Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
6.8
3.9
Journals
Sensors
Special Issues
Advances in Mobile Robot Perceptions, Planning, Control and Learning
sensors-logo
Submit to Sensors Review for Sensors Propose a Special Issue

Journal Menu

► Journal Menu

Journal Browser

► Journal Browser
share announcement

Advances in Mobile Robot Perceptions, Planning, Control and Learning

A special issue of Sensors (ISSN 1424-8220). This special issue belongs to the section "Sensors and Robotics".

Deadline for manuscript submissions: closed (20 November 2023) | Viewed by 19276

Special Issue Editors

Dr. Yingbai Hu

E-Mail Website
Guest Editor
Department of Informatics, Technical University of Munich, 85748 Munich, Germany
Interests: optimization control; imitation learning; reinforcement learning; motion planning
Special Issues, Collections and Topics in MDPI journals
Dr. Chao Zeng

E-Mail Website
Guest Editor
Department of Informatics, University of Hamburg, 22527 Hamburg, Germany
Interests: learning from demonstration; compliant manipulation; robot learning and control
Special Issues, Collections and Topics in MDPI journals
Dr. Shu Li

E-Mail Website1 Website2
Guest Editor
School of Electrical Engineering, Liaoning University of Technology, Jinzhou 121001, China
Interests: intelligent control of wheeled mobile robots; intelligent control theory and applications
Special Issues, Collections and Topics in MDPI journals
Prof. Dr. Alois Christian Knoll

E-Mail Website
Guest Editor
Department of Informatics, Technical University of Munich, 85748 Munich, Germany
Interests: cognitive, medical, and sensor-based robotics; multiagent systems; data fusion; adaptive systems; multimedia information retrieval; model-driven development of embedded systems with applications of automotive software and electric transportation; simulation systems for robotics and traffic
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear Colleagues,

With the increasing demand for mobile robots, such as lunar rovers, unmanned driving vehicles, rescue robots, and delivery robots, in the fields of aerospace, terrain, surface, and underwater, there is a growing interest in the development of new technologies that can be used to advance state-of-the-art mobile robots. A reliable mobile manipulation consists of at least three core parts: the perception of  the environment, motion planning, and control. When detecting the environment through sensing (LiDAR, radar, camera, GPS, IMU, etc.), it is important to design a planner through various theoretical approaches (machine learning, reinforcement learning, convex/nonconvex optimization, evolutionary computation, potential field methods, etc.) to find the optimal trajectory of a mobile robot while avoiding static and/or dynamic obstacles. Additionally, it is necessary to design a robust controller (sliding mode control, model predictive control, adaptive neural network control, etc.) for the robot in perturbed environments, such as complex terrains and external contact forces.

It is expected that mobile robots can tackle the designed tasks (grasping, autonomous driving, etc.) under diverse and unstructured environmental conditions, but it brings challenges for sensing, planning, and control. For this reason, the perception, implementation, modeling, control, and learning of mobile robots have become urgent issues. We welcome original research contributions and state-of-the-art reviews from both theoretical and experimental works which promote further research activities in this area.

The main topics of this Special Issue include, but are not limited to, the following:

  • Mobile robot intelligent perception and control;
  • Visual or haptic control with sensor feedback;
  • Human–robot interaction or teleoperation control;
  • Motion planning and navigation indoors or outdoors;
  • Machine learning for object detection, recognition, and tracking;
  • Reinforcement/imitation/transfer learning for mobile robots;
  • Multimodal learning for mobile robots;
  • Advanced modeling and sensors for mobile manipulation;
  • Applications of mobile manipulation.

Dr. Yingbai Hu
Dr. Chao Zeng
Dr. Shu Li
Prof. Dr. Alois Christian Knoll
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Sensors is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Related Special Issue

Published Papers (15 papers)

Download All Papers
Order results
Result details
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:

Research

16 pages, 3590 KiB  
Article
HapticWhirl, a Flywheel-Gimbal Handheld Haptic Controller for Exploring Multimodal Haptic Feedback
by Jose Luis Berna Moya, Anke van Oosterhout, Mark T. Marshall and Diego Martinez Plasencia
Sensors 2024, 24(3), 935; https://doi.org/10.3390/s24030935 - 31 Jan 2024
Viewed by 602
Abstract
Most haptic actuators available on the market today can generate only a single modality of stimuli. This ultimately limits the capacity of a kinaesthetic haptic controller to deliver more expressive feedback, requiring a haptic controller to integrate multiple actuators to generate complex haptic [...] Read more.
Most haptic actuators available on the market today can generate only a single modality of stimuli. This ultimately limits the capacity of a kinaesthetic haptic controller to deliver more expressive feedback, requiring a haptic controller to integrate multiple actuators to generate complex haptic stimuli, with a corresponding complexity of construction and control. To address this, we designed a haptic controller to deliver several modalities of kinaesthetic haptic feedback using a single actuator: a flywheel, the orientation of which is controlled by two gimbals capable of rotating over 360 degrees, in combination with a flywheel brake. This enables the controller to generate multiple haptic feedback modalities, such as torque feedback, impact simulation, low-frequency high-amplitude vibrations, inertial effects (the sensation of momentum), and complex haptic output effects such as the experience of vortex-like forces (whirl effects). By combining these diverse haptic effects, the controller enriches the haptic dimension of VR environments. This paper presents the device’s design, implementation, and characterization, and proposes potential applications for future work. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

21 pages, 3955 KiB  
Article
VP-SOM: View-Planning Method for Indoor Active Sparse Object Mapping Based on Information Abundance and Observation Continuity
by Jiadong Zhang and Wei Wang
Sensors 2023, 23(23), 9415; https://doi.org/10.3390/s23239415 - 26 Nov 2023
Viewed by 482
Abstract
Active mapping is an important technique for mobile robots to autonomously explore and recognize indoor environments. View planning, as the core of active mapping, determines the quality of the map and the efficiency of exploration. However, most current view-planning methods focus on low-level [...] Read more.
Active mapping is an important technique for mobile robots to autonomously explore and recognize indoor environments. View planning, as the core of active mapping, determines the quality of the map and the efficiency of exploration. However, most current view-planning methods focus on low-level geometric information like point clouds and neglect the indoor objects that are important for human–robot interaction. We propose a novel View-Planning method for indoor active Sparse Object Mapping (VP-SOM). VP-SOM takes into account for the first time the properties of object clusters in the coexisting human–robot environment. We categorized the views into global views and local views based on the object cluster, to balance the efficiency of exploration and the mapping accuracy. We developed a new view-evaluation function based on objects’ information abundance and observation continuity, to select the Next-Best View (NBV). Especially for calculating the uncertainty of the sparse object model, we built the object surface occupancy probability map. Our experimental results demonstrated that our view-planning method can explore the indoor environments and build object maps more accurately, efficiently, and robustly. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

15 pages, 1030 KiB  
Article
Trajectory Tracking Control of Transformer Inspection Robot Using Distributed Model Predictive Control
by Lai Wei, Guofei Xiang, Congjun Ma, Xuejian Jiang and Songyi Dian
Sensors 2023, 23(22), 9238; https://doi.org/10.3390/s23229238 - 17 Nov 2023
Cited by 1 | Viewed by 738
Abstract
To overcome the difficulty in tracking the trajectory of an inspection robot inside a transformer, this paper proposes a distributed model predictive control method. First, the kinematics and dynamics models of a robot in transformer oil are established based on the Lagrange equation. [...] Read more.
To overcome the difficulty in tracking the trajectory of an inspection robot inside a transformer, this paper proposes a distributed model predictive control method. First, the kinematics and dynamics models of a robot in transformer oil are established based on the Lagrange equation. Then, by using the nonlinear model predictive control method and following the distributed control theory, the motion of a robot in transformer oil is decoupled into five independent subsystems. Based on this, a distributed model predictive control (DMPC) method is then developed. Finally, the simulation results indicate that a robot motion control system based on DMPC achieves high tracking accuracy and robustness with reduced computing complexity, and it provides an effective solution for the motion control of robots in narrow environments. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

20 pages, 6390 KiB  
Article
Leader–Follower Approach for Non-Holonomic Mobile Robots Based on Extended Kalman Filter Sensor Data Fusion and Extended On-Board Camera Perception Controlled with Behavior Tree
by Arpit Joon and Wojciech Kowalczyk
Sensors 2023, 23(21), 8886; https://doi.org/10.3390/s23218886 - 01 Nov 2023
Cited by 1 | Viewed by 682
Abstract
This paper presents a leader–follower mobile robot control approach using onboard sensors. The follower robot is equipped with an Intel RealSense camera mounted on a rotating platform. Camera observations and ArUco markers are used to localize the robots to each other and relative [...] Read more.
This paper presents a leader–follower mobile robot control approach using onboard sensors. The follower robot is equipped with an Intel RealSense camera mounted on a rotating platform. Camera observations and ArUco markers are used to localize the robots to each other and relative to the workspace. The rotating platform allows the expansion of the perception range. As a result, the robot can use observations that are not within the camera’s field of view at the same time in the localization process. The decision-making process associated with the control of camera rotation is implemented using behavior trees. In addition, measurements from encoders and IMUs are used to improve the quality of localization. Data fusion is performed using the EKF filter and allows the user to determine the robot’s poses. A 3D-printed cuboidal tower is added to the leader robot with four ArUco markers located on its sides. Fiducial landmarks are placed on vertical surfaces in the workspace to improve the localization process. The experiments were performed to verify the effectiveness of the presented control algorithm. The robot operating system (ROS) was installed on both robots. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

16 pages, 10558 KiB  
Article
Real-Time Road Intersection Detection in Sparse Point Cloud Based on Augmented Viewpoints Beam Model
by Di Hu, Kai Zhang, Xia Yuan, Jiachen Xu, Yipan Zhong and Chunxia Zhao
Sensors 2023, 23(21), 8854; https://doi.org/10.3390/s23218854 - 31 Oct 2023
Viewed by 635
Abstract
Road intersection is a kind of important navigation landmark, while existing detection methods exhibit clear limitations in terms of their robustness and efficiency. A real-time algorithm for road intersection detection and location in large-scale sparse point clouds is proposed in this paper. Different [...] Read more.
Road intersection is a kind of important navigation landmark, while existing detection methods exhibit clear limitations in terms of their robustness and efficiency. A real-time algorithm for road intersection detection and location in large-scale sparse point clouds is proposed in this paper. Different from traditional approaches, our method establishes the augmented viewpoints beam model to perceive the road bifurcation structure. Explicitly, the spatial features from point clouds are jointly extracted in various viewpoints in front of the robot. In addition, the evaluation metrics are designed to self-assess the quality of detection results, enabling our method to optimize the detection process in real time. Considering the scarcity of datasets for intersection detection, we also collect and annotate a VLP-16 point cloud dataset specifically for intersections, called NCP-Intersection. Quantitative and qualitative experiments demonstrate that the proposed method performs favorably against the other parallel methods. Specifically, our method performs an average precision exceeding 90% and an average processing time of approximately 88 ms/frame. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

23 pages, 138297 KiB  
Article
Online Multi-Contact Motion Replanning for Humanoid Robots with Semantic 3D Voxel Mapping: ExOctomap
by Masato Tsuru, Adrien Escande, Iori Kumagai, Masaki Murooka and Kensuke Harada
Sensors 2023, 23(21), 8837; https://doi.org/10.3390/s23218837 - 30 Oct 2023
Viewed by 1428
Abstract
This study introduces a rapid motion-replanning technique driven by a semantic 3D voxel mapping system, essential for humanoid robots to autonomously navigate unknown territories through online environmental sensing. Addressing the challenges posed by the conventional approach based on polygon mesh or primitive extraction [...] Read more.
This study introduces a rapid motion-replanning technique driven by a semantic 3D voxel mapping system, essential for humanoid robots to autonomously navigate unknown territories through online environmental sensing. Addressing the challenges posed by the conventional approach based on polygon mesh or primitive extraction for mapping, we adopt semantic voxel mapping, utilizing our innovative Extended-Octomap (ExOctomap). This structure archives environmental normal vectors, outcomes of Euclidean Cluster Extraction, and principal component analysis within an Octree structure, facilitating an O(log N) efficiency in semantic accessibility from a position query x∈R3. This strategy reduces the 6D contact pose search to simple 3D grid sampling. Moreover, voxel representation enables the search of collision-free trajectories online. Through experimental validation based on simulations and real robotic experiments, we demonstrate that our framework can efficiently adapt multi-contact motions across diverse environments, achieving near real-time planning speeds that range from 13.8 ms to 115.7 ms per contact. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

20 pages, 7767 KiB  
Article
A Novel AGV Path Planning Approach for Narrow Channels Based on the Bi-RRT Algorithm with a Failure Rate Threshold
by Bin Wu, Wei Zhang, Xiaonan Chi, Di Jiang, Yang Yi and Yi Lu
Sensors 2023, 23(17), 7547; https://doi.org/10.3390/s23177547 - 30 Aug 2023
Cited by 5 | Viewed by 947
Abstract
The efficiency of the rapidly exploring random tree (RRT) falls short when efficiently guiding targets through constricted-passage environments, presenting issues such as sluggish convergence speed and elevated path costs. To overcome these algorithmic limitations, we propose a narrow-channel path-finding algorithm (named NCB-RRT) based [...] Read more.
The efficiency of the rapidly exploring random tree (RRT) falls short when efficiently guiding targets through constricted-passage environments, presenting issues such as sluggish convergence speed and elevated path costs. To overcome these algorithmic limitations, we propose a narrow-channel path-finding algorithm (named NCB-RRT) based on Bi-RRT with the addition of our proposed research failure rate threshold (RFRT) concept. Firstly, a three-stage search strategy is employed to generate sampling points guided by real-time sampling failure rates. By means of the balance strategy, two randomly growing trees are established to perform searching, which improves the success rate of the algorithm in narrow channel environments, accelerating the convergence speed and reducing the number of iterations required. Secondly, the parent node re-selection and path pruning strategy are integrated. This shortens the path length and greatly reduces the number of redundant nodes and inflection points. Finally, the path is optimized by utilizing segmented quadratic Bezier curves to achieve a smooth trajectory. This research shows that the NCB-RRT algorithm is better able to adapt to the complex narrow channel environment, and the performance is also greatly improved in terms of the path length and the number of inflection points. Compared with the RRT, RRT* and Bi-RRT algorithms, the success rate is increased by 2400%, 1900% and 11.11%, respectively. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

18 pages, 5618 KiB  
Article
Improved A* Algorithm for Path Planning of Spherical Robot Considering Energy Consumption
by Hao Ge, Zhanfeng Ying, Zhihua Chen, Wei Zu, Chunzheng Liu and Yicong Jin
Sensors 2023, 23(16), 7115; https://doi.org/10.3390/s23167115 - 11 Aug 2023
Cited by 6 | Viewed by 1114
Abstract
Spherical robots have fully wrapped shells, which enables them to walk well on complex terrains, such as swamps, grasslands and deserts. At present, path planning algorithms for spherical robots mainly focus on finding the shortest path between the initial position and the target [...] Read more.
Spherical robots have fully wrapped shells, which enables them to walk well on complex terrains, such as swamps, grasslands and deserts. At present, path planning algorithms for spherical robots mainly focus on finding the shortest path between the initial position and the target position. In this paper, an improved A* algorithm considering energy consumption is proposed for the path planning of spherical robots. The optimization objective of this algorithm is to minimize both the energy consumption and path length of a spherical robot. A heuristic function constructed with the energy consumption estimation model (ECEM) and the distance estimation model (DEM) is used to determine the path cost of the A* algorithm. ECEM and DCM are established based on the force analysis of the spherical robot and the improved Euclidean distance of the grid map, respectively. The effectiveness of the proposed algorithm is verified by simulation analysis based on a 3D grid map and a spherical robot moving with uniform velocity. The results show that compared with traditional path planning algorithms, the proposed algorithm can minimize the energy consumption and path length of the spherical robot as much as possible. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

15 pages, 383 KiB  
Communication
Adaptive Output Feedback Control for Nonholonomic Chained Systems with Integral Input State Stability Inverse Dynamics
by Yang Li, Linxing Xu, Xiuli Wang and Cunsong Wang
Sensors 2023, 23(14), 6351; https://doi.org/10.3390/s23146351 - 13 Jul 2023
Viewed by 895
Abstract
This paper investigates a class of nonholonomic chained systems with integral input-to-state stable (iISS) inverse dynamics subject to unknown virtual control directions and parameter uncertainty included in drift terms. First, the system is divided into two interconnected subsystems according to the system’s structure. [...] Read more.
This paper investigates a class of nonholonomic chained systems with integral input-to-state stable (iISS) inverse dynamics subject to unknown virtual control directions and parameter uncertainty included in drift terms. First, the system is divided into two interconnected subsystems according to the system’s structure. Second, one controller is designed using a switch strategy for state finite escape. Then, another controller and adaptive law are designed by combining a reduced-order state observer and backstepping method after input-state scaling. Finally, simulation results validate the feasibility of the proposed control algorithm. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

21 pages, 11121 KiB  
Article
Deep-Reinforcement-Learning-Based Object Transportation Using Task Space Decomposition
by Gyuho Eoh
Sensors 2023, 23(10), 4807; https://doi.org/10.3390/s23104807 - 16 May 2023
Viewed by 1088
Abstract
This paper presents a novel object transportation method using deep reinforcement learning (DRL) and the task space decomposition (TSD) method. Most previous studies on DRL-based object transportation worked well only in the specific environment where a robot learned how to transport an object. [...] Read more.
This paper presents a novel object transportation method using deep reinforcement learning (DRL) and the task space decomposition (TSD) method. Most previous studies on DRL-based object transportation worked well only in the specific environment where a robot learned how to transport an object. Another drawback was that DRL only converged in relatively small environments. This is because the existing DRL-based object transportation methods are highly dependent on learning conditions and training environments; they cannot be applied to large and complicated environments. Therefore, we propose a new DRL-based object transportation that decomposes a difficult task space to be transported into simple multiple sub-task spaces using the TSD method. First, a robot sufficiently learned how to transport an object in a standard learning environment (SLE) that has small and symmetric structures. Then, a whole-task space was decomposed into several sub-task spaces by considering the size of the SLE, and we created sub-goals for each sub-task space. Finally, the robot transported an object by sequentially occupying the sub-goals. The proposed method can be extended to a large and complicated new environment as well as the training environment without additional learning or re-learning. Simulations in different environments are presented to verify the proposed method, such as a long corridor, polygons, and a maze. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

19 pages, 7544 KiB  
Article
TeCVP: A Time-Efficient Control Method for a Hexapod Wheel-Legged Robot Based on Velocity Planning
by Junkai Sun, Zezhou Sun, Jianfei Li, Chu Wang, Xin Jing, Qingqing Wei, Bin Liu and Chuliang Yan
Sensors 2023, 23(8), 4051; https://doi.org/10.3390/s23084051 - 17 Apr 2023
Cited by 2 | Viewed by 1327
Abstract
Addressing the problem that control methods of wheel-legged robots for future Mars exploration missions are too complex, a time-efficient control method based on velocity planning for a hexapod wheel-legged robot is proposed in this paper, which is named time-efficient control based on velocity [...] Read more.
Addressing the problem that control methods of wheel-legged robots for future Mars exploration missions are too complex, a time-efficient control method based on velocity planning for a hexapod wheel-legged robot is proposed in this paper, which is named time-efficient control based on velocity planning (TeCVP). When the foot end or wheel at knee comes into contact with the ground, the desired velocity of the foot end or knee is transformed according to the velocity transformation of the rigid body from the desired velocity of the torso which is obtained by the deviation of torso position and posture. Furthermore, the torques of joints can be obtained by impedance control. When suspended, the leg is regarded as a system consisting of a virtual spring and a virtual damper to realize control of legs in the swing phase. In addition, leg sequences of switching motion between wheeled configuration and legged configuration are planned. According to a complexity analysis, velocity planning control has lower time complexity and less times of multiplication and addition compared with virtual model control. In addition, simulations show that velocity planning control can realize stable periodic gait motion, wheel-leg switching motion and wheeled motion and the operation time of velocity planning control is about 33.89% less than that of virtual model control, which promises a great prospect for velocity planning control in future planetary exploration missions. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

17 pages, 3316 KiB  
Article
Indoor Robot Path Planning Using an Improved Whale Optimization Algorithm
by Qing Si and Changyong Li
Sensors 2023, 23(8), 3988; https://doi.org/10.3390/s23083988 - 14 Apr 2023
Cited by 2 | Viewed by 1479
Abstract
An improved whale optimization algorithm is proposed to solve the problems of the original algorithm in indoor robot path planning, which has slow convergence speed, poor path finding ability, low efficiency, and is easily prone to falling into the local shortest path problem. [...] Read more.
An improved whale optimization algorithm is proposed to solve the problems of the original algorithm in indoor robot path planning, which has slow convergence speed, poor path finding ability, low efficiency, and is easily prone to falling into the local shortest path problem. First, an improved logistic chaotic mapping is applied to enrich the initial population of whales and improve the global search capability of the algorithm. Second, a nonlinear convergence factor is introduced, and the equilibrium parameter A is changed to balance the global and local search capabilities of the algorithm and improve the search efficiency. Finally, the fused Corsi variance and weighting strategy perturbs the location of the whales to improve the path quality. The improved logical whale optimization algorithm (ILWOA) is compared with the WOA and four other improved whale optimization algorithms through eight test functions and three raster map environments for experiments. The results show that ILWOA has better convergence and merit-seeking ability in the test function. In the path planning experiments, the results are better than other algorithms when comparing three evaluation criteria, which verifies that the path quality, merit-seeking ability, and robustness of ILWOA in path planning are improved. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

22 pages, 9987 KiB  
Article
A Novel Real-Time Autonomous Crack Inspection System Based on Unmanned Aerial Vehicles
by Kwai-Wa Tse, Rendong Pi, Yuxiang Sun, Chih-Yung Wen and Yurong Feng
Sensors 2023, 23(7), 3418; https://doi.org/10.3390/s23073418 - 24 Mar 2023
Cited by 4 | Viewed by 2168
Abstract
Traditional methods on crack inspection for large infrastructures require a number of structural health inspection devices and instruments. They usually use the signal changes caused by physical deformations from cracks to detect the cracks, which is time-consuming and cost-ineffective. In this work, we [...] Read more.
Traditional methods on crack inspection for large infrastructures require a number of structural health inspection devices and instruments. They usually use the signal changes caused by physical deformations from cracks to detect the cracks, which is time-consuming and cost-ineffective. In this work, we propose a novel real-time crack inspection system based on unmanned aerial vehicles for real-world applications. The proposed system successfully detects and classifies various types of cracks. It can accurately find the crack positions in the world coordinate system. Our detector is based on an improved YOLOv4 with an attention module, which produces 90.02% mean average precision (mAP) and outperforms the YOLOv4-original by 5.23% in terms of mAP. The proposed system is low-cost and lightweight. Moreover, it is not restricted by navigation trajectories. The experimental results demonstrate the robustness and effectiveness of our system in real-world crack inspection tasks. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

16 pages, 3534 KiB  
Article
Constant Force-Tracking Control Based on Deep Reinforcement Learning in Dynamic Auscultation Environment
by Tieyi Zhang, Chao Chen, Minglei Shu, Ruotong Wang, Chong Di and Gang Li
Sensors 2023, 23(4), 2186; https://doi.org/10.3390/s23042186 - 15 Feb 2023
Viewed by 1461
Abstract
Intelligent medical robots can effectively help doctors carry out a series of medical diagnoses and auxiliary treatments and alleviate the current shortage of social personnel. Therefore, this paper investigates how to use deep reinforcement learning to solve dynamic medical auscultation tasks. We propose [...] Read more.
Intelligent medical robots can effectively help doctors carry out a series of medical diagnoses and auxiliary treatments and alleviate the current shortage of social personnel. Therefore, this paper investigates how to use deep reinforcement learning to solve dynamic medical auscultation tasks. We propose a constant force-tracking control method for dynamic environments and a modeling method that satisfies physical characteristics to simulate the dynamic breathing process and design an optimal reward function for the task of achieving efficient learning of the control strategy. We have carried out a large number of simulation experiments, and the error between the tracking of normal force and expected force is basically within ±0.5 N. The control strategy is tested in a real environment. The preliminary results show that the control strategy performs well in the constant force-tracking of medical auscultation tasks. The contact force is always within a safe and stable range, and the average contact force is about 5.2 N. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

17 pages, 1160 KiB  
Article
Variational Information Bottleneck Regularized Deep Reinforcement Learning for Efficient Robotic Skill Adaptation
by Guofei Xiang, Songyi Dian, Shaofeng Du and Zhonghui Lv
Sensors 2023, 23(2), 762; https://doi.org/10.3390/s23020762 - 09 Jan 2023
Viewed by 2487
Abstract
Deep Reinforcement Learning (DRL) algorithms have been widely studied for sequential decision-making problems, and substantial progress has been achieved, especially in autonomous robotic skill learning. However, it is always difficult to deploy DRL methods in practical safety-critical robot systems, since the training and [...] Read more.
Deep Reinforcement Learning (DRL) algorithms have been widely studied for sequential decision-making problems, and substantial progress has been achieved, especially in autonomous robotic skill learning. However, it is always difficult to deploy DRL methods in practical safety-critical robot systems, since the training and deployment environment gap always exists, and this issue would become increasingly crucial due to the ever-changing environment. Aiming at efficiently robotic skill transferring in a dynamic environment, we present a meta-reinforcement learning algorithm based on a variational information bottleneck. More specifically, during the meta-training stage, the variational information bottleneck first has been applied to infer the complete basic tasks for the whole task space, then the maximum entropy regularized reinforcement learning framework has been used to learn the basic skills consistent with that of basic tasks. Once the training stage is completed, all of the tasks in the task space can be obtained by a nonlinear combination of the basic tasks, thus, the according skills to accomplish the tasks can also be obtained by some way of a combination of the basic skills. Empirical results on several highly nonlinear, high-dimensional robotic locomotion tasks show that the proposed variational information bottleneck regularized deep reinforcement learning algorithm can improve sample efficiency by 200–5000 times on new tasks. Furthermore, the proposed algorithm achieves substantial asymptotic performance improvement. The results indicate that the proposed meta-reinforcement learning framework makes a significant step forward to deploy the DRL-based algorithm to practical robot systems. Full article
(This article belongs to the Special Issue Advances in Mobile Robot Perceptions, Planning, Control and Learning)
►▼ Show Figures

Figure 1

Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-15
Back to TopTop