Interactive Collaborative Robotics

This paper describes a macro/micro robot manipulation system consisting of an electromagnet couple and an industrial manipulator. The system has capability of motion 6 Degrees of Freedom (6 DOF) providing the same amount of DOFs to the manipulation of the microrobot. A custom-design mechanism which is attached onto the tip of the industrial manipulator provides the required magnetic field profiles for force and torque generation using coil couple and their ability sliding motion in linear directions. This combination provides reprogrammable working space for the microrobot manipulation. Robot Operating System (ROS) based programming integrates all the subsystem software. Visual feedback assures the real-time microrobot position and orientation data. Close-loop motion control of the microrobot was tested using custom designed constrained paths take part in a plane. Experiment were presented for specific motions of the microrobot to show the microrobot motion abilities. The results are promising which may orient to the applications like micro assembly and micromanipulation.

In this article, a microservice architecture of cyber-physical space was considered and in particular localization service was implemented. We propose an architecture of proactive localization service, which allows predicting the activity of the tracked object in cyber-physical space. To solve the position prediction problem, we tested machine learning methods and contrasted the results of the trained models. Three machine learning algorithms were tested (artificial neural network, random forest, and decision tree) with two different datasets. The reason of high/low prediction accuracy were identified. As a result of testing, the best result has neural network. In this case mean absolute error is 8.2 and 11.7 m respectively for dataset №1 and №2, while random forest has 13 and 14 m error. The architecture of the service was developed using containerization technologies and special tools for their deployment and management.

In this paper, we implement a quadcopter assembly with control and navigation module. The project also includes the design of the control panel for the operator which consists of a set of the microcontroller and the glove equipped with sensors and buttons. The panel has a touch screen which displays current parameters such as vehicle status, including information about orientation and geographical coordinates. The concept of quadcopter control is based on the movement of the operator hand. In addition, we have included the object detection for detecting the objects from the quadcopter view of point. To detect an object, we need to have some idea of where the object may be and how the image is divided into segments. It creates a kind of chicken and egg problem, where we must recognize the shape (and class) of the object knowing its location and recognize the location of the object knowing its shape. Some visual characteristics such as clothing and the human face, they can be part of the same subject, but it is difficult to recognize this without recognizing the object first.

Technological advances in recent history allow usage of robot manipulators in every aspects of manufacturing. Integration of robot manipulators into the machining operations not only increases the quality of the end products but also decreases the time required for their machining operations. In terms of delicacy in these operations, robotic grinding can be given as one of the most important applications of the field. Thus this study focuses on the trajectory planning problem of a PRR planar serial redundant manipulator that is proposed to be utilized for surface finishing. Throughout the study kinematic representation of PRR manipulator was given in detail and its kinematic analysis was carried out along with the direct and inverse tasks. After the kinematic equations were obtained, a desired end effector trajectory was given in order to simulate real surface finishing application on a plane. Required joint position functions were taken as polynomial functions. In the light of this, coefficients of the polynomials were solved to approximate the desired trajectory. At the end of the study comparisons between the desired and generated trajectories were plotted graphically.

The paper describes the modeling, development and testing results of a multi-agent control system for a group of ground-based robots. The developed model of collective behavior was verified in the Matlab system, then transferred to the ROS framework and tested in the virtual Gazebo environment. Such a development path allowed us to obtain a workable code at the computer simulation stage and then use it to ensure the functioning of a group of autonomous ground-based robots in natural conditions. The paper describes all the main stages of the project: development of a mathematical model of the multi-agent control system, implementation in ROS system using Python language and verification by control of real robot group (group of eduMIP robots). Comparison of the results of computer simulation and the physical experiments made it possible to conclude that they are sufficiently consistent, which in turn confirms the correctness of the constructed group control model. Thus, within the framework of the SMTU project, the task of developing models and group control algorithms for a simple multi-agent system has been solved. The purpose of the next stage of work is the adaptation and development of proven models and algorithms for the control system of a group of underwater robots.

Robots are used to solve routine, monotonous, difficult and dangerous tasks; therefore, agriculture is one of the largest spheres for using robotic systems. One of the main problems faced by developers of autonomous robotic systems is the navigation of the robot in space. This paper presents a solution to the problem of navigation, based on the maintenance of a constant radio signal between the UAV or mobile platform and control system. Radio communication is maintained by building a mesh network based on LoRa data transmission technology modules throughout the entire path of the robot. Navigation system is a mesh network based on the radio beacon. Three methods for determining the coordinates of additional module location were considered. These methods are intended for organizing mush network from modules that are not connected to each other. Analysis of considered methods was presented. Methods are not designed for arbitrary movement of robotic systems. Each of the presented methods has its advantages and disadvantages, the first two methods have the main advantage being the smallest number of modules used to connect all radio modules to the network, but with a decrease in the number of modules there is a problem of reducing system reliability. The third method solves this problem by clustering and can withstand the failure of a large number of additional modules, and the system itself becomes more like a mesh network.

Collaboration humans and robots in close proximity in a shared workspace is a stimulating feature of Industry 4.0. Collaborative robotics solution provides for the production processes the benefits due to the characteristic properties of robots. These properties include high level of accuracy, speed and repeatability. In combination with the flexibility and cognitive skills of humans the process of human-robot interaction achieves an efficient human-robot collaboration. Today’s topical issue research include problems of developing safer robots in human-machine systems. In considering the issue of safety human-robot interaction should be taken into account regulations and standards. A key challenge of human-robot collaboration is deciding how to distribute functions between human and robot to provide efficient interaction. This paper is a review on history of the robotic automation and modern research including Russian experience.

This study focuses on gesture recognition in mobile interaction settings, i.e. when the interacting partners are walking. This kind of interaction requires a particular coordination, e.g. by staying in the field of view of the partner, avoiding obstacles without disrupting group composition and sustaining joint attention during motion. In literature, various studies have proven that gestures are in close relation in achieving such goals.Thus, a mobile robot moving in a group with human pedestrians, has to identify such gestures to sustain group coordination. However, decoupling of the inherent -walking- oscillations and gestures, is a big challenge for the robot. To that end, we employ video data recorded in uncontrolled settings and detect arm gestures performed by human-human pedestrian pairs by adopting a signal processing approach. Namely, we exploit the fact that there is an inherent oscillatory motion at the upper limbs arising from the gait, independent of the view angle or distance of the user to the camera. We identify arm gestures as disturbances on these oscillations. In doing that, we use a simple pitch detection method from speech processing and assume data involving a low frequency periodicity to be free of gestures. In testing, we employ a video data set recorded in uncontrolled settings and show that we achieve a detection rate of 0.80.

Describing the design of the specialized handling robot for installation of cross ties in the subway. For the reliable reproduction of the required trajectories at high movement speed of manipulator links the task was set to develop its dynamic model, and the particular microprocessor system of the dynamic control of the manipulator. The equations of a three-coordinate manipulator dynamics in the angular system of coordinates are obtained in a differential and vector form of record providing the solution of the return problem of dynamics. The structure of a microprocessor system for the dynamic control of the three-unit manipulator with angular system of coordinates is developed. It is expedient to use the three-coordinate manipulators with angular system of coordinates in a limited space of a subway. When the operating impact effects are formed of the on drives of the specialized robot for the installation of cross ties the most versatile and rational method is that of dynamic control, developing the dynamic model of the manipulator on the basis of its algorithm and Lagrange’s method.

The paper deals with the problem of the distribution of roles in coalition robots with limited communications. A formal formulation of the task of role distribution in the coalition of mobile robots is given. An analysis of existing approaches to the distribution of roles in groups of robots is given, such as solving the assignment problem by the Kuhn-Munkres algorithm, using the game theory apparatus, applying the methods of probability theory, and the method of propagating the control wave using a local conversion mechanism. An iterative approach to the distribution of roles in a group of robots, based on the strategy of decentralized control and the principles of swarm interaction, is proposed. A method for the distribution of roles in coalitions of mobile robots and an algorithm that implements this method for a separate coalition robot in the distribution of roles based on the proposed approach are described. The results of the study of the proposed approach, carried out with the help of computer simulation in coalitions of 100 robots in the distribution of three roles, are presented. The estimation of the error of the distribution of roles using the proposed algorithmically implemented method has been made and compared with the known approaches. The areas of possible practical application of the developed approach are shown.

This paper proposes an assistive control scheme for rehabilitation robots. The design objective is to develop a control strategy to follow a predefined path by maximizing the active participation of impaired subjects. For this purpose, we develop an impedance controller for velocity tracking and gain the assist-as-needed property by means of modulating the target impedance. The given method neither requires a compliant robot nor needs the dynamic model of the system. The approach also provides motion timing freedom due to working in the velocity domain. The performance of the controller is evaluated through experimental tests.

In this paper a new application of the distributed ledger technology is proposed. The swarm robotics is a rapidly developing area due to the numerous advantages of the swarm. Yet there can be situations when some additional functional tasks should be relocated from one robot to another, for example, if there is a need to offload one of the robots. For such resource allocation tasks the robot reliability must be taken into account. This causes the importance of the robot workload logging in the swarm, because it is not sufficient to take into account the current workload to estimate the reliability level. In the paper the technique of the distributed-ledger-based workload logging is presented as well as information propagation methods are considered and the peculiarities of the robot swarm has been taken into account.

The paper proposes an approach to driver support in vehicle cabin oriented to dangerous states determination and recommendation generation. To determine dangerous states, we propose to analyze images from smartphone front-facing camera as well as analyze information from accessible sensors. We identified two main dangerous states that are most important to identify to prevent the possible accidents in the roads: drowsiness and distraction. In scope of the approach determined dangerous states are used to generate recommendations for the driver to notify him/her about drowsiness and distraction. Since the attention of the driver should be in the road during the driver, we propose the human-computer interaction interface that is based on speech recognition to interact the driver. Using the interface, the driver interacts with the driver support system to increase the quality of the dangerous state determination and recommendation generation.

Humanoid robots are created to facilitate many facets of daily life, both in scenarios when humans and robots collaborate and when robot completely replaces human. One of such more important cases is the household assistance for older people. When a robot operates in home environments the needs to interact with various household objects, of different shape and size. A humanoid end-effector is typically modeled to have from two to five configuration of fingers designed specifically for grasping. By making fingers flexible and using dexterous arm one could operate objects in many different configurations. If one chooses to provide a finger control by actuating each of the finger’s phalanxes by using separate motor, humanoid hand becomes costly and overall size of the hand will significantly increase to accommodate necessary hardware and wiring. To address this issue, many engineers prefer to employ mimic joints to reduce a cost and size, while keeping acceptable levels of finger’s dexterity. This paper presents a study on household objects pick and place task being implemented for AR-601M humanoid robot that is using mimic joints in his fingers. Experiments were conducted in a Gazebo simulation with 5 model objects, which were created to be representations of real typical household items.

This paper presents a new algorithm for hierarchical case-based behavior planning in a coalition of agents – HierMAP. The considered algorithm, in contrast to the well-known planners HEART, PANDA, and others, is intended primarily for use in multi-agent tasks. For this, the possibility of dynamically distributing agent roles with different functionalities was realized. The use of a psychologically plausible approach to the representation of the knowledge by agents using a semiotic network allows applying HierMAP in groups in which people participate as one of the actors. Thus, the algorithm allows us to represent solutions of collaborative problems, forming human-interpretable results at each planning step. Another advantage of the proposed method is the ability to save and reuse experience of planning – expansion in the field of case-based planning. Such extension makes it possible to consider information about the success/ failure of interaction with other members of the coalition. Presenting precedents as a special part of the agent’s memory (semantic network on meanings) allows to significantly reduce the planning time for a similar class of tasks. The paper deals with smart relocation tasks in the environment. A comparison is made with the main hierarchical planners widely used at present.

The current paper deals with the swarm robots reliability. Cloud robotics, fog robotics and the Internet of the Robotic Things are the fast growing scientific fields nowadays, yet the terms “cloud” and “fog” relate to the network facilities and devices rather than the swarm. In this paper the new approach is proposed, to place a fog-like structure into the swarm and so to affect the reliability of those robots, which need the reliability correction. To show the potential of the approach proposed, simple models are developed, as well as some simulations have been made. Based on the simulation results, “greedy” and “egoistic” strategies are proposed to affect the robots reliability.

Biped humanoid robot dynamics is approximated by dynamics of 3D linear inverted pendulum, which can be derived from dynamics of ordinary 3D inverted pendulum. Based on this approximation of biped robot dynamics we can generate walking patterns, which are specified by step parameters, that specify desired zero moment point (ZMP) trajectory. To track desired center of mass (CoM) trajectory, modified foot positions are calculated by minimizing an error function in closed form. Different forces acting on biped robot are taken into account during calculation of robot dynamics for plain surface as well for uneven terrain. Some walking patterns and walking primitives are expressed algorithmically for different 3D cases in terms of ground-fixed coordinate frame. Some dynamic constraints applicable in this setting are presented.

Structural synthesis of parallel manipulators is the first step of design process when new parallel manipulators are needed to be designed. Structural synthesis can be obtained by using structural synthesis formulations. Furthermore, structural synthesis of parallel manipulators can be quickly solved by computer programs. These type of programs are referred as “Automatic Sketching of Mechanism and Manipulators”. In this paper, a new Graphical User Interface (GUI) is presented to obtain structural synthesis of parallel manipulators. Two distinct mobility formulations are used to calculate joints on the limbs of parallel manipulators. Apart from other automatic sketching of mechanisms and manipulators, variable general constrained loops were used for the first time in this study. Manipulators with single platform are used in the example designs for the first program runs. After the total number of joints are calculated by the program and they can be placed manually or automatically. This makes proposed program to be user friendly and easy to use. Algorithm can be extended for the design of multi-platform parallel manipulators.

This paper analyzes modern solutions in wireless charging system development. The main energy transfer standards, prototypes and experimental assemblies and their characteristics are considered. We present circuits and engineering solutions of the transmitting and receiving parts of the developed wireless energy transfer system. The results of testing the prototype are discussed, as well as obtained dependencies of the efficiency from the transmitted power and characteristic curve for various energy transfer distances. The presented prototype ensures magnetic insulation and features energy transfer up to 76.47% and can be used in swarm robotic systems for balancing energy resources of swarm and wireless power transfer between agents of swarm and in other different robotic complexes.

Mobile robots, operated in hard environments, are investigated. It is shown, that in order to ensure the required reliability parameters of the onboard equipment, it should have the fault-tolerance properties. The task of designing fault-tolerant assembles can be properly solved only if there is an adequate model of reliability parameters estimation. A two-stage method of reliability parameters estimation is proposed, in which at the first stage the “lifetime” of one unit in a complex failure-restoration cycle is determined, and at the second stage, the “lifetime” of the fault-tolerant assemble as a whole, is determined. To solve the problem of the second stage “lifetime” density sampling procedure is envisaged. The method of evaluation of the fault-tolerant assembles reliability parameters, with use of discrete model, is worked out.

Autonomous exploration and coverage in 3D environments recently has became a rapidly developing research field. Emerging 3D reconstruction methods, designed specifically for exploration and coverage, allows capturing an environment in a greater details. However, not much work addresses certain difficulties inherent to dense clutter environments. We observed those difficulties and made an attempt that seeks to expand the applicability of such methods to more demanding scenarios. Automating the process of testing and evaluation by designing a dense clutter environment generation algorithm (DCEGen) allows us to measure comparative performance of available algorithms. We focus on path-planning algorithms used in an unmanned ground vehicles. The algorithm was implemented and verified using Gazebo simulator.

The present paper deals with the topical issues of the control system development for electric drives of the powered (active) lower-limb exoskeleton device designed for industrial applications. The article explores ways to assist in walking and running modes. Use of special types of the measuring complex allowing to evaluate the operator’s activity and to form the driving signals for electric drives. Particular attention is paid to the simulation of non-linear properties of the electric drive and the measuring system. To assess the results of numerical simulations, a comprehensive criterion for evaluating the quality indicators of the control system, was developed and optimization possibilities for the control system parameters were studied. A comparative analysis of different modes of human moving in assistive industrial device has been presented, and a few proposals for their practical application have been outlined in the article.

The paper provides a brief overview of the state-of-the-art of development of tethered high-altitude unmanned telecommunication platforms. It is described an architecture, scientific and technical issues and the designing principles of such platforms which can be used to expand the capabilities of ground-based electronic warfare. The experience of Institute of Control Sciences (ICS RAS) on the development and implementing the tethered platform is also presented.

This study developed the basic principles and algorithms for presenting the semantics of simple extended sentences describing static scenes based on multi-agent cognitive architectures. It is shown that software agents can represent word patterns and concepts corresponding to these words. And such multi-agent algorithms can form the basis for modeling the semantics of relations between concepts, representing various semantic classes in the composition of the sentence.

In this work, homogeneous gripping mechanical devices for connecting small-sized modular autonomous robots are described. The review of solutions of gripping and holding industrial mechanisms, and robotic switching mechanisms, as well as docking devices of space vehicles is given. The design of the connecting gripping mechanism of Mobile Autonomous Reconfigurable System (MARS) for coaxial conjugation of robots is proposed and it works in conjunction and passive modes to form modular structures. In the passive state, the working body of the mechanism forms a geometrical figure, which is suitable for connecting with identical device. In this case working body opens gripping mechanism and capture similar mechanism in the passive state. The proposed mechanism is based on the iris diaphragm, which prevents the uncontrolled displacement of the working body and stops the mechanism. It excludes the accidental rotation around the axis of the autonomous unit during operation in the formed structure. Infrared sensors (IR) were used for the concentric alignment of the axes of the connection devices. Sensors of this type estimate the distance and deflection angle of the opposed mechanism.

Increased availability of unmanned aerial vehicles (UAVs), also known as drones, have advanced research and industrial interests, establishing this platform as a promising part of future emerging technologies. In this paper, we present an overview of optical wireless communications (OWC) related to UAV, which benefit from both optical data rates and mobility of UAV. Since drones are usually employed in outdoor scenarios, the UAV based free-space optical (FSO) communication system is considered, as the FSO represents the optical wireless signal transmission from infra-red band spectrum in outdoor environments. A brief recapitulation of main studies in the field of UAV-aided cooperation within FSO systems is presented. Optical channel modeling is presented in details, taking into account both misalignment and positioning of the transceivers.

In this paper, problematic issues in ensuring the cybersecurity of autonomous unmanned objects were considered. Moreover, prerequisites that determine the need for external monitoring systems were identified. The type and statistical characteristics used for the analysis and classification of sound signals were also shown. The proposed approach to the analysis of the cybersecurity condition of an autonomous object is based on classification methods and allows the identification of the current status based on digitized acoustic information processing. An experiment aimed at obtaining statistical information on various types of unmanned object maneuvers with various arrangements of an audio recorder was conducted. The data obtained was processed using two-layer feed-forward neural networks with sigmoid hidden neurons. Hence, the problem of identifying the cybersecurity condition of autonomous unmanned objects on the basis of processing acoustic signal information obtained through side channels was solved. Digitized information from an acoustic sensor (microphone) located statically in the experiment area was classified more accurately than from the microphone located directly on the autonomous object. With a minimum time of statistical information accumulation using the proposed approach, it becomes possible to identify differences in maneuvers performed by the unmanned object and, consequently, the cybersecurity condition of the object with a probability close to 0.7. The proposed approach for processing signal information can be used as an additional independent element to determine the cybersecurity condition of autonomous objects of unmanned systems. This approach can be quickly adapted using various mathematical tools and machine learning methods to achieve a given quality probabilistic assessment.

Medicine is a perspective area for collaborative robotics. The paper presents the collaborative robot as a surgeon’s assistant, accompanying the operation, submitting the necessary tools and performing other auxiliary actions. Such a robot must be mobile, have a manipulator, means of visual communication, an autonomous navigation system in the operating room, and an interactive system for interaction with the surgeon. The last task is considered in the paper. At the voice request of the surgeon, the robot have to find the necessary medical tool on the desktop and transmit it to the surgeon. This operation involves three steps: firstly, at the voice request, the robot must determine which tool is required by the surgeon; on the second step- to find the right tool on the desktop and take it; and on the third – to hand the tool to the surgeon. In the paper the neural networks technology is proposed to solve the recognition problems aroused at two first stages.

The paper considers an algorithm for calculating the motion parameters for a group of mobile robots performing a joint transport task. The motion of this group of mobile robots is considered on a plane surface. Trajectory of motion passes over different zones with various soil properties. Rectilinear motion and motion along the arc of known radius are considered. The algorithm was successfully tested in previously developed special software for debugging and modelling.

We study the navigation problem for a robot moving amidst static and dynamic obstacles and rely on a hierarchical approach to solve it. First, the reference trajectory is planned by the safe interval path planning algorithm that is capable of handling any-angle translations and rotations. Second, the path following problem is treated as the constrained control problem and the original flatness-based approach is proposed to generate control. We suggest a few enhancements for the path planning algorithm aimed at finding trajectories that are more likely to be followed by a robot without collisions. Results of the conducted experimental evaluation show that the number of successfully solved navigation instances significantly increases when using the suggested techniques.

The locomotion of legged robot is often controlled by predefined gaits, and this approach works well when all joints and motors are operating normally. However, walking legged robots usually have high risk of being damaged during operation, causing the breakdown of the robotic joints. In this paper, we introduce a reinforcement learning based approach for the legged robot to generate real-time locomotion response to the emergence of locomotion breakdown. Our approach detects the functionality of the available joints, substitutes the pre-defined gaits with proper gait function accordingly, and upgrades the gait-generation function by Q-Learning for the proper locomotion.

The extractive summartization methods try to summarize a single or multiple documents based on informative sentences exactly as they appear in source(s). One method to choose these sentences is to use users’ query, which could be problematic in many cases, specially in scientific context. One way to tackle this challenge is to gather more information about the user and his preferences. Therefore, in this paper we propose a novel framework to use the users’ feedbacks and a social robotics platform, Nao robot, has been adapted as an interacting agent. This agent has multiple communication channels and could learn the user model and adapt to his/her needs via reinforcement learning approach. The whole approach is then studied in terms of how much it is able to adapt based on user’s feedback, and also in terms of interaction time.