive. In this paper, we propose an approach to considerably increase the optical tracking volume without adding cameras. First, when the target becomes no longer visible by at least two cameras we propose a single-view tracking mode which requires only one camera. Furthermore, we propose to rely on controlled cameras able to track the UAV all around the volume to provide 6DoF tracking data through multi-view registration. This is achieved by using a visual servoing scheme. The two methods can be combined in order to maximize the tracking volume. We propose a proof-of-concept of such an optical tracking system based on two consumer-grade cameras and a pan-tilt actuator and we used this approach on UAV localization.

With the increasing demand of humans and robots collaborating in a confined workspace, it is essential that robots can react instantaneously to unforeseen events while respecting comfort of human co-workers. As the human pays especial attention to the end-effector movement rather than individual joints in such tasks, we argue that it is necessary to limit exactly these end-effector dynamics directly on SE(3), regardless of any specific coordinate system. Only then it is possible to constrain the robot end-effector in its absolute dynamic values, to enhance human comfort. Common online trajectory generators (OTGs) are either limited to single degrees of freedom (DOFs) or consider decoupled multi-DOF mainly in joint space. While these strategies can be directly applied to Cartesian coordinate systems by using Euler angles for orientation, constraining the change in Euler angles does not lead to the correct angular velocities. In this work we present a novel OTG with constraints on translational and rotational magnitudes on SE(3). Simulations as well as experiments on a KUKA LWR4+ show the potential contribution towards improved Human-Robot-Collaboration.

An objective performance measure for movement tasks is widely regarded as having utmost relevance for the therapy of movement disorders. Existing systems typically rely on human experts, which is known to produce substantial inter- and intra-rater variability. Present solutions are either based on simple features or invasive motion capture techniques. They typically work on a specific motion task only and fail to generalize to other tasks. In addition, they often require manual offline pre- and post-processing. In this paper we present a novel approach to compute a continuous and objective performance measure online during a patient session, without tedious and time-consuming pre- or post-processing steps. Our approach is able to generalize between different motion capture devices and different motion tasks. It runs on live motion data extracted with a non-invasive marker-less off-the-shelf vision-based tracking system as well as on data extracted from an inertial measurement unit suit. In the experiments we show that our approach is competitive with an offline approach as well as with the Unified Parkinson's Disease Rating Scale. Our approach is robust with respect to motion execution speed and it outperforms the offline approach regarding movement task generalization. We show promising results to track the current state of a Parkinson's subject online during a therapy session.

Using multiple cooperative robots is advantageous for time critical Search and Rescue (SaR) missions as they permit rapid exploration of the environment and provide higher redundancy than using a single robot. A considerable number of applications such as autonomous driving and disaster response could benefit from merging mapping data from several agents. Online multi-robot localization and mapping has mainly been addressed for robots equipped with cameras or 2D LiDARs. However, in unstructured and ill-lighted real-life scenarios, a mapping system can potentially benefit from a rich 3D geometric solution. In this work, we present an online localization and mapping system for multiple robots equipped with 3D LiDARs. This system is based on incremental sparse pose-graph optimization using sequential and place recognition constraints, the latter being identified using a 3D segment matching approach. The result is a unified representation of the world and relative robot trajectories. The complete system runs in real-time and is evaluated with two experiments in different environments: one urban and one disaster scenario. The system is available open source and easy-to-run demonstrations are publicly available.

The main contribution of this paper is an invariant extended Kalman filter (EKF) for visual inertial navigation systems (VINS). It is demonstrated that the conventional EKF based VINS is not invariant under the stochastic unobservable transformation, associated with a translation and a rotation about the gravitational direction. This can lead to inconsistent state estimates as the estimator does not obey a fundamental property of the physical system. To address this issue, we use a novel uncertainty representation to derive a Right Invariant error extended Kalman filter (RIEKF-VINS) that preserves this invariance property. RIEKF-VINS is then adapted to the multi-state constraint Kalman filter framework to obtain a consistent state estimator. Both Monte Carlo simulations and real-world experiments are used to validate the proposed method.

This work uses the intermediary quaternions in the design of a backstepping control technique with integral properties in order to perform an autonomous trajectory tracking using a quadcopter vehicle. Nowadays, in order to determine the orientation of a vehicle, most of the inertial systems of aircrafts can give directly the rotation matrix and taking advantage of this fact, the intermediary quaternions can be determined in a simple way from this matrix. Moreover, one specific orientation corresponds to only one intermediary quaternion and this helps to cope the unwinding phenomenon presented when working with the classical quaternions. The proposed control algorithm is validated numerically and experimentally when the quadrotor follows a circular trajectory. In addition, during the simulation part, some external perturbations and white noise were added in order to test the robustness of the algorithm.

Inspired by findings of sensorimotor coupling in humans and animals, there has recently been a growing interest in the interaction between action and perception in robotic systems [1]. Here we consider perception and action as two serial information channels with limited information-processing capacity. We follow [2] and formulate a constrained optimization problem that maximizes utility under limited information-processing capacity in the two channels. As a solution we obtain an optimal perceptual channel and an optimal action channel that are coupled such that perceptual information is optimized with respect to downstream processing in the action module. The main novelty of this study is that we propose an online optimization procedure to find bounded-optimal perception and action channels in parameterized serial perception-action systems. In particular, we implement the perceptual channel as a multi-layer neural network and the action channel as a multinomial distribution. We illustrate our method in a NAO robot simulator with a simplified cup lifting task.

Autonomous vehicles rely on precise high definition (HD) 3D maps for navigation. This paper presents the mapping component of an end-to-end system for crowdsourcing precise 3D maps with semantically meaningful landmarks such as traffic signs (6 dof pose, shape and size) and traffic lanes (3D splines). The system uses consumer grade parts, and in particular, relies on a single front facing camera and a consumer grade GPS. Using real-time sign and lane triangulation on-device in the vehicle, with offline sign/lane clustering across multiple journeys and offline Bundle Adjustment across multiple journeys in the backend, we construct maps with mean absolute accuracy at sign corners of less than 20 cm from 25 journeys. To the best of our knowledge, this is the first end-to-end HD mapping pipeline in global coordinates in the automotive context using cost effective sensors.

We demonstrate a flexible, electrostatic adhesive gripper designed to controllably grasp and manipulate soft goods in space. The 8-fingered gripper has 50 cm2 of active electrodes operating at 3 kV. It generates electrostatic adhesion forces up to 3.5 N (0.70 kPa) on Ge-coated polyimide film and 1.2 N on MLI blanket, a film composite used for satellite thermal insulation. Extremely low-force gripper engagement (0.08 N) and release (0.04 N) of films is ideal for micro-gravity. Individual fingers generate shear adhesion forces up to 4.76 N (5.04 kPa) using electrostatic adhesive and 45.0 N (47.6 kPa) with a hybrid electrostatic / gecko adhesive. To simulate a satellite servicing task, the gripper was mounted on a 7-DoF robot arm and performed a supervised grasp, manipulate, and release sequence on a hanging, Al-coated PET film.

The magnetic nanoparticles (MNPs) have been emerging as effective drug carriers. The electromagnetic actuated systems can steer the MNPs in vascular network for targeted drug delivery (TDD) systems. However, the real-time monitoring of MNPs remained as an open challenge which limits the application of these systems. In this paper, for the first time a 2D real-time imaging scheme has been developed for the MNPs guidance system. The proposed 2D navigation system based on magnetic particle imaging can allow real-time monitoring of MNPs and provides promising potentials for simultaneous monitoring and steering. In the experiments, MNPs of 45 to 65 nm diameter and 5 nm core (Iron) could be successfully monitored in real-time (update rate of 2 Hz) in 2D. The proposed 2D monitoring system can provide dynamic tracking of MNPs in targeted drug delivery and makes a MPI-based navigation system more feasible.

Recent advances in sampling-based motion planning have exploited concepts similar to those used in the heuristic graph search community, such as computing heuristic cost-to-go estimates and using state-space abstractions to derive them. Following this trend, we explore how the concept of search effort can be exploited to find plans quickly. Most previous work in motion planning attempts to find plans quickly by preferring states with low cost-to-go. Recent work in graph search suggests that following search effort-to-go estimates can yield faster planning. In this paper, we demonstrate how this idea can be adapted to the context of kinodynamic motion planning. Our planner, BEAST, uses estimates of effort that are learned on-line to guide the expansion of a motion tree toward states through which a plan is estimated to be easy to find. We present results with four different simulated vehicles (car, hovercraft, blimp and quadrotor) in six different environments indicating that BEAST is able to find solutions much more quickly and has a higher success rate than previous methods. We see this work as further strengthening the algorithmic connections between motion planning and heuristic graph search.

Image and point cloud registration methods compute the relative pose between two images. Commonly used registration algorithms are iterative and rely on the assumption that the motion between the images is small. In this work, we propose a fast pose estimation technique to compute a rough estimate of large motions between depth images, which can be used as initialization to dense registration methods. The main idea is to explore the properties given by planar surfaces with co-visibility and their normals from two distinct viewpoints. We present, in two decoupled stages, the rotation and then the translation estimation, both based on the normal vectors orientation and on the depth. These two stages are efficiently computed by using low resolution depth images and without any feature extraction/matching. We also analyze the limitations and observabilty of this approach, and its relationship to ICP point-to-plane. Notably, if the rotation is observable, at least five degrees of freedom can be estimated in the worst case. To demonstrate the effectiveness of the method, we evaluate the initialization technique in a set of challenging scenarios, comprising simulated spherical images from the Sponza Atrium model benchmark and real spherical indoor sequences.