Search

Robotic arm performance varies due to normal and abnormal events. Normal events may include degradation of equipment, motors, mechanical system joints, and gears, while abnormal events may occur such as faulty episodes. In this paper, we address positional performance degradation that can be stopped and redressed if suitable required action is achieved. The Tool Center Point (TCP) position measurement devices are expensive, hence unavailable to every robot. Some industrial processes are critically sensitive to target tool position such as assembly, pin and past, and material handling. We propose a data driven artificial intelligence tool to detect anomalies and degradation of the robotic arm for a positional health assessment without the need for special advanced sensors. TCP deviation is predicted using deep machine learning models that train on time series of historical data of the robot’s performance. Statistical thresholds are calculated to detect the robotic arm’s degradation and anomalies by performing residual analysis. An alarm system is built by applying the proposed monitoring tool online.

In this paper, we propose a novel adaptive control scheme for robotic systems by incorporating the parameter error into the adaptive law. By carrying out filter operations, the robotic system is linearly parameterized without using the measurements of acceleration. Then a new adaptive algorithm is introduced to guarantee that the parameter error and control error exponentially converge to zero. In particular, we provide an intuitive method to verify the standard PE condition for the parameter estimation. The robustness against disturbances is also studied and comparisons to several adaptive laws are provided. Simulations with a realistic robot arm are presented to validate the improved performance.

Thyroid imaging using radioactive tracers is a common task in clinics and is usually performed using 2D gamma cameras (scintigraphy). In this work we present a setup for 3D imaging of the thyroid using a mini gamma camera mounted on a robotic arm. Several images are acquired moving the mini gamma camera along a trajectory around the thyroid. Afterwards, a tomographic reconstruction computes a 3D SPECT-like image of the thyroid. First results are shown of a thyroid phantom using a conventional statistical reconstruction scheme (MLEM) and using a sparse regularization approach based on total variation.

Online learning of vision-based robot control requires appropriate activation strategies during operation. In this chapter we present such a learning approach with applications to two areas of vision-based robot control. In the first setting, selfevaluation is possible for the learning system and the system autonomously switches to learning mode for producing the necessary training data by exploration. The other application is in a setting where external information is required for determining the correctness of an action. Therefore, an operator provides training data when required, leading to an automatic mode switch to online learning from demonstration. In experiments for the first setting, the system is able to autonomously learn the inverse kinematics of a robotic arm. We propose improvements producing more informative training data compared to random exploration. This reduces training time and limits learning to regions where the learnt mapping is used. The learnt region is extended autonomously on demand. In experiments for the second setting, we present an autonomous driving system learning a mapping from visual input to control signals, which is trained by manually steering the robot. After the initial training period, the system seamlessly continues autonomously. Manual control can be taken back at any time for providing additional training.

The contributions for this book have been gathered over several years from conferences held in the series of Mechatronics and Machine Vision in Practice, the latest of which was held in Ankara, Turkey. The essential aspect is that they concern practical applications rather than the derivation of mere theory, though simulations and visualization are important components. The topics range from mining, with its heavy engineering, to the delicate machining of holes in the human skull or robots for surgery on human flesh. Mobile robots continue to be a hot topic, both from the need for navigation and for the task of stabilization of unmanned aerial vehicles. The swinging of a spray rig is damped, while machine vision is used for the control of heating in an asphalt-laying machine. Manipulators are featured, both for general tasks and in the form of grasping fingers. A robot arm is proposed for adding to the mobility scooter of the elderly. Can EEG signals be a means to control a robot? Can face recognition be achieved in varying illumination?"

This paper justifies the need for more research in the area of ‘assistive robotics’ and it describes the design and performance of a manipulator arm and gripper, mounted on an electric scooter, that can enable frail elderly and people with disabilities to collect and retrieve heavy objects, such as shopping items, located on high shelves. Known by the acronym ‘ESRA’, short for ‘Electric Scooter Robot Arm’, this device can be attached to a mobile vehicle like an electric scooter or an electric-powered wheelchair. This paper discusses the basic mechanical design of the ESRA manipulator and gripper, joint control, important design tools and failure analysis that were used for its development, and overall performance of the prototype arm. The object retrieval performance of this manipulator can also be enhanced using machine vision techniques for automatic object detection and gripper guidance. Different end-effector tools can be mounted on the end of the ESRA to enable the operator to perform a wide variety of common household chores and manipulation tasks.

Recently we have shown developments on capacitive tactile proximity sensors (CTPS) and their applications. In this work we give an overview of these developments and put them into a more general perspective, emphasizing what the common grounds are for the different applications, i.e., preshaping and grasping, haptic exploration as well as collision avoidance and safe human-robot interaction. We discuss issues related to signal processing and the design of a smart skin for the robot arm and its end-effector. On a higher level we discuss the concept of proximity servoing and its use for the above mentioned applications.

The integration of robotics into the field of gynecologic surgery initially involved a voice-recognition system and robotic control of laparoscopic camera movement (AESOP® Endoscope Positioner (Computer Motion Inc., Goleta, CA)). This technology further evolved with the ZEUS® Surgical System (Computer Motion Inc., Goleta, CA) to include two robotic arms, thereby allowing the surgeon to operate from a console away from the operative table. The da Vinci Surgical System® (Intuitive Surgical, Sunnyvale, CA, USA), which was approved for gynecologic surgery in 2005 by the United States Food and Drug Administration, incorporates three or four robotic arms to allow surgeon control of the visual field and up to three laparoscopic instruments.

Motion planners for robots in unknown and dynamic environments often assume known obstacle geometry and use that to predict unknown motions of obstacles through tracking, but such an assumption may not be realistic. In [1], we introduced a collision-free perceiver (CFP) that can detect guaranteed collision-free trajectory segments in the unknown configuration-time (CT) space of a robot without assuming known obstacle geometry or motion. However, such a guarantee by the CFP is at the expense of a finite period for perception and processing of each collision-free CT point. In this paper, we address how to incorporate the CFP, taking into account its finite processing time, into real-time motion planning to enable a robot of high degree of freedom to plan and move at the same time in an unknown and unpredictable environment while minimizing unsafe stops when the robot may collide with an obstacle. The approach was implemented and tested in experiments with a real 7-DOF robot arm and a stereo-vision sensor, indicating the potential of the approach.

This paper presents the concept and implementation of an encountered-type haptic interface that enables interaction with soft round-shaped objects. The core part of the system is a silicone balloon whose size is controlled by a pneumatic actuator in accordance with the size of a virtual object being touched. The balloon is mounted on a position-controlled robotic arm, and its position is controlled so that it follows a user’s hand. The user encounters the balloon only when the proxy representing the hand in the virtual world is in contact with the virtual object. Our new haptic system allow the user not only to grasp a soft object with variable size, but also to follow the contour of the virtual object. The accuracy and control bandwidth of the rendering result was measured and shows that the radius of the balloon was accurately controlled with high speed. The potential of the system is also demonstrated by an example interaction scenario: interaction with human avatar’s arm.

Detecting hydraulic excavators in videos can increase the confidence coefficient of illegal construction in nationalized land. Hydraulic Excavators have multifarious working postures making them a difficult target using even state of the art object recognition algorithms. The contribution of this paper is to propose an inverse ”V” model for hydraulic excavator detection. In this paper, we describe an hydraulic excavator detection system based on inverse ”V” feature of mechanical arm which is formed by boom and dipper and show a detection system. Then a real-time video processing method is presented which is used for monitoring illegal construction activities on a land of state-ownership.

Brain-Computer Interfaces (BCIs) are devices that translate a user’s brain activity into messages or commands (Wolpaw et al., 2002; Pfurtscheller et al., 2010; Wolpaw and Wolpaw, 2012). BCIs have four components: a signal acquisition system that records the user’s brain activity; a signal processing module that finds meaningful patterns within this brain activity that reflect a user’s intent; an output device such as a monitor, wheelchair or robotic arm; and an operating protocol that controls the interaction among these different modules and people such as the user or an operator. Signal acquisition may involve sensors placed on or in the brain (through neurosurgery), or noninvasive methods such as scalp-mounted electrode caps or functional MRI.

In organic farming, eradicating weeds is a problem because great importance is attached to the ecosystem, and consequently agricultural chemicals are not used. Therefore, a great deal of heavy labor is needed for eradicating weeds. The aim of this research is to design and implement a robot arm for weeding mounted on a robot in a paddy field. It is hoped that this research will result in rice field managers being able to remove weeds remotely by means of being able to observe the rice field via a stereo camera. This task will make a big reduction of time and labor for rice farming.

This paper discusses the project which is one of the Beijing City College Students’ Scientific Research and Entrepreneurial Action Plan Project named "unmanned area fetching trolley", The project used infrared sensor to measure the distance between the car and the obstacle, The MCU would analyze and process this signal, and generate two different PWM signals respectively to drive motors when the distance is less than a certain value, so as to realize avoiding obstacle automatically. After Hall sensor detected the PWM wave corresponding to the actual speed of the car, the PID module would compare the PWM wave of the actual speed and the PWM wave which is generated by MCU, so as to determine whether the output is desired and achieve closed-loop control, thus can make the control of speed is more precise. The mechanical arm is composed of five steering, the upper monitor control it through the wireless module, each instruction can make the rotation angle of each actuator accurate to 1 degree, so the mechanical arm is flexible and precise. The project ultimately achieved the desired effect that avoid obstacle automatically, upper monitor control accurately. After expansion and improving, it can be used for automatic weeding, automatic cleaning, automatic patrol, automatic sowing, automatic harvesting and so on.

Extraction of rock samples using robotic arms is an essential technique for planetary exploration. By sampling and analyzing the rocks, we can gain knowledge, not only about the planetaly body itself but also about other bodies and the space environment as a whole. In conventional missions, images of the surface are captured and sent to Earth so that the human operator can deside a sampling strategy. However, in such missions, communication delays with Earth are inevitable, and a large number of scientists and engineers are involved in the decisions, which makes the procedure time-consuming, expensive and limited in efficiency. For these reasons, automated extraction processes are required to increase the scientific return. In this study, a novel intelligent manipulator system which samples the rocks semi-automatically by measuring environmental stiffness is proposed. To make such a system autonomous in an unknown environment, vision-based approaches are usually preferred. However, information obtained from cameras is difficult to process. In this paper, another technique based on measuring applied forces is presented, and we demonstrate its effectiveness for the detection of buried rocks.

This paper presents a tele-control system comprised of a robot hand/arm and an operator. In our system, the angle of the finger of the robot hand is controlled according to the angle of the operator’s finger, and the position of the robot arm is controlled according to the position of the operator’s arm. Simultaneously, the operator feels the environmental force as detected by the touch sensor attached to the robot hand, resulting in so-called bilateral master/slave control. To date, there have been a few studies of bilateral master/slave systems that use a multi-fingered humanoid robot hand in communication networks with delay caused by physical distance. The purpose of our study was to achieve tele-operation with bilateral master/slave control between the operator’s hand/arm and the multi-fingered humanoid robot hand/arm using the Internet with delay caused by distance. For our experiment, the operator was in the USA, and the multi-fingered humanoid robot hand/arm and object were in Japan. Using our system, the operator could successfully grasp and move an object while he felt the reaction force from the object. This technology can applicable to tele-operation such as feeling the hardness and weight of the stone which exist in the moon, on the ground etc.

Dynamic optimization problems require constant tracking of the optimum. A solution for such a problem has to be adjustable in order to remain optimal as the optimum changes. The manner of changing design parameters to predefined values is dealt with in the field of control. Common control approaches do not consider the optimality of the design, in terms of the objective function, while adjusting to the new solution. This study highlights the issue of the optimality of adaptation, and defines a new optimization problem – ”Optimization of Adaptation”. It is a multiobjective problem that considers the cost of the adaptation and the optimality while the adaptation takes place. An evolutionary algorithm is proposed in order to solve this problem, and it is demonstrated, first, with an academic example, and then with a real life application of a robotic arm control.

This paper studies the ergonomic issues of a new multimodal application based on the combination of an electrooculography interface and a brain-computer interface to interact with a robot arm. This application allows performing pick and place operations in a specific designed environment. Eight able-bodied volunteers tested the application and were requested to fill a survey about several aspects. To critically discuss and analyze the results, questions about comfort, reliability, usability and usefulness were answered by the participants. The results of the survey show the usefulness of the system and the application, and its relevance when using it with handicapped people with severe neuromuscular disorders. However, the use of multimodality also shows some limitations.

In order to draw a multi-joint robot arm, the first thing that we must do is to know how to create surfaces of the 3D basic parts in MATLAB TM . The 3D basic parts may include but are not limited to a cylinder and its variations, a sphere or an ellipsoid, a rectangular block, and a torus. Since in MATLAB TM , each 3D part drawing is always initiated at the base coordinate frame, we need to be not only aware of how to create it, but also familiar of its data structure in order to modify it and/or send it to a desired destination (both position and orientation). The detailed instructions, procedures and examples can be found either in the MATLAB TM online help manual, or refer to the literature [1]-[4]. Let us here start with a surface of cylinder drawing.

We have recently established simultaneous 7 degree-of-freedom (DoF) brain-computer interface (BCI) control of a robotic arm. Using signals recorded from single units of monkeys with implanted chronic microelectrode arrays, we can now demonstrate brain control of a prosthetic arm that exhibits the following features: (1) simultaneous 7-degree of freedom (DoF) brain control over 3-D robot hand translation, 3-D rotation, and finger aperture, (2) integrated kinematic (movement) and dynamic (force) control of a brain-controlled prosthetic robot through a novel impedance-based movement controller, (3) simplified methods for constructing cortical extraction models based only on observation of the moving robot, and (4) a generalized method for training subjects to use complex prosthetic robot devices using a novel form of operator-machine shared control.

Joint stiffness control for safety is widely investigated and many devices are invented both active and passive control of joint stiffness. However, none of them are satisfied both larger torque endurance and active control. The variable yielding torque spring-clutch (VSC) can guarantee both active control and large torque endurance. This paper explains about initial motivation and detail procedures from design objective to making prototype and testing.

Learning how to control arm joints for goal-directed reaching tasks is one of the earliest skills that need to be acquired by Developmental Robotics in order to scaffold into tasks of higher Intelligence. Motor Babbling seems as a promising approach toward the generation of internal models and control policies for robotic arms. In this paper we propose a mechanism for learning sensory-motor associations using layered arrangement of Self-Organizing Neural Network (SOINN) and joint-egocentric representations. The robot starts off by random exploratory motion, then it gradually shift into more coordinated, goal-directed actions based on the measure of error-change. The main contribution of this research is in the proposition of a novel architecture for online sensory-motor learning using SOINN networks without the need to provide the system with a kinematic model or a preprogrammed joint control scheme. The viability of the proposed mechanism is demonstrated using a simulated planar robotic arm.

Different types of nuclear imaging systems have been used in the past, starting with pre-operative gantry-based SPECT systems and gamma cameras for 2D imaging of radioactive distributions. The main applications are concentrated on diagnostic imaging, since traditional SPECT systems and gamma cameras are bulky and heavy. With the development of compact gamma cameras with good resolution and high sensitivity, it is now possible to use them without a fixed imaging gantry. Mounting the camera onto a robot arm solves the weight issue, while also providing a highly repeatable and reliable acquisition platform. In this work we introduce a novel robotic setup performing scans with a mini gamma camera, along with the required calibration steps, and show the first SPECT reconstructions. The results are extremely promising, both in terms of image quality as well as reproducibility. In our experiments, the novel setup outperformed a commercial fhSPECT system, reaching accuracies comparable to state-of-the-art SPECT systems.

The use of robots in industrial applications has been widespread from handling tasks to processes. The finishing processes include operations such as deburring, grinding and polishing. Within these processes, there is a need to control the contact force between the workpiece and the robot. The force control can be implemented in a passive or active form, either in the robot arm or with an external device. This paper presents the development of a device to be used in robotic finishing applications. It provides a semi-active system that limits and maintains the contact force and is used in conjunction with the robot controller. The device uses a pneumatic driven linear axis fitted with a position sensor. The architecture of the developed system and experimental results regarding the performance of the built prototype are presented. The implementation of different control strategies to adjust the robot path velocity are proposed and discussed.

A novel distributed adaptive control algorithm of a pair of humanoid robot arm system (Bristol-Elumotion-Robotic-Torso II (BERT II)) is proposed, analysed and simulated. Two humanoid arms are subjected to a distributed synchronisation control with a virtual leader-following trajectory to be followed, serving a potential application for a smooth cooperative task. The approach presented here is inspired by multi-agent theory. Graph theoretical concept such as Laplacian matrix is used to represent mutual communication between the two arms (regarded as agent nodes) with one of the arm ‘pinned’ to a virtual leader (leader node). The stability of the proposed algorithm is analysed through Lyapunov technique. The algorithm features a sliding-mode term based adaptive law that captures directly the parameter estimation error. An auxiliary filtered regression vector and filtered computed torque is introduced. This allows the definition of another auxiliary matrix, a filtered regression matrix, which facilitates the introduction of the sliding term into the adaptation law. Parameter error convergence to zero can be guaranteed within finite-time with a Persistent-Excitation (PE) or Sufficient Richness condition for the virtual leader’s trajectory.

This chapter describes a novel methodology for behavior learning of an agent, called Coaching. The proposed method is an interactive and iterative learning method which allows a human trainer to give a subjective evaluation to the robotic agent in real time, and the agent can update the reward function dynamically based on this evaluation simultaneously. We demonstrated that the agent is capable of learning the desired behavior by receiving simple and subjective instructions such as positive and negative. The proposed approach is also effective when it is difficult to determine a suitable reward function for the learning situation in advance. We have conducted several experiments with a simulated and a real robot arm system, and the advantage of the proposed method is verified throughout those experiments.

Our aim is to reconstruct the brain-body loop of stroke patients via an EEG-driven robotic system. After the detection of motor command generation, the robotic arm should assist patient’s movement at the correct moment and in a natural way. In this study we performed EEG measurements from healthy subjects performing discrete spontaneous motion. An EEG analysis based on the temporal correlation of the brain activity was employed to determine the onset of single motion motor command generation.

The presented research work introduces the safe limitation of contact forces between an industrial robot and its human operator during physical collaboration. While estimating these contact forces with a physics algorithm in the robot control software, they can be easily limited before getting harmful. The developed algorithms allow the user to stop and move the robot arm with moderate effort and therefore he feels safe during interaction. Through bypassing and redesigning the position feed-back controllers of the robot, the drive output torques can be held under the critical threshold calculated via Newton-Euler dynamics.

To provide a rich learning experience, an intelligent tutoring agent should be able to take into account past and present events, and to learn from its interactions with learners to continuously improve the assistance it provides. Until now, the learning capabilities of tutoring agents in educational technologies have been generally very limited. In this chapter, we address this challenge with Conscious-Emotional-Learning Tutoring System (CELTS), a cognitive tutoring agent, whose architecture is inspired by the latest neuroscientific theories. CELTS implements human-like learning capabilities such as episodic, emotional, procedural and causal learning. The agent is used in a simulation-based tutoring system for learning the complex task of operating the Canadarm2 robotic arm on the International Space Station (ISS). Experimental evaluations showed that the agent’s learning capabilities considerably enhance its adaptation to learners during interactions and consequently improves learners’ performance.

Active respiratory motion management has received increasing attention in the past decade as a means to reduce the internal margin (IM) component of the clinical target volume (CTV)—planning target volume (PTV) margin typically added around the gross tumor volume (GTV) during radiation therapy of thoracic and abdominal tumors. Engineering and technical developments in linear accelerator design and respiratory motion monitoring respectively have made the delivery of motion adaptive radiation therapy possible through real-time control of either dynamic multileaf collimator (MLC) motion (gantry based linear accelerator design) or robotic arm motion (robotic arm mounted linear accelerator design).

In order to model an open serial-chain robot arm kinematically, we need not only to represent a 6 d.o.f. rigid motion in Cartesian space for the purpose of task/path description, but also to determine a complete transformation along the entire serial-chain robot body in joint space for the purpose of joint values determination. During the 1950’s, Denavit and Hartenberg had independently developed a wonderful procedure to uniquely determine all the joint parameters plus variables for an open serial-chain robotic manipulator. Today, almost everyone recognizes that the D-H procedure is the best approach to uniquely modeling robotic kinematics, which is commonly called the D-H Convention [1, 2, 4].

The paper discusses a new method of tracking and controlling robots that interact with humans (natural interaction) to provide assistance services in manufacturing tasks. Using depth sensors the robots are able to assist the human operator and to avoid collisions. Natural interaction is implemented using a depth sensor which monitors the activity outside and inside the robot system workspace. The sensor extracts depth data from the environment and then uses the processing power of a workstation in order to detect both humans and robot arms. This is done by detecting skeletons which represent the position and posture of the humans and manipulators. Using skeleton tracking, a software agent is able to monitor the movements of the human operators and robotic arms and to detect possible collisions in order to stop the robot motion at the right time. Also the agent can interpret the posture (or full body gesture) of the human operator in order to send basic commands to the robot.

The problem of collision-free path planning for redundant manipulator with the given desired pose is considered. For orientation, quaternion representation is applied to give singularity-free orientation control. A hybrid algorithm of combining the Jacobian pseudoinverse algorithm with Rapidly-Exploring Random Tree RRT method is proposed. The nearest configuration point to the task space with the goal pose is taken as the expanded node of the tree. The best expanding direction is calculated with the Jacobian pseudoinverse control algorithm. The velocity of the manipulator's end-effector is restricted by using the bisection gradient-decent extend algorithm to avoid the occurrence of joint velocity mutation. The simulation results from a 7 degrees of freedom robotic arm show the rapidity and validity of the proposed hybrid algorithm.

In this paper, a method to automate industrial manufacturing processes using an intelligent multi-camera system to assist a robotic arm on a production line is presented. The examined assembly procedure employs a volumetric method for the initial estimation of object’s properties and an octree decomposition process to generate the path plans for the robotic arm. Initially, the object is captured by four cameras and its volumetric representation is produced. Thereafter, a quality check with its respective CAD model is performed and the final details of the 3D model are refined. An octree decomposition technique is utilized afterwards to facilitate the automatic generation of the assembly path plans and translate them to a sequence of movements for the robotic arm. The algorithm is fast, computationally simple and produces an assembly sequence that can be translated to any major robotic programming language. The proposed algorithm is assessed and preliminary experimental results are discussed.

Recently, many studies on educational system are conducted. In this paper, a unique educational experiment system is proposed for undergraduate students to be able to efficiently learn basic mechatronics techniques. The system is composed of three subsystems. The first subsystem is used to learn input/output port operations, periodically LED lights ON/OFF and a stepping motor control. The second subsystem is effective to learn AD transformation for several sensor information, DA transformation for DC motor control and a PID control method. Further, the third subsystem is designed by using a robot arm with four-DOFs to learn PWM (Pulse Width Modulation) control of a DC motor and force control of an end-effector. The effectiveness of the proposed system was confirmed through experimental instructions in Tokyo University of Science, Yamaguchi.

We propose a skill learning and inference framework, which includes five processing modules as follows: 1) human demonstration process, 2) autonomous segmentation process, 3) process of learning dynamic movement primitives, 4) process of learning Bayesian networks, 5) process of constructing motivation graph and inferring skills. Based on the framework, the robot learns and infers situation-adequate and goal-oriented skills to cope with uncertainties and human perturbations. To validate the framework, we present the experimental results when using a robot arm that performs a daily-life task.