Search results

Autonomous obstacle avoidance is important in unmanned surface vehicle (USV) navigation. Although the result of obstacle detection is often inaccurate because of the inherent errors of LIDAR, conventional methods typically emphasize on a single obstacle-avoidance algorithm and neglect the limitation of sensors and safety in a local region. Conventional methods also fail in seamlessly integrating local and global obstacle avoidance algorithms. This paper aims to present a cooperative manoeuvring approach including both local and global obstacle avoidance.

The global algorithm used in our USV is the Artificial Potential Field-Ant Colony Optimization (APF-ACO) obstacle-avoidance algorithm, which plans a relative optimal path on the specified electronic map before the cruise of USV. The local algorithm is a multi-layer obstacle-avoidance framework based on a single LIDAR to present an efficient solution to USV path planning in the case of sensor errors and collision risks. When obstacles are within a layer, the USV uses a corresponding obstacle-avoidance algorithm. Then the USV moves towards the global direction according to fuzzy rules in the fuzzy layer.

The presented method offers a solution for obstacle avoidance in a complex environment. The USV follows the global trajectory planed by the APF-ACO algorithm. While, the USV can bypass current obstacle in the local region based on the multi-layer method effectively. This fact was validated by simulations and field trials.

The method presented in this paper takes advantage of algorithm integration that remedies errors of obstacle detection. Simulation and experiments were also conducted for performance evaluation.

Dynamic movement primitives (DMPs) is a general robotic skill learning from demonstration method, but it is usually used for single robotic manipulation. For cloud-based robotic skill learning, the authors consider trajectories/skills changed by the environment, rebuild the DMPs model and propose a new DMPs-based skill learning framework removing the influence of the changing environment.

The authors proposed methods for two obstacle avoidance scenes: point obstacle and non-point obstacle. For the case with point obstacles, an accelerating term is added to the original DMPs function. The unknown parameters in this term are estimated by interactive identification and fitting step of the forcing function. Then a pure skill despising the influence of obstacles is achieved. Using identified parameters, the skill can be applied to new tasks with obstacles. For the non-point obstacle case, a space matching method is proposed by building a matching function from the universal space without obstacle to the space condensed by obstacles. Then the original trajectory will change along with transformation of the space to get a general trajectory for the new environment.

The proposed two methods are certified by two experiments, one of which is taken based on Omni joystick to record operator’s manipulation motions. Results show that the learned skills allow robots to execute tasks such as autonomous assembling in a new environment.

This is a new innovation for DMPs-based cloud robotic skill learning from multi-scene tasks and generalizing new skills following the changes of the environment.

Robotic arms play a crucial role in various industrial operations, such as sorting, assembly, handling and spraying. However, traditional robotic arm control algorithms often struggle to adapt when faced with the challenge of dynamic obstacles. This paper aims to propose a dynamic obstacle avoidance method based on reinforcement learning to address real-time processing of dynamic obstacles.

This paper introduces an innovative method that introduces a feature extraction network that integrates gating mechanisms on the basis of traditional reinforcement learning algorithms. Additionally, an adaptive dynamic reward mechanism is designed to optimize the obstacle avoidance strategy.

Validation through the CoppeliaSim simulation environment and on-site testing has demonstrated the method's capability to effectively evade randomly moving obstacles, with a significant improvement in the convergence speed compared to traditional algorithms.

The proposed dynamic obstacle avoidance method based on Reinforcement Learning not only accomplishes the task of dynamic obstacle avoidance efficiently but also offers a distinct advantage in terms of convergence speed. This approach provides a novel solution to the obstacle avoidance methods for robotic arms.

This paper aims to outline and implement a novel hybrid controller in humanoid robots to map an optimal path. The hybrid controller is designed using the Owl search algorithm (OSA) and Fuzzy logic.

The optimum steering angle (OS) is used to deal with the obstacle located in the workspace, which is the output of the hybrid OSA Fuzzy controller. It is obtained by feeding OSA's output, i.e. intermediate steering angle (IS), in fuzzy logic. It is obtained by supplying the distance of obstacles from all directions and target distance from the robot's present location.

The present research is based on the navigation of humanoid NAO in complicated workspaces. Therefore, various simulations are performed in a 3D simulator in different complicated workspaces. The validation of their outcomes is done using the various experiments in similar workspaces using the proposed controller. The comparison between their outcomes demonstrates an acceptable correlation. Ultimately, evaluating the proposed controller with another existing navigation approach indicates a significant improvement in performance.

A new framework is developed to guide humanoid NAO in complicated workspaces, which is hardly seen in the available literature. Inspection in simulation and experimental workspaces verifies the robustness of the designed navigational controller. Considering minimum error ranges and near collaboration, the findings from both frameworks are evaluated against each other in respect of specified navigational variables. Finally, concerning other present approaches, the designed controller is also examined, and major modifications in efficiency have been reported.

This study aims to compare traffic sign (TS) and obstacle detection for autonomous vehicles using different methods. The review will be performed based on the various methods, and the analysis will be done based on the metrics and datasets.

In this study, different papers were analyzed about the issues of obstacle detection (OD) and sign detection. This survey reviewed the information from different journals, along with their advantages and disadvantages and challenges. The review lays the groundwork for future researchers to gain a deeper understanding of autonomous vehicles and is obliged to accurately identify various TS.

The review of different approaches based on deep learning (DL), machine learning (ML) and other hybrid models that are utilized in the modern era. Datasets in the review are described clearly, and cited references are detailed in the tabulation. For dataset and model analysis, the information search process utilized datasets, performance measures and achievements based on reviewed papers in this survey.

Various techniques, search procedures, used databases and achievement metrics are surveyed and characterized below for traffic signal detection and obstacle avoidance.

This paper aims to facilitate the research and development of resilient navigation approaches, explore the robustness of adversarial training to different interferences and promote their practical applications in real complex environments.

In this paper, the authors first summarize the real accidents of self-driving cars and develop a set of methods to simulate challenging scenarios by introducing simulated disturbances and attacks into the input sensor data. Then a robust and transferable adversarial training approach is proposed to improve the performance and resilience of current navigation models, followed by a multi-modality fusion-based end-to-end navigation network to demonstrate real-world performance of the methods. In addition, an augmented self-driving simulator with designed evaluation metrics is built to evaluate navigation models.

Synthetical experiments in simulator demonstrate the robustness and transferability of the proposed adversarial training strategy. The simulation function flow can also be used for promoting any robust perception or navigation researches. Then a multi-modality fusion-based navigation framework is proposed as a light-weight model to evaluate the adversarial training method in real-world.

The adversarial training approach provides a transferable and robust enhancement for navigation models both in simulation and real-world.

In the vast majority of the individual robot installations, the robot arm is just one piece of a complex puzzle of components, such as grippers, jigs or external axis, that together compose an industrial robotic cell. The success of such installations is very dependent not only on the selection of such components but also on the layout and design of the final robotic cell, which are the main tasks of the system integrators. Consequently, successful robot installations are often empirical tasks owing to the high number of experimental combinations that could lead to exhaustive and time-consuming testing approaches.

A newly developed optimized technique to deal with automatic planning and design of robotic systems is proposed and tested in this paper.

The application of a genetic-based algorithm achieved optimal results in short time frames and improved the design of robotic work cells. Here, the authors show that a multi-layer optimization approach, which can be validated using a robotic tool, is able to help with the design of robotic systems.

The usage of the proposed approach can be valuable to industrial corporations, as it allows for improved workflows, maximization of available robotic operations and improvement of efficiency.

To date, robotic solutions lack flexibility to cope with the demanding industrial environments. The results presented here formalize a new flexible and modular approach, which can provide optimal solutions throughout the different stages of design and execution control of any work cell.

In the past decade, more than 700 million people are affected by some kind of disability or handicap. In this context, the research interest in assistive robotics is growing up. For people with mobility impairments, daily life operations, as dressing or feeding, require the assistance of dedicated people; thus, the use of devices providing independent mobility can have a large impact on improving their life quality. The purpose of this paper is to present the development of a robotic system aimed at assisting people with this kind of severe motion disabilities by providing a certain level of autonomy.

The system is based on a hierarchical architecture where, at the top level, the user generates simple and high-level commands by resorting to a graphical user interface operated via a P300-based brain computer interface. These commands are ultimately converted into joint and Cartesian space tasks for the robotic system that are then handled by the robot motion control algorithm resorting to a set-based task priority inverse kinematic strategy. The overall architecture is realized by integrating control and perception software modules developed in the robots and systems environment with the BCI2000 framework, used to operate the brain–computer interfaces (BCI) device.

The effectiveness of the proposed architecture is validated through experiments where a user generates commands, via an Emotiv Epoc+ BCI, to perform assistive tasks that are executed by a Kinova MOVO robot, i.e. an omnidirectional mobile robotic platform equipped with two lightweight seven degrees of freedoms manipulators.

The P300 paradigm has been successfully integrated with a control architecture that allows us to command a complex robotic system to perform daily life operations. The user defines high-level commands via the BCI, letting all the low-level tasks, for example, safety-related tasks, to be handled by the system in a completely autonomous manner.

The Group of Unmanned Assistant Robots Deployed in Aggregative Navigation by Scent (GUARDIANS) multi‐robot team is to be deployed in a large warehouse in smoke. The team is to assist firefighters search the warehouse in the event or danger of a fire. The large dimensions of the environment together with development of smoke which drastically reduces visibility, represent major challenges for search and rescue operations. The GUARDIANS robots act alongside a firefighter and guide and accompany the firefighters on the site while indicating possible obstacles and the locations of danger and maintain communications links. The purpose of this paper is to focus on basic navigation behaviours of multi‐robot or human‐robot teams, which have to be achieved without central and on‐line control in both categories of GUARDIANS robots' tasks.

In order to fulfill the aforementioned tasks, the robots need to be able to perform certain behaviours. Among the basic behaviours are capabilities to stay together as a group, that is, generate a formation and navigate while keeping this formation. The control model used to generate these behaviours is based on the so‐called social potential field framework, which the authors adapt to the specific tasks required for the GUARDIANS scenario. All tasks can be achieved without central control, and some of the behaviours can be performed without explicit communication between the robots.

The GUARDIANS environment requires flexible formations of the robot team: the formation has to adapt itself to the circumstances. Thus, the application has forced the concept of a formation to be re‐defined. Using the graph‐theoretic terminology, it can be said that a formation may be stretched out as a path or be compact as a star or wheel. The developed behaviours have been implemented in simulation environments as well as on real ERA‐MOBI robots commonly referred to as Erratics. Advantages and shortcomings of the model, based on the simulations as well as on the implementation with a team of Erratics are discussed.

This paper discusses the concept of a robot formation in the context of a real world application of a robot team (Swarm).

In response to these shortcomings, this paper proposes a dynamic obstacle detection and tracking method based on multi-feature fusion and a dynamic obstacle recognition method based on spatio-temporal feature vectors.

The existing dynamic obstacle detection and tracking methods based on geometric features have a high false detection rate. The recognition methods based on the geometric features and motion status of dynamic obstacles are greatly affected by distance and scanning angle, and cannot meet the requirements of real traffic scene applications.

First, based on the geometric features of dynamic obstacles, the obstacles are considered The echo pulse width feature is used to improve the accuracy of obstacle detection and tracking; second, the space-time feature vector is constructed based on the time dimension and space dimension information of the obstacle, and then the support vector machine method is used to realize the recognition of dynamic obstacles to improve the obstacle The accuracy of object recognition. Finally, the accuracy and effectiveness of the proposed method are verified by real vehicle tests.

The paper proposes a dynamic obstacle detection and tracking method based on multi-feature fusion and a dynamic obstacle recognition method based on spatio-temporal feature vectors. The accuracy and effectiveness of the proposed method are verified by real vehicle tests.