Moving Personal Robots in Real-Time Using Primitive Motions

Personal robotics is a new and attractive use of robotic technologies. In this paper we study one of its important topics—real-time motion planning of personal robots. We propose to use primitive motions and their combinations to make this possible. The primitive motion has a unified pattern and is borrowed from the motion pattern of human hands observed by previous researchers. The pattern is simple yet powerful and can form complex trajectories. A reflexive motion control scheme is proposed to activate appropriate primitive motions for a desired motion. As a result, real-time motion planning of personal robots becomes possible by using discrete and sparse human commands. Experiments results are presented to show the effectiveness of the proposed scheme.     

Posture control of robot manipulators with fuzzy voice commands using a fuzzy coach–player system

This paper presents a method of controlling robot manipulators with fuzzy voice commands. Recently, there has been some research on controlling robots using information-rich fuzzy voice commands such as 'go little slowly' and learning from such commands. However, the scope of all those works was limited to basic fuzzy voice motion commands. In this paper, we introduce a method of controlling the posture of a manipulator using complex fuzzy voice commands. A complex fuzzy voice command is composed of a set of fuzzy voice joint commands. Complex fuzzy voice commands can be used for complicated maneuvering of a manipulator, while fuzzy voice joint commands affect only a single joint. Once joint commands are learned, any complex command can be learned as a combination of some or all of them, so that, using the learned complex commands, a human user can control the manipulator in a complicated manner with natural language commands. Learning of complex commands is discussed in the framework of fuzzy coach–player model. The proposed idea is demonstrated with a PA-10 redundant manipulator.     

Whole-Body Motion Generation Integrating Operator's Intention and Robot's Autonomy in Controlling Humanoid Robots

This paper introduces a framework for whole-body motion generation integrating operator's control and robot's autonomous functions during online control of humanoid robots. Humanoid robots are biped machines that usually possess multiple degrees of freedom (DOF). The complexity of their structure and the difficulty in maintaining postural stability make the whole-body control of humanoid robots fundamentally different from fixed-base manipulators. Getting hints from human conscious and subconscious motion generations, the authors propose a method of generating whole-body motions that integrates the operator's command input and the robot's autonomous functions. Instead of giving commands to all the joints all the time, the operator selects only the necessary points of the humanoid robot's body for manipulation. This paper first explains the concept of the system and the framework for integrating operator's command and autonomous functions in whole-body motion generation. Using the framework, autonomous functions were constructed for maintaining postural stability constraint while satisfying the desired trajectory of operation points, including the feet, while interacting with the environment. Finally, this paper reports on the implementation of the proposed method to teleoperate two 30-DOF humanoid robots, HRP-1S and HRP-2, by using only two 3-DOF joysticks. Experiments teleoperating the two robots are reported to verify the effectiveness of the proposed method.     

Atmospheric perturbation and control of a shuttle/tethered satellite

A simplified model of the strong atmospheric perturbation of a small satellite attached to the orbiter by a long, straight tether predicts undesirable uncontrolled libration motions similar to those from sophisticated models. The simplified model is used to compare performance limitations of two simple control systems that rely only on tether tension for three-dimensional control. Because the effects on libration of both the changing kinematic and atmospheric drag torques caused by changing length are largely predictable, the controller that models these changes has substantially better performance. In particular, the altitude variation of the tethered satellite above an oblate earth can be controlled substantially better with a controller that uses a length command adapted to these predictable changes than one that only reacts to tension changes. Also, small amplitude out-of-plane librations, that are coupled only to second-order with changing length, can be damped substantially faster by the controller that periodically commands properly phased length changes than by the controller that only reacts to the second-order changes in tension.     

Reduced order extended command governor

Extended command governors (ECGs) are add-on schemes that modify set-point commands as necessary to ensure that imposed state and control constraints are not violated by closed-loop systems designed for set-point tracking. In this paper, we propose a reduced order ECG for systems with dynamics decomposable into slow and fast state variables. We demonstrate that ECG implementation can be based on slow states only, thus reducing the computational complexity. This is achieved by introducing additional constraints, and by slightly tightening the original constraints. We demonstrate that the proposed ECG maintains the response properties of the conventional ECG, including the convergence to the nearest feasible command in finite time in the case of constant reference commands. The results are also shown to apply to conventional command governors. For the case when the reduced order state is not directly measured, a formulation of the result in the presence of a state observer is developed.     

Progressive learning and its application to robot impedancelearning

An approach to learning control using an excitation scheduling technique is developed and applied to an impedance learning problem for fast robotic assembly. Traditional adaptive and learning controls incur instability depending on the reference inputs provided to the system. This technique avoids instability by progressively increasing the level of system excitation. Called progressive learning, it uses scheduled excitation inputs that allow the system to learn quasistatic parameters associated with slow input commands first, followed by the learning of dynamic parameters excited by fast input commands. As learning progresses, the system is exposed to a broader range of input excitation, which nonetheless does not incur instability and unwanted erratic responses. In robotic assembly, learning starts with a slow, quasistatic motion and goes to a fast, dynamic motion. During this process, the stiffness terms involved in the impedance controller are learned first, then the damping terms and finally by the inertial terms. The impedance learning problem is formulated as a model-based, gradient following reinforcement learning. The method allows the suppression of excessive parameter changes and thereby stabilizes learning. By gradually increasing the motion speed command, the internal model as well as the control parameters can be learned effectively within a focused, local area in the large parameter space, which is then gradually expanded as speed increases. Several strategies for motion speed scheduling are also addressed.     

Learning of a controller for non-recurring fast movements

In this paper a learning method is described which enables a conventional industrial robot to accurately execute the teach-in path in the presence of dynamical effects and high speed. After training the system is capable of generating positional commands that in combination with the standard robot controller lead the robot along the desired trajectory. The mean path deviations are reduced to a factor of 20 for our test configuration. For low speed motion the learned controllers' accuracy is in the range of the resolution of the positional encoders. The learned controller does not depend on specific trajectories. It acts as a general controller that can be used for non-recurring tasks as well as for sensor-based planned paths. For repetitive control tasks accuracy can be even increased. Such improvements are caused by a three level structure estimating a simple process model, optimal a posteriori commands, and a suitable feedforward controller, the latter including neural networks for the representation of nonlinear behaviour. The learning system is demonstrated in experiments with a Manutec R2 industrial robot. After training with only two sample trajectories the learned control system is applied to other totally different paths which are executed with high precision as well.     

Telerobotic visual servoing

This paper addresses the problem of integrating the human operator with autonomous robotic visual tracking and servoing modules. A CCD camera is mounted on the end-effector of a robot and the task is to servo around a static or moving rigid target. In manual control mode, the human operator, with the help of a joystick and a monitor, commands robot motions in order to compensate for tracking errors. In shared control mode, the human operator and the autonomous visual tracking modules command motion along orthogonal sets of degrees of freedom. In autonomous control mode, the autonomous visual tracking modules are in full control of the servoing functions. Finally, in traded control mode, the control can be transferred from the autonomous visual modules to the human operator and vice versa. This paper presents an experimental setup where all these different schemes have been tested. Experimental results of all modes of operation are presented and the related issues are discussed. In certain degrees of freedom (DOF) the autonomous modules perform better than the human operator. On the other hand, the human operator can compensate fast for failures in tracking while the autonomous modules fail. Their failure is due to difficulties in encoding an efficient contingency plan.     

An approach to enlarge learning space coverage for robot learningcontrol

In robot learning control, the learning space for executing general motions of multijoint robot manipulators is quite large. Consequently, for most learning schemes, the learning controllers are used as subordinates to conventional controllers or the learning process needs to be repeated each time a new trajectory is encountered, although learning controllers are considered to be capable of generalization. In this paper, we propose an approach for larger learning space coverage in robot learning control. In this approach, a new structure for learning control is proposed to organize information storage via effective memory management. The proposed structure is motivated by the concept of human motor program and consists mainly of a fuzzy system and a cerebellar model articulation controller (CMAC)-type neural network. The fuzzy system is used for governing a number of sampled motions in a class of motions. The CMAC-type neural network is used to generalize the parameters of the fuzzy system, which are appropriate for the governing of the sampled motions, to deal with the whole class of motions. Under this design, in some sense the qualitative fuzzy rules in the fuzzy system are generalized by the CMAC-type neural network and then a larger learning space can be covered. Therefore, the learning effort is dramatically reduced in dealing with a wide range of robot motions, while the learning process is performed only once. Simulations emulating ball carrying under various conditions are presented to demonstrate the effectiveness of the proposed approach     

A study on the use of tactile instructions for developing robot’s motions

Developing motions for humanoid robots is time consuming. However, sport and dance instructors can easily adjust their students?? postures by simple touches. This suggests the possibility of exploiting touch for motion development, and allows us to propose a methodology based on this concept. To realize such a system, it is required to define how the robot should interpret touches. We propose a supervised learning approach to cope with this issue, and verify its feasibility experimentally. We then study the data collected by the algorithm, and show that the system is practical both for motion development and for studying human-robot tactile communication. In particular, we present considerations on the sparsity that characterize the whole process and suggest how sparsity can be exploited for efficient interpretation of tactile instructions.     

The manipulator language with the characteristic of a functional programming language

This article describes the motion-oriented robot language IML (interactive manipulator language) with the characteristic of a functional programming language. The main functions of IML are (1) It is possible to describe the iterative motions without using a loop. (2) The user-defined procedures (commands) can be called by specifying the command name. (3) It is possible to describe robot motion (a sequence of the position and orientation of a robot hand) and force magnitude in the task-oriented Cartesian coordinate system (task coordinate system) suitable for robot tasks. Furthermore, it is possible to describe the translation and rotation of the coordinate system without syntactic distinction. (4) As the teaching data can be easily embedded in the language and can be played back in the force control mode, complex task programming becomes easy. In IML, as the user-defined command and the teaching data can be used just as the builtin system command; the system can be extended easily and naturally. In this article these features are described, and it is shown that these functions can be realized with the unified representation by introducing the concept of a functional programming language. The effectiveness of IML was confirmed by actually making a robot perform some tasks programmed in IML.     

A discrete event approach to the command sequence replanning and control in a teleprogramming system

A command sequence replanning and control method, which enables the slave system to autonomously recover from error conditions, is proposed in a telerobot system. A task model described as a form of the controlled Petri net (CPN) is used as a prior knowledge for the slave system to carry out the given task successfully without the operator's aid even in unexpected error conditions. The CPN model incorporates the contact states and transitional motions between them that possibly exist in the task execution process. The motion command is automatically generated from the master system and transmitted to the slave whenever the contact state changes in the master model. Referring to the CPN model and the given motion commands, the slave system detects if the actual contact state is unexpected and then plans the recovery path from the unexpected state. The feasibility of the command sequence replanning and control algorithm is verified through an example to perform a simple part‐mating task. © 2001 John Wiley & Sons, Inc.     

Robot motion design using bunraku emotional expressions – focusing on Jo-Ha-Kyū in sounds and movements*

One of the UNESCO intangible cultural heritages Bunraku puppets can play one of the most beautiful puppet motions in the world. The Bunraku puppet motions can express emotions without the so-called ‘Uncanny Valley.’ We try to convert these emotional motions into robot affective motions so that robots can interact with human beings more comfortable. In so doing, in the present paper, we present a robot motion design framework using Bunraku affective motions that are based on the so-called ‘Jo-Ha-Kyū,’ and convert a few simple Bunraku motions into a robot motions using one of deep learning methods. Our primitive experiments show that Jo-Ha-Kyū can be incorporated into robot motion design smoothly, and some simple affective robot motions can be designed using our proposed framework.     

Design and implementation of a linear jerk filter for a computerized numerical controller

The proposed filter is placed in front a buffered digital differential analyzer, and is formed by combining three modified moving average filters. It can export the position commands to ensure smooth and accurate motion of a tool with a linear jerk change. These output commands can guarantee contour accuracy despite the error in the chord height. The acceleration and jerk can be designed simply by specifying the number of registers. The filter can be implemented using three circular buffers to simplify the arithmetic and reduce the computation time. The high precision motion commands are confirmed by installing the filtering algorithm in a digital signal processor of a computerized numerical controller. The radius of the command trajectory does not become distorted at high speed 30 m/min. The commands filtered by the linear jerk filter stabilize the beginning and end of the actual motion of the machine table.     

基于语音指令的远程控制机器人系统的设计与实现

通过对语音识别、网络通信与智能机器人运动控制的综合应用,成功地设计和实现了语音远程控制智能机器人的系统。操作者将语音指令输入到客户端,识别后通过socket机制将指令传送给智能机器人,使之按指令完成运动。实验验证了语音远程控制智能机器人运动的准确性和实时性。     

Learning and remembering interactive commands in a text-editing task

Users of interactive computer systems often experience difficulty in learning and remembering the command vocabulary needed to communicate with the system. This study investigates how task and vocabulary differences affect initial learning and subsequent memory for commands used in a simple editing task. Systems with semantically specific terms were learned no more quickly than systems with semantically general terms, but the nature of the command vocabulary induced different learning strategies. Users of the specific vocabulary made less use of help (provided in the form of a command menu and definitions of operations) than did usersof the general command vocabulary. However, users ofthe specific vocabulary appeared to take more time actively considering options before deciding to consult HELP. These strategy differences were reflected in users' memory for the commands and the task operations 2 weeks later. In addition, the learning strategies adopted were dependent on user's predispositions as measured by individual difference questionnaires.     

Motion Sickness Simulation Based on Sensorimotor Control

Sensorimotor control is an essential mechanism for human motions, from involuntary reflex actions to intentional motor skill learning, such as walking, jumping, and swimming. Humans perform various motions according to different task goals and physiological sensory perception; however, most existing computational approaches for motion simulation and generation rarely consider the effects of human perception. The assumption of perfect perception (i.e., no sensory errors) of existing approaches restricts the generated motion types and makes dynamical reactions less realistic. We propose a general framework for sensorimotor control, integrating a balance controller and a vestibular model, to generate perception‐aware motions. By exploiting simulated perception, more natural responses that are closer to human reactions can be generated. For example, motion sickness caused by the impairments in the function of the vestibular system induces postural instability and body sway. Our approach generates physically correct motions and reasonable reactions to external stimuli since the spatial orientation estimation by the vestibular system is essential to preserve balance. We evaluate our framework by demonstrating standing balance on a rotational platform with different angular speeds and duration. The generated motions show that either faster angular speeds or longer rotational duration cause more severe motion sickness. Our results demonstrate that sensorimotor control, integrating human perception and physically‐based control, offers considerable potential for providing more human‐like behaviors, especially for perceptual illusions of human beings, including visual, proprioceptive, and tactile sensations.     

INNOVATIVE INTELLIGENT NAVIGATION SYSTEM BY APPLYING FUZZY LOGIC AND CELL ASSEMBLIES IN RESCUING ROBOT

In this study, an intelligent navigation system is developed by using fuzzy logic and Cell Assemblies (CAs) approaches, together with various sensors, in accordance with mimicking human behavior. Because disaster areas can be extremely dangerous and might cover large areas, it is important to make the robot's movement safer and more efficient. Therefore, two models with different functions are designed. The “robot navigation” model is applied to a narrow or dark place by using a laser range finder whereas the “intelligent cognitive” model uses intelligent direction change in an open area by integrating a vision camera. The intelligent cognitive model that is implemented by CAs with fatiguing Leaky Integrate and Fire (fLIF) neurons can not only absorb the ideals of working and long-term memories to explain the phenomenon of biology neurons, but can also imitate human cognitive processes. Additionally, it is not difficult to combine and expand this model to become multifunctional, which mimics human learning. The main contribution of this study is that the clever combination of sensors significantly reduces the frequency of use of vision and image processing. Furthermore, the combined system can avoid potential risks and efficiently shorten the motion time by importing the vision camera. This system has been tested in several simulated environment schemes and the experimental results have proven that it can produce correct action commands relative to different schemes and can also improve the motion path to effectively meet the requirements of disaster relief.