In-hand manipulation has attracted attention because of its potential for performing dexterous manipulation tasks. Few successful examples using real robotic fingers have been reported because model-based approaches have been assumed. A gradient descent-based iterative learning control is one of the typical methods for improving the control performance without the need for a precise model. However, the learning performances deteriorate greatly owing to the stochastic uncertainties, and the learning rates have to be determined manually. We propose a novel iterative learning scheme with adaptive learning rate methods for dexterous in-hand manipulation. The proposed scheme not only eliminates the need for a precise model and manual tuning of a learning rate but also is robust to stochastic uncertainties and insensitive to hyperparameters. The validity of the proposed iterative learning scheme is demonstrated through several experiments.

In this study, we propose a novel iterative learning scheme, which generates time-series data comprising the impedance value for human-robot cooperative work, where the human operator moves the end-effector from an initial position to a goal position. The proposed learning scheme iteratively updates time-series data comprising the variable impedance value to minimize a task-oriented cost function. Two types of noise reduction technique are used in the proposed learning scheme, which reduce the noise in the time direction and in the trial direction. The validity of the proposed method was verified based on experiments.

This paper proposes an iterative learning control scheme for whole-arm object manipulation by using a robotic hand system composed of a velocity-controlled finger and a torque-controlled finger. The novel iterative learning control scheme consists of the PD-type iterative learning control scheme for the velocity-controlled finger to follow the desired trajectory and the optimization-based iterative learning control scheme for the torque-controlled finger to maintain equilibrium. The switching between the two controllers is made to achieve compliant manipulation. The validity of the proposed scheme is demonstrated through experiments.

The present paper discusses a planning and control strategy for robotic whole-arm manipulation of a slippery polygonal object by taking into account the estimated bounds of the frictional coefficient. In this study, randomized planning methods are first proposed in order to generate contact state transitions. Second, a novel control strategy that can switch manipulation between quasi-static, dynamic, and caging manipulation modes is proposed. Solving the forward dynamics problem, which is formulated as a complementarity problem, generates the desired trajectory and thus ensures that undesirable contact modes do not occur within the estimated bounds of the frictional coefficient. The validity of the proposed methods is then demonstrated through simulations and experiments.

This paper reports on our initial efforts to achieve robotic nonprehensile catching. First, we show the importance of nonprehensile catching in a robotic catching task. Robotic nonprehensile catching can be achieved by caging or partial caging with gravity. Since an object is restricted within a bounded region via nonprehensile catching, accurate finger positioning is not required because the robotic hand should be moved to an appropriate region where caging is achieved. Second, control strategies for nonprehensile catching are proposed. The robotic arm motion is generated on-line by using visual sensory feedback in response to the object motion. Noting that the object passes from the outside to the inside of the workspace of the robotic arm, the motion phases of which are divided into tracking and catching based on the specified region in a 3-D space. Finally, the validity of the resulting methods is demonstrated by experiments.

The present paper introduces a planning and control strategy for whole-arm manipulation of a slippery polygonal object. Randomized planning methods are first proposed in order to generate contact state transitions, which help not only to reduce the amount of calculation required, but also to handle a hybrid system composed of a continuous system and a discrete system, which has a large search space and complicated constraints. Second, a novel control strategy, which can switch manipulation modes among quasi-static, dynamic, and caging manipulation depending on the situation, is proposed. This strategy not only can cope with changes in the mechanics of the system caused by contact state transitions, but also can increase the manipulation feasibility and reliability. The validity of the proposed methods is verified through simulations and experiments.

The present paper proposes the learning control method for the throwing manipulation which can control not only the position but also the orientation of the polygonal object more accurately and robustly by low-degree-of-freedom robotic arm. We show experimentally the validity of the proposed control method with the one-degree-freedom robotic arm. We also demonstrate the usefulness of the throwing manipulation by applying it to sorting task and assembly task on experiments.

This paper proposes a throwing manipulation strategy for a robot with one revolute joint. The throwing manipulation enables the robot not only to manipulate the object to outside of the movable range of the robot, but also to control the position of the object arbitrarily in the vertical plane even though the robot has only one degree of freedom. In the throwing manipulation, the robot motion is dynamic and quick, and the contact state between the robot and the object changes. These make it difficult to obtain the exact model and solve its inverse problem. In addition, since the throwing manipulation requires more powerful actuators than the static manipulation, we should set the control input by taking consideration of the performance limits of the actuators. The present paper proposes the control strategy based on the iteration optimization learning to overcome the above problems and verifies its effectiveness experimentally.

The present paper proposes the motion planning based on “the dynamics shaping” for a robotic arm to hit the target robustly toward the desired direction, of which the concept is to shape the robot dynamics appropriately in order to accomplish the desired motion. According to the linear system theory, the positional error of the end-point converges onto near the singular vector corresponding to its maximum singular value of the output controllability matrix of the robotic arm. Therefore, if we can control the direction of the singular vector by applying the dynamics shaping, we will be able to control the direction of the positional error of the end-effector caused by the disturbance. We propose a novel motion planning based on the dynamics shaping and verify numerically and experimentally that the robotic arm can robustly hit the target toward the desired direction with a simple open-loop control system even though the disturbance is applied.

The present paper introduces an evaluation of the manipulation performance of a robotic arm with respect to control accuracy and mechanical efficiency, taking into consideration the effects of gravity and the dynamic process between inputs and outputs in the robotic system. A measure based on the output controllability ellipsoid is proposed, which shows the relationship between the end-effector motion and the joint driving force. Computer simulations demonstrate the effects of gravity on the manipulation performance and the benefits of the proposed measure for the trajectory planning.

We have proposed a self-reconfigurable parallel robot, which can be configured to 4R and 5R closed kinematic chains. This paper proposes a parallel mechanism mobile robot by mounting it on a crawler mechanism. The combined mobile robot gains some useful functionalities other than just locomotion as an advantage of its parallel mechanism, such as getting over a bump by controlling its center of gravity and carrying an object by making use of its shape. Furthermore, the robot can form three-dimensional structures with other such robots and reach a certain height which the single one cannot. We have developed two such robots that are self-contained. This paper analyzes the motions of the functionalities and verifies them experimentally using the robots.

We have developed a self-reconfigurable robot which can form a 5R closed kinematic chain, only two of whose joints are actuated. This paper discusses its dynamic rolling. When it rolls, it has one DOF of its absolute orientation besides two DOFs of its shape. We show that the absolute orientation is subject to an acceleration constraint, not a velocity constraint. Therefore, the dynamics of the rolling motion needs to be formulated to control it. This paper proposes a controller which can reduce its negative acceleration caused by gravity. The shape and orientation of the robot cannot be controlled simultaneously. This paper proposes a control strategy switching shape and orientation controllers. We verify the effectiveness of the control strategy by simulations and experiments.