Showing papers in "Robotics and Autonomous Systems in 2018"

TL;DR: The proposed adaptive variable impedance control for force tracking has the capability to track the dynamic desired force and compensate for uncertainties in environment and can achieve better force tracking performance than the constant impedance control.

TL;DR: A critical review of the major contributions to RMP in dynamic environments, which includes artificial potential field based, artificial intelligence based, probabilistic based RMP and applications in areas of Agent systems and computer geometry.

TL;DR: This paper proposes a novel RGB-D data-based motion removal approach that is on-line and does not require prior-known moving-object information, such as semantics or visual appearances, and integrates the approach into the front end of anRGB-D SLAM system.

TL;DR: An extensive study of aerial vehicles and manipulation/interaction mechanisms in aerial manipulation is presented and the shortcomings of current aerial manipulation research are highlighted and a number of directions for future research are suggested.

TL;DR: The success rate of robot path planning and the optimal extent of the robot path are effectively improved by the improved A* algorithm.

TL;DR: The development of a robot-integrated smart home (RiSH) which can be used for research in assistive technologies for elderly care and the operation of the various components in the RiSH are evaluated.

TL;DR: Theoretical and experimental evaluation of the proposed algorithms have been made and compared to the latest state of the art motion planning algorithms under different challenging environmental conditions and have proven their remarkable improvement in efficiency and convergence rate.

TL;DR: A novel learning algorithm called “Reset-free Trial-and-Error” (RTE) is introduced that breaks the complexity by pre-generating hundreds of possible behaviors with a dynamics simulator of the intact robot, and allows complex robots to quickly recover from damage while completing their tasks and taking the environment into account.

TL;DR: A deep RL framework for adaptive control applications of AUVs based on an actor-critic goal-oriented deep RL architecture, which takes the available raw sensory information as input and as output the continuous control actions which are the low-level commands for the AUV’s thrusters.

TL;DR: A power augmentation and rehabilitation exoskeleton based on a novel actuator that has been shown to reduce the amount of muscles effort needed to perform a number of simple grasps and an output force mathematical model for it has been developed.

TL;DR: A generative model that learns a bidirectional mapping between human whole-body motion and natural language using deep recurrent neural networks (RNNs) and sequence-to-sequence learning and is capable of generating correct and detailed natural language descriptions from human motions.

TL;DR: In this paper, a state-of-the-art graphical model for semantic segmentation is extended to incorporate boat pitch and roll measurements from the on-board inertial measurement unit (IMU) and a stereo verification algorithm that consolidates tentative detections obtained from the segmentation.

TL;DR: A learning-from-demonstration framework that integrates force sensing and variable impedance control to learn force-based variable stiffness skills is proposed and validated in simulation using 2D and 7D systems and a couple of real scenarios.

TL;DR: The effectiveness of the proposed algorithm with respect to task completion time, resource consumption, and communication overhead is theoretically analyzed and is also demonstrated from the simulation of the delivery mission.

TL;DR: An innovative hybrid encoder that integrates deep Boltzmann machines (DBM) and auto-encoders (AE) is designed that combines the greedy learning features of DBM with the dimensionality reduction capacity of AE to accurately and reliably detect the presence of obstacles.

TL;DR: This paper is investigating asymmetric bellow flexible pneumatic actuator (ABFPA) as a bending joint made of suitable rubber material in the construction of a novel underactuated multi-jointed, multi-fingered soft robotic hand for prosthetic application.

TL;DR: It is shown that some type of cyber-attacks on Real Time Location Systems, specifically Denial of Service and Spoofing, can be detected by a system built using Machine Learning techniques.

TL;DR: A new trajectory generation framework for robotic table tennis that does not involve a fixed hitting plane is introduced and a free-time optimal control approach is used to derive two different trajectory optimizers, Focused Player and Defensive Player, which encode two different play-styles.

TL;DR: A generic framework that integrates an autonomous obstacle detection module and a reinforcement learning (RL) module to develop reactive obstacle avoidance behavior for a UAV and shows that the proposed saliency detection algorithm performs better than state-of-the-art, and the RL algorithm can learn the avoidance behavior from the manual experiences.

TL;DR: A novel approach to analytically compute the path in an efficient and effective manner is introduced, showing that the method can adapt in real time the robot’s path in order to avoid several types of obstacles, while producing a map of the surroundings.

TL;DR: An online walking-pattern generation algorithm with footstep adjustment that helps a humanoid robot (DRC-HUBO+) to regain balance following disturbance, i.e., from strong pushing or stepping on unexpected obstacles.

TL;DR: A robust feedback linearization controller is developed to deal with this highly coupled and nonlinear dynamics of the proposed tri-rotor UAV, which linearizes the dynamics globally using geometric transformations to produce a linear model that matches the Jacobi linearization of the non linear dynamics at the operating point of interest.

TL;DR: A novel transformable wheel-legged robot, namely Land Devil Ray (LDR), is proposed for search and rescue mission in complex terrains, and has excellent performance in maneuverability, stability, maximum obstacle-negotiation height and mode switch process.

TL;DR: A novel method to estimate a head pose from a monocular camera based on multi-task learning deep neural network that uses a small grayscale image that outperforms state-of-the-art approaches quantitatively and qualitatively with an average head pose mean error of less than 4° in real-time.

TL;DR: A control framework for robot-assisted minimally invasive general surgery (RA-MIS) for physical human–robot collaboration using a redundant 7-DoF serial robot is proposed, demonstrating the feasibility of the null-space compliance control approach while preserving the accuracy of the surgical task.

TL;DR: An obstacle avoidance and motion planning scheme for a hexapod robot that takes advantage of the superior mobility of the legged robot and fulfills requirements of walking stability and kinematic feasibility simultaneously is presented.

TL;DR: The results show that an embedded FPGA based SoC architecture is an interesting alternative for a SLAM algorithm implementation using the hardware–software co-design approach and the system meets performance requirements of a robot to operate in real-time constraints.

TL;DR: A square-shaped rotational-flow adsorption unit is designed and a soft skirt structure is designed to improve the robot’s load ability and obstacle-surmounting ability and shows that the prototype robot can move stably on coarse walls and can pass over large grooves and bulges easily.

TL;DR: Only two simple sensors were used in a soft pneumatic gripper (SPG) control system to make it possess innervated-like ability and demonstrate that the SPG is able to maintain stable grasping state for a long time based on size recognition ability no matter what material the object is.

TL;DR: The state of the art of the most common motion planners is reviewed, and a set of benchmarks are presented with the aim to provide not only a theoretical review but also a qualitative/quantitative comparison of the algorithms.