Real Robot Applications Research Articles

6-DOFs Robot Placement Based on the Multi-Criteria Procedure for Industrial Applications

Robot acceptance is rapidly increasing in many different industrial applications. The advancement of production systems and machines requires addressing the productivity complexity and flexibility of current manufacturing processes in quasi-real time. Nowadays, robot placement is still achieved via industrial practices based on the expertise of the workers and technicians, with the adoption of offline expensive software that demands time-consuming simulations, detailed time-and-motion mapping activities, and high competencies. Current challenges have been addressed mainly via path planning or robot-to-workpiece location optimization. Numerous solutions, from analytical to physical-based and data-driven formulation, have been discussed in the literature to solve these challenges. In this context, the machine learning approach has proven its superior performance. Nevertheless, the industrial environment is complex to model, generating extra training effort and making the learning procedure, in some cases, inefficient. The industrial problems concern workstation productivity; path-constrained minimal-time motions, considering the actuator’s torque limits; followed by robot vibration and the reduction in its accuracy and lifetime. This paper presents a procedure to find the robot base location for a prescribed task within the robot’s workspace, complying with multiple criteria. The proposed hybrid procedure includes analytical, physical-based, and data-driven modeling to solve the optimization problem. The contribution of the algorithm, for a given user-defined task, is the search for the best robot base location that enables the target points, maximizing the manipulability, avoiding singularities, and minimizing energy consumption. Firstly, the established method was verified using an anthropomorphic robot that considers different levels of a priori kinematics and system dynamics knowledge. The feasibility of the proposed method was evaluated through various simulations for small- and medium-sized robots. Then, a commercial offline program was compared, considering three scenarios and fourteen robots demonstrating an energy reduction in the 7.6–13.2% range. Moreover, the unknown joint dependency in real robot applications was investigated. From 11 robot positions for each active joint, a direct kinematic was appraised with an automatic DH scheme that generates the 3D workspace with an RMSE lower than 65.0 µm. Then, the inverse kinematic was computed using an ANN technique tuned with a genetic algorithm showing an RMSE in an S-shape task close to 702.0 µm. Finally, three experimental campaigns were performed with a set of tasks, repetitions, end-effector velocity, and payloads. The energy consumption reduction was observed in the 12.7–22.9% range. Consequently, the proposed procedure supports the reduction in workstation setup time and energy saving during industrial operations.

Overview of image-based 3D reconstruction technology

Three-dimensional (3D) reconstruction technology is the key technology to establish and express the objective world by using computer, and it is widely used in real 3D, automatic driving, aerospace, navigation and industrial robot applications. According to different principles, it is mainly divided into methods based on traditional multi-view geometry and methods based on deep learning. This paper introduces the above methods from the perspective of three-dimensional space representation. The feature extraction and stereo matching theory of traditional 3D reconstruction methods are the theoretical basis of 3D reconstruction methods based on deep learning, so the paper focuses on them. With the development of traditional 3D reconstruction methods and the development of deep learning related theories, the explicit deep learning 3D reconstruction method represented by MVSNet and the implicit 3D reconstruction method represented by NeRF have been gradually developed. At the same time, the dataset and evaluation indicators for 3D reconstruction were introduced. Finally, a summary of image based 3D reconstruction was provided.

SLAMICP Library: Accelerating Obstacle Detection in Mobile Robot Navigation via Outlier Monitoring following ICP Localization.

The Iterative Closest Point (ICP) is a matching technique used to determine the transformation matrix that best minimizes the distance between two point clouds. Although mostly used for 2D and 3D surface reconstruction, this technique is also widely used for mobile robot self-localization by means of matching partial information provided by an onboard LIDAR scanner with a known map of the facility. Once the estimated position of the robot is obtained, the scans gathered by the LIDAR can be analyzed to locate possible obstacles obstructing the planned trajectory of the mobile robot. This work proposes to speed up the obstacle detection process by directly monitoring outliers (discrepant points between the LIDAR scans and the full map) spotted after ICP matching instead of spending time performing an isolated task to re-analyze the LIDAR scans to detect those discrepancies. In this work, a computationally optimized ICP implementation has been adapted to return the list of outliers along with other matching metrics, computed in an optimal way by taking advantage of the parameters already calculated in order to perform the ICP matching. The evaluation of this adapted ICP implementation in a real mobile robot application has shown that the time required to perform self-localization and obstacle detection has been reduced by 36.7% when obstacle detection is performed simultaneously with the ICP matching instead of implementing a redundant procedure for obstacle detection. The adapted ICP implementation is provided in the SLAMICP library.

Condition-invariant and compact visual place description by convolutional autoencoder

AbstractVisual place recognition (VPR) in condition-varying environments is still an open problem. Popular solutions are convolutional neural network (CNN)-based image descriptors, which have been shown to outperform traditional image descriptors based on hand-crafted visual features. However, there are two drawbacks of current CNN-based descriptors: (a) their high dimension and (b) lack of generalization, leading to low efficiency and poor performance in real robotic applications. In this paper, we propose to use a convolutional autoencoder (CAE) to tackle this problem. We employ a high-level layer of a pre-trained CNN to generate features and train a CAE to map the features to a low-dimensional space to improve the condition invariance property of the descriptor and reduce its dimension at the same time. We verify our method in four challenging real-world datasets involving significant illumination changes, and our method is shown to be superior to the state-of-the-art. The code of our work is publicly available at https://github.com/MedlarTea/CAE-VPR.

Intrinsic motivation learning for real robot applications.

Intrinsic motivation learning for real robot applications.

Unsupervised Adversarial Domain Adaptation for Sim-to-Real Transfer of Tactile Images

Transferring optical tactile skills learned from simulated environments to the real world benefits many robotic tactile applications, which can reduce the cost of data collection. However, the models purely trained on simulated data are often difficult to generalize well to the unseen real world due to the domain gap between the faultless training images and testing images with unpredictable manufacturing defects or natural wear. In this paper, we propose an Adaptively Correlation-attentive and Task-related Network (ACTNet) for tactile image transfer, a novel unsupervised adversarial domain adaptation method to narrow the domain gap for pixel-level tactile perception tasks. An adaptively correlative attention mechanism is introduced to improve the generator, which is capable of leveraging global information and focusing on salient regions. We also construct a task-related constraint loss based on the tactile image classification task to facilitate the zero-shot sim-to-real transfer of the tactile image classifier. Extensive experiment results on tactile image generation and tactile image classification indicate that our method outperforms the state-of-the-art method. Classification accuracy of 92.85% in the real-world environment is achieved without using any real-world classification labels.

Self-Supervised Reinforcement Learning for Active Object Detection

Active object detection (AOD) offers significant advantage in expanding the perceptual capacity of a robotics system. AOD is formulated as a sequential action decision process to determine optimal viewpoints to identify objects of interest in a visual scene. While reinforcement learning (RL) has been successfully used to solve many AOD problems, conventional RL methods suffer from (i) sample inefficiency, and (ii) unstable outcome due to inter-dependencies of action type (direction of view change) and action range (step size of view change). To address these issues, we propose a novel self-supervised RL method, which employs self-supervised representations of viewpoints to initialize the policy network, and a self-supervised loss on action range to enhance the network parameter optimization. The output and target pairs of self-supervised learning loss are automatically generated from the policy network online prediction and a range shrinkage algorithm (RSA), respectively. The proposed method is evaluated and benchmarked on two public datasets (T-LESS and AVD) using on-policy and off-policy RL algorithms. The results show that our method enhances detection accuracy and achieves faster convergence on both datasets. By evaluating on a more complex environment with a larger state space (where viewpoints are more densely sampled), our method achieves more robust and stable performance. Our experiment on real robot application scenario to disambiguate similar objects in a cluttered scene has also demonstrated the effectiveness of the proposed method.

Pose Estimation for Robot Manipulators via Keypoint Optimization and Sim-to-Real Transfer

Keypoint detection is an essential building block for many robotic applications like motion capture and pose estimation. Historically, keypoints are detected using uniquely engineered markers such as checkerboards or fiducials. More recently, deep learning methods have been explored as they have the ability to detect user-defined keypoints in a marker-less manner. However, different manually selected keypoints can have uneven performance when it comes to detection and localization. An example of this can be found on symmetric robotic tools where DNN detectors cannot solve the correspondence problem correctly. In this work, we propose a new and autonomous way to define the keypoint locations that overcomes these challenges. The approach involves finding the optimal set of keypoints on robotic manipulators for robust visual detection and localization. Using a robotic simulator as a medium, our algorithm utilizes synthetic data for DNN training, and the proposed algorithm is used to optimize the selection of keypoints through an iterative approach. The results show that when using the optimized keypoints, the detection performance of the DNNs improved significantly. We further use the optimized keypoints for real robotic applications by using domain randomization to bridge the reality gap between the simulator and the physical world. The physical world experiments show how the proposed method can be applied to the wide-breadth of robotic applications that require visual feedback, such as camera-to-robot calibration, robotic tool tracking, and end-effector pose estimation. As a way to encourage further research in this topic, we establish the “Robot Pose” dataset, comprising calibration and tracking problems and ground truth data, available online ¹.

An Atlas for the Inkjet Printing of Large-Area Tactile Sensors.

This review aims to discuss the inkjet printing technique as a fabrication method for the development of large-area tactile sensors. The paper focuses on the manufacturing techniques and various system-level sensor design aspects related to the inkjet manufacturing processes. The goal is to assess how printed electronics simplify the fabrication process of tactile sensors with respect to conventional fabrication methods and how these contribute to overcoming the difficulties arising in the development of tactile sensors for real robot applications. To this aim, a comparative analysis among different inkjet printing technologies and processes is performed, including a quantitative analysis of the design parameters, such as the costs, processing times, sensor layout, and general system-level constraints. The goal of the survey is to provide a complete map of the state of the art of inkjet printing, focusing on the most effective topics for the implementation of large-area tactile sensors and a view of the most relevant open problems that should be addressed to improve the effectiveness of these processes.

Empirical study of future image prediction for image-based mobile robot navigation

Recent image-based robotic systems use predicted future state images to control robots. Therefore, the prediction accuracy of the future state image affects the performance of the robot. To predict images, most previous studies assume that the camera captures the entire scene and that the environment is static. However, in real robot applications, these assumptions do not always hold. For example, if a camera is attached to a mobile robot, its view changes from time to time.In this study, we analyzed the relationship between the performance of the image prediction model and the robot’s behavior, controlled by an image-based navigation algorithm. Through mobile robot navigation experiments using front-faced and omni-directional cameras, we discussed the capabilities of the image prediction models and demonstrated their performance when applied to the image-based navigation algorithm.Moreover, to adapt to the dynamic changes in the environment, we studied the effectiveness of directing the camera to the ceiling. We showed that robust navigation can be achieved without using images from cameras directed toward the front or the floor, because these views can be disturbed by moving objects in a dynamic environment.

Overcoming some drawbacks of Dynamic Movement Primitives

Dynamic Movement Primitives (DMPs) is a framework for learning a point-to-point trajectory from a demonstration. Despite being widely used, DMPs still present some shortcomings that may limit their usage in real robotic applications. Firstly, at the state of the art, mainly Gaussian basis functions have been used to perform function approximation. Secondly, the adaptation of the trajectory generated by the DMP heavily depends on the choice of hyperparameters and the new desired goal position. Lastly, DMPs are a framework for ‘one-shot learning’, meaning that they are constrained to learn from a unique demonstration. In this work, we present and motivate a new set of basis functions to be used in the learning process, showing their ability to accurately approximate functions while having both analytical and numerical advantages w.r.t. Gaussian basis functions. Then, we show how to use the invariance of DMPs w.r.t. affine transformations to make the generalization of the trajectory robust against both the choice of hyperparameters and new goal position, performing both synthetic tests and experiments with real robots to show this increased robustness. Finally, we propose an algorithm to extract a common behavior from multiple observations, validating it both on a synthetic dataset and on a dataset obtained by performing a task on a real robot.

Nonsingular fast terminal second-order sliding mode for robotic manipulators based on feedback linearization

In this manuscript, a new robust nonsingular fast terminal second-order sliding mode based on input/output feedback linearization is developed for high precision joint position tracking of robotic manipulators subject to perturbations and unknown dynamics. The proposed approach reduces chattering, eliminates the problem of singularity in classical terminal sliding mode and allows fast convergence during the so-called sliding phase, which is suitable for real robot applications. Moreover, since a simple linear equivalent system is derived using feedback linearization, the design procedure doesn’t require anymore the computation of the time derivative of the dynamics. Thus, the real-time implementation is simplified, and the control torque inputs that are fed into the system will not be risky for the controlled system. Simulation and experimental results are respectively conducted on a two rigid arm robot and on an industrial 7-DOF robot arm to demonstrate the effectiveness of the developed control scheme, its simplicity and superiority over other nonlinear controller based on second-order sliding mode.

Multi-Channel Interactive Reinforcement Learning for Sequential Tasks.

The ability to learn new tasks by sequencing already known skills is an important requirement for future robots. Reinforcement learning is a powerful tool for this as it allows for a robot to learn and improve on how to combine skills for sequential tasks. However, in real robotic applications, the cost of sample collection and exploration prevent the application of reinforcement learning for a variety of tasks. To overcome these limitations, human input during reinforcement can be beneficial to speed up learning, guide the exploration and prevent the choice of disastrous actions. Nevertheless, there is a lack of experimental evaluations of multi-channel interactive reinforcement learning systems solving robotic tasks with input from inexperienced human users, in particular for cases where human input might be partially wrong. Therefore, in this paper, we present an approach that incorporates multiple human input channels for interactive reinforcement learning in a unified framework and evaluate it on two robotic tasks with 20 inexperienced human subjects. To enable the robot to also handle potentially incorrect human input we incorporate a novel concept for self-confidence, which allows the robot to question human input after an initial learning phase. The second robotic task is specifically designed to investigate if this self-confidence can enable the robot to achieve learning progress even if the human input is partially incorrect. Further, we evaluate how humans react to suggestions of the robot, once the robot notices human input might be wrong. Our experimental evaluations show that our approach can successfully incorporate human input to accelerate the learning process in both robotic tasks even if it is partially wrong. However, not all humans were willing to accept the robot's suggestions or its questioning of their input, particularly if they do not understand the learning process and the reasons behind the robot's suggestions. We believe that the findings from this experimental evaluation can be beneficial for the future design of algorithms and interfaces of interactive reinforcement learning systems used by inexperienced users.

Multi-Channel Interactive Reinforcement Learning for Sequential Tasks.

The ability to learn new tasks by sequencing already known skills is an important requirement for future robots. Reinforcement learning is a powerful tool for this as it allows for a robot to learn and improve on how to combine skills for sequential tasks. However, in real robotic applications, the cost of sample collection and exploration prevent the application of reinforcement learning for a variety of tasks. To overcome these limitations, human input during reinforcement can be beneficial to speed up learning, guide the exploration and prevent the choice of disastrous actions. Nevertheless, there is a lack of experimental evaluations of multi-channel interactive reinforcement learning systems solving robotic tasks with input from inexperienced human users, in particular for cases where human input might be partially wrong. Therefore, in this paper, we present an approach that incorporates multiple human input channels for interactive reinforcement learning in a unified framework and evaluate it on two robotic tasks with 20 inexperienced human subjects. To enable the robot to also handle potentially incorrect human input we incorporate a novel concept for self-confidence, which allows the robot to question human input after an initial learning phase. The second robotic task is specifically designed to investigate if this self-confidence can enable the robot to achieve learning progress even if the human input is partially incorrect. Further, we evaluate how humans react to suggestions of the robot, once the robot notices human input might be wrong. Our experimental evaluations show that our approach can successfully incorporate human input to accelerate the learning process in both robotic tasks even if it is partially wrong. However, not all humans were willing to accept the robot's suggestions or its questioning of their input, particularly if they do not understand the learning process and the reasons behind the robot's suggestions. We believe that the findings from this experimental evaluation can be beneficial for the future design of algorithms and interfaces of interactive reinforcement learning systems used by inexperienced users.

Online Reconfiguration of Distributed Robot Control Systems for Modular Robot Behavior Implementation

The use of autonomous robots in areas that require executing a broad range of different tasks is currently hampered by the high complexity of the software that adapts the robot controller to different situations the robot would face. Current robot software frameworks facilitate implementing controllers for individual tasks with some variability, however, their possibilities for adapting the controllers at runtime are very limited and don’t scale with the requirements of a highly versatile autonomous robot. With the software presented in this paper, the behavior of robots is implemented modularly by composing individual controllers, between which it is possible to switch freely at runtime, since the required transitions are calculated automatically. Thereby the software developer is relieved of the task to manually implement and maintain the transitions between different operational modes of the robot, what largely reduces software complexity for larger amounts of different robot behaviors. The software is realized by a model-based development approach. We will present the metamodels enabling the modeling of the controllers as well as the runtime architecture for the management of the controllers on distributed computation hardware. Furthermore, this paper introduces an algorithm that calculates the transitions between two controllers. A series of technical experiments verifies the choice of the underlying middleware and the performance of online controller reconfiguration. A further experiment demonstrates the applicability of the approach to real robotics applications.

An image vision and automatic calibration system for universal robots

The rapid development of technology is causing the replacement of many traditional manufacture industries by automation. Robotic arms are now commonly used in many sectors. The requirements for robotic arms in different sectors are quite different; but, in general, robotic arms not only save cost and man power, they also improve safety. The aim of this study was an investigation of the integration of image identification with robotic arms. To do this, the Denavit–Hartenberg transformation matrix was used to analyze the mechanical kinematics of the joints of each robotic arm axis. This allowed the spatial relationships between the Cartesian 3 D coordinate system and the joint of each axis to be determined and communication between the robotic arm and image identification to be established. A robotic arm prototype platform with automatic image identification and calibration was constructed using a quick and robust method. Several of the variables that exist in real robotic arm applications have been solved in this study: primarily the accuracy error that occurs when repeated gripping of workpieces is done and a movement and placement track is set up. Deviations occur frequently and if image identification can be applied to offset repeated inaccuracy, the quality of finished products will be more consistent. The Universal Robots 5 six-axis robot and cameras, which use a filter, Sobel computation, and Hough transformation computer vision image processing technology, together comprise a system that can automatically identify the workpiece and carry out loading and unloading accurately.

SoEasy: A Software Framework for Easy Hardware Control Programming for Diverse IoT Platforms.

Many Internet of Things (IoT) applications are emerging and evolving rapidly thanks to widespread open-source hardware platforms. Most of the high-end open-source IoT platforms include built-in peripherals, such as the universal asynchronous receiver and transmitter (UART), pulse width modulation (PWM), general purpose input output (GPIO) ports and timers, and have enough computation power to run embedded operating systems such as Linux. However, each IoT platform has its own way of configuring peripherals, and it is difficult for programmers or users to configure the same peripheral on a different platform. Although diverse open-source IoT platforms are widespread, the difficulty in programming those platforms hinders the growth of IoT applications. Therefore, we propose an easy and convenient way to program and configure the operation of each peripheral using a user-friendly Web-based software framework. Through the implementation of the software framework and the real mobile robot application development along with it, we show the feasibility of the proposed software framework, named SoEasy.

Learning movement primitive libraries through probabilistic segmentation

Movement primitives are a well-established approach for encoding and executing movements. While the primitives themselves have been extensively researched, the concept of movement primitive libraries has not received similar attention. Libraries of movement primitives represent the skill set of an agent. Primitives can be queried and sequenced in order to solve specific tasks. The goal of this work is to segment unlabeled demonstrations into a representative set of primitives. Our proposed method differs from current approaches by taking advantage of the often neglected, mutual dependencies between the segments contained in the demonstrations and the primitives to be encoded. By exploiting this mutual dependency, we show that we can improve both the segmentation and the movement primitive library. Based on probabilistic inference our novel approach segments the demonstrations while learning a probabilistic representation of movement primitives. We demonstrate our method on two real robot applications. First, the robot segments sequences of different letters into a library, explaining the observed trajectories. Second, the robot segments demonstrations of a chair assembly task into a movement primitive library. The library is subsequently used to assemble the chair in an order not present in the demonstrations.

Online 3D LIDAR Monte Carlo localization with GPU acceleration

PurposeThis paper aims to focus on real-world mobile systems, and thus propose relevant contribution to the special issue on “Real-world mobile robot systems”. This work on 3D laser semantic mobile mapping and particle filter localization dedicated for robot patrolling urban sites is elaborated with a focus on parallel computing application for semantic mapping and particle filter localization. The real robotic application of patrolling urban sites is the goal; thus, it has been shown that crucial robotic components have reach high Technology Readiness Level (TRL).Design/methodology/approachThree different robotic platforms equipped with different 3D laser measurement system were compared. Each system provides different data according to the measured distance, density of points and noise; thus, the influence of data into final semantic maps has been compared. The realistic problem is to use these semantic maps for robot localization; thus, the influence of different maps into particle filter localization has been elaborated. A new approach has been proposed for particle filter localization based on 3D semantic information, and thus, the behavior of particle filter in different realistic conditions has been elaborated. The process of using proposed robotic components for patrolling urban site, such as the robot checking geometrical changes of the environment, has been detailed.FindingsThe focus on real-world mobile systems requires different points of view for scientific work. This study is focused on robust and reliable solutions that could be integrated with real applications. Thus, new parallel computing approach for semantic mapping and particle filter localization has been proposed. Based on the literature, semantic 3D particle filter localization has not yet been elaborated; thus, innovative solutions for solving this issue have been proposed. Recently, a semantic mapping framework that was already published was developed. For this reason, this study claimed that the authors’ applied studies during real-world trials with such mapping system are added value relevant for this special issue.Research limitations/implicationsThe main problem is the compromise between computer power and energy consumed by heavy calculations, thus our main focus is to use modern GPGPU, NVIDIA PASCAL parallel processor architecture. Recent advances in GPGPUs shows great potency for mobile robotic applications, thus this study is focused on increasing mapping and localization capabilities by improving the algorithms. Current limitation is related with the number of particles processed by a single processor, and thus achieved performance of 500 particles in real-time is the current limitation. The implication is that multi-GPU architectures for increasing the number of processed particle can be used. Thus, further studies are required.Practical implicationsThe research focus is related to real-world mobile systems; thus, practical aspects of the work are crucial. The main practical application is semantic mapping that could be used for many robotic applications. The authors claim that their particle filter localization is ready to integrate with real robotic platforms using modern 3D laser measurement system. For this reason, the authors claim that their system can improve existing autonomous robotic platforms. The proposed components can be used for detection of geometrical changes in the scene; thus, many practical functionalities can be applied such as: detection of cars, detection of opened/closed gate, etc. […] These functionalities are crucial elements of the safe and security domain.Social implicationsImprovement of safe and security domain is a crucial aspect of modern society. Protecting critical infrastructure plays an important role, thus introducing autonomous mobile platforms capable of supporting human operators of safe and security systems could have a positive impact if viewed from many points of view.Originality/valueThis study elaborates the novel approach of particle filter localization based on 3D data and semantic mapping. This original work could have a great impact on the mobile robotics domain, and thus, this study claims that many algorithmic and implementation issues were solved assuming real-task experiments. The originality of this work is influenced by the use of modern advanced robotic systems being a relevant set of technologies for proper evaluation of the proposed approach. Such a combination of experimental hardware and original algorithms and implementation is definitely an added value.

Adaptive Partial Shortcuts: Path Optimization for Industrial Robotics

The quality of a path generated from an automated motion planning algorithm is of considerable importance, particularly when used in a real world robotic application. In this work a new path optimization algorithm, called the Adaptive Partial Shortcut algorithm, is presented. This algorithm optimizes paths as a post process to motion planning, and is designed specifically for use on industrial manipulators. The algorithm optimizes a robot's degrees of freedom independently allowing it to produce manipulator paths of particularly high quality. This new algorithm utilizes an adaptive method of selecting the degree of freedom to optimize at each iteration, giving it a high level of efficiency. Tests conducted show the effectiveness of the algorithm; over a range of different test paths, the adaptive algorithm was able to generate solutions with a 60 % reduction in collision checks compared to the original partial shortcut approach.