Data-Driven Human-Robot Interaction Without Velocity Measurement Using Off-Policy Reinforcement Learning

In this paper, we present a novel data-driven design method for the human-robot interaction (HRI) system, where a given task is achieved by cooperation between the human and the robot. The presented HRI controller design is a two-level control design approach consisting of a task-oriented performance optimization design and a plant-oriented impedance controller design. The task-oriented design minimizes the human effort and guarantees the perfect task tracking in the outer-loop, while the plant-oriented achieves the desired impedance from the human to the robot manipulator end-effector in the inner-loop. Data-driven reinforcement learning techniques are used for performance optimization in the outer-loop to assign the optimal impedance parameters. In the inner-loop, a velocity-free filter is designed to avoid the requirement of end-effector velocity measurement. On this basis, an adaptive controller is designed to achieve the desired impedance of the robot manipulator in the task space. The simulation and experiment of a robot manipulator are conducted to verify the efficacy of the presented HRI design framework.      

一种基于三维建图和虚拟现实的人机交互系统

以提高人机共融水平为目的,以救援机器人为背景,提出并实现基于三维建图和虚拟现实(VR)技术的人机交互系统.在该系统中,救援机器人基于多线激光雷达和惯性测量单元(IMU)实时构建环境的三维点云地图,并将建图结果增量式地表示为3D-NDT地图,实时传输至操作台的虚拟现实系统中可视化;同时,操作人员利用虚拟现实系统的交互设备生成机器人的控制指令,控制机器人运动,构成一个完整的人在回路的人机交互系统.该系统在将机器人环境实时在虚拟现实中可视化的基础上,可以给操作人员以极强的沉浸感,有利于操作人员更直接地理解机器人所处环境.此外,该系统作为一种新的人机交互方式,为提高人与机器人的自然交互水平提供了新思路,对促进人机交互技术的发展具有重要意义.     

Assisting manual welding with robot

This paper presents a first attempt to assist manual welding with a physically interactive robot. An interactive control scheme is developed to suppress the vibrations of torch during the welding of novice welders. The torch is attached to the end-effector of a haptic-robot. Human and robot act together on the welding torch: the human controls the direction and speed; the robot suppresses the sudden and abrupt motions. The control scheme is developed by experimenting with an air-paint-brush. The painting process emulates the actual welding. Such an emulating environment is useful to surmount the difficulties of experimentation with actual welding. The impedance parameters of the control scheme are investigated. A damping value is determined for an effective vibration suppression and minimum human effort. A variable impedance control scheme is applied to ease the manipulation of the torch while not welding. The results of real welding of novice welders with and without robot assistance are presented. There is a considerable improvement in the performance of the welders when they are assisted with the robot.     

Teaching the User By Learning From the User:Personalizing Movement Control in Physical Human-robot Interaction

This paper proposes a novel approach for physical human-robot interactions (pHRI), where a robot provides guidance forces to a user based on the user performance. This framework tunes the forces in regards to behavior of each user in coping with different tasks, where lower performance results in higher intervention from the robot. This personalized physical human-robot interaction (p2HRI) method incorporates adaptive modeling of the interaction between the human and the robot as well as learning from demonstration (LfD) techniques to adapt to the users' performance. This approach is based on model predictive control where the system optimizes the rendered forces by predicting the performance of the user. Moreover, continuous learning of the user behavior is added so that the models and personalized considerations are updated based on the change of user performance over time. Applying this framework to a field such as haptic guidance for skill improvement, allows a more personalized learning experience where the interaction between the robot as the intelligent tutor and the student as the user, is better adjusted based on the skill level of the individual and their gradual improvement. The results suggest that the precision of the model of the interaction is improved using this proposed method, and the addition of the considered personalized factors to a more adaptive strategy for rendering of guidance forces.      

Intuitive robot teleoperation for civil engineering operations with virtual reality and deep learning scene reconstruction

Robotic teleoperation, i.e., manipulating remote robotic systems at a distance, has gained its popularity in various industrial applications, including construction operations. The key to a successful teleoperation robot system is the delicate design of the human-robot interface that helps strengthen the human operator’s situational awareness. Traditional human-robot interface for robotic teleoperation is usually based on imagery data (e.g., video streaming), causing the limited field of view (FOV) and increased cognitive burden for processing additional spatial information. As a result, 3D scene reconstruction methods based on point cloud models captured by scanning technologies (e.g., depth camera and LiDAR) have been explored to provide immersive and intuitive feedback to the human operator. Despite the added benefits of applying reconstructed 3D scenes in telerobotic systems, challenges still present. Most 3D reconstruction methods utilize raw point cloud data due to the difficulty of real-time model rendering. The significant size of point cloud data makes the processing and transfer between robots and human operators difficult and slow. In addition, most reconstructed point cloud models do not contain physical properties such as weight and colliders. A more enriched control mechanism based on physics engine simulations is impossible. This paper presents an intelligent robot teleoperation interface that collects, processes, transfers, and reconstructs the immersive scene model of the workspace in Virtual Reality (VR) and enables intuitive robot controls accordingly. The proposed system, Telerobotic Operation based on Auto-reconstructed Remote Scene (TOARS), utilizes a deep learning algorithm to automatically detect objects in the captured scene, along with their physical properties, based on the point cloud data. The processed information is then transferred to the game engine where rendered virtual objects replace the original point cloud models in the VR environment. TOARS is expected to significantly improve the efficiency of 3D scene reconstruction and situational awareness of human operators in robotic teleoperation.     

Optimization of criterion for objective evaluation of HRI performance that approximates subjective evaluation: a case study in robot competition

The standard approach to evaluate the performance of human-robot interaction (HRI) is subjective evaluation, for example, using questionnaires. Because such subjective evaluation is time-consuming, an alternative evaluation method based on only objective factors (i.e. human reaction behavior) is required for autonomous learning by robots and for scoring in robot competitions, etc. To this end, we aimed to investigate the extent to which subjective evaluation results can be approximated using objective factors. We observed and stored HRI history data through a robot-competition task in which the robot was required to generate comprehensible and unambiguous natural language expressions and gestures to guide inexpert users in virtual everyday environments. In addition, to acquire subjective evaluation results, we asked third-parties to evaluate the HRI performance by reviewing the stored interaction histories. From the results of a case study of robot-competition, we demonstrate the necessity for an objective method for the evaluation of HRI performance and a determination process for an effective evaluation criterion. The results of a multiple linear regression analysis to estimate the subjective evaluation results reveal that the subjective evaluation of HRI can indeed be reasonably approximated base on objective factors.     

Robust human position estimation in cooperative robotic cells

Robots are increasingly present in our lives, sharing the workspace and tasks with human co-workers. However, existing interfaces for human-robot interaction / cooperation (HRI/C) have limited levels of intuitiveness to use and safety is a major concern when humans and robots share the same workspace. Many times, this is due to the lack of a reliable estimation of the human pose in space which is the primary input to calculate the human-robot minimum distance (required for safety and collision avoidance) and HRI/C featuring machine learning algorithms classifying human behaviours / gestures. Each sensor type has its own characteristics resulting in problems such as occlusions (vision) and drift (inertial) when used in an isolated fashion. In this paper, it is proposed a combined system that merges the human tracking provided by a 3D vision sensor with the pose estimation provided by a set of inertial measurement units (IMUs) placed in human body limbs. The IMUs compensate the gaps in occluded areas to have tracking continuity. To mitigate the lingering effects of the IMU offset we propose a continuous online calculation of the offset value. Experimental tests were designed to simulate human motion in a human-robot collaborative environment where the robot moves away to avoid unexpected collisions with de human. Results indicate that our approach is able to capture the human’s position, for example the forearm, with a precision in the millimetre range and robustness to occlusions.     

Evaluating the Robot Personality and Verbal Behavior of Domestic Robots Using Video-Based Studies

Robots are increasingly being used in domestic environments and should be able to interact with inexperienced users. Human–human interaction and human–computer interaction research findings are relevant, but often limited because robots are different from both humans and computers. Therefore, new human–robot interaction (HRI) research methods can inform the design of robots suitable for inexperienced users. A video-based HRI (VHRI) methodology was here used to carry out a multi-national HRI user study for the prototype domestic robot BIRON (BIelefeld RObot companioN). Previously, the VHRI methodology was used in constrained HRI situations, while in this study HRIs involved a series of events as part of a 'hometour' scenario. Thus, the present work is the first study of this methodology in extended HRI contexts with a multi-national approach. Participants watched videos of the robot interacting with a human actor and rated two robot behaviors (Extrovert and Introvert). Participants' perceptions and ratings of the robot's behaviors differed with regard to both verbal interactions and person following by the robot. The study also confirms that the VHRI methodology provides a valuable means to obtain early user feedback, even before fully working prototypes are available. This can usefully guide the future design work on robots, and associated verbal and non-verbal behaviors.     

Caregiving role in human–robot interaction: A study of the mediating effects of perceived benefit and social presence

This study investigates whether assigning a caregiving role to a robot or to its human interactant has psychological effects on the quality of human–robot interaction (HRI). College students interacted with a social robot in a between-subjects experiment (N = 60) with two manipulated conditions: one where the robot played the role of an ophthalmologist (with participants serving as patients) and one where participants played the role of the ophthalmologist (with the robot serving as the patient). Results suggest that being a recipient of caregiving acts leads users to form more positive perceptions of the robot than being an ostensible caregiver to the robot. Results also indicate that perceived benefit of being in a relationship with the robot mediates the effects of the caregiving role on relationship satisfaction with—and trust towards—the robot while feelings of social presence mediate the effects on humanlike-ness and intelligence of the robot. These findings demonstrate the applicability of the Computers Are Social Actors (CASA) paradigm to the context of HRI.     

Intelligent welding system technologies: State-of-the-art review and perspectives

Welding systems are being transformed by the advent of modern information technologies such as the internet of things, big data, artificial intelligence, cloud computing, and intelligent manufacturing. Intelligent welding systems (IWS), making use of these technologies, are drawing attention from academic and industrial communities. Intelligent welding is the use of computers to mimic, strengthen, and/or replace human operators in sensing, learning, decision-making, monitoring and control, etc. This is accomplished by integrating the advantages of humans and physical systems into intelligent cyber systems. While intelligent welding has found pilot applications in industry, a systematic analysis of its components, applications, and future directions will help provide a unified definition of intelligent welding systems. This paper examines fundamental components and techniques necessary to make welding systems intelligent, including sensing and signal processing, feature extraction and selection, modeling, decision-making, and learning. Emerging technologies and their application potential to IWS will also be surveyed, including Industry 4.0, cyber-physical system (CPS), digital twins, etc. Typical applications in IWS will be surveyed, including weld design, task sequencing, robot path planning, robot programming, process monitoring and diagnosis, prediction, process control, quality inspection and assessment, human-robot collaboration, and virtual welding. Finally, conclusions and suggestions for future development will be proposed. This review is intended to provide a reference of the state-of-the-art for those seeking to introduce intelligent welding capabilities as they modernize their traditional welding stations, systems, and factories.     

A survey of Tactile Human–Robot Interactions

Robots come into physical contact with humans in both experimental and operational settings. Many potential factors motivate the detection of human contact, ranging from safe robot operation around humans, to robot behaviors that depend on human guidance. This article presents a review of current research within the field of Tactile Human–Robot Interactions (Tactile HRI), where physical contact from a human is detected by a robot during the execution or development of robot behaviors. Approaches are presented from two viewpoints: the types of physical interactions that occur between the human and robot, and the types of sensors used to detect these interactions. We contribute a structure for the categorization of Tactile HRI research within each viewpoint. Tactile sensing techniques are grouped into three categories, according to what covers the sensors: (i) a hard shell, (ii) a flexible substrate or (iii) no covering. Three categories of physical HRI likewise are identified, consisting of contact that (i) interferes with robot behavior execution, (ii) contributes to behavior execution and (iii) contributes to behavior development. We populate each category with the current literature, and furthermore identify the state-of-the-art within categories and promising areas for future research.     

A visual path-following learning approach for industrial robots using DRL

Manufacturing companies are in constant need for improved agility. An adequate combination of speed, responsiveness, and business agility to cope with fluctuating raw material costs is essential for today’s increasingly demanding markets. Agility in robots is key in operations requiring on-demand control of a robot’s tool position and orientation, reducing or eliminating extra programming efforts. Vision-based perception using full-state or partial-state observations and learning techniques are useful to create truly adaptive industrial robots. We propose using a Deep Reinforcement Learning (DRL) approach to solve path-following tasks using a simplified virtual environment with domain randomisation to provide the agent with enough exploration and observation variability during the training to generate useful policies to be transferred to an industrial robot. We validated our approach using a KUKA KR16HW robot equipped with a Fronius GMAW welding machine. The path was manually drawn on two workpieces so the robot was able to perceive, learn and follow it during welding experiments. It was also found that small processing times due to motion prediction (3.5 ms) did not slow down the process, which resulted in smooth robot operations. The novel approach can be implemented onto different industrial robots to carry out different tasks requiring material deposition.     

基于动态贝叶斯网络的人机交互过程参与度评估方法

在人机交互过程中,基于动态贝叶斯网络与领域专家知识,提出一种参与度评估方法,重点解决训练样本不足的问题．首先,对人机交互过程中的社会信号与动态贝叶斯网络的推理过程进行了描述,并建立了参与度评估模型的拓扑结构;然后,针对基于数据驱动模型参数化的局限性,通过设计语言变量集来收集专家建议,并对这些建议进行模糊化以及解模糊处理,完成了评估模型的构建;最后,通过机器人平台NAO并设计交互场景对评估模型进行了验证．实验表明,参与度评估结果与交互对象的实际行为相一致,能够使机器人在人机交互过程中正确识别出交互对象的参与意图．与已有的参与度评估方法相比,所提出的方法不受训练样本的限制,具有较高的实用价值和较强的泛化能力．     

Human-Oriented Interaction With an Anthropomorphic Robot

A very important aspect in developing robots capable of human-robot interaction (HRI) is the research in natural, human-like communication, and subsequently, the development of a research platform with multiple HRI capabilities for evaluation. Besides a flexible dialog system and speech understanding, an anthropomorphic appearance has the potential to support intuitive usage and understanding of a robot, e.g., human-like facial expressions and deictic gestures can as well be produced and also understood by the robot. As a consequence of our effort in creating an anthropomorphic appearance and to come close to a human- human interaction model for a robot, we decided to use human-like sensors, i.e., two cameras and two microphones only, in analogy to human perceptual capabilities too. Despite the challenges resulting from these limits with respect to perception, a robust attention system for tracking and interacting with multiple persons simultaneously in real time is presented. The tracking approach is sufficiently generic to work on robots with varying hardware, as long as stereo audio data and images of a video camera are available. To easily implement different interaction capabilities like deictic gestures, natural adaptive dialogs, and emotion awareness on the robot, we apply a modular integration approach utilizing XML-based data exchange. The paper focuses on our efforts to bring together different interaction concepts and perception capabilities integrated on a humanoid robot to achieve comprehending human-oriented interaction.     

How to include User eXperience in the design of Human-Robot Interaction

In recent years Human-Robot Collaboration (HRC) has become a strategic research field, considering the emergent need for common collaborative execution of manufacturing tasks, shared between humans and robots within the modern factories. However, the majority of the research focuses on the technological aspects and enabling technologies, mainly directing to the robotic side, and usually neglecting the human factors. This work deals with including the needs of the humans interacting with robots in the design in human-robot interaction (HRI). In particular, the paper proposes a user experience (UX)-oriented structured method to investigate the human-robot dialogue to map the interaction with robots during the execution of shared tasks, and to finally elicit the requirements for the design of valuable HRI. The research adopted the proposed method to an industrial case focused on assembly operations supported by collaborative robots and AGVs (Automated Guided Vehicles). A multidisciplinary team was created to map the HRI for the specific case with the final aim to define the requirements for the design of the system interfaces. The novelty of the proposed approach is the inclusion of typically interaction design tools focusing in the analysis of the UX into the design of the system components, without merely focusing on the technological issues. Experimental results highlighted the validity of the proposed method to identify the interaction needs and to drive the interface design.     

人机交互中的个性化情感模型

人与机器人的交互过程中,情感因素的引入能够使人机交流更加自然和谐.因此,完整的人工情感模型的建立是首要解决的问题.基于情感能量理论基础,首先,提出了心境自发转移和刺激转移模型.其次,结合情绪自发转移的马尔可夫链模型和刺激转移的HMM模型,将心境和情绪的自发和刺激转移过程统一在一个框架下.最后,将完整的人工情感模型软件化并应用于儿童玩伴机器人上,在接受非结构化环境与用户的信息输入后,个性化的情感软件模块产生输出,实现针对儿童用户的玩伴机器人个性化交互,通过应用验证了该模型的有效性.     

康复机器人的同步主动交互控制与实现

提出了一种适用于康复机器人的人机交互控制方法. 结合一款具有平面并联结构的上肢康复机器人, 实现了与用户(患者)运动意图同步的、柔顺的主动康复训练. 在训练中, 利用自适应频率振荡器, 从表面肌电信号(Surface electromyography, sEMG)中获取运动模式信息, 然后结合运动模式和期望的正常运动轨迹, 生成与主动运动意图同步的参考训练轨迹. 本文通过仿真和实际实验对所提出的方法进行了验证, 振荡器可以在2~5s内快速实现与用户主动运动意图的同步, 然后利用阻抗控制器给予柔顺的辅助. 通过调节阻抗参数, 可以为患者的运动训练提供不同程度的辅助.     

直接脑控机器人接口技术

直接脑控机器人接口(Brain-controlled robot interface, BCRI)是一种新型的人-机器人接口技术, 是脑-机器接口/脑-计算机接口(Brain-machine interface, BMI/Brain-computer interface, BCI)在机器人控制领域的重要应用和研究方向. 研究者相继在Nature、Science和其他重要国际期刊上报道了相关的实验研究和开发, 目前已成为国际前沿研究热点. 本文主要围绕BCRI中的控制策略、BMI/BCI模块与机器人多层控制模块的适应和融合、BCRI中的脑信号自适应分类算法以及人、BMI/BCI模块和机器人控制系统的三边自适应展开论述, 分析了目前的研究情况、存在的局限和面临的若干重要问题, 指出进一步的研究思路和方向.