Next Wave in Robotics

This paper describes how the user can determine the exactness of his motion by using proposed tutoring system based on virtual robot without any other’s help. Sensors in the user’s cloth measure the positions of the joints. The main PC gathers these data and shows the user’s motion by using virtual robot. Tutoring system in PC compares between the database of exemplary motion and the user’s motion and gives feedback to the user.

In order to achieve flexibility and reusability in mobile robot development, we present a component-based robotic framework with its adaptive component model, MiniROS. It provides data, signal and service port as inter-component interfaces so that developers can design complex robot behaviors with data-based, control-based or service-based paradigm. In addition, a strategy is introduced to allow member components of a robot to automatically adjust the structure and parameters by themselves according to the environment and tasks. Finally we show an example application using our component-based model.

In this paper an integrated system for human robot interaction is presented. It is demonstrated that safety features in human robot interaction can be engineered by combining a robotic arm, equipped with a compliant controller, an anthropomorphic robot hand and a spoken language communication system. A simplified human-robot interaction scenario, based on a typical care robot situation, is exploited to show that safety can be enhanced by the monitoring of torques and motor currents to establish contact with the environment. Furthermore, spoken language is utilised to resolve potentially dangerous contact situations.

Aiming at the gain planning of SCUT-I humanoid robot, a stair motion pattern, controlled by step lengths, step period and other walking parameters, is designed. And stair restrict condition is described. According to the parameters and restrict equations, hip and ankle trajectories are planned based on IPM (inverted pendulum model). Finally, according to the model parameters of SCUT-I robot, the robot model is built and simulated by matlab6.5, which verifies the effectiveness and stability of this stairs walking gait generation method.

A real-time path generation based on the elastic band technique is presented to find a collision-free trajectory for an autonomous small-scale helicopter flying through cluttered, dynamic three-dimensional (3D) environments. The dynamic path is followed by the adaptive trajectory tracking controller augmented with the radial basis function neural networks (RBFNN). The effectiveness and merit of the proposed method are exemplified by performing three simulation scenarios: static obstacle avoidance, dynamic obstacle avoidance and terrain following.

The development of a versatile kit to raise student interest in learning the implementation skills of intelligent mobile robots is presented in this paper. The kit is capable of solving micromouse mazes, line mazes, and following line tracks with different curvatures at different speed settings. It is first devised to be used in various project-oriented hands-on laboratory courses for students in the department of electronic engineering of Lunghwa University of Science and Technology, and introductory workshops for vocational high school students and teachers with electronic and information engineering backgrounds. To enhance the learning outcomes, contests can also be organized for students to see how well the techniques learned in the laboratory are applied in their mobile robots.

In order to guarantee that grasping with robot fingers are safe when interacting with a human or a touched object, the robot fingers have to be compliant. In this study, a novel active and robust compliant control technique is proposed by employing an Integral Sliding Mode Control (ISMC). The ISMC allows us to use a model reference approach for which a virtual mass-spring damper can be introduced to enable compliant control. The performance of the ISMC is validated for the constrained underactuated BERUL (Bristol Elumotion Robot fingers) fingers. The results show that the approach is feasible for compliance interaction with objects of different softness. Moreover, the compliance results show that the ISMC is robust towards nonlinearities and uncertainties in the robot fingers in particular friction and stiction.

This paper proposes a generation method of facial expressions using fuzzy measure and fuzzy integral for robotic heads. Human’s emotion state can be represented by a fuzzy measure which can effectively deal with ambiguity. Because facial expressions are usually ambiguous such that it is difficult to discern emotions and assign a sharp boundary to each emotion. In this method, users can adjust the personality of robot by assignign fuzzy measure to every set of emotions. The partial evaluation values of the current emotion state are obtained from a difference between the ideal basic emotion states and the current emotion state. The Choquet integral of the partial evaluation values with respect to the fuzzy measure is calculated to decide which emotion should occur. The effectiveness of the proposed method is demonstrated through computer simulations and experiments with a robotic head with 19 degrees of freedom, developed in RIT Lab., KAIST.

In this paper, a parallel link joint module (PLJM) is developed to achieve three degree-of-freedom (DOF) joint motions with respect to a fixed center of rotation. The proposed 3-DOF spatial PLJM is configured with three linear screw actuators, a base platform, a movable platform, a central bar, and seven ball joints. Due to similar motion characteristics, the PLJM can be used to construct 3-DOF joints of humanoid robots such as hip, ankle, and shoulder joints. Meanwhile, the motions of neck, waist, and wrists can be also constructed using the PLJM. In addition to parallel kinematics mechanism features, the linear screw actuator based PLJM also provides power saving benefits. That is, the PLJM only consumes energy when the PLJM works. As a consequence, humanoid robots constructed using the PLJM may provide longer service time when compared to rotary based joints. Structurally, a PLJM is desired as the connection of a 3-DOF joint and a follower limb; therefore, the PLJM may simplify the mechanical structure of humanoid robots. In addition to the PLJM design, the kinematics, trajectory controls, and interactive sensor integrations are implemented in this paper. Several interesting experiments are demonstrated to verify our approaches.

This paper proposes an ensemble artificial neuro-molecular system for motion recognition for a wearable sensor system with 3-axis accelerometers. Human motions can be distinguished through classification algorithms for the wearable sensor system of two 3-axis accelerometers attached to both forearms. Raw data from the accelerometers are pre-processed and forwarded to the classification algorithm designed using the proposed ensemble artificial neuro-molecular(ANM) system. The ANM system is a kind of bio-inspired algorithm like neural network. It is composed of many artificial neurons that are linked together according to a specific network architecture. For comparison purpose, other algorithms such as artificial neuro-molecular system, artificial neural networks support vector machine, k-nearest neighbor algorithm and k-means clustering, are tested. In experiments, eight kinds of motions are randomly selected in a daily life to test the performance of the proposed system and to compare its performance with that of existing algorithms.

This article focuses on how to use simple image processing techniques to realize a dynamic object detecting and tracking surveillance system on a SoPC. We try to mount a camera on a two-dimensional rotation machinery so as to dynamically search the environment by controlling the rotation of this machinery. In detection mode, the system rotates the machinery along a predefined path to capture images with fixed time interval and compare the images with their corresponding previously recorded reference images for determining if any intrusion objects appear. Once an intrusion object is detected, the system switches to tracking mode. In tracking mode, successive images are compared to find the most possible area in the image where the object locates. The color which occupies biggest region in this possible area in the image is finally recognized as the feature of the intrusion object. The resulting system has functions including intrusion detecting, object tracking, warning message sending, and internet remote watching and all these functions have been experimentally proven that they works well on the SoPC system simultaneously.

This paper presents an intelligent motion controller for four-wheeled omnidirectional mobile robots with four independent driving wheels equally spaced at 90 degrees from one another by using ant colony optimization (ACO). The optimal parameters of motion controller are obtained by minimizing the performance index using the metaheuristic ACO algorithm. These optimal parameters are used in the ACO motion controller to obtain better performance for four-wheeled omnidirectional mobile robots to achieve both trajectory tracking and stabilization. Simulation results are conducted to show the effectiveness and merit of the proposed ACO-based intelligent motion controller for four-wheeled omnidirectional mobile robots.

This paper presents an adaptive hierarchical decoupling sliding-mode speed controller for an electric unicycle. A completely dynamic model of the electric unicycle moving in a flat terrain is derived using Lagrangian mechanics. With the model, an aggregated hierarchical sliding-mode control is used to accomplish robust self-balancing and velocity control (regulation) of the electric unicycle incorporating with viscous and static frictions. Computer simulations and experimental results are conducted for illustration of the effectiveness and applicability of the proposed control method.

A cooperative multi-robot system is proposed to solve the problem of dynamic patrol planning. Each mobile robot has its own patrol mission in the beginning. The patrol mission of each robot needs to be updated when the number of mobile robots increases or decreases during patrol. From the results of the simulation, it is clear that the proposed approach demonstrates several advantages, such as decreased time complexity, a lower routing path cost, improved balance of workload among robots, the potential to scale to a large number of robots, and adaptability with regards to environmental perturbations introduced by changes in the number of robots in patrol.

In this paper, a robot with a manipulator, two vision systems, and a three-wheeled omni-directional mobile platform is designed and implemented. A shoulder, an elbow, a wrist, and a movable gripper are designed to let the robot manipulator have 5 degrees of freedom (DOF). The forward kinematics and inverse kinematics are applied to control the manipulator. The forward kinematics is obtained by the D-H (Denavit-Hartenberg) coordinate. The D-H matrix is a coordinate transformed matrix from one coordinate frame to the next one. The inverse kinematics is constructed by the geometry. Two vision systems are applied to search the object and control the manipulator to pick up it. A static camera is used to construct one vision system to get the environment information. An eye-in-hand camera is used to construct the other vision system to track a specific object. Some experimental results are described. The robot manipulator is controlled to draw a picture by using the direct kinematics and geometric inverse kinematics.

A sensor system is developed in the paper to measure the position and orientation of planar robots. In the sensor system, a monocular vision is integrated with a detection method for abstracting the scale- and orientation-invariant image features. Instead of using multiple cameras, a monocular vision is utilized as the only sensing device to reduce the computation cost. The scale- and orientation-invariant method is employed to guarantee a robust detection and description of features abstracted from an image. Experiment is carried out on a free-moving monocular camera to verify the performances of the proposed system.

A robot state estimation algorithm based on the vision feedback is proposed in the paper. The algorithm consists of an image feature detector and an extended Kalman filter (EKF) based estimator. The detected image features are scale-invariant and provide a robust representation of moving objects and static landmarks in the environment. The recursive EKF-based estimator is utilized to determine the pose and velocity of moving robots. Experiments are carried out on a hand-held binocular camera to verify the performances of the proposed state estimation algorithm. The results show that the integration of the image feature detector and the state estimator is efficient in highly dynamic environments.

This paper describes the implementation of a fast and accurate gesture recognition system. Image sequences are used to train a standard SVM to recognize Yes, No, and Neutral gestures from different users. We show that our system is able to detect facial gestures with more than 80% accuracy from even small input images.

This paper presents a simulation-based analysis of chaotic circuits. In the simulation and experiments, the inductor is replaced with one operational amplifier (OP) and other linear components. The negative resistance is realized with two

μ

A741 OPs. At first, we discuss the low loss chaotic circuit. Then, we proposed a lossless chaotic circuit. Finally, we verify those simulations with experimental results. Applications of chaotic circuits to robot will also be described.

Background subtraction is one of several image segmentation techniques. This technique is used in conditions where the background is boring and static, such as in video surveillance. The codebook model is one of the latest and best techniques utilized for background subtraction. Implementing this technique for robotic soccer vision is a good idea. However, the robotic soccer application needs very fast and robust image pre-processing for image segmentation. We slightly modified the codebook algorithm to get the best performance to be implemented in robotic soccer vision. The result of the experiment shows that the performance of the algorithm becomes better.

Threaded C is a meta-language that is based on C, but is annotated with thread, monitor thread, and semaphore markup. Threaded C uses the runtime provided by the Freezer OS, a small, memory-efficient embedded kernel. The combination of Freezer OS and Threaded C allows the simple expression of common control problems in robotics. The system is geared especially towards robotics education, as it matches the mental map that children have of how control structures should work.

The lamination process of Printed Circuit Board (PCB) is a complex engineering system. Because there are many factors and uncertainties in the process system, it’s hard for the optimization of successful lamination ratio. In this paper, the Taguchi method for lamination process optimization is proposed. Based on fundamental analysis, the major factors including board temperature, lamination temperature, lamination press, and lamination time are figured out. From the scale level of the major factors, orthogonal arrays L18 are engaged for the Taguchi experiments. Based on the orthogonal arrays L18 of the Taguchi experiments, the best successful ratio is analyzed to select the level of major factors. The selected level does again Taguchi experiment to improve the successful ratio. Experiment results show that the successful ratio improves very well.

Based on fuzzy control to adjust attractive factor and repulsive factor of potential function, a method is proposed to navigate a soccer robot from a given initial configuration to a ball and avoid an opponent robot. The potential field method has some problems that include oscillations in the presence of obstacles. Thus, this paper suggests use fuzzy control to adjust attractive factor and repulsive factor. By the attractive factor and repulsive factor of the potential function, the navigation path of the mobile robot is smooth and the shortest. To show the feasibility of the proposed method, the simulation results are included in the following illustrations.

This paper proposes Box-Cox transformation-based annealing robust fuzzy neural networks (ARFNNs) that can be used effectively for function approximated problem with skewness noises. In order to overcome the skewness noises problem, the Box-Cox transformation that its object is usually to make residuals more homogeneous in regression, or transform data to be normally distributed has been added to the annealing robust fuzzy neural networks. That is, the proposed approach uses Box-Cox transformation for skewness noises problem and support vector regression (SVR) for the number of rule in the simplified fuzzy inference systems. After the initialization, an annealing robust learning algorithm (ARLA) is then applied to adjust the parameters of the Box-Cox transformation-based annealing robust fuzzy neural networks. Simulation results show that the proposed approach has a fast convergent speed and more generalization capability for the function approximated problem with skewness noises.

A fuzzy PID controller is proposed to asymptotically stabilize a three-dimensional overhead crane using a hybrid optimization approach in this article. In the proposed algorithm, the PID gains are adaptive then the fuzzy PID controller has more flexibility and capability than the conventional ones with fixed gains. To tune the fuzzy PID controller simultaneously, a hybrid optimization procedure integrating genetic algorithm (GA) and particle swarm optimization (PSO) method is adopted. The simulation results illustrate that the proposed controller with few fuzzy rules can effectively perform the asymptotical stability of the prototype overhead crane.

In order to make the robot lighter, minimize the size, and have good heat dissipation, the technique of high power density DC-DC converter design is necessary. It can minimize the internal circuit in the robots, improve heat dissipation, and have high efficiency. Therefore, adoption of a forward converter synchronous rectification is used. This technique can diminish the loss of diode forward so it is able to increase the conversion efficiency. The active clamp reset transformer is used on primary-side to make Transformer demagnetization. At the same time, the power MOSFET can achieve the goal of zero voltage switching function, improving the efficiency of the converter. Finally, the forward converter is made with operating frequency of 300k Hz and 100W output power, and the maximum efficiency is 92.4%.

To improve stability of the robot power supply when perform task without voltage drop or sufficiency; therefore, a design of high-efficiency energy-saving method for power management is presented. It can be used in rechargeable lithium batteries and monitoring current output of different voltage so that the power contained inside the robot is able to be used for robot vehicles. By a micro-controller monitors the Robot vehicles use the CAN Bus protocol to receive the fuel cell internal parameter values and transmission power values of the robot vehicles, in the QGVA 5.7-inch touch screen display, so that drivers can immediately to control the operational status of the robot vehicles.

A hydrogen fuel cell power supply system which includes pressure valve, solenoid valves, flow controllers, temperature sensors and Proton Exchange Membrane Fuel Cell (PEMFC) is presented in the paper. The power supply system builds a control circuit loop includes a charge-discharge circuit with buck conversion topology. The feedback control signal of the power supply system by digital signal processor (dsPIC). This structure can achieve the goal of the small fuel cell power supply system stability, have high efficiency fuel cell, high efficiency converters, and to provide energy to a small robot.

This simulator is developed for Android Soccer Tournament (AndroSot) in FIRA. Due to the robot soccer game presents a dynamic and complex environment, it provides a challenging platform for multi-agent research. However, if there were some problems occurred in the robot actions and image processing algorithm, it is very difficult to run or test strategy systems. In order to solve these issues, a humanoid robot soccer competition’s strategy simulation system is proposed, which provides developer to test the feasibility and advancement of the game strategy. In this simulator, strategies which compiled to DLL files may be explicitly loaded at run-time.

The heavy weight of a big humanoid robot body makes its dynamics considerable. Therefore, the walking pattern derived according to only kinematics is difficult for a big humanoid robot walking. In this paper, the dynamics of a big humanoid robot is approached to an inverted pendulum for its walking pattern. The inverted pendulum approach regards the humanoid body as a point mass to simplify walking pattern parameters. However, parameter uncertainty is not avoided, and usually joint servo motors cannot provide enough torque for the big humanoid robot walking. In this paper, the actual servo motor move trajectories in experiments are read back to compare with the desired ones. The errors between the desire and actual trajectories are used to modify the designed trajectories. Experimental results show that the big humanoid robot can walk very well after modifying.

With the rapid technological advancement in vision systems, the advantages of vision based system as the sensing device have influenced the quality of the Robot Soccer game. This study presents an approach to produce a balanced image in terms of brightness under un-uniform lighting environment through the control of lens iris opening. Existing image processing module and CCD camera were used to obtain data of HSL (color system) properties in three un-uniform lighting conditions. These data were used to deduce the percentage of optimum iris lens aperture to increase efficiency of soccer robot detection under the un-uniform lighting conditions. The optimum range of iris lens opening was identified and the robots were detected successfully for all the un-uniform lighting conditions.

Soccer Robotics is the field that combines multiple important discipline including artificial intelligence and mobile robotics. The framework for the complementary purposes of research and education in multi-agent robotics arises for exploration of new technology. Preventive maintenance is essential thing to preserve aging of Robot Soccer Mirosot and also to decrease the costing to maintain performance of the robot. This paper will be discussing a framework classified as model-specific to execute preventive maintenance for Robot Soccer as follow the standard of procedure. A plan schedule of preventive maintenance established quarterly will be considered the physical robot situated in a real environment and it’s reactive, proactive and communicative. The primary goal of this framework preventive maintenance it will be use as standard of procedure to anyone during handling maintenance of Robot Soccer Mirosot. It also to prevent the failure of Robot Soccer before it actually occurs. It is designed to preserve and enhance equipment reliability by replacing defect components before they actually fail.

In order for a mobile robot to perform its assigned tasks, it often requires a representation of its environment such as knowledge of how to navigate in its environment, and a method for determining its position in the environment. A major problem in computer vision and machine learning is to achieve a good feature as it can largely determine the performance of a vision system. A good feature should be informative, invariant to noise or a given set of transformations, and fast to compute. Also, in certain settings sparsity of the feature response, either across images or within a single image, is desired. Our objective of this paper is to obtain optimal features as well as determining the optimal class of angle in order to estimate mobile robot orientation single or unified images from two camera orientations. We introduce feature selection process before classifying features based on support vector machine classifier. We achieve better accuracy rate by only reducing its feature number from 30 features down to only 17 features on unified images. Furthermore, we also find that only 5 classes of robot angles are sufficient to estimate robot orientation correctly.

In this paper, we create a passing, obstacle avoiding and shooting strategy for robotic soccer coordination. Based on a predefined-scenario in a robotic soccer game, we simulate a mini case study which involves two robots and a ball. We modify role, act and behavior method to meet the game requirements. About 61% of the testing achieved the shooting of a goal by manipulating and redesigning the strategy. The shortest goal shooting time was about 5 seconds. We hope to improve this initial strategy in the future.