ICRA 2018 Papers

Although existing inertial motion-capture systems work reasonably well (≤10° error in Euler angles), their accuracy suffers when sensor positions change relative to the associated body segments (±60° mean error and 120° standard deviation). We attribute this performance degradation to undermined calibration values, sensor movement latency and displacement offsets. The latter specifically leads to incongruent rotation matrices in kinematic algorithms that rely on rotational transformations. To overcome these limitations, we propose to employ machine-learning techniques. In particular, we use multi-layer perceptrons to learn sensor-displacement patterns based on 3 hours of motion data collected from 12 test subjects in the lab over 215 trials. Furthermore, to compensate for calibration and latency errors, we directly process sensor data with deep neural networks and estimate the joint angles. Based on these approaches, we demonstrate up to 69% reduction in tracking errors.

This paper presents a viable approach for incorporating collision avoidance strategies into existing guidance and control systems on marine vessels. We propose a method that facilitates the use of simulation-based Model Predictive Control (MPC) for collision avoidance (COLAV) on marine vessels. Any COLAV strategy to be applied in real traffic must adhere to the international regulations for preventing collisions at sea (COLREGS). The proposed MPC COLAV method does not rely on an accurate model of the guidance system to achieve vessel behaviors that are compliant with the COLREGS. Rather, it depends on transitional costs in the MPC objective for collision avoidance maneuvers that are being executed by the marine vessel. Hence, it is straightforward to implement the MPC COLAV on different vessels without specific knowledge of the vessel's guidance strategy. Moreover, it offers the possibility to switch between different (possibly application specific) guidance strategies on the same vessel while running the same MPC COLAV algorithm. We present results from full scale experiments that show the viability of our method in different collision avoidance scenarios.

We present a low-drift visual odometry algorithm that separately estimates rotational and translational motion from lines, planes, and points found in RGB-D images. Previous methods estimate drift-free rotational motion from structural regularities to reduce drift in the rotation estimate, which is the primary source of positioning inaccuracy in visual odometry. However, multiple orthogonal planes are required to be visible throughout the entire motion estimation process; otherwise, these VO approaches fail. We propose a new approach to estimate drift-free rotational motion jointly from both lines and planes by exploiting environmental regularities. We track the spatial regularities with an efficient SO(3)-manifold constrained mean shift algorithm. Once the drift-free rotation is found, we recover the translational motion from all tracked points with and without depth by minimizing the de-rotated reprojection error. We compare the proposed algorithm to other state-of-the-art visual odometry methods on a variety of RGB-D datasets (including especially challenging pure rotations) and demonstrate improved accuracy and lower drift error.

Dynamic tactile displays represent a class of haptic devices of particular interest to the blind and visually impaired. Despite many years of research and development efforts, the low-cost, low-power, compact tactile display remains elusive. This paper describes a low-cost electromechanical actuator array based on low-force, low-displacement bistable actuators that can be realized using inexpensive, commercially available components and high-volume manufacturing processes. The array is applicable to several types of tactile displays; as an initial proof of concept, we demonstrate a 6-dot Braille cell with a 3mm pitch that requires less than 50mJ per actuator per transition.

Robust cross-seasonal localization is one of the major challenges in long-term visual navigation of autonomous vehicles. In this paper, we exploit recent advances in semantic segmentation of images, i.e., where each pixel is assigned a label related to the type of object it represents, to attack the problem of long-term visual localization. We show that semantically labeled 3D point maps of the environment, together with semantically segmented images, can be efficiently used for vehicle localization without the need for detailed feature descriptors (SIFT, SURF, etc.), Thus, instead of depending on hand-crafted feature descriptors, we rely on the training of an image segmenter. The resulting map takes up much less storage space compared to a traditional descriptor based map. A particle filter based semantic localization solution is compared to one based on SIFT-features, and even with large seasonal variations over the year we perform on par with the larger and more descriptive SIFT-features, and are able to localize with an error below 1 m most of the time.

In this paper we combine Gaussian process regression and impedance control, to illicit robust, anthropomorphic, adaptive control of a powered ankle prosthesis. We learn the non-linear manifolds which guide how locomotion variables temporally evolve, and regress that surface over a velocity range to create a manifold. The joint set of manifolds, as well as the temporal evolution of the gait-cycle duration is what we term a locomotion envelope. Current powered prostheses have problems adapting across speeds. It is likely that humans rely upon a control strategy which is adaptable, can become more robust and accurate with more data and provides a nonparametric approach which allows the strategy to grow with the number of observations. We demonstrate such a strategy in this study and successfully simulate locomotion well beyond our training data. The method we propose is based on common physical features observed in numerous human subjects walking at different speeds. Based on the derived locomotion envelopes we show that ankle power increases monotonically with speed among all subjects. We demonstrate our methods in simulation and human experiments, on a powered ankle foot prosthesis to demonstrate the effectiveness of the method.

Autonomous vehicles (AVs) require accurate metric and topological location estimates for safe, effective navigation and decision-making. Although many high-definition (HD) roadmaps exist, they are not always accurate since public roads are dynamic, shaped unpredictably by both human activity and nature. Thus, AVs must be able to handle situations in which the topology specified by the map does not agree with reality. We present the Variable Structure Multiple Hidden Markov Model (VSM-HMM) as a framework for localizing in the presence of topological uncertainty, and demonstrate its effectiveness on an AV where lane membership is modeled as a topological localization process. VSM-HMMs use a dynamic set of HMMs to simultaneously reason about location within a set of most likely current topologies and therefore may also be applied to topological structure estimation as well as AV lane estimation. In addition, we present an extension to the Earth Mover's Distance which allows uncertainty to be taken into account when computing the distance between belief distributions on simplices of arbitrary relative sizes.

This paper describes the design of a new integrity risk prediction/monitoring methodology for robot localization that uses feature extraction and data association algorithms. The work specifically addresses incorrect association faults when employing a local nearest neighbor data association algorithm. This approach is more efficient and easier to implement than previous work. The methodology is tested in simulation, showing that the computed upper bound on integrity risk is a performance metric capable of providing warnings when the safety of the system cannot be guaranteed.

This paper describes a novel behavior-based local navigation approach for rough off-road scenarios. Trajectory candidates are generated based on vehicle kinematics and dynamics as well as the desired global trajectory. In contrast to on-road local navigation approaches, the work at hand proposes the use of a shiftable elevation grid map instead of occupancy maps since traversability in rough terrains does not only depend on location, but also on the robot's orientation. The traversability is evaluated by determining tire contact points with the terrain to take various different safety and efficiency aspects like underbody collisions and rollover risk into account. By exploiting the behavior-based control paradigm, the navigation approach can be easily extended and its robustness is shown in experimental evaluations using an Unimog U5023.

This paper presents a method for localizing primitive shapes in a dense point cloud computed by the RGB-D SLAM system. To stably generate a shape map containing only primitive shapes, the primitive shape is incrementally modeled by fusing the shapes estimated at previous frames in the SLAM, so that an accurate shape can be finally generated. Specifically, the history of the fusing process is used to avoid the influence of error accumulation in the SLAM. The point cloud of the shape is then updated by fusing the points in all the previous frames into a single point cloud. In the experimental results, we show that metric primitive modeling in texture-less and unprepared environments can be achieved online.

This paper presents the design of a multirotor unmanned aerial vehicle (UAV) capable of landing semiautomatically on a power line while carrying a payload. The vehicle then rolls along the line to perform an inspection. Special attention is given to the vehicle's onboard vision system, which consists of a monocular camera and LiDAR used together to compute the pose of the vehicle relative to the power line. Landing assistance is provided to the pilot by a position-based visual controller that aligns and keeps the vehicle centered along the power line. The pilot remains in control of vertical and longitudinal movement during descent. The proposed approach was tested on a full-scale test line and shows promise for future applications of high value to the electric industry such as non-destructive testing of power transmission lines.

Localization is an important task for mobile service robots in indoor spaces. In this research, we propose a novel technique for indoor localization using a spherical camera. Spherical cameras can obtain a complete view of the surroundings allowing the use of global environmental information. We take advantage of this in order to estimate camera position and the orientation with respect to a known 3D line map of an indoor environment, using a single image. We robustly extract 2D line information from the spherical image via spherical-gradient filtering and match it to 3D line information in the line map. Our method requires no information about the 3D-2D line correspondences. In order to avoid a complicated six degrees of freedom (6 DoF) search for position and orientation, we use a Manhattan world assumption to decompose the line information in the image. The 6 DoF localization process is divided into two phases. First, we estimate the orientation by extracting the three principle directions from the image. Then, the position is estimated by robustly matching the distribution of lines between the image and the 3D model via a spherical Hough representation. This decoupled search can robustly localize a spherical camera using a single image, as we demonstrate experimentally.

To date, insect scale aerial robots have required wire tethers for providing power due to the challenges of integrating the required high-voltage power electronics within their severely constrained weight budgets. In this paper we present a significant milestone in the achievement of flight autonomy: the first wireless liftoff of a 190 mg aerial vehicle. Our robot is remotely powered using a 976 nm laser and integrates a complete power electronics package weighing a total of 104 mg, using commercially available components and fabricated using a fast-turnaround laser based circuit fabrication technique. The onboard electronics include a lightweight boost converter capable of producing high voltage bias and drive signals of over 200 V at up to 170 Hz and regulated by a microcontroller performing feedback control. We present our system design and analysis, detailed description of our fabrication method, and results from flight experiments.

Operating in rough, unstructured terrain is an essential requirement for any truly field-deployable ground robot. Search-and-rescue, border patrol and agricultural work all require operation in environments with little established infrastructure for easy navigation. This presents challenges for sensor-based navigation such as vision, where erratic motion and feature-poor environments test feature tracking and hinder the performance of repeat matching of point features. For vision-based route-following methods such as Visual Teach and Repeat (VT&R), maintaining similar visual perspective of salient point features is critical for reliable odometry and accurate localisation over long periods. In this paper, we investigate a potential solution to these challenges by integrating a gimbaled camera with VT&R on a Grizzly Robotic Utility Vehicle (RUV) for testing at high speeds and in visually challenging environments. We examine the benefits and drawbacks of using an actively gimbaled camera to attenuate image motion and control viewpoint. We compare the use of a gimbaled camera to our traditional fixed stereo configuration and demonstrate cases of improved performance in Visual Odometry (VO), localisation and path following in several sets of outdoor experiments.

State-of-the-art robotic perception systems have achieved sufficiently good performance using Inertial Measurement Units (IMUs), cameras, and nonlinear optimization techniques, that they are now being deployed as technologies. However, many of these methods rely significantly on vision and often fail when visual tracking is lost due to lighting or scarcity of features. This paper presents a state-estimation technique for legged robots that takes into account the robot's kinematic model as well as its contact with the environment. We introduce forward kinematic factors and preintegrated contact factors into a factor graph framework that can be incrementally solved in real-time. The forward kinematic factor relates the robot's base pose to a contact frame through noisy encoder measurements. The preintegrated contact factor provides odometry measurements of this contact frame while accounting for possible foot slippage. Together, the two developed factors constrain the graph optimization problem allowing the robot's trajectory to be estimated. The paper evaluates the method using simulated and real sensory IMU and kinematic data from experiments with a Cassie-series robot designed by Agility Robotics. These preliminary experiments show that using the proposed method in addition to IMU decreases drift and improves localization accuracy, suggesting that its use can enable successful recovery from a loss of visual tracking.

This paper presents a control strategy synthesis method for dynamical systems with differential constraints while satisfying a set of given rules in consideration of their importances. A special attention is given to situations where all rules cannot be met in order to fulfill a given task. Such dilemmas compel us to make a decision on the degree of satisfaction of each rule including which rule should be maintained or not. In this work, we propose a learning-based model predictive control method in order to solve this problem, where a key insight is to combine a learning method and traditional control scheme so that the designed controller behaves close to human experts. A rule is represented as a signal temporal logic (STL) formula. A robustness slackness, a margin to the satisfaction of the rule, is learned from expert's demonstrations using Gaussian process regression. The learned margin is used in a model predictive control procedure, which helps to decide how much to obey each rule, even ignoring specific rules. In track driving simulation, we show that the proposed method generates human-like behavior and efficiently handles dilemmas as human teachers do.

One of the main open challenges in visual odometry (VO) is the robustness to difficult illumination conditions or high dynamic range (HDR) environments. The main difficulties in these situations come from both the limitations of the sensors and the inability to perform a successful tracking of interest points because of the bold assumptions in VO, such as brightness constancy. We address this problem from a deep learning perspective, for which we first fine-tune a deep neural network with the purpose of obtaining enhanced representations of the sequences for VO. Then, we demonstrate how the insertion of long short term memory allows us to obtain temporally consistent sequences, as the estimation depends on previous states. However, the use of very deep networks enlarges the computational burden of the VO framework; therefore, we also propose a convolutional neural network of reduced size capable of performing faster. Finally, we validate the enhanced representations by evaluating the sequences produced by the two architectures in several state-of-art VO algorithms, such as ORB-SLAM and DSO.

Deep Reinforcement Learning (DRL) has been applied successfully to many robotic applications. However, the large number of trials needed for training is a key issue. Most of existing techniques developed to improve training efficiency (e.g. imitation) target on general tasks rather than being tailored for robot applications, which have their specific context to benefit from. We propose a novel framework, Assisted Reinforcement Learning, where a classical controller (e.g. a PID controller) is used as an alternative, switchable policy to speed up training of DRL for local planning and navigation problems. The core idea is that the simple control law allows the robot to rapidly learn sensible primitives, like driving in a straight line, instead of random exploration. As the actor network becomes more advanced, it can then take over to perform more complex actions, like obstacle avoidance. Eventually, the simple controller can be discarded entirely. We show that not only does this technique train faster, it also is less sensitive to the structure of the DRL network and consistently outperforms a standard Deep Deterministic Policy Gradient network. We demonstrate the results in both simulation and real-world experiments.

In the automation of many kinds of processes, the observable outcome can often be described as the combined effect of an entire sequence of actions, or controls, applied throughout the process execution. In these cases, strategies to optimise control policies for individual stages of the process are not applicable, and instead the whole policy needs to be optimised at once. On the other hand, the cost to evaluate the policy's performance might also be high, being desirable that a solution can be found with as few interactions as possible with the real system. We consider the problem of optimising control policies to allow a robot to complete a given race track within a minimum amount of time. We assume that the robot has no prior information about the track or its own dynamical model, just an initial valid driving example. Localisation is only applied to monitor the robot and to provide an indication of its position along the track's centre axis. With that in mind, we propose a method for finding a policy that minimises the time per lap while keeping the vehicle on the track using a Bayesian optimisation (BO) approach over a reproducing kernel Hilbert space. We apply an algorithm to search more efficiently over high-dimensional policy-parameter spaces with BO, by iterating over each dimension individually, in a sequential coordinate descent-like scheme. Experiments demonstrate the performance of the algorithm against other methods in a simulated car racing environment.