Modeling the motion of experts in sports is important in demonstrating their motion to beginners. A primary approach to this objective is to study the dynamics of a transient response from one position to another. Hence, finger-motion analysis is a primitive but crucial step to be considered. The state-space representation with a step input is used to eliminate the personal habits of experts from their motion and to present simple models to learners according to their learning stages. Based on the deterministic realization and N4SID algorithms, two algorithms are developed to demonstrate simple motions according to the order of the state-space representation. They are applied to the finger motion captured by the Leap Motion Controller.

View full abstract

Low illumination image enhancement is a difficult but scientific task. With the image brightness increasing, the noises are amplified, and with the contrast and detail increasing, the false information is generated. To solve this problem, a multi-branch attention network is proposed to process low-light images directly without additional operations. The proposed network is composed with enhancement module (EM) and Convolutional Block Attention Module (CBAM). The attention module can make the CNN network structure gradually focus on the weak light area in the image, and the enhancement module can fully highlight the multi-branch feature graph under the guidance of attention. In this way, the overall quality of the picture will be greatly improved, including contrast, brightness, etc. Through a large number of experiments, our model can produce better visual effects, and also achieve good results in quantitative indicators.

View full abstract

PID control is widely used in process systems represented by chemical processes and petroleum refining processes. The reason is that PID control has a simple structure. However, most of the existing systems are non-linear systems, and it is difficult to always obtain good control results with fixed PID control. Therefore, in this study, we propose a method of tuning the PID gain according to the deviation (control error) of the control result, and verify the effectiveness of this method through experiments. For self-tuning PID control using a local linear model, we propose a program that performs PID tuning only when the deviation occurs with a certain magnitude. A simulation is performed on the Hammerstein model, which is a non-linear system. As a result of the experiment, the number of PID gain changes could be significantly reduced.

View full abstract

In the near future, deep-sea mineral resource development will be possible to be commercialized. To minimize the CAPEX of the project, estimation of deposit quality is critically important and exploratory drilling is indispensable for estimating the amount of resources. To reduce the cost of exploratory drilling, we are going to develop a system of seabed drilling which can move on the seabed by itself without a support vessel on the sea. This exploration platform must be able to move along with the undulations of the seafloor and have a structure that supports the reaction force of the drilling. The authors have been studying an eight-legged drilling platform. This paper introduces our tank test system and some results of tank experiment.

View full abstract

In 1993, Joel Feinstein found one of perfect plays for Othello (Reversi) with board size of 2n×2n (n = 2,3) , which showed that the second player(White) is the winner, respectively. We have been analyzing Othello with a board size of 4×2n (n = 3,4,…) since 2015. As results, we found one of perfect plays for each on the 4×2n (n = 3,4,5) boards, which showed that the first player (Black) is the winner, respectively. Recently, we have finished the analysis of the 4×12 board. In this paper, we will introduce the analysis results of 4×12 board Othello and summarize the results that have been found so far.

View full abstract

In the field of robotics education, understanding the design, manufacturing, and control of robots is an important aspect of understanding the development of robots. Therefore, various university-level robotics education courses have been created to cover these topics. However, students typically learn about design, manufacturing, and control in different classes. This can potentially lower their understanding of robotics development. There are various reasons why these topics are taught in different classes, but the main reason is the difficulty of manufacturing. Manufacturing education parts often requires a machine such as a CNC, and in the case of a class for many students, it takes a lot of time to manufacture objects designed by students. Recently, low-cost 3D printers have become available, which are considered a solution for reducing manufacturing time. In this paper, I propose a year-round robotics class that covers design, manufacturing, and control. Then I introduce the developed 5-axis robotic arm for education, which can be manufactured using 3D a printer, as a result of the proposed class. The evaluation of the developed arm showed that the maximum payload (with a safety factor of 2.5) was 2.2 [kg], and the maximum repeatability was ± 0.201 [mm].

View full abstract

In food industry, shortage of workers is one of a serious problem. Automation of food handling is a critical nowadays. To alleviate the damage caused by food picking operation by robotic hand, we propose non-contact acoustic impedance estimation with ultrasonic wave, which should be preceded before the picking for optimization of grasp stiffness. We assume the correlation between hardness and acoustic impedance. That is, the harder the medium is, the larger the impedance should be. As cooked food, the unevenness of hardness should be considered. We made samples with media of different acoustic impedances, and experimented with ultrasonic to find the relation between reflection coefficient and hardness. As the result, reflection coefficient increased by 0.03-0.15 at the boundary of two media as the bottom medium is switched from urethane to Aluminum.

View full abstract

We have developed a fall risk alert system using a smartphone with a LiDAR sensor, Apple’s iPhone 12 Pro, as an electronic travel aid (ETA) for visually impaired to reduce fall risks from the edge of station platform and the stairs. The proposed system generates the height grid map generated from 3D point clouds and posture from the smartphone, and classify the walking area to four areas, walkable, positive obstacle, negative obstacle and occlusion areas. In this paper, we describe the algorithm and the experimental results of the environmental recognition system for the fall risk alert. In the experiments, the accuracy, stability, and time delay are evaluat-ed in approaching a fall risk area, such as stairs. The results show the smartphone has the possibility to be the ETA to support the visually impaired and reduce the fall risks.

View full abstract

Currently, Japan is facing serious headwinds from declining birthrates as well as an aging society, and their economic consequences are widely known. Among many possible strategies to address the aging and shrinking population, the revitalizing the local economy in each prefecture become more prominent with an emphasis on spurring tourism. To shed light on this issue, this manuscript focuses on determining empirically the interrelationship between tourists and tourism destination characteristics using regression modelling. More specifically, the inter-linkages among tourists and tourism destinations is conducted by developing a 4-cell model. The results finds support for the importance of the development of public traffic among different tourism destinations as a promising economic revitalization strategy

View full abstract

We proposed a new tactile presentation device for a palm to tele-operate a robot hand mounted on a rescue robot. The proposed device presents a tactile sensation on the back of the operator’s hand rather than the palm to avoid interference. The required specification for the device were determined through subject experiments conducted in our previous studies. In particular, we developed a prototype device through suction stimulation with the required specifications, confirmed that the device can present tactile sensation correctly, and discussed the results.

View full abstract

The counseling service system is a case management tool used by professional guidance counse-lors and full-time teacher-counselor. However, even after multiple maintenances, system satisfac-tion and willingness to use it cannot be improved. This study used the modified Delphi method and analytic hierarchy process to construct the service quality index of the counseling service system by adding expert suggestions from different backgrounds. The study found that the service quality indicators of the counseling service system can be divided into five criteria and 21 sub-criteria to explore, among which were "perfect confidentiality mechanism", "web design & layout meet the needs of the work", and "functional operation is simple and convenient" most valued by experts and users. Based on the research results, this study puts forward substantive implications for academics and management.

View full abstract

In this paper, the distributed optimization problem for multi-agent system (MAS) formation control of unmanned aerial vehicles (UAVs) is suggested here. The situation that the internal states of a single UAV can be made fully available was aimed at, the internal optimal control law of a single UAV is designed using the optimal control theory. To cope with the obstacle that each agent in the system can only communicate with its neighbor agents, the distributed formation control law of the system is introduced based on the communication topology of the system, and further the stability of the system is analyzed in helps of graph theory. The validity of the suggested scheme is verified by both the numerical simulation and UAV platform.

View full abstract

The story units in this study are the units described in Common Lisp based on the types of folktales analyzed by Seki et al. We have been developing story units for our integrated narrative generation system (INGS) as a type of narrative technique that generates a new narrative based on the synthesis, transformation, expansion, and so on, of a narrative structure. The story units function in the INGS with conceptual dictionaries. In this paper, we combine these story units with a verb conceptual dictionary and a noun conceptual dictionary. The constitutional elements of each story unit are verb and noun concepts; therefore, combining them with a noun conceptual dictionary enables the substantial function of story units based on the types of Japanese folktales as a group of narrative techniques in the INGS.

View full abstract

In quality inspection, which is the final stage of the manufacturing process, there is an operational manual on how the object should be inspected. In this study, we followed the gaze of the workers to search for the "way of looking" at the objects to be inspected, which is not described in the manual. As a result of experiment, we found a difference of the reciprocating eye movements between skilled worker and the beginner. It is suggested that this indicates a way of looking at inspection that has been unconscious until now.

View full abstract

With the vigorous advancement of intelligent manufacturing, smart home has gradually entered people's daily life. At present, smart window always have the problem of low intelligence and a few functions. And almost all kinds of traditional windows have the problem of inconvenient cleaning outside. The paper proposed a new mechanical structure design of the window, so that the window can be turned by nearly 160 degrees, which is convenient for cleaning the both sides of the window. In terms of intelligent systems, a variety of sensors combined with Internet modules are used to realize the remote intelligent control (turning with the wind, and automatically closing windows when there is no one at home when bad weather). Window adds a child mode, which is based on machine vision. In this mode, when the window detects that there is a child nearby, the window will be automatically closed to effectively avoid the accident of children falling from the building caused by climbing the window.

View full abstract