Search

In this paper, we propose a method to improve the recognition performance of a probabilistic model through entropy analysis after transforming time-varying reaction force/torque signals into frequency components. To perform tasks that require physical interaction, it is important for robots to recognize reaction force/torque during the interactions between robots and environments. However, the reaction force/torque measured by F/T sensor contains a lot of noise signals due to the sensitivity of the sensor. Therefore, the recognition performance depends on noise signals included in training and/or test dataset. To solve this problem, the reaction force/torque signals are transformed from time domain into frequency domain by fast Fourier transform. Then, some task-relevant frequency components are selected based on their entropy analysis, after which they are used to learn a hidden Markov model. To evaluate our proposed method, several robot manipulation tasks are performed using an open dataset including reaction force/torque signals: approaching, transferring, and positioning.

In this paper, we propose a method to model social interaction between a human and a virtual avatar. To this end, two human performers fist perform social interactions according to the Learning from Demonstration paradigm. Then, the relative relevance of all joints of both performers should be reasonably modeled based on human demonstrations. However, among all possible combinations of relative joints, it is necessary to select only some of the combinations that play key roles in social interaction. We select such significant features based on the joint motion significance, which is a metric to measure the significance degree by calculating both temporal entropy and spatial entropy of all human joints from a Gaussian mixture model. To evaluate our proposed method, we performed experiments on five social interactions: hand shaking, hand slapping, shoulder holding, object passing, and target kicking. In addition, we compared our method to existing modeling methods using different metrics, such as principal component analysis and information gain.

In this paper, we propose an approach for a robot to adapt time scale of motions to guarantee temporal constraints of robot tasks. To this end, an selective Gaussian mixture model is first presented based on motion significance. Also, an selective Gaussian mixture regression is then presented to select time indexes for regression of robot motions from Gaussian mixtures based on motion significance. Motion significance indicates the relative significance of every motion frame, which is defined as a set of data points acquired at a time index of multiple motion trajectories. Such motion significance is measured by combining both temporal entropy and spatial entropy of a motion frame, based on the analysis of Gaussian mixtures. To evaluate our proposed method, we perform the experiment in which a robot draws a figure on a whiteboard. In addition, we discuss how to develop the adaptive time scaling by adjusting piecewise Gaussian distributions involved in Gaussian mixtures.

Understanding human hand usage is one of the richest information source to recognize human manipulation actions. Since humans use various tools during actions, grasp recognition gives important cues to figure out humans' intention and tasks. Earlier studies analyzed grasps with positions of hand joints by attaching sensors, but since these types of sensors prevent humans from naturally conducting actions, visual approaches have been focused in recent years. Convolutional neural networks require a vast annotated dataset, but, to our knowledge, no human grasping dataset includes ground truth of hand regions. In this paper, we propose a grasp recognition method only with image-level labels by the weakly supervised learning framework. In addition, we split the grasp recognition process into two stages that are hand localization and grasp classification so as to speed up. Experimental results demonstrate that the proposed method outperforms existing methods and can perform in real-time.

In human-human joint assembly work, two human workers complete the assembly of a toy-car by inserting a green wheel and a blue wheel into its wheelbases together. At this time, the green wheel should first be inserted, after which the blue wheel is allowed to be inserted into tis wheelbase. In this work, we propose a learning method for a robot to interact with a human worker instead of a human worker. That is, a human worker is replaced by a robot. In this human-robot joint assembly work, the robot has to complete the assembly of a toy-car with the rest of two humans. For this, a unified framework is used with the following four processes; the process (i) - of segmenting motion trajectories using a Gaussian Mixture Model (GMM) [1], the process (ii) - of modeling motion primitives; here, such motion primitives are represented as Dynamic Movement Primitives (DMPs), the process (iii) - of learning motion causalities; To find pre- and post-conditions for a task execution, we use to find what a robot should direct its attention in motion trajectories from motion trajectories of a robot and to find what a robot should direct its attention in motion trajectories from all possible object-object motion pairs and object-robot motion pairs. Here, pre- and postconditions indicate what have to be checked to activate motion primitives and what have been changed after executing motion primitives, respectively. To obtain pre- and post-conditions, significant variables are selected based on spatial entropies of all motion pairs. These motion causalities are represented Bayesian networks including significant variables. Finally, the process (iv) - of selecting motion primitives according to current and goal situations by using the motivation graph proposed in [2]. To evaluate our proposed method, a toy-car assembling task is performed by inserting a green wheel and a blue wheel into their wheelbases.

This paper describes the design of a new 1-axis torque sensor that consists of two different type spokes. The sensor has three trapezoidal shape spokes and three support spokes. It serves a double purpose that the sensitivity and stiffness for driving torque. We suggest trapezoid-typed spokes rather than conventional rectangular-typed spoke for sensitivity, and support spokes are used to tolerate to decrease the deformation. The sensor is verified by finite element model analysis and experiments. The result showed the linear relationship between the applied weight and the measured value.

This paper presents a simulation based test method for functional and performance evaluation of robotic components and modules. In the proposed test method, function test procedure consists of unit, state, and interface tests which assess if the functional specifications of robot component 1 or module 2 are met. As for performance test, simulation environment provides a down scaled virtual work space for performance test of robot module accommodating virtual devices in conformity with the detailed performance specifications of real robot components. The proposed method can be used for verification of reliability of robot modules and components which prevents faults of them before their usage in real applications. In addition, the developed test system can be extended to support various test conditions implying possible cost saving for additional tests.

An auto-recharging system for a mobile robot can help the robot to perform its tasks constantly and without human intervention. For implementing the system, a docking mechanism is required. This paper presents a new docking mechanism with a localization error-compensation capability. The proposed mechanism uses the combination of mechanical structure and magnetic forces between the docking connectors. It is a structure to improve the allowance ranges of lateral and directional docking errors, in which the robot is able to dock into the docking station. Consequently, this mechanism reduces dependency of a robot control and allows easy docking with only mechanical configuration. In this paper, the superiority of the proposed mechanism is verified with experimental results.

In recent years, collision safety has been one of the most important issues for service robots. To ensure collision safety assurance, a passive compliance method is preferred to an active one because it can provide faster and more reliable responses to dynamic collision. Since both positioning accuracy and collision safety are equally important, a robot arm should have very low stiffness when subjected to a collision force greater than the one causing human injury, but maintain very high stiffness otherwise. In order to realize these ideal features, a novel safe joint mechanism (SJM) composed of linear springs and a modified slider-crank mechanism is proposed in this paper. The SJM has the advantages of variable stiffness which can be achieved only by passive mechanical elements. Various experiments on static and dynamic collisions show the high stiffness of the SJM against an external force of less than the critical impact force, but an abrupt drop in the stiffness when the external force exceeds this critical force, which guarantees positioning accuracy and collision safety. Furthermore, the critical impact force can be set to any value depending on the application and the environment.

Electric power equipment must be operated in the state of high reliability with its responsibility in the seamless power supply. Therefore, persistent maintenance is very important. There are many maintenance methods well known in industrial business. Among them, the reliability-centered maintenance (RCM) is a state of the art maintenance technique. This paper shows a RCM method for electric power equipments. This reliability-centered maintenance (RCM) has been applied in the fields of aerospace, nuclear power plants and rail system, etc. In this paper, we design the FMEA flow for power transformer and evaluate the importance level based on the fault information. The importance level is applied to the inspection period, fault detection method and operator selection. The results of the system are provided as a worksheet containing the information that needs to be maintained.

This study investigates the dynamic characteristics of synchronous reluctance motor (SynRM), with segmental rotor structure, using finite element method in which the moving mesh technique is considered. The focus of this paper is the effective of on-line parameter identification system for position sensorless control of a SynRM under saturation and iron loss. Comparisons are given with angle of the observer and those of proposed FEM & Preisach model of synchronous reluctance motor, respectively. The position sensorless control using identified motor parameters is realized, and the efficiency of the on-line parameter identification system is verified by experimental results.

It is necessary to modify the state-of-the-art of speed control theory because of the phase asymmetry in the linear induction motor (LIM) and for the constant speed control of mover using single vector control inverter system, it is important that primary stack is located in appropriated intervals in the 3D conveyer system using LIM. The dynamic characteristic analysis method of the vector controlled LIM using coupled FEM and control algorithm taking into account the movement is proposed. The focus of this paper is the analysis relative to selecting primary stack intervals in order to constant speed control in the 3D conveyer system using LIM.

A safe robot arm can be achieved by either passive or active compliance. The passive compliance systems composed of purely mechanical elements often provide faster and more reliable responses for dynamic collision than the active ones involving sensors and actuators. Since both positioning accuracy and collision safety are important, a robot arm should exhibit very low stiffness when subjected to the collision force greater than the one causing injury to humans, but maintain very high stiffness otherwise. To implement these requirements, a novel safe link mechanism (SLM), which consists of linear springs, a double-slider mechanism and shock absorbing modules, is proposed in this research. The main contribution of SLM lies in its variable stiffness capability implemented only by passive mechanical elements. Various experiments for static and dynamic collision show that the stiffness of SLM is kept very high for the external force less than the critical impact force, but it drops abruptly as the external force exceeds the critical force, thus guaranteeing the collision safety. Furthermore, the critical impact force can be set to any value depending on the applications.

In development of a human-like robot, a motor control has to manage many motors simultaneously. This paper proposes the method of motor control which makes the human-like robot have facial expressions and behaviors similar to a human. It applied the behavior-based control and motor arbiter for performing continuous behaviors by synthesizing unit behaviors of an upper body and a facial expression. The behavior-based control selects the proper behavior and facial expression about a sensor input of a robot and an output of a high task. The motor arbiter exists between the behavior and the motor driver, and arbitrates interfered outputs generated from when one or more behaviors are executed concurrently. The behavior and facial expression of the human-like robot are expressive of smooth and continuous behaviors by using the behavior-based control and motor arbiter which synthesize several unit behaviors.

This study investigates the dynamic characteristics of synchronous reluctance motor (SynRM), with segmental rotor structure, using finite element method in which the moving mesh technique is considered. The focus of this paper is the sensorless vector control parameters estimation of SynRM under saturation and iron loss. Extensive experimental work has been conducted to verify the accuracy of proposed analysis method.

In this paper, a coordinated scheduling strategy is proposed to reduce end-to-end response time in distributed control systems based on the CAN (controller area network). For the strategy, the analysis of practical end-to-end response time in the worst case is performed priory with considering real-time multi-task operating systems. The end-to-end response time is composed of response time in real-time multi-task scheduling of the operating systems, response time in the CAN communications, and delay caused by asynchronous timing between the operating systems and the CAN communications. The coordinated scheduling strategy is focused on the reduction of the delay caused by asynchronous timing between operating systems and network communications. An online deadline assignment strategy is proposed for the coordination. In the strategy, periods are defined as hard deadlines, and the deadlines of the DMS (deadline monotonic scheduling) are defined as soft deadlines with online assignments for the coordination. The performance enhancement with the coordinated scheduling strategy is validated through simulation tests, by performance comparison between a system with the proposed coordinated scheduling strategy and the system without the coordinated scheduling strategy.

This paper proposes the architecture of a 3 dimensional Web monitoring system for monitoring and control of processes from the soft real-time point of view. Cosmo player is used as a VRML viewer in this paper. The proposed architecture uses Java API for the server API, push methods and Java applet for periodic and continuous updates for monitored data, VRML and CSS (cascaded style sheet) and EAI (external authoring interface) specification for 3 dimensional representation of processes.

A omni-directional self-propulsive trowelling robot is proposed. The proposed robot with two rotary trowels does not require any mechanism such as wheels to obtain driving forces. When the robot flattens a concrete floor with its two rotating trowels, the unbalanced friction forces occur between the trowels and the concrete floor. These friction forces are used to move the robot. Thus, the robot can move in any direction by controlling the two rotary trowels properly. In this paper, firstly the driving force for each trowel is computed by deriving the friction force between the trowel and the concrete floor. Secondly, the relationship between the driving force for the robot and the control variable of the robot is derived. Finally, the basic motions of the robot are realized by using the obtained relationship. This paper figures out how the concrete floor finishing robot with two trowels moves and will contribute to realizing it.

In this work, we have studied the variation of optical characteristics in Te/sub 100-x/Ge, binary system thin films with compositional ranges of x=10 to x=60. To obtain the optimum composition capable of utilizing as an optical recording material, the transmittance and reflectance changes and the contrast ratio between amorphous and crystalline films are investigated using a diode laser with wavelength of 780 nm. It was found that the as-deposited amorphous thin films prepared by thermal evaporation are crystallized by annealing around the crystalline temperature T/sub c/ for each film and then crystalline phases are analyzed using XRD. The reflectance of crystalline films increased the comparison with the amorphous films. The contrast ratio showed 1.45/spl sim/2.1. The amorphous thin films expect with composition of x=50, 60 at% and the crystalline thin films with composition 7=10, 40 at% showed the transmittance change in 80%RH/66/spl deg/C environments. As the results, the Te/sub 50/Ge/sub 50/, Te/sub 40/Ge/sub 60/ thin film can be estimated at the optimum media for optical recording in our study.

This paper analyses an architecture of network-based control system and deals with its application using the CAN protocol. It is difficult to determine an optimal network architecture for a specific system. We propose an architecture of the network-based control system applicable to a specific AGV system with a manipulator arm. The fixed number of periodic messages to be occurred is pre-defined in this system. To determine whether the proposed system architecture is effective or not, we perform a traffic analysis for the real-time communication of all messages. Through simulation, the range of transmission speed is found satisfying the required conditions, and the number of possible sensors to be added is investigated for improving the system performance, when the sampling periods of analog sensors are determined under a fixed condition that the transmission speed is over 500 Kbits/sec.