In this study, we developed a human attention model for smooth human-robot interaction. The model consists of the saliency map generation module and manipulation map generation module. The manipulation map describes top-down factors, such as human face, hands and gaze in the input image. To evaluate the proposed model, we applied the model to a magic video, and measured human gaze points during watching the video. Based on the experimental results, the proposed model can better explain human attention than the saliency map.

For smooth interaction between human and robot, the robot should have an ability to manipulate human attention and behaviors. In this study, we developed a visual attention model for manipulating human attention by a robot. The model consists of two modules, such as the saliency map generation module and manipulation map generation module. The saliency map describes the bottom-up effect of visual stimuli on human attention and the manipulation map describes the top-down effect of face, hands and gaze. In order to evaluate the proposed attention model, we measured human gaze points during watching a magic video, and applied the attention model to the video. Based on the result of this experiment, the proposed attention model can better explain human visual attention than the original saliency map.

Recently, flexible and light actuators that mimic muscle fibers have been actively researched. Here, we focused on an electroactive polymer (EAP) dielectric elastomer. To construct the actuator, the dielectric elastomer is rolled into a tube, and extended by applying a voltage across its electrodes. In this paper, we experimentally obtained a static characteristic model of the electrostrictive rubber actuator, and proposed a control method. We then incorporated the actuator into a 1-DOF manipulator, and constructed a controller from a mechanical equilibrium model of this manipulator. Finally, the constructed controller was tested in a series of experiments. The angle of the arm fell below the desired angle because of friction in the joint and the spring characteristics of actuators. However, we confirmed that the controller reduced the influence of the load by torque feedback. We also investigated the influence of joint stiffness.

A teleoperation system for a crane is developed. The crane uses a small actuator and large actuator in each direction (x and y). The large actuator is controlled by input signals from a laser pointer interface, and the small actuator is used as a regulator and is feedback-controlled by a wire angle sensor. To avoid overshoots in the suspended object at the target point and to eliminate the effect of the trembling of the irradiated light by the laser pointer, the velocity commands sent to the large actuator are designed by accumulating cubic functions, each of which is based on the displacement of the irradiated light in a sampling period. The effectiveness of the developed control system and its control algorithm were verified through experimental results.

A time optimal control algorithm for quadruped robots is developed and its effectiveness is verified by experiments and simulations. The derivation of the optimal solution consists of two steps that are the speeding up of supporting legs and that of swinging legs. The way of the speeding up of swinging legs is very similar to that of a two link manipulator developed by Bobrow. In this paper, first, a method of using torque redundancy of the robot is proposed for speeding up of supporting legs with the constraints due to ZMP, the power limits of joint actuators and the limits of available friction forces. Then, an algorithm for obtaining the fastest walking pattern in a trot gait period is developed by combining the result of the supporting legs and that of the swinging legs. The obtained walking pattern is installed in a quadruped robot SONY ERS-7 and its effectiveness is verified by experiments.

A tele-operation system for a crane is developed. The crane consists of a small actuator and a large actuator in x and y direction respectively. The large actuators are controlled by input signals from an operator and the small actuators are used as a regulator and feedback controlled by a wire angle sensor. Since a motion of the suspended object can be approximated as a second order system, overshoots of the object inevitably happen at the moment when the operator give a stop command to the system. To avoid this overshoots, a novel user interface is developed where operators indicate via points or target point of the object on a monitor screen from the camera imaging the scene below the crane. The system interpolates the via points by using four order functions and give the velocity commands to the crane until the inputs of the via points are stopped. The final point is regarded as the target point. From experimental results, the effectiveness of the developed control system and its control algorithm is verified.

A reliable control method is proposed for multiple loop control systems. If a feedback loop fails (e.g., as a result of a sensor breakdown), the control system develops an unstable fluctuation. To cope with this problem, the proposed method uses an equivalent transfer function (ETF) for active redundancy compensation after loop failure. The ETF is used to maintain the transfer function of the entire system the same before and after loop failure. In this study, the characteristics of a reliable control system that uses an ETF were examined experimentally. A position servo system including a disturbance observer is used in the experiment. In case of feedback loop failure, the control system could not be stabilized using the disturbance observer alone; however, the ETF stabilized the control system after feedback loop failure.

An omni-directional moving platform with four wheels is developed and its controller is designed for tracking desired trajectories. The four wheels of the platform are torque controlled by DC servo motors independently. The generalized coordinate of the platform is three, therefore the system has one torque redundancy and one odometry sensor redundancy which can be used for improving the control performance of the moving platform. First, ways of exploiting odometry sensor redundancy is discussed. Then a criterion for exploiting the torque redundancy is proposed to make larger the margin of driving force, and fundamental simulations and experiments are done to show the control characteristics of the developed mobile platform.

A force control device with high power and high resolution is developed. It consists of a spring and a small force actuator which are connected in parallel between the base plate and the output plate. By installing this device into a tip of a high power industrial robot, large forces with high resolution can be achieved under the existence of static friction. In this paper, the principle of force control device is verified by using a simple two d.o.f. experimental setup. From experimental results, the effectiveness of the developed control device and its control algorithm is verified.

In this paper, the sampling real-time QMDP value method is proposed and applied to a goalkeeper task of a soccer robot. Soccer is a challenging task for an autonomous robot that has poor computing resources and sensors, and a good subject in the study of decision making in dynamic environments. A robot frequently decides its behavior without enough observation of the environment. In the proposed method, the risk of a score is solved beforehand toward every set of position of the goalkeeper and motion of the ball. The shortage of observation is calculated and represented by particle filters. The proposed method uses the risk function and the particle filters so as to choose appropriate behavior of the goalkeeper. The experiment with an actual robot suggests that the method can decide actions reflexively toward the motion of the ball.

Time optimal control method for quadruped walking robots are developed and installed into a practical robot system. Each leg is modeled as a two link manipulator whose time optimal control theory has already been established by Bobrow. However, in legged robot systems, each leg supports their body weight, and the reaction forces from its ground must be inside of their friction corn. Moreover, the ZMP (zero moment point) of the robot is constrained for stable walk. Therefore, time optimal control inputs must be designed considering these constraints. SONY ERS-7 is used as a quadruped walking robot and a fundamental experiment is done. From the experimental results, the effectiveness of the developed control algorithm is verified

Two control methods of controlling a redundant manipulator mounted on a base plate suspended by six wires are proposed. By suspending the base plate by six linearly independent wires, the plate doesn't swing even if the reaction forces are applied to the plate from the moving manipulator. However, if the reaction forces become too large due to fast motion, some wires become loose and the base plate may turn over. To avoid the overturn, two ways to exploit the redundancy of the manipulator are proposed and their effectiveness is verified by simulations and experiments

We have proposed the real-time QMDP method for decision making of a robot under uncertain state recognition. This method evaluates every action and chooses the best one with a particle filter for estimation and a state-value function of dynamic programming. Different from our past work, this paper applies it to a complicated decision making task that yields local maxima and discontinuity on the state-value function. We then verify whether the method can choose proper actions or not in such a condition. As an example, total behavior of a goalkeeper for robot soccer is planned by using value iteration. This task contains three strategies, which are related to three kinds of local maxima respectively. Simulations, experiments and actual games have suggested that the method can decide actions effectively according as uncertain result of state estimation.

A sensor system for measuring tilting angles of a wire which suspends a swinging mass is developed. Two CCD cameras are placed toward the wire on a horizontal plane, and obtain the images of the moving wire respectively. Through the simultaneously obtained two wire images, the angles between the wire and a vertical line are calculated. By feedback control using wire angles measured by the developed sensor, a mass suspended by a two dimensional crane on a horizontal plane is positioned without residual vibration at a target point. Since the developed sensor system is placed near the upper end of the wire regardless of wire length, it is easy to install the developed system in conventional cranes. By using this sensor system, it is expected to automate heavy object positioning, handling and so on.

As a part of research on autonomous loading system by wheel loader, a method for appropriate arrangement of bucket trajectory for scooping motion will be proposed. In this paper, relation between resistance force and advancing direction of the bucket is analyzed theoretically. Advancing direction is dominant factor for resistance force on the bucket during scooping. Based on this analysis, simple rules for bucket trajectory arrangement are proposed. In scooping procedure, scooped volume is estimated using 3D model obtained with stereo-vision system. The shape of the pile is measured and converted into pile model prior to scooping. The developed method and system are installed on experimental scale model of wheel loader. The results show good performance for different condition of pile.

A method of acquiring reaction force distributions on a sole of a walking humanoid robot from ground is developed. The tactile sensors are embedded on the soles of a humanoid robot for the acquisition of the ground reaction force distributions. The tactile sensor consists of a pressure conductive rubber sheet put between upper and lower electrode arrays and was able to measure the reaction forces from 0.00 MPa to 0.05 MPa at all of the sensing elements. Force distribution data including 30 /spl times/ 16 points can be obtained at a period of 20 ms by using RTLinux. It is shown that the ground reaction force distributions can experimentally be obtained by a humanoid robot "HOAP-I" and the locus of ZMP along walking of a humanoid robot is given. This information is useful for constructing feedback control loops or generating a new walking gait.

In order for a legged mobile manipulator to work, it is essential to change its posture for stable standing in response to the force at the end effector. We define 'firm standing' as keeping a stable posture in working without moving the positions of the point of contact of legs and end effector. We analyze firm standing and show required degree of freedom for it. We also experiment on it by using redundant d.o.f. in 2-D space.

Focuses on a method of estimating contact force dynamically between environment and an object, which is grasped by the soft planar fingers with tactile sensors. The external contact force acting on the grasped object causes the tactile sensors to generate the displacement distributions and the force distributions. We present a method of dynamic tactual image acquisitions for both distributions at the pseudo video rate. Experimental results show the measurement of the tactual image flows due to the deflection of an object and the magnitude and orientation of force vectors operating on the finger according to the external contact force. The contact position of the external force acting on the grasped object is estimated within 15% from the tactual image distributions.

A control method of positioning a serial link manipulator suspended by a crane type robot with multiple wires is proposed. The manipulator is fixed on a base plate suspended by three wires. Since the manipulator is suspended by wires, the position and orientation of the base may change according to external forces applied to the manipulator or the change of the manipulator pose. Thus, the inverse kinematics becomes very complicated, and moreover, the mechanism does not seem to be suitable for tasks requiring manipulator tip forces. In this paper, it is shown that the inverse kinematics becomes very simple by using a redundant DOF manipulator, and a method of exploiting the manipulator for tasks requiring forces is described.

Describes the sensing behavior of a high compliant tactile sensor on an object's surface and a formulation of the tactual sensing using methods of differential geometry. The tactile sensor has been manufactured with a function of high compliance, which is achieved with silicone rubber placed on a sensing mechanism organized for the detection of contact patterns. The mechanism is constructed as an optical wave guide which is capable of transmitting the contact pattern to a CCD camera. From the contact pattern, the tactile sensor detects the displacement distribution and the force distribution corresponding to the object's surface shape. By using a method derived from differential geometry, the geometrical contact mechanisms between the sensor and the surface of an object are discussed. The first fundamental form and the second fundamental form are defined by the displacement distribution generated by the tactual sensing. The local shape of the object's surface in, the neighborhood of a contact point is determined from the Gaussian curvature and the mean curvature given by the first fundamental form and the second fundamental form defined by the displacement distribution. Experimental results for estimating an object's local surface shape are presented for a tactile sensor manufactured for trial experiments.

A method of determining optimal grasping forces for a parallel two-fingered hand with compliant 2D tactile sensors is described. The tactile sensors can get 2D patterns of pressure distribution and have many potential applications such as shape recognition of a grasped object, assembly tasks and so on. In order to achieve assembly tasks using robot hands, information about the position of contact point and the contact force between the grasped object and its environment is essential. When contact force is applied to a grasped object from the environment, the pressure patterns of tactile sensors change according to not only amplitude of the applied force but also the contact position. Thus, a way to derive the contact position and force by tactile sensor data has been studied until now, and it is proven that the maximum measurable contact force from the environment depends on not only sensor characteristics but also grasping force. In the paper a way to determine the optimal grasping force to measure the largest contact force is proposed and some fundamental experiments for testing the derived grasping forces are done.

This paper proposes methods of fusing tactual and visual data for a hand-eye cooperation system, and realizes the methods using a developed tactile sensor. Since tactual and visual data convey information with different characteristics, smart fusion of these two sensor data is expected to provide useful additional information. In this paper, visually identified edges and planar regions or edges from tactual data are utilized and three fusion methods are introduced (1) measurement of a 3D edge when a tactile sensor contacts a plane, (2) measurement of a 3D edge when a tactile sensor contacts a plane at the edge and (3) measurements of a cylinder when a tactile sensor contacts the cylinder at an edge.

Presents a method of sensor fusion with aspect information acquired with visual and tactile sensing. The sensor fusion system is made up on the geodesic dome for the visual and tactile sensing system. The visual sensing system discriminates a position and an orientation of the CCD camera on the geodesic dome. If an image acquired with the CCD camera is identified for the model images at the facet in the geodesic dome, the position and orientation of the camera corresponds with the facet on the geodesic dome. Then, the camera and object are represented with the geometrical relationships in the coordinated system. The tactual images are given with the tactile sensor which makes contact with the object surface at the touching direction of a finger. If the position and orientation of the hand are given, the tactile images on the finger are given with the estimation by computer simulation.

This paper proposes a simple method for handling objects by cooperation of multiple position-controlled robots. In order to use position controllers for cooperation of multiple robots, it is necessary to install mechanical compliance in each robot to avoid excessive inner forces due to the mutual positioning errors. First, the compliance needed for the cooperation of position-controlled robots is reviewed. Using compliant mechanisms with free joints, cooperative control system between two industrial robots are developed. As the cooperative system with mechanical compliance has kinematic redundancy, a way to use the redundancy is proposed. Some fundamental experiments verify the effectiveness of the proposed strategy.

This paper proposes a simple method for carrying a large object by cooperation of multiple mobile manipulators with position controllers. It is necessary for each robot to have compliance in order to avoid excessive inner forces due to the mutual positioning errors. Thus manipulators on mobile platforms are used as free joint mechanisms by locking some joints and making the rest joints free. These free joints play the role of mechanical compliance. First, compliance needed for cooperation among position-controlled robots is shown. Second, the statics and error characteristics of cooperative systems including compliant mechanisms are analyzed. Some feedback control laws for platforms moving on rough ground are proposed and discussed.

This paper proposes a new strategy for manipulation of a rigid object by cooperation of multiple position-controlled manipulators using force control devices. The force control device consists of springs and a position-controlled micro-manipulator. Force control is achieved by controlling deflections of each spring by the micro-manipulator. A cooperative control algorithm considering the flexibility of macro/micro manipulators is proposed. After the differential equation of the error vector system is derived, a method of designing feedback gain matrices is shown for linearizing the system. To verify the efficiency of the algorithm, experiments are done on carrying an object by two industrial robots.

This paper proposes a simple method for handling a single object by cooperation of two manipulators controlled by PID position controllers. Flexibility is essential to avoid excessive inner forces caused by the mutual positioning error between two manipulators. First, the way to design these flexible mechanisms is shown. Second, the closed loop structure consisting of two manipulators and the installed flexible mechanisms is kinematically analyzed. As an example, a cooperative system between a 6-DOF robot and a conventional crane is designed and an inverse kinematics algorithm is proposed. The method proposed here is also very effective for cooperation in distributed autonomous robotic systems.>

This paper proposes a strategy for handling a single object by cooperation of multiple robots controlled by PID position controllers. Since such a cooperative system has mechanical closed loops, installation of free joints or flexibility among robots is necessary to avoid excessive inner forces caused by the positioning error of each robot. First, the essential kinematic conditions to achieve the cooperation among position-controlled robots are shown. Second, the relationship between the positioning error of the handling object and that of each robot is analyzed. Finally, a mechanism for the cooperation of three mobile robots is proposed and its kinematic characteristics and statics are analyzed. The strategy proposed here is very effective and practical for cooperation of not only conventional industrial robots but also distributed autonomous mobile robot systems with non-holonomic constraints.>

The manipulability for wire suspended mechanisms is discussed. For kinematics of a wire driven mechanism, we need not only the geometrical constraints of wire lengths but also the force constraints, such as the wire tension which should be always greater than or equal to zero. First, the sufficient and necessary mechanical conditions corresponding to the force constraints are shown. Under these conditions, we can obtain the manipulability of a wire driven mechanism using almost the same methods as for a popular multilink manipulator. Second, a more practical evaluation index "set of manipulating forces" is introduced and the way of calculation of this index is explained. We also derive both manipulability indexes as for wire suspended mechanisms in the gravity field by modeling the gravitational force as a wire having tension corresponding to gravity. Numerical examples are given to explain the proposed indexes and to show their validity. The method is expected to be effective in designing and task planning of wire driven systems including wire suspended mechanisms.>

This paper proposes a new algorithm for the manipulation of a rigid object by cooperative control of multiple flexible manipulators. Assuming that the flexibility exists only at the force sensor attached to the tip of the manipulator, each manipulator is modeled as a rigid manipulator with six springs at the tip. A cooperative control algorithm considering the flexibility is proposed. After the differential equation of the error vector of the system is derived, the method of determining the feedback gain matrices is shown for linearizing the system. To verify the efficiency of the algorithm, the experiments are done on carrying an object held by two parallel flexible beams.

The dynamics of a robot is highly non- linear because of coupling torques between joints. Since coupling torques result in disturbance to control system of a robot, the elimination of the coupled part of the torques has been studied in robot control. On the contrary, in this paper, we propose a method of utilizing the coupling torques for the the purpose of control of robots. We intro- duce a concept of dynamic redundant degrees of freedom, which can modify characteristics of a robot dynamics task-independently. We utilize dynamic redundant degrees of freedom both to minimize joint torques in the least square sense or to generate great momentary force at the end-effec- tor of a robot. The effectiveness of utilizing dynamic redundancy'' is verified by both simula- tions and experiments.