This paper is concerned with the presentation of a parallel compliance adaptation method for systems equipped with rotary motion mechanisms towards obtaining energy efficiency in cyclic tasks over a reasonable range of task frequency variations. In this work, we first introduce a variable parallel elastic actuator (VPEA) design for implementation on uni-directional joints that can respond in line with

This work considers the problem of chaotic path planning, using an improved memory tehnique to boost performance. In this application, the dynamics of two simple chaotic maps are first used to generate a pseudo-random bit generator. Using this as a source, a series of navigation commands are generated and used by an autonomous robot to explore an area, while maintaining a random and unpredictable motion

This work presents a novel software architecture to autonomously identify and evaluate the gymnastic activity that people are carrying out. It is composed of three different interconnected layers. The first corresponds to a Multilayer Perceptron (MLP) trained from a set of angular magnitudes derived from the information provided by the OpenPose library. This library works frame by frame, so some postures

In-vehicle human-machine interface (HMI) mainly refers to the T-shaped panel system with instruments, center console, gear lever and other components installed. For intelligent vehicles, the high level of intelligent interconnection may to some extent make drivers lack situational safety awareness and reduce the usability of the system. Thus, this study attempted to establish a relationship between

This paper presents a trajectory generation algorithm with which a three-dimensional (3D) biped robot can perform aperiodic gaits by modifying only a small set of gait parameters. In addition to the double support phase (DSP), the gait can transit smoothly from one single support phase (SSP) to another. We decouple the sagittal and coronal dynamics, firstly. In the sagittal plane, the linear inverted

Replacement of lightning arrester is one of the common tasks in live-line maintenance, and peg-in-hole assembly is a very difficult operation for a robot, because there are visual inaccuracy and force model uncertainty in the process of assembly. This paper presents a new implementation approach fusing signals of vision detection and fuzzy force to realize the high efficiency peg-in-hole assembly by

The spread of the use of robotic devices in neuro-rehabilitation therapies requires the availability of lightweight, easy-to-use, cost-effective and versatile systems. RobHand has been designed with these goals in mind. It is a hand exoskeleton especially suitable for patients suffering from spasticity in the fingers since it is easy to place in the hand and, from an underactuated design, allows both

We present a real-time stereo visual-inertial-SLAM system which is able to recover from complicated kidnap scenarios and failures online in realtime. We propose to learn the whole-image-descriptor in a weakly supervised manner based on NetVLAD and decoupled convolutions. We analyze the training difficulties in using standard loss formulations and propose an allpairloss and show its effect through extensive

Robot grasping and manipulation require estimation of 3D object poses. Recently, a number of methods and datasets for vision-based pose estimation have been proposed. However, it is unclear how well the performance measures developed for visual pose estimation predict success in robot manipulation. In this work, we introduce an approach that connects error in pose and success in robot manipulation

We propose a representation for the set of forces a robot can counteract using full system dynamics: the residual force polytope. Given the nominal torques required by a dynamic motion, this representation models the forces which can be sustained without interfering with that motion. The residual force polytope can be used to analyze and compare the set of admissible forces of different trajectories

This paper proposes a unified vision-based manipulation framework using image contours of deformable/rigid objects. Instead of explicitly defining the features by geometries or functions, the robot automatically learns the visual features from processed vision data. Our method simultaneously generates – from the same data – both visual features and the interaction matrix that relates them to the robot

In order to possess a significant degree of autonomy, a robot must be able to perceive its environment and store a representation of that environment for use in tasks such as localisation, navigation, collision avoidance, and higher decision making. It must do this subject to constraints on memory and processing power typical of the embedded computer systems commonly found on small robotic devices

When data are organized in matrices or arrays of higher dimensions (tensors), classical regression methods first transform these data into vectors, therefore ignoring the underlying structure of the data and increasing the dimensionality of the problem. This flattening operation typically leads to overfitting when only few training data is available. In this paper, we present a mixture-of-experts model

We investigate a multi-agent planning problem, where each agent aims to achieve an individual task while avoiding collisions with other agents. Each agent’s task is expressed as a Time-Window Temporal Logic (TWTL) specification defined over a discretized environment. We propose a distributed receding horizon algorithm for online planning of agent trajectories. We show that under mild assumptions on

Modern laser SLAM (simultaneous localization and mapping) and structure from motion algorithms face the problem of processing redundant data. Even if a sensor does not move, it still continues to capture scans that should be processed. This paper presents the novel filter that allows dropping 2D scans that bring no new information to the system. Experiments on MIT and TUM datasets show that it is possible

Accurate prediction of acoustic communication performance is an important capability for marine robots. In this paper, we propose a model-based learning methodology for the prediction of underwater acoustic communication performance. The learning algorithm consists of two steps: (i) estimation of the covariance matrix by evaluating candidate functions with estimated parameters using detrended measurements;and

In the last decades, visual target tracking has been one of the primary research interests of the Robotics research community. The recent advances in Deep Learning technologies have made the exploitation of visual tracking approaches effective and possible in a wide variety of applications, ranging from automotive to surveillance and human assistance. However, the majority of the existing works focus

This paper introduces a concept of safe path planning for UAV’s autonomous operation in an urban environment where GNSS-positioning may become unreliable or even unavailable. If the operation environment is a priori known and geo-localized, it is possible to predict a GNSS satellite constellation and hence to anticipate its signal occlusions at a given point and time. Motivated from this, our main

Hand-assistive devices are used to help post-stroke victims encumbered with hand impairments perform activities of daily living (ADL). Unlike robotic rehabilitation devices used in restricted medical conditions for designated periods, hand-assistive devices are designed to be portable and to be used for extended periods by individuals engaging in ADL. Several hand-assistive device designs have been

In the problem of driving forces/torques distribution of the redundantly actuated parallel mechanisms (PMs), although numerous optimization analysis methods, including the minimum input torque method and minimum energy consumption method have been proposed so far, however, the actual control modes of the actuators were not taken into account among the existing methods for the above problem. Therefore

Imitation learning (IL) is a popular method used to train machine learning models that are capable of acting on their environment based on expert examples. Two types of IL models are inverse reinforcement learning (IRL) and behavioral cloning (BC). Models trained under IRL traditionally perform better than those trained under BC due to compounding covariate shift associated with the latter, which typically

We present an integrated Task-Motion Planning (TMP) framework for navigation in large-scale environments. Of late, TMP for manipulation has attracted significant interest resulting in a proliferation of different approaches. In contrast, TMP for navigation has received considerably less attention. Autonomous robots operating in real-world complex scenarios require planning in the discrete (task) space

Constraint-based control approaches offer a flexible way to specify robotic manipulation tasks and execute them on robots with many degrees of freedom. However, the specification of task constraints and their associated priorities usually requires a human-expert and often leads to tailor-made solutions for specific situations. This paper presents our recent efforts to automatically derive task constraints

Snake robots have advantages of terrain adaptability over wheeled mobile robots and traditional articulated robot arms because of their limbless thin body structure and high flexibility. They have extensive applications in tasks such as rescue, disaster recovery, inspection and minimally invasive surgery. Current research on snake robots is mainly focused on snake-like locomotion and the embodiment

Position resolution is a major problem in teleoperation applications with significant disparity between master and slave workspace. While rate mode control is suitable for slave free motion operations, it poses significant stability and performance challenges for task manipulation. In this paper, we propose a hybrid control scheme that offers seamless transition between position and rate control modes

A key feature in the context of simultaneous localization and mapping is loop-closure detection, a process determining whether the current robot’s environment perception coincides with previous observation. However, in long-term operations, both computational efficiency and memory requirements involved in an autonomous robot operation in uncontrolled environments, are of particular importance. The

A robotic system for the reduction of fractured femur bone is proposed in this research to help orthopedics during the labor intensive bone reduction procedures and also save them from radiation stimulated environment. Fractured femur reduction is a good candidate for robotics application owing to its elongated anatomy and strong counteracting forces from surrounding muscles. However, the robot forces

Programming robots to perform complex manipulation tasks is difficult because many tasks require sophisticated controllers that may rely on data such as manipulability ellipsoids, stiffness/damping and inertia matrices. Such data are naturally represented as Symmetric Positive Definite (SPD) matrices to capture specific geometric characteristics of the data, which increases the complexity of hard-coding

This paper presents a method for planning optimal trajectories with a team of Unmanned Aerial Vehicles (UAVs) performing autonomous cinematography. The method is able to plan trajectories online and in a distributed manner, providing coordination between the UAVs. We propose a novel non-linear formulation for this challenging problem of computing multi-UAV optimal trajectories for cinematography; integrating

This paper proposed an image-based visual servoing (IBVS) control law for a quadrotor that is equipped with a single monocular camera attached to its bottom. For control purposes virtual reticle plane (VRP) algorithm is used to track the relative 3D position of the quadrotor to the tilting and moving target landing platform within the range of the camera’s field of view (FOV). In this article, the

Real-time robotic grasping, supporting a subsequent precise object-in-hand operation task, is a priority target towards highly advanced autonomous systems. However, such an algorithm which can perform sufficiently-accurate grasping with time efficiency is yet to be found. This paper proposes a novel method with a 2-stage approach that combines a fast 2D object recognition using a deep neural network

For sensory data fusion, a calibration method between 3D light detection and ranging (LiDAR) and color camera based on ranging statistical characteristics and improved RANSAC algorithm is proposed. The multi-frame LiDAR point cloud data of the calibration triangular board are recorded. The scanned points with close angles are defined a cluster within same degrees. Furthermore, accurate points are preserved

In this paper, we present an online nonlinear Model Predictive Control (MPC) method for collision-free, deadlock-free navigation by multiple autonomous nonholonomic Wheeled Mobile Robots (WMRs). Our proposed method solves a nonlinear constrained optimization problem at each time step over a specified horizon to compute a sequence of optimal control inputs that drive the robots to target poses along

During execution, activity durations may vary from those predicted in the generated schedule. In this article we study (re) scheduling invocation, execution during rescheduling, and flexible execution to enable a high level of responsiveness to uncertainty in activity execution duration. We discuss these methods theoretically in the context of an embedded scheduler and practically in the context of

Swarms of autonomous agents are useful in many applications due to their ability to accomplish tasks in a decentralized manner, making them more robust to failures. Due to the difficulty in running experiments with large numbers of hardware agents, researchers often make simplifying assumptions and remove constraints that might be present in a real swarm deployment. While simplifying away some constraints

Autonomous unmanned systems and robots must be able to actively leverage all available information sources — including imprecise but readily available semantic observations provided by human collaborators. This work develops and validates a novel active collaborative human–machine sensing solution for robotic information gathering and optimal decision making problems, with an example implementation

This paper describes a system for service robots that combines ontological knowledge reasoning and human–robot interaction to interpret natural language commands and successfully perform household chores, such as finding and delivering objects. Knowledge and context reasoning is essential for providing more efficient service robots, given their diverse and continuously changing environments. Moreover

Dexterous grasping is one of the most fundamental abilities of robots to implement various manipulation tasks. Robots should have the same ability as humans to plan various grasp types for dexterous grasping. This paper addresses the problem of the adaptability of grasp planning. A novel adaptive grasp planning framework is designed to adapt to various grasp types rather than a single one. In this

Most current bipedal robots were modeled with an assumption that there is no slip between the stance foot and ground. This paper relaxes that assumption and undertakes a comprehensive study of the compass gait biped on slippery ground. It presents in detail the control of a biped that allows for foot slipping, and shows that feasible gaits fail on slippery ground for two causes: falling backward or

Balancing two-wheeled autonomous vehicles at low forward speeds is one of the primary challenges in the development of such vehicles. Gyrostabilizers can be used as actuators to make the balance; however, conventional gyros are not typically able to maintain constant moments and directions to stabilize against constant ‘heel’. In this paper, we present an innovative gyrostabilizer including a twin-flywheel

Simultaneous localization and mapping is a fundamental process in robot navigation. We focus on LiDAR to complete this process in ground robots traveling on complex terrain by proposing GR-LOAM, a method to estimate robot ego-motion by fusing LiDAR, inertial measurement unit (IMU), and encoder measurements in a tightly coupled scheme. First, we derive a odometer increment model that fuses the IMU and

This paper presents an approach to planning under uncertainty in resource-constrained environments. We describe our novel method for online plan modification and execution monitoring, which augments an existing plan with pre-computed plan fragments in response to observed resource availability. Our plan merging algorithm uses causal structure to interleave actions, creating solutions online using observations

Robots still cannot perform everyday manipulation tasks, such as grasping, with the same dexterity as humans do. In order to explore the potential of supervised deep learning for robotic grasping in unstructured and dynamic environments, this work addresses the visual perception phase involved in the task. This phase involves the processing of visual data to obtain the location of the object to be

The paper at hand introduces a novel system for producing an enhanced semantic map that leverages a reconstruction approach of street-view scenes using computer vision and machine learning techniques. Focusing on the recognition and localization of objects/entities, the composed map combines semantic information from publicly available, yet of lower accuracy, satellite images, with more detailed data

In this study, a new convex optimization (CO) approach to time-optimal trajectory planning (TOTP) is described, which considers both torque and jerk limits. The key insight of the approach is that the non-convex jerk limits are transformed to linear acceleration constraints and indirectly introduced into CO as the linear acceleration constraints. In this way, the convexity of CO will not be destroyed

Point clouds registration is a fundamental step of many point clouds processing pipelines; however, most algorithms are tested on data that are collected ad-hoc and not shared with the research community. These data often cover only a very limited set of use cases; therefore, the results cannot be generalized. Public datasets proposed until now, taken individually, cover only a few kinds of environment

Human-joint-position estimation is crucial for patient-transfer robots. However, high accuracy and real-time property are difficult to achieve simultaneously. To tackle the problem, we develop a new convolutional neural network (CNN), containing two levels of subnetworks, to fuse the information in color and depth images. The first-level subnetwork generates two-dimensional (2D) human joint positions

Object manipulation planning in clutter suffers from perception uncertainties due to occlusion, as well as action constraints required by collision avoidance. Partially observable Markov decision process (POMDP) provides a general model for planning under uncertainties. But a manipulation task usually have a large action space, which not only makes task planning intractable but also brings significant

Triangle mesh maps for robotic applications are becoming increasingly popular, but are not yet effectively supported in the Robot Operating System (ROS). We introduce the Mesh Tools package consisting of message definitions, RViz plugins and tools, as well as a persistence layer. These tools make annotated triangle maps available in ROS and allow to publish, edit and inspect such maps within the existing

Inspired by the abilities of amoeba to alter their shape, a continuous-track robot called Ourobot has been developed that is able to adapt its shape to the environment. Using tactile sensors at the outer hull of the robot, the outline of the terrain and collisions with obstacles can be detected. Thus, the robot is able to locomote in uneven terrain and climb steep slopes. Since the shape adaption is

The manipulation of an object into a desired location in a cluttered and restricted environment requires reasoning over the long-term consequences of an action while reacting locally to the multiple physics-based interactions. We present Visual Receding Horizon Planning (VisualRHP) in a framework which interleaves real-world execution with look-ahead planning to efficiently solve a short-horizon approximation

In this paper, we propose a two-stage learning framework for visual navigation in which the experience of the agent during exploration of one goal is shared to learn to navigate to other goals. We train a deep neural network for estimating the robot’s position in the environment using ground truth information provided by a classical localization and mapping approach. The second simpler multi-goal Q-function

We consider the problem of distributed goal conflict resolution in multi-robot systems while remaining resilient to intermittent communication losses between robots. Our proposed approach uses a spatial approximation technique called α-shape to represent the regions that have been explored by robots followed by a O(logn) algorithm that incrementally combines and shares the α-shape information between

Autonomous driving and driver assistance require a continuous and reliable perception of the road boundaries, namely curbs and berms, including also other minor, or not so minor, obstacles in the neighborhood of the car. This paper proposes to use a 4-layer LIDAR placed close to the ground to capture measurements of the road ahead of the car and allow the detection of the boundaries. This setup provides

With the advancement of the autonomous mobile robots applied to Warehouses and the creation of the Robotic Mobile Fulfillment System after the market implementation of the Kiva Robots, it is necessary to carry out a deeper approach of the researches carried out to this date. The objective of this survey is to provide a unified and accessible presentation of the basic concepts of a warehouse system

Developing safe autonomous robotic applications for outdoor agricultural environments is a research field that still presents many challenges. Simultaneous Localization and Mapping can be crucial to endow the robot to localize itself with accuracy and, consequently, perform tasks such as crop monitoring and harvesting autonomously. In these environments, the robotic localization and mapping systems

This paper presents a state-of-the-art filter that reduces the complexity in object detection, tracking and mapping applications. Existing edge detection and tracking methods are proposed to create suitable autonomy for mobile robots, however, many of them face overconfidence and large computations at the entrance to scenarios with an immense number of landmarks. The method in this work, the Line–Circle–Square

Multi-UAV system is widely used in surveillance, search and rescue, and industrial inspection. Multi-UAV trajectory planning is crucial for the multi-UAV system, but multi-UAV trajectory planning often needs to consider many constraints, such as trajectory smoothness, obstacle collisions, mutual collisions, dynamic limits, time-consuming, and trajectory length. It is a challenge to balance these constraints

Inferring the depth of images is a fundamental inverse problem within the field of Computer Vision since depth information is obtained through 2D images, which can be generated from infinite possibilities of observed real scenes. Benefiting from the progress of Convolutional Neural Networks (CNNs) to explore structural features and spatial image information, Single Image Depth Estimation (SIDE) is