Autonomous Robotic Applications Research Articles

CESDQL: Communicative experience-sharing deep Q-learning for scalability in multi-robot collaboration with sparse reward

Owing to the massive transformation in industrial processes and logistics, warehouses are also undergoing advanced automation. The application of Autonomous Mobile Robots (a.k.a. multi-robots) is one of the important elements of overall warehousing automation. The autonomous collaborative behaviour of the multi-robots can be considered as employment on a control task and, thus, can be optimised using multi-agent reinforcement learning (MARL). Consequently, an autonomous warehouse is to be represented by an MARL environment. An MARL environment replicating an autonomous warehouse poses the challenge of exploration due to sparse reward leading to inefficient collaboration. This challenge aggravates further with an increase in the number of robots and the grid size, i.e., scalability. This research proposes Communicative Experience-Sharing Deep Q-Learning (CESDQL) based on Q-learning, a novel hybrid multi-robot communicative framework for scalability for MARL collaboration with sparse rewards, where exploration is challenging and makes collaboration difficult. CESDQL makes use of experience-sharing through collective sampling from the Experience (Replay) buffer and communication through Communicative Deep recurrent Q-network (CommDRQN), a Q-function approximator. Through empirical evaluation of CESDQL in a variety of collaborative scenarios, it is established that CESDQL outperforms the baselines in terms of convergence and stable learning. Overall, CESDQL achieves 5%, 69%, 60%, 211%, 171%, 3.8% & 10% more final accumulative training returns than the closest performing baseline by scenario, and, 27%, 10.33% & 573% more final average training returns than the closest performing baseline by the big-scale scenario.

Improved 3D laser point cloud reconstruction for autonomous mobile robot applications by using SVM-R technique

PurposeIn autonomous mobile robots, high-level accuracy and precision in 3D perception are required for object detection, shape estimation and obstacle distance measurement. However, the existing methods suffer from limitations like inaccurate point clouds, noise in sensor data and synchronization problems between 2D LiDAR and servomotor. These factors can lead to the wrong perception and also introduce noise during sensor registration. Thus, the purpose of this study is to address these limitations and enhance the perception in autonomous mobile robots.Design/methodology/approachA new sensor mounting structure is developed for 3D mapping by using a 2D LiDAR and servomotor. The proposed method uses a support vector machine regression (SVM-R) technique to optimize the waypoints of the servomotor for the point cloud reconstruction process and to obtain a highly accurate and detailed representation of the environment.FindingsThe study includes an analysis of the SVM-R model, including Linear, radial basis function (RBF) and Polynomial kernel. Results show that the Linear kernel performs better with the lowest 3.67 mean absolute error (MAE), 26.24 mean squared error (MSE) and 5.12 root mean squared error (RMSE) values than the RBF and Polynomial kernels. The 2D to 3D point cloud reconstruction shows that the proposed method with a new sensor mounting structure performs better in perception accuracy and achieves an error of 0.45% in measuring the height of the target objects whereas in previous techniques the error was very large.Originality/valueThe study shows the effectiveness of SVM-R in the 3D point cloud reconstruction process and exhibits remarkable performance for object height measurement. Further, the proposed technique is applicable for future advanced visual applications and has a superior performance over the other conventional methods.

Real-Time Semantic Segmentation of 3D LiDAR Point Clouds for Aircraft Engine Detection in Autonomous Jetbridge Operations

This paper presents a study on aircraft engine identification using real-time 3D LiDAR point cloud segmentation technology, a key element for the development of automated docking systems in airport boarding facilities, known as jetbridges. To achieve this, 3D LiDAR sensors utilizing a spinning method were employed to gather surrounding environmental 3D point cloud data. The raw 3D environmental data were then filtered using the 3D RANSAC technique, excluding ground data and irrelevant apron areas. Segmentation was subsequently conducted based on the filtered data, focusing on aircraft sections. For the segmented aircraft engine parts, the centroid of the grouped data was computed to determine the 3D position of the aircraft engine. Additionally, PointNet was applied to identify aircraft engines from the segmented data. Dynamic tests were conducted in various weather and environmental conditions, evaluating the detection performance across different jetbridge movement speeds and object-to-object distances. The study achieved a mean intersection over union (mIoU) of 81.25% in detecting aircraft engines, despite experiencing challenging conditions such as low-frequency vibrations and changes in the field of view during jetbridge maneuvers. This research provides a strong foundation for enhancing the robustness of jetbridge autonomous docking systems by reducing the sensor noise and distortion in real-time applications. Our future research will focus on optimizing sensor configurations, especially in environments where sea fog, snow, and rain are frequent, by combining RGB image data with 3D LiDAR information. The ultimate goal is to further improve the system’s reliability and efficiency, not only in jetbridge operations but also in broader autonomous vehicle and robotics applications, where precision and reliability are critical. The methodologies and findings of this study hold the potential to significantly advance the development of autonomous technologies across various industrial sectors.

AN IMPROVED COVERAGE PATH PLANNING FOR SERVICE ROBOTS BASED ON BACKTRACKING METHOD

Service robots are of paramount importance in both industrial and residential environments. Nevertheless, these robots frequently encounter obstacles in the form of perilous elements present in their surroundings, which may result in their malfunction and hinder their optimal performance. Hence, autonomous robot applications heavily rely on coverage path planning algorithms, which facilitate the execution of area coverage operations in a streamlined and economical manner. The article presents a coverage path planning algorithm for service robots that is based on a backtracking procedure. The proposed method attempts to strike a balance between the safety of the cleaning automaton and the extent of coverage. Following straight-forward instructions, the mobile robot (MR) initially generates a spiral or zigzag trajectory. It examines each position for possible backtracking locations. When obstacles and visited positions obstruct the path, the improved A* algorithm is utilized to return to known unvisited regions. Subsequently, the robot proceeds to sweep the regions until no remaining back-tracking sites remain. Simulation of the method on the Robot Operating System (ROS) verifies that it is possible to restrict the movement of high-risk robots.

Development of High-Temperature Permanent Magnet Synchronous Motor for Autonomous Robotic Navigation: A Design and Finite Element Analysis Approach

The traditional industrial robots come with the prime mover, i.e. Electric Motors (EM) which ranges from a few hundred too few kilo watts of power ratings. However, for autonomous robotic navigation systems, we require motors which are light weighted with the aspect of high torque and power density. This aspect is very critical, when the EMs in robotic navigations are subjected to harsh high temperature survival conditions, where the sustainability of the performance metrics of the electromagnetic system of the EMs degrade with the prevailing high temperature conditions. Hence, this research work address and formulate the design methodology to develop a 630 W High Temperature PMSM (HTPMSM) in the aspect of high torque and power density, which can be used for the autonomous robotic navigation systems under high temperature survival conditions of 200°C. Two types of rotor configurations i.e. the Surface Permanent Magnet type (SPM) and the Interior Permanent Magnet type (IPM) of HTPMSM are examined for its optimal electromagnetic metrics under the temperature conditions of 200°C. The 630 W HTPMSM is designed to deliver the rated torque of 2 Nm within the volumetric & diametric constraints of D x L which comes at 80 x 70 mm at the rated speed of 3000 rpm with the survival temperature of 200°C with the target efficiency of greater than 90%. The FEM based results are validated through the hardware prototypes for both SPM and IPM types, and the results confirm the effectiveness of the proposed design methodology of HTPMSM for sustainable autonomous robotic navigation applications.

Improved nonlinear model predictive control with inequality constraints using particle filtering for nonlinear and highly coupled dynamical systems

Abstract Motion planning and controller design are challenging tasks for highly coupled and nonlinear dynamical systems such as autonomous vehicles and robotic applications. Nonlinear model predictive control (NMPC) is an emerging technique in which sampling-based methods are used to synthesize the control and trajectories for complex systems. In this study, we have developed the sampling-based motion planning algorithm with NMPC through Bayesian estimation to solve the online nonlinear constrained optimization problem. In the literature, different filtration techniques have been applied to extract knowledge of states in the presence of noise. Due to the detrimental effects of linearization, the Kalman filter with NMPC only achieves modest effectiveness. Moving horizon estimation (MHE), on the other hand, frequently relies on simplifying assumptions and lacks an effective recursive construction. Additionally, it adds another optimization challenge to the regulation problem that has to be solved online. To address this problem, particle filtering is implemented for Bayesian filtering in nonlinear and highly coupled dynamical systems. It is a sequential Monte Carlo method that involves representing the posterior distribution of the state of the system using a set of weighted particles that are propagated through time using a recursive algorithm. For nonlinear and strongly coupled dynamical systems, the novel sampling-based NMPC technique is effective and simple to use. The efficiency of the suggested method has been assessed using simulated studies.

Analysis of kinematic characteristics of a mobile caterpillar robot with a SCARA-type manipulator

Automation and robotization of various production and technological processes in many industries is one of the leading trends in the development of modern society. Industrial robots have recently become quite widespread, and it is almost impossible to imagine any modern production in the fields of mechanical engineering (machine building), instrumentation, pharmaceuticals, food, chemical industries, etc., without robotic complexes. Over the past few decades, another area of robotics has emerged: autonomous mobile robots. It combines research in mechanics, electronics, and computer technologies, including artificial intelligence. Among the most common applications of autonomous mobile robots are the performance of various technological operations in places that are dangerous to human life (radiation, biological or chemical contamination) or uninhabitable (space, sea depths, volcanic craters, etc.). Mobile robots have also proven themselves in rescue operations during cataclysms and natural disasters, anti-terrorist operations, military operations, mine clearance, etc. Given the urgency of the issue of mobile robotics development, this article proposes a new design of an autonomous robotic complex built on the basis of a tracked chassis and equipped with a SCARA-type manipulator. The main task of the developed robot is to perform various technological operations in places where human presence is dangerous or impossible, in particular, when performing demining tasks. In the course of the research, the kinematics of the manipulator was analyzed in detail to determine its working area, and the kinematic parameters of the tracked chassis were experimentally tested while it was moving over rough terrain. The obtained results can be used to further improve the design and control system of the robot and manipulator and in the process of determining the specific technological tasks that will be assigned to this robotic platform.

Motion planning and control of an autonomous mobile robot

The application of autonomous mobile robots (AMRs) has gradually become crucial in smart factories due to the advantages of improving production efficiency and reducing labour costs. Motion planning has been a key part of AMR control development. This paper presents motion planning and position tracking control systems of an omnidirectional wheel AMR powered by a hybrid fuel cell and battery power source. First, the kinematical and dynamic models of the AMR are introduced. The navigation system comprises three loops, with the first loop being motor control, the second loop being position tracking control, and a motion planning layer. The position data of the AMR for feedback control is obtained through sensor fusion of data from the inertial measurement unit (IMU) sensor, encoder sensor, and ranging sensor with simultaneous localisation and mapping (SLAM) algorithm. The motion planning is then applied to obtain an optimal path with the shortest distance and collision-free movement. In addition, the tracking algorithm is designed to drive the AMR to follow the optimal path and achieve high accuracy. The experimental results show a 30% improvement in tracking accuracy compared to traditional approaches and 8 hours of continuous working, which is promising for industrial applications, and the results are satisfactory in terms of both accuracy and efficiency requirements.

A novel directional sampling-based path planning algorithm for ambient intelligence navigation scheme in autonomous mobile robots

Path planning algorithms determine the performance of the ambient intelligence navigation schemes in autonomous mobile robots. Sampling-based path planning algorithms are widely employed in autonomous mobile robot applications. RRT*, or Optimal Rapidly Exploring Random Trees, is a very effective sampling-based path planning algorithm. However, the RRT* solution converges slowly. This study proposes a directional random sampling-based RRT* path planning algorithm known as DR-RRT* to address the slow convergence issue. The novelty of the proposed method is that it reduces the search space by combining directional non-uniform sampling with uniform sampling. It employs a random selection approach to combine the non-uniform directional sampling method with uniform sampling. The proposed path planning algorithm is validated in three different environments with a map size of 384*384, and its performance is compared to two existing algorithms: RRT* and Informed RRT*. Validation is carried out utilizing a TurtleBot3 robot with the Gazebo Simulator and the Robotics Operating System (ROS) Melodic. The proposed DR-RRT* path planning algorithm is better than both RRT* and Informed RRT* in four performance measures: the number of nodes visited, the length of the path, the amount of time it takes, and the rate at which the path converges. The proposed DR-RRT* global path planning algorithm achieves a success rate of 100% in all three environments, and it is suited for use in all kinds of environments.

Pose Graph Relocalization with Deep Object Detection and BIM-Supported Object Landmark Dictionary

Indoor localization is a prerequisite for autonomous robot applications in the construction industry. However, traditional localization techniques rely on low-level features and do not exploit construction-related semantics. They also are sensitive to environmental factors such as illumination and reflection rate, and therefore suffer from unexpected drifts and failures. This study proposes a pose graph relocalization framework that utilizes object-level landmarks to enhance a traditional visual localization system. The proposed framework builds an object landmark dictionary from Building Information Model (BIM) as prior knowledge. Then a multimodal deep neural network (DNN) is proposed to realize 3D object detection in real time, followed by instance-level object association with false-positive rejection, and relative pose estimation with outlier removal. Finally, a keyframe-based graph optimization is performed to rectify the drifts of traditional visual localization. The proposed framework was validated using a mobile platform with red-green-blue-depth (RGB-D) and inertial sensors, and the test scene was an indoor office environment with furnishing elements. The object detection model achieved 62.9% mean average precision (mAP). The relocalization technique reduced translational drifts by 64.67% and rotational drifts by 41.59% compared with traditional visual–inertial odometry.

Reinforcement Learning for Mobile Robotics Exploration: A Survey.

Efficient exploration of unknown environments is a fundamental precondition for modern autonomous mobile robot applications. Aiming to design robust and effective robotic exploration strategies, suitable to complex real-world scenarios, the academic community has increasingly investigated the integration of robotics with reinforcement learning (RL) techniques. This survey provides a comprehensive review of recent research works that use RL to design unknown environment exploration strategies for single and multirobots. The primary purpose of this study is to facilitate future research by compiling and analyzing the current state of works that link these two knowledge domains. This survey summarizes: what are the employed RL algorithms and how they compose the so far proposed mobile robot exploration strategies; how robotic exploration solutions are addressing typical RL problems like the exploration-exploitation dilemma, the curse of dimensionality, reward shaping, and slow learning convergence; and what are the performed experiments and software tools used for learning and testing. Achieved progress is described, and a discussion about remaining limitations and future perspectives is presented.

Adaptive Multi-Robot Communication System and Collision Avoidance Algorithm for Precision Agriculture

In precision farming technology, the interest of the researchers has been focused on the applications of autonomous mobile robots for agricultural operations such as planting, inspection, spraying, and harvesting. However, each autonomous robot generally performs a single agricultural task. In this context, complete autonomy in precision farming can be achieved by using coordinated multi-robot systems that can easily and safely cooperate to accomplish agricultural tasks. The efficiency of the multi-robot system depends on the number of robots, the size of the robots, the distance between each robot, the instant location and heading angle of the robots, and the size of the farmland. This paper describes the development of wireless Robot to Robot (R2R) communication system architecture and the collision avoidance algorithm for multi-robot precision farming applications. The developed system uses the fusion of a digital compass and GPS receiver for wirelessly broadcasting the spatial and temporal data of the mobile robots through WiFi. In this study, WiFi broadcasting was chosen for reasons such as the advantages of long wireless signal range and strength, not being easily affected by weather and dust, low cost, and so on. The proposed system realizes the real-time wireless broadcasting of the mobile robot information for eliminating the collision of mobile robots and improving the level of safety management. The results show that the system has flexible, reliable, and adaptable solution, and thus can increase the efficiency of the multi-robot system in precision farming applications.

A bio-inspired algorithm in image-based path planning and localization using visual features and maps

With the growing applications of autonomous robots and vehicles in unknown environments, studies on image-based localization and navigation have attracted a great deal of attention. This study is significantly motivated by the observation that relatively little research has been published on the integration of cutting-edge path planning algorithms for robust, reliable, and effective image-based navigation. To address this gap, a biologically inspired Bat Algorithm (BA) is introduced and adopted for image-based path planning in this paper. The proposed algorithm utilizes visual features as the reference in generating a path for an autonomous vehicle, and these features are extracted from the obtained images by convolutional neural networks (CNNs). The paper proceeds as follows: first, the requirements for image-based localization and navigation are described. Second, the principles of the BA are explained in order to expound on the justifications for its successful incorporation in image-based navigation. Third, in the proposed image-based navigation system, the BA is developed and implemented as a path planning tool for global path planning. Finally, the performance of the BA is analyzed and verified through simulation and comparison studies to demonstrate its effectiveness.

Model-Predictive Control for Omnidirectional Mobile Robots in Logistic Environments Based on Object Detection Using CNNs.

Object detection is an essential component of autonomous mobile robotic systems, enabling robots to understand and interact with the environment. Object detection and recognition have made significant progress using convolutional neural networks (CNNs). Widely used in autonomous mobile robot applications, CNNs can quickly identify complicated image patterns, such as objects in a logistic environment. Integration of environment perception algorithms and motion control algorithms is a topic subjected to significant research. On the one hand, this paper presents an object detector to better understand the robot environment and the newly acquired dataset. The model was optimized to run on the mobile platform already on the robot. On the other hand, the paper introduces a model-based predictive controller to guide an omnidirectional robot to a particular position in a logistic environment based on an object map obtained from a custom-trained CNN detector and LIDAR data. Object detection contributes to a safe, optimal, and efficient path for the omnidirectional mobile robot. In a practical scenario, we deploy a custom-trained and optimized CNN model to detect specific objects in the warehouse environment. Then we evaluate, through simulation, a predictive control approach based on the detected objects using CNNs. Results are obtained in object detection using a custom-trained CNN with an in-house acquired data set on a mobile platform and in the optimal control for the omnidirectional mobile robot.

Visual Place Recognition in Changing Environments with Sequence Representations on the Distance-Space Domain

Navigating in a perpetually changing world can provide the basis for numerous challenging autonomous robotic applications. With a view to long-term autonomy, visual place recognition (vPR) systems should be able to robustly operate under extreme appearance changes in their environment. Typically, the utilized data representations are heavily influenced by those changes, negatively affecting the vPR performance. In this article, we propose a sequence-based technique that decouples such changes from the similarity estimation procedure. This is achieved by remapping the sequential representation data into the distance-space domain, i.e., a domain in which we solely consider the distances between image instances, and subsequently normalize them. In such a way, perturbations related to different environmental conditions and embedded into the original representation vectors are avoided, therefore the scene recognition efficacy is enhanced. We evaluate our framework under multiple different instances, with results indicating a significant performance improvement over other approaches.

Enabling Edge-Cloud Video Analytics for Robotics Applications

Emerging deep learning-based video analytics tasks demand computation-intensive neural networks and powerful computing resources on the cloud to achieve high accuracy. Due to the latency requirement and limited network bandwidth, edge-cloud systems adaptively compress the data to strike a balance between overall analytics accuracy and bandwidth consumption. However, the degraded data leads to another issue of poor tail accuracy, which means the extremely low accuracy of a few semantic classes and video frames. Autonomous robotics applications especially value the tail accuracy performance but suffer using the prior edge-cloud systems. We present Runespoor, an edge-cloud video analytics system to manage the tail accuracy and enable emerging robotics applications. We train and deploy a super-resolution model tailored for the tail accuracy of analytics tasks on the server to significantly improves the performance on hard-to-detect classes and sophisticated frames. During online operation, we use an adaptive data rate controller to further improve the tail performance by instantly adjusting the data rate policy according to the video content. Our evaluation shows that Runespoor improves class-wise tail accuracy by up to 300%, frame-wise 90%/99% tail accuracy by up to 22%/54%, and greatly improves the overall accuracy and bandwidth trade-off.

Explainable Goal-driven Agents and Robots - A Comprehensive Review

Recent applications of autonomous agents and robots have brought attention to crucial trust-related challenges associated with the current generation ofartificial intelligence (AI)systems. AI systems based on the connectionist deep learning neural network approach lack capabilities of explaining their decisions and actions to others, despite their great successes. Without symbolic interpretation capabilities, they are ‘black boxes’, which renders their choices or actions opaque, making it difficult to trust them in safety-critical applications. The recent stance on the explainability of AI systems has witnessed several approaches toeXplainable Artificial Intelligence (XAI); however, most of the studies have focused on data-driven XAI systems applied in computational sciences. Studies addressing the increasingly pervasive goal-driven agents and robots are sparse at this point in time. This paper reviews approaches on explainable goal-driven intelligent agents and robots, focusing on techniques for explaining and communicating agents’ perceptual functions (e.g., senses, vision) and cognitive reasoning (e.g., beliefs, desires, intentions, plans, and goals) with humans in the loop. The review highlights key strategies that emphasize transparency, understandability, and continual learning for explainability. Finally, the paper presents requirements for explainability and suggests a road map for the possible realization of effective goal-driven explainable agents and robots.

Object-Aware View Planning for Autonomous 3-D Model Reconstruction of Buildings Using a Mobile Robot

Active mapping ranks among the most critical applications of autonomous mobile robots. In this article, we focus on 3D model reconstruction of buildings using an outdoor ground robot equipped with a solid-state LiDAR and a gimbal. A view-planning approach is proposed to generate and evaluate view sequences with reference to the characteristics of solid-state LiDAR and the robot configuration. To reconstruct the specific buildings while ignoring unrelated objects in the environment, we introduce object awareness into the map representation, and frontier- and sampling-based strategies are combined to plan the view sequences for the effective observation of a target building. Instead of greedily maximizing information gain, frontier clusters indicating the mapped object’s incomplete surfaces are used to evaluate the sampled views, and sampling the minimum view set generates view sequences able to efficiently cover the clusters. Moreover, using object awareness to partition the area of sampled views around the building facilitates a divide-and-conquer strategy that overcomes local characteristics and guides a complete reconstruction. Simulation and real-world experiments revealed that our approach allows the thorough and efficient active reconstruction of various buildings and even outperforms state-of-the-art approaches.

Liquid crystal elastomer composite-based photo-oscillator for microrobots

In recent years, photoactive materials have attracted extensive interest in microrobots for their attractive abilities of untethered and tunable control with light. Conventional photo-oscillators based on Azo-containing liquid crystal network require complex surface alignment techniques and light sources with specific wavelengths and polarity, which limits their application in controlled autonomy. Here, we report a facile strategy to create self-oscillating microrobots powered and controlled by a wide spectrum of constant light. The oscillators are composed of a layer of candle soot (CS)-containing liquid crystal elastomer (LCE) attached to a layer of polydimethylsiloxane (PDMS). The strip-shaped oscillators with one end fixed can execute self-sustained oscillation through a self-shadowing mechanism. LCEs with CS as an excellent photo-absorber provides superior photothermal actuation, while PDMS with low viscoelasticity accelerates the actuation-recovery cycle of the oscillator. Our LCE composite photo-oscillators show tunable frequencies and amplitudes by structural and light intensity modulation, showing potential for autonomous soft robotic applications.

Application and Development of Autonomous Robots in Concrete Construction: Challenges and Opportunities

Updated concrete construction robots are designed to optimize equipment operation, improve safety, enhance workspace awareness, and further ensure a proper working environment for construction workers. The importance of concrete construction robots has been constantly highlighted, as they have a profound impact on construction quality and efficiency. Autonomous vehicle driving monitoring has been widely employed in concrete construction robots; however, they lack clear relevance to the key functions in the building process. This paper aims to bridge this knowledge gap by systematically classifying and summarizing the existing concrete construction robots, analyzing their existing problems, and providing direction for their future development. The prescription criteria and selection of robots depend on the concrete construction process, which includes six common functional levels: distribution, leveling and compaction, floor finishing, surface painting, 3D printing, and surveillance. Misunderstood functions and the improper adjustment of construction robots may lead to increased cost, reduced effectiveness, and restricted application scenarios. Our review identifies current commercial and recently studied concrete construction robots to facilitate the standardization and optimization of robotic construction design. Moreover, this study may be able to guide future research and technology development efforts for autonomous robots in concrete construction.