A new concept of human-machine interface is proposed. In human-human communications, especially in face to face communication, sub-verbal and non-verbal messages have more importance than messages transported by vocabulary in communications. On the other hand, in traditional man-machine interfaces, machines only understand pre-defined operations, and never understand operators' sub-verbal and non-verbal messages. This causes some difficulties of machine usability. Grasping-and-moving are fundamental hand operations necessary for performing tasks using modern man-machine interfaces. However, behavioral information associated with grasping, such as force, posture, etc. has not been utilized for traditional interfaces, although behavioral information may have potential utility in developing task-adaptive and operator-adaptive interfaces; it is known that how people grasp devices depends on the purpose of their tasks. As an example of typical pointing devices, this paper describes a computer mouse whose approaching speed, or gain, is adjusted depending on the nature of the task. This paper suggests that the grasping-and-moving tasks should consist of two phases. One is “approaching phase” in which the operator moves the mouse pointer close to the target with smaller grasping force. The other is “positioning phase” that follows “approaching phase” and in which the operator makes a fine adjustment of the pointer position to locate it within the target area with larger grasping force. Proposed new computer mouse in this paper has a sensor to detect the operator's grasping force and uses it to adjust the gain of the mouse. Two sets of experiments were conducted; one for confirming the assumption that the grasping force should be small while approaching the target and large while positioning, and the other for confirming that the proposed mouse can reduce the time needed to set the pointer onto various kinds of objects.

Currently, vehicle's communication infrastructures are conducted in several loations, and they provide some traffic and signal information for vehicles. Using this information, automatic driving technologies are widely studied, and they focus on the avoidances of stopping by red signal and car crashes. They have deep insight for each research fields, however, there are no proposal for realizing both avoidance at same time. In this paper, we propose action determination learning scheme for realizing driving support considering both avoidance. Our scheme is based on Deep Reinforcement Learning similar with latest crash avoidance scheme. From simulation evaluation, the numbers of both crashes and stopping by red signal becomes lower by time goes on, and effective parameter tuning is revealed at same time.

In this paper, a framework has been examined for quantitatively analyzing the relation between the workplace environment and intellectual concentration, through “factors” that connect between them, in order to improve intellectual concentration in the office. Specifically, “human characteristics” have been focused on and the factors affecting intellectual concentration was categorized into two groups. They are dynamic factors (arousal, mood, fatigue, stressor assessment) which vary during work and static factors (reference value, environmental sensitivity) which do not change. Using these factors, the measurement method and the quantification method have been examined, and EFiC framework (Environment-Factor-intellectual Concentration) has been proposed for deriving the mechanism of intellectual concentration affected by the workplace environment. In order to confirm the effectiveness of the EFiC framework, it was applied to the measurement data acquired in a past experiment of the intellectual concentration affected by lighting environment. As the result, concrete suggestions to improve the operating environment based on the characteristics of people were obtained. By applying this framework to measurement experiments of various intellectual concentration, it is expected that effective suggestions for improving intellectual concentration will be obtained. In addition, in order to expand this framework in general purpose, it is conceivable to consider more effective measurement method and analysis method.

This paper describes a procedure to design potential functions for path following control of port-Hamiltonian systems. The conventional path following control method needs to find a time invariant potential function which takes its minimum on the desired path. It is difficult to find such a function for a complex path, since it has to satisfy additional several constraints. Inspired by the results of the existing trajectory tracking control method of port-Hamiltonian systems, we propose an improved path following control method in which a potential function for path following control is acquired by solving simple partial differential equations.

A model reduction method of interconnected linear systems is proposed via generalized mixed H₂/H_∞ balanced realizations, where an H₂ Linear Matrix Inequality (LMI) and an H_∞ LMI have the same positive definite diagonal matrix. A feature of the model reduction is that the reduced order model preserves the same interconnected structure as the original model. It is shown that the H₂ and the H_∞ norms of the reduced order model are bounded by the values which appear in the H₂ and H_∞ LMIs. Upper bounds of the H₂ and the H_∞ norms of the error system between the original model and the reduced order model are also established. Its effectiveness is demonstrated through a numerical example.

This study proposes a design method of dynamic quantizer with communication rate constraints. Dynamic quantizers are well known as an effective method for the signal quantization in the meaning of the noise shaping. The noise shaping characteristics of the dynamic quantizers are determined by the matrices A, B and C. In the previous studies, A, B and C are given as time-invariant matrices. In this paper, we propose a dynamic quantizer form including periodically time-varying matrices for improving the control performance under the communication rate constraints. Analysis method of periodically time-varying dynamic quantizer is reduced to that of dynamic quantizers using time invariant matrices by using lifting technique. The effectiveness of the designed periodically time-varying dynamic quantizer is assessed via numerical examples.

A practical heuristic approach to Node Edge Arc Routing Problem (NEARP or MCGRP) and an aplication to a newspaper delivery problem is proposed. In this approach, the creation of neighborhood based on one dimensional data model and heuristic optimization technique using Simulated Annealing (SA) are adopted. The rates of adopting rules in producing neighbors and the values of parameters in the SA procedure are revised from them used in the method which was developed in the authors' previous study. Computational experiments are examined on a set of benchmark NEARP problems (CBMix series). In two problems over twenty-three benchmark problems, the proposed method overcomes the best-known solutions in smaller computing time. A newspaper delivery problem in a district in Japan is modeled as NEARP and solved by the proposed method. Computational results on the problem are presented and the effectiveness of the proposed practical approach is demonstrated.

This paper proposes a reliable data transfer method using WSN to collect information of livestock in smart livestock farming. Evaluation results showed the proposed method reduces the power consumption by 73% at the same data reachability as the conventional method.