When manipulating an object with multiple effectors such as in multidigit grasping or multiagent collaboration, forces and torques (i.e., wrench) applied to the object at different contact points generally do not fully contribute to the resultant object wrench, but partly compensate each other. The current literature, however, lacks a physically plausible decomposition of the applied wrench into its manipulation and internal components. We formulate the wrench decomposition as a convex optimization problem, minimizing the Euclidean norms of manipulation forces and torques. Physical plausibility in the optimization solution is ensured by constraining the internal and manipulation wrench by the applied wrench. We analyze specific cases of three-fingered grasping and 2-D beam manipulation, and show the applicability of our method to general object manipulation with multiple effectors. The wrench decomposition method is then extended to quantification of measures that are important in evaluating physical human-human and human-robot interaction tasks. We validate our approach via comparison to the state of the art in simulation and via application to a human-human object transport study.

While convolutional neural networks (CNNs) have been successfully applied to many challenging classification applications, they typically require large datasets for training. When the availability of labeled data is limited, data augmentation is a critical preprocessing step for CNNs. However, data augmentation for wearable sensor data has not been deeply investigated yet.

For seamless physical human-robot interaction (pHRI), estimating human intention is essential. Most system identification approaches to pHRI model the human as a black box without prior assumptions about the underlying behavioral structure. However, integrating a priori knowledge about behavioral characteristics of the human provides superior prediction performance. In this work we present a novel method for human behavior prediction during physical interaction that incorporates an empirically supported human motor control model. The arm dynamics of the human are modeled as a mechanical impedance that follows a latent desired trajectory. We adopt a Bayesian perspective setting Gaussian Process (GP) priors on impedance parameters and the desired trajectory, which allows regression about human behavior from observed trajectories and interaction forces. The proposed impedance-based GP model is validated in simulation and in an experiment with human participants to demonstrate its prediction performance and generalization capability.

Our emotion estimation refers to estimation of a reader's sentiments and feelings after they read a narrative (i.e., post-reading emotion types). The emotion types manifest in a reader's feelings combine to form diverse and various patterns. To elicit this phenomenon, we here attempt to estimate readers' emotional patterns based on their cognitive appraisals of the work. This paper describes our cognitive appraisal methodology, variation in readers' emotion types due to differences in their cognitive appraisals, and report and discuss the results of emotion estimation experiments in which a reader performs a cognitive appraisal of the text of a narrative.

While research has demonstrated how vibrotactile devices can be effectively used to guide human behavior, efficient mappings of vibration patterns for spatial guidance in time-critical dynamical tasks have not yet been understood. In this paper, we contrast two types of action-dependent, haptic stimulus designs to demonstrate the different effects of vibrotactile feedback on the human control performance. A wireless bracelet is used to provide patterns of vibrotactile stimuli in real-time, representing either optimal hand velocity or acceleration for the stabilization of an inverted pendulum. The optimal control behavior is supplied by a linear quadratic regulator. The analyses of the participants' stabilization and learning behavior revealed a significant improvement caused by the additional velocity-dependent feedback. The results are consistent with previous research, which indicates that the human sensory-motor system is generally more sensitive to velocity than acceleration information. In summary, the present paper suggests how human-centric vibrotactile stimuli should be designed and how they can be effectively transmitted to the human user for time-critical behavioral guidance.

This article presents the implementation and experimental validation of a control system dedicated to human robot physical interaction during object handovers. Our control law defines the trajectory of the robotic arm towards a previously unknown handover location based on the Dynamic Movement Primitives formalism adapted for human robot object handover. The control law was deployed on a Kuka Light-Weight Arm equipped with the Azzurra anthropomorphic hand. We employed an industrial-like setting involving three different human postures for object handover in order to evaluate the performance and user experience of our control law and its generalizability. The evaluation was conducted over 1000 object handover trials between the human and robot partners, and the kinematic data and subjective experience were gathered for each trial. The outcomes of this evaluation validate the current implementation and guide the next steps towards more efficient and fluid human robot interaction.

This article presents the design and execution of the experiments used to develop and evaluate a robot prototype system for fluent Human-Robot object handover interactions. A key aspect of our experimental methodology is the deep integration between Human-Robot and Human-Human object handover experiments. This provides a solid baseline and knowledge base for the prototype evaluation, both in terms of movement dynamics and in subjective user evaluation.

The consultation algorithm for computer shogi determines each move by voting of multiple computer shogi players. In this research, we discuss some methods that optimize weights of player vote during a shogi game. To change the weight of player's vote, we use two kinds of criterion. One is the adoption times of each player vote and the other is evaluation value of the board position. Experimental results suggest the validity of our proposed method.

Grasping is one of the most common forms of dexterity. So far, most research has focused on slow-varying loads which can be resisted by anticipatory grip adjustments. There are common cases, however, when a rapid, unexpected increase in the load occurs and where the central nervous system must re-adjust the grip dynamically to prevent slippage. During such events, the central nervous system reactively updates the grip force to minimize further escape of an object. While existing theories postulates that the shear strain of the finger pads caused by the load force is a primary source of information for detecting a new load condition, vibrations induced by even minute object slip in the hand might more effectively signal the occurrence of unwanted movement of the object relatively to the hand. With the help of a high-sensitivity force sensor interposed in the load-path of a fast traction-creating device, we recorded the fluctuations of the force projected onto the fingertip when a rapid perturbation was applied to a grasped object. These fluctuations are indicative of slip. The results highlight the existence of a correlation between the amplitude of the vibrations and the grip force modulation, when textural features are present. The study provides promising evidence that the central nervous system exploits vibrations to detect the onset of unwanted movement of an object relatively to the hand to optimally scale the grip force in response to unexpected, rapid load variations.

This article presents the current state of an ongoing work on Human-Robot interaction in which two partners collaborate during an object hand-over interaction. The manipulator control is based on the Dynamic Movement Primitives model, specialized for the object hand-over context. The proposed modifications enable finer control of the dynamic of the DMP to align it to human control strategies, where the contributions of the feedforward and feedback parts of the control are different to the original DMP formulation. Furthermore, the proposed scheme handles moving goals. With these two modifications, the model no longer requires an explicit estimation of the exchange position and it can generate motion purely reactively to the instantaneous position of the human hand. The quality of the control system is evaluated through an extensive comparison with ground truth data related to the object interaction between two humans acquired in the context of the European project CogLaboration which envisages an application in an industrial setting.

Tagging has emerged as a powerful mechanism that enables users to find and understand entities. However, there are three types of issues in traditional tagging systems. In this paper, we explore and seek a tag-algorithm that predicts tags of users and contents with a degree of relevance, which we called the tag ratio. We described our algorithm and evaluated them by the Naive Bayes classifier. Experiment results showed that all rating's precision of sought continuous number content's tag were better than raw binary valued content's tag's.

This paper shows a new design and prototyping method for a bending pneumatic rubber actuator and its application to a soft-bodied manta swimming robot. The design is based on optimal design using non-linear finite element method, in which geometrical and material non-linearity are considered and fabrication process is based on a rapid and efficient prototyping system using a CAD/CAM based rubber molding process. In this paper, the characteristics of several possible actuators are analyzed and evaluated to lead to an optimal actuator design. The actuator works very well with smooth and soft motion. The manta swimming robot in which the developed actuators are embedded is also designed based on non-linear finite element method. The developed manta swimming robot is made only of rubber and it swims in water smoothly as if it was a living fish. The experimental results of the manta robot motion show that good agreement with those of analytical results.

Recently, there has been significant research interest in homogeneous modular robots that can transform (i.e. reconfigure their overall shape). However, many of the proposed transformation mechanisms are too expensive and complex to be practical. The transformation process is also normally slow, and therefore the mechanisms are not suitable for situations where frequent, quick reconfiguration is required. To solve these problems, we have studied a transformable mobile robot composed of multiple homogeneous gear-type units. Each unit has only one actuator and cannot move independently. But when engaged in a swarm configuration, units are able to move rapidly by rotating around one another. The most important problem encountered when developing our multi-module robot was determining how units should join together. We designed a passive attachment mechanism that employs a single, six-pole magnet carried by each unit. Motion principles for the swarm were confirmed in simulation, and based on these results we constructed a series of hardware prototypes. In our teleoperation experiments we verified that a powered unit can easily transfer from one stationary unit to another, and that the swarm can move quickly in any direction while transforming.

This paper describes the compiler that we have developed in this study for high-speed simulation of cellular automata (CA). CA is a powerful method to understand and simulate the behavior of complex systems. CA is temporal and spatial discrete system. Updating many cells' state valves at high speed is needed for a high precision simulation. SIMD instructions can accelerate CA simulations because each cell can be updated simultaneously in CA theory. The compiler generates a high speed simulator using MMX technology. MMX technology is one of SIMD technologies and enables to process many operations by a single instruction. The compiler generates a simulator from the description of local rules in DORA language. We designed DORA language for cording local rules with simple description naturally. We ran the performance test of the simulator generated by our compiler. The results showed that the simulators generated by our compiler runs about 3 to 11 times faster than ones with C compiler.

The objective is to propose an immune distributed competitive problem solver using MHC and an immune network and to verify its validity by means of computer simulations. Our algorithm solves the division-of-labor issues and problems for each agent work domain in a multi-agent system (MPS) by two immune functions. First, the Major Histocompatibility Complex (MHC) distinguishes "self" from "nonself", used in the process of eliminating states of competition. Second, the immune network that produces specific antibodies by modification of immune cells is used to produce adaptive behaviors for agents. Then, to investigate the validity of the proposed method, this algorithm is applied to the "N-th agent travelling salesmen problem (n-TSP)" as a typical case problem for multi-agent systems. The effectiveness of solving multi-agent system problems in this way is clarified through some simulations.

In the field of artificial intelligence, the development of a method for the acquisition of game-strategies is one of the important issues. We deal with the acquisition of a game-strategy by using a competitive co-evolution approach as a search method. The competitive co-evolution is a mechanism of interactive improvement of solutions. However, in practical games, it is not easy to acquire the best strategy by applying the competitive co-evolution model because of the complex strategy space. Therefore, to design the acquisition system of stronger game-strategy, we propose an improved competitive co-evolution model that introduces the concept of "package" as a set of good strategies. Creating a good package needs to collect some good strategies to defeat various kinds of opponents. We apply the algorithm to some games to show its effectiveness and efficiency.

As the neural networks or genetic algorithms, adaptive algorithms become popular and these techniques are applied to many kinds of optimization problems. The immune system is one of the adaptive biological system whose functions are to identify and to eliminate foreign material. In this paper, we propose an optimization algorithm based on immune model and applied to the n-th agents' travelling salesman problem called n-TSP. Some computer simulations are designed to investigate the performance of the immune algorithm. The results of simulations represent that the immune algorithm shows good performance for the combinatorial optimization problems.