Search

Abstract McKibben pneumatic actuators (MPAs) are soft actuators that exert tension by applying compressed air to expand a rubber tube. Although electro-pneumatic regulators can control air pressure, most are large and expensive. This study utilizes a dynamic quantizer to control the MPA with a small solenoid valve that can only open and close the valve instead of an electro-pneumatic regulator. A dynamic quantizer is one of the quantizers that converts continuous signals to discrete signals. Our previous study confirmed that tension control of MPA under isometric conditions could be realized using a dynamic quantizer. However, it is often necessary to control the length of the MPA as well as the tension of the MPA. This study implements a dynamic quantizer to control the length of the MPA with a small solenoid valve. Numerical simulations and experimental tests verify the effectiveness of the proposed method. The results of the numerical simulations and experimental tests confirmed that the length of the MPA can be controlled using the dynamic quantizer.

Abstract The McKibben pneumatic actuator (MPA) is a soft actuator used for performing various practical functions in robots. Particularly, many dynamic robots have been realized using MPAs. However, there is a trade-off between torque generated by MPA and the range of motion of the joint. In this study, we focus on the jumping motion of a leg-type robot and use an elliptical pulley whose moment arm changes depending on the robot’s posture. To confirm the effectiveness of the elliptical pulley, the relationship between the knee joint pulley (patella) shape and jumping height was analyzed by simulation, and the shape of the patella maximizing jumping height was determined. It was shown that an elongated elliptical patella shape is more effective for the jumping motion than a circular one. Furthermore, the effectiveness of the analytically determined patella shape was confirmed by experiments using an actual robot.

This study focuses on autonomous robot navigation. In general, robot controllers are designed based on an accurate model of the environment. In contrast to this, we adopted an approach where information is written into the environment and used to operate the robot controller. This method can reduce the computational complexity and energy cost and has the advantage of expanding the system to consider multiple robots. However, the mechanism by which information is to be written into an environment in order to operate the robot in a real space with obstacles or walls has not yet been clarified. Thus, in this study, we adopted acoustic devices in the design to consider situations where robot controllers must interact with obstacles. First, we examined the acoustic field created by interactions between sound waves and obstacles using the finite-difference time-domain method and verified it in the real world. Then, we apply the designed acoustic field to an robot controller. The experimental results demonstrated successful robot navigation. The results also show that complicated sensor systems for the robot are not required to be designed to capture obstacles in working places. At the same time, our approach enable autonomous navigation even when the robot cannot visually recognize the target position.

In this study, we discovered a phenomenon in which a quadruped robot without any sensors or microprocessor can autonomously generate the various gait patterns of animals using actuator characteristics and select the gaits according to the speed. The robot has one DC motor on each limb and a slider-crank mechanism connected to the motor shaft. Since each motor is directly connected to a power supply, the robot only moves its foot on an elliptical trajectory under a constant voltage. Although this robot does not have any computational equipment such as sensors or microprocessors, when we applied a voltage to the motor, each limb begins to adjust its gait autonomously and finally converged to a steady gait pattern. Furthermore, by raising the input voltage from the power supply, the gait changed from a pace to a half-bound, according to the speed, and also we observed various gait patterns, such as a bound or a rotary gallop. We investigated the convergence property of the gaits for several initial states and input voltages and have described detailed experimental results of each gait observed.

A swarm system is a branch of a collection of robotic systems composed of a lot of “simple” robots. In general, group cooperative behaviors require a considerable amount of information, numerous sensors and complicated networks, which makes the designing of their control system much difficult; therefore, a more natural and sensorless-based interaction is expected to aid the building of an easy-to-understand control scheme for elaborate swarm systems. This paper inspired by the collective motions of army ants focuses on a group behavior in which individuals can cooperate with each other using its own body. Army ants cooperate with one another to overcome large gaps, such as the gap between the tree and its leaf, wherein some ants form a pillar by connecting with one another, such that other ants can walk on it to cross the gap. This paper presents a sensorless swarm system in which the mechanical structure and the physical interaction of the system is only utilized. The collaborative obstacle-crossing task is considered, and the robotic swarm system, comprising the individual robot along with the mechanical interaction mechanism, is constructed. Through an obstacle climbing experiment and the passing of a gap, the validity of the mechanical interaction approach is examined.

In the logistic system, it is necessary to reduce the weight of the structure to realize energy-saving and highly efficient conveyance of a stacker crane that conveys packages in a warehouse. On the other hand, a lightweight stacker crane becomes a flexible structure, and its movement and vibration control are very important and has been discussed by many studies from long ago. Many designs of dynamic FF controllers have been discussed. In this paper, we propose model following two degree of freedom control using a simple reference model that can simultaneously control vibration and motion by utilizing the internal integral characteristics of the model. We discuss FF command and reference signal considering constraints of actuator. As an FB controller with 2 degrees of freedom control, vibration sensor-less active control using motor position-controlled FB torque is proposed and its effectiveness is verified by simulation and experiment.

In general, small robots can operate in narrow spaces, and large robots can traverse over rough terrains. Although a few robots with adjustable wheel diameter have been developed, they require multiple actuators for rotating the wheels and changing the wheel diameter. This paper proposes a wheel robot that can adjust the wheel diameter according to the wheel rotation. The robot uses a centrifugal force-aided diameter adjustable mechanism; the wheel diameter is expanded/contracted by a circular pantograph via the centrifugal force. Thus, the wheel diameter can be changed depending on the drive of the wheel. We adopt the circular pantograph mechanism to maintain the round shape of the wheel during expansion and contraction. We conducted experiments with the developed robot and found that the wheel diameter appropriately changes with the wheel acceleration and deceleration. The results show that the proposed mechanism improves the performance when climbing a step and operating in a narrow space. Further, the robot also had mechanical flexibility, which helps reduce the impact when facing a step. These characteristics could be achieved because the wheel was passively and adaptively deformed according to the running environment.

In biological research, various experiments such as behavioral experiments and physiological ones are often conducted with pharmacologically treated animals. In such experiments, it is necessary to inject the same volume of solution into numerous small animals, such as insects to prepare several experimental subjects. However, repeating manual injections is burdensome, and it is also difficult to maintain injection quality and consistency. We have developed a microinjection system that can support and semiautomate the injections of small animals. The system consists of two cameras, a micromanipulator, a syringe pump, and a structural framework all operated from a personal computer to quickly inject the same volume of liquid solutions at the same position and depth into small animals. The microinjection system has sufficient extensibility for it to be used in a variety of applications.

This paper is concerned with the demonstration of scallop-like swimming with the aid of robot toy; using jet propulsion by clapping the shells quickly. Previous works from biological approach have indicated what structures contributes to generate the jet propulsion. This paper tries to perform quick underwater locomotion by mimicking the structures of scallop based on constructive approach: development of scallop-like robot with autonomous generation of jet propulsion. After designing the structure and actuation mechanism by mimicking the scallop, we focus on the effect of the mantle property on swimming. Firstly we confirm that developed robot can perform scallop-like swimming. Secondly we also compare the swimming speed by changing the property of designed mantles.

A McKibben-type pneumatic actuator (MPA) has been widely used for various robots, and dynamic motion with MPA has been realized with a simple control method. However, motion analysis of robot using MPA has not been done enough yet due to complicated output characteristics of MPA. The purpose of our research is to establish a motion analysis method of a robot using MPA to realize more dynamic and stable movement. For that purpose, in this paper, we propose a new model of MPA suitable for motion analysis which is not too complicated. The proposed model is derived from linearly approximating the model proposed in our previous study about the length of MPA, and its coefficient can be theoretically obtained. First, verification experiments of the proposed model were performed on MPA of three lengths, and it was confirmed that the correlation coefficient between the proposed model and experimental results was sufficiently large. Next, we verified the effectiveness of the proposed model when the model was applied to an actual robot.MPA was mounted on a two-dimensional leg robot, realizing bending and stretching exercise was performed. It is confirmed that high tracking performance could be realized by pressure inputs design based on the proposed model. The results of these verification experiments indicate that the proposed model is sufficiently effective.

Torque-unit Manipulator(TUM) is a design concept of (space) manipulator and a Multibody system with nonholonomic constraints. Each joint of TUM is free and a device which is called “torque unit” is attached on each link. The “torque unit” can be made easily of a rotary actuator and a disc. Position-controllability of the links of TUM has been shown with conventional control laws for manipulators. However a problem arises that the discs in torque units usually have residual angular velocities when the links approach desired positions. The reason of the problem is that TUM is a kind of nonholonomic system, hence the angular velocity of each disc at each time depends on the motion history of the links. Controllability of the angular velocities of the discs to desired values has been considered by planning motion trajectories of links. Then, a control scheme for the TUM from zero-velocity initial state to zero-velocity final state has been given. However the trajectories of the links were not smooth enough. Then, a condition is made clear for the trajectories of the links to satisfy for bringing the angular velocity of a disc to zero in this paper. Also trajectories of the links are planned which is smooth enough as an example of application of the condition.

This paper tries to clarify how rocking motion and body structure effect on the emergence of walking and the changes of the gait such as walking speed and the length of stride. Humans naturally exhibit adaptive performance in various environments. Moreover, they are able to adjust the gait for their purposes. The adaptive locomotion is generated not only with leg swing motion, but also the lateral rocking of the body, rotation of the center of mass, etc. Most previous researches have mainly mentioned the direct drive of the legs. On the other hand, one previous research has indicated interesting motion observation that the body motion in lateral plane will get to a dominant factor in the phase of low speed walking. According to the previous result, we propose an idea that the walking can be generated by an excitation of lateral rocking of the body without direct actuations of legs. Therefore, in this paper, we especially focus our attention on three-dimensional quasi-passive walking provoked by the rocking motion of the robot body in lateral plane. After developing a robot with an oscillator that excites rocking motion of the robot body in lateral plane, we confirmed the emergence of walking in the experiment. Next, we investigated the gait difference by changing the parameters of the rocking motion (amplitude, period) and the robot body structures (length of the leg, offset of the legs to forward). The experimental analyses indicate that the walking speed, robustness and consumption of energy are mainly correlation to the period of lateral rocking and the offset of the legs to forward. In summery, an appropriate combination between the period of rocking motion and structures of the robot body leads to control of the gait.

In this research, we study on posture control of a two-link torque unit manipulator under the condition of viscous friction. The torque unit is a module which consists of an electric motor and a reaction wheel, and it can be set at any arbitrary position as actuators to control a rotating link system with free joints. In this paper, we propose a control method using a kinetic energy of the reaction wheel. The viscous friction cannot be ignored in the real system. There arises a problem such that angular momentum remains in the reaction wheel of the torque units by the influence of the viscous friction. Then, we clarify the relation between angular momentum of the reaction wheels and rotation angles of the links. First, we try to control the first link to an arbitrary reference position without regard for the angle of the second link. The angle of the first link can be controlled by changing the viscous friction between the first link and the base. Next, we try to control the second link to an arbitrary reference position by using the residual angular momentum of the reaction wheel on the second link. Finally, we propose a control method based on these ideas, and we confirm the feasibility of the proposed method through the numerical simulations.

In this paper, we investigate gathering control of autonomous multiple agents by single mobile robot. This work was motivated by so-called sheepdog harnessing, in which a mobile robot is developed to gather a flock of ducks or sheep and maneuver them to a specified goal position. Based on discrete model of sheepdog system, we apply two different types of control methods to the system. After defining some evaluation indices (harness time, dispersion of flock), we analyze relationship among the parameters of the robot, the sheep and the control algorithms (center-targeting control, tangent-targeting control) by numerical simulations. Moreover, we propose an improved control method to maneuver separate groups of sheep to a goal position based on on-line clustering.