Robotics and Autonomous Systems

In recent years, social robots have been increasingly used in the cultural sector to make artworks and exhibits more appealing to a larger audience. On the other hand, visitors, more accustomed to using novel technologies, are increasingly ...

Display Omitted

• The anthropomorphic behaviors in the robot enhance the naturalness of the interaction
• Robots with anthropomorphic behaviors can enhance the experience of museum visitors
• The use of backchannel communication significantly improves ...

Autonomous inspection and maintenance tasks with robots in oil and gas refineries require moving along pipelines and manipulation dexterity in cluttered environments. This paper investigates the problem of controlling a wheeled mobile manipulator ...

• Pipe inspection requires moving along pipelines and dexterity in cluttered environments.
• Novel over-actuated hybrid platform to inspect the structures while stabilizing on the pipe.
• Exploiting the arm dynamic in an optimization ...

Accurate detection and tracking of vulnerable road users and traffic objects represent vital tasks for autonomous driving and driving assistance systems. The recent developments in the field rely heavily on deep learning techniques, which require ...

• In-depth analysis of different performance metrics of recent SOTA object detectors of YOLO series.
• Benchmarked these detectors on the BDD100K.
• Demonstrated the potential usage of NvDCF/DeepSORT for ADAS.
• Explored tracking with ...

Powered exoskeletons have been shown to significantly reduce physical workload during occupational tasks. Due to this they have great potential impact on future labor practices. However, powered exoskeleton controllers must first be developed ...

• Whole Body Control for powered exoskeletons balances multiple task objectives.
• Shared Autonomy framework seamlessly integrates human and robot control.
• User “acceptance” of exoskeleton control objectives enables intuitive task ...

Autonomous navigation in rough and dynamic 3D environments is a major challenge for modern robotics. This paper presents a novel traversability analysis and path planning technique that processes 3D point cloud maps to compute terrain gradient ...

• Novel 3D traversability analysis technique based on terrain gradient information.
• Promotes efficient planning and autonomous navigation in forest environments.
• Minimizes mechanical effort, fostering fuel economy, and reduced wear ...

We present an autonomous approaching scheme for mobile robot traversing obstacle stairwells, which overcomes the restricted field of vision caused by obstacles. The scheme consists of three parts: stair localization, structural parameter ...

• We develop the double scales normal and selection (DSNS) algorithm to identify stair from sparse point clouds.
• We develop the multi-plane joint recognition (MPJR) algorithm to estimate the structural parameter.
• We propose the ...

• A novel PCVT is applied to the construction vehicle to improve transmission efficiency.
• A double forgetting factors least squares method is designed for identifying change of working conditions.
• The adaptive shift strategies based ...

A new power-cycling variable transmission (PCVT) is proposed and applied to construction vehicles to improve transmission efficiency. To adapt to changes in external loads, construction vehicles need to shift gears continuously. However, the ...

• A variable stiffness prosthetic knee joint is designed and evaluated.
• A knee structure-based variable stiffness parallel elastic actuation is proposed.
• An adaptive variable stiffness adjustment strategy is proposed.
• Knee joint ...

Active prosthetic knee joints are proven to provide outstanding actuating performance in various walking situations. Currently, most active prosthetic knee joints are directly actuated via motors, which have the disadvantages of high-power ...

• A two-objective optimization technique concerning MRS productivity and stiffness.
• Robot base position and task assignment are optimized simultaneously.
• Position switch provided by robot mobility is fully considered.
• Specially ...

Flexible and efficient multi-robot systems are being increasingly used in the field of aerospace manufacturing. The task planning considers solely the objective of productivity has been well studied in previous works, however, when dealing with ...

This paper proposes an optimization-based adaptive local planner to address the aggressive motion planning problem of an autonomous mobile robot in realistic semi-structured scenarios. The goal is to enable the robot to perform autonomous ...

• A real-time less-conservative method to avoid static structured obstacles.
• A collision-aware velocity governor to balance cautious and aggressive behaviour.
• An unified optimization-based navigation framework with practical ...

• A analytical differential kinematic model is introduced for continuum robots.
• A feedback controller is developed based on the kinematic model, with tension supervision.
• The controller versatility is experimentally validated on ...

Generic and high-performance feedback control is still challenging for tendon-driven continuum robots. Conventional model-based controllers, based on the piecewise constant curvature (PCC) assumption, explicitly require the arc parameters (...

In the research and development of soft rehabilitation gloves (SRGs), individualized support considering users’ morphological and biomechanical characteristics is the next critical step toward use in rehabilitation practice. Compared with single-...

• Personalized support is the next critical step for soft rehabilitation gloves (SRGs).
• Modular soft actuators enable support for each joint by only adjusting connectors.
• A new modular soft actuator improved the performance by ...

A service robot can provide a smoother interaction experience if it has the ability to proactively detect whether a nearby user intends to interact, in order to adapt its behavior e.g. by explicitly showing that it is available to provide a ...

• A machine learning approach to predict intention to interact with a robot.
• The method can be trained in a self-supervised way.
• Real-life experiments shows the effectiveness to detect interaction intentions with high accuracy and ...

Aeronautical performance is enhanced by increasingly complex and lightweight composite parts. The improvement of their manufacturing must ensure the quality measured by Non-Destructive Testing. To evaluate the health of aeronautical parts that ...

Display Omitted

• Calibration of a redundant robotic cell dedicated to the X-ray tomography system
• Development of new tools including a large-sized test phantom and a calibration comb
• The methodology enhances the process capability, leading to ...

With the development of Unmanned Surface Vehicles (USVs), the perception of inland waterways has become significant to autonomous navigation. RGB cameras can capture images with rich semantic features, but they would fail in adverse weather and ...

• A novel and effective clutter removal algorithm in the point cloud level for 4D mmWave radar.
• A novel and high-performance panoptic perception model based on graph convolution network and feature-level fusion.
• An effective method ...

The explosive growth of civil applications of Unmanned Aerial Vehicles (UAVs) calls for control algorithms that enable safe and trustworthy operations, especially in complex environments to facilitate real-world deployment. Although Model-Based ...

• A physics-informed neural network (PINN) is proposed for quadrotor dynamical modeling.
• The law of conservation of momentum is embedded as learning bias for PINN training.
• A custom implementation of quadrotor ground effect is made ...

• RRT is prone to local optimization problems in narrow areas.
• The purpose of improving RRT is to improve convergence speed and algorithm performance in narrow areas.
• Propose cross-test for narrow area localization, using the ...

The RRT algorithm based on random sampling is widely used in planning problems with non-holonomic constraints. In indoor environments, narrow passages exist in the map, and the RRT method suffers from slow convergence speed and poor path quality. ...

• Efficient model design with shared components for pedestrian detection, tracking and attribute recognition.
• Achieving state-of-the-art pedestrian multiple attribute recognition.
• The utilization of multiple partially-labeled ...

Safe interaction with pedestrians is important for autonomous vehicles in urban environments. Comprehensive understanding of pedestrian behavior is a challenge for autonomous driving, as different pedestrian attributes have different ...

This work presents a new depth- and semantics-aware conditional generative model, named TITAN-Next, for cross-domain image-to-image translation in a multi-modal setup between LiDAR and camera sensors. The proposed model leverages scene semantics ...

• Domain translation between LiDAR and RGB space.
• Improvement of over 23% compared with previous state of the art.
• Sensor Failure in autonomous vehicles.
• Depth estimation.