Hiroyuki Ogata

Abstract The injection molding market is expected to grow with the increasing demand from various industries, such as automotive and furniture. However, weld lines in the injection molded part reduce the structural strength and affects its appearance. This study evaluates the effect of the weld line on the structure by shifting the position of the weld line outside the area where the load is applied by manipulating the opening time of the injection gates. We examined the effect of weld lines on the strength of a chair through coupled resin flow and structural analyses. The analysis revealed that the position of the weld lines is crucial for structural strength if the knockdown factor (KDF) is low. Additionally, controlling the opening time of the gate and shifting the position of the weld line improved the structure's appearance, number of weld line and avoid the lower meeting angle. However, depending on the material, the effect may be weak. For example, when the weld line location did not change much despite changing the gate timing, little change was observed in the number of weld lines. Hence, the improvement strategy of weld line should be considered for material types and strength. Highlights Injection molding growth driven by automotive, packaging, furniture demand. Weld lines impact structural strength and appearance. Shifting weld line position outside load area improves the structural appearance and strength. Effects of shifting weld line position depended on the material types. Weld line position is crucial for low knockdown factor and affects strength.

Under the Synecoculture environment, in which various plants are raised in mixed and dense vegetation, automatic maintenance of the field is difficult because of difficulties in separating each harvest. In this project, the situation in which one plant dominates the other plants is called “dominant situation”, and such dominant plants are to be cut. So, in this paper, we propose a method for detecting dominant plants from RGB images using deep learning. First, we partition the original image into small blocks. We perform VGG16 for each small block to predict the number of plants. If the number of the small blocks in each of which the number of plants is less than two exceeds the threshold, the original image is judged as a candidate of “dominant situation”. If the original image is judged as the candidate, similarity between dominant small blocks is computed using AKAZE, and if the similarity is high, the small blocks are judged to be in dominant situation. Experimental results show that high accuracies for estimating dominant situations are achieved.

Skiing is a difficult sport for beginners to start compared to other sports because of the severe restriction of time and place, and the high risk of injuries such as falling. Therefore, a ski training system that integrates VR and hardware technology is developed. In this paper, we evaluate its effectiveness through experiments. Specifically, we verify and compare the training effects of two systems with different feedback methods: “a system that feeds back the changes in somatosensory perception as VR images” and “a system that feeds back the changes in somatosensory perception by increasing or decreasing the gauge”. By comparing these two systems, we examine whether the learning effect of the system can be improved by incorporating VR technology into the ski learning system, and obtained promising experimental results.

This paper proposes a method for quantitatively evaluating sprinting motions using the videos of runners. Specifically, this paper explores the coordination between physical motions, which has been recognized as very important in sprinting. After detecting and normalizing the joint coordinates from sprinting videos, the cross-correlations of two windowed time-series data are calculated using the windowing cross-correlation function, and the coordination between the motions of the two joints is quantified. Experiments that use 20 subjects are conducted. As a result of classifying the cross-correlation obtained from the subjects' data into two clusters using k-means clustering, conditions in which the obtained cluster includes a high percentage of inexperienced sprinters are found. To verify whether the motions corresponding to these conditions are valid as the evaluation criterion of sprinting, Spearman's rank correlation coefficients between cross-correlations and 30-m time records are calculated. The results show a weak correlation with respect to the coordination between the elbow and knee motions. Therefore, it can be said that the cross-correlation corresponding to the coordination can be used as a quantitative criterion in sprinting.

This paper proposes a method that can classify foot strike types using a deep learning model and can extract unknown factors, which enables to evaluate running motions without being influenced by biases of sports experts, using the contribution degree of input values (CDIV). Accelerometers are attached to the runner's body, and when the runner runs, a fixed camera observes the runner and acquires a video sequence synchronously with the accelerometers. To train a deep learning model for classifying foot strikes, we annotate foot strike acceleration data for RFS (Rearfoot strike) or non-RFS objectively by watching the video. To interpret the unknown factors extracted from the learned model, we calculate two CDIVs: the contributions of the resampling time and the accelerometer value to the output (foot strike type). Experiments on classifying unknown runners' foot strikes were conducted. As a common result to sport science, it is confirmed that the CDIVs contribute highly at the time of the right foot strike, and the sensor values corresponding to the right and left tibias contribute highly to classifying the foot strikes. Experimental results show the right tibia is important for classifying foot strikes. This is because many of the training data represent difference between the two foot strikes in the right tibia. As a conclusion, our proposed method could extract unknown factors from the classifier and could interpret the factors that contain similar knowledge to the prior knowledge of experts, as well as new findings that are not included in conventional knowledge.

To identify the difference in performers' motions, this paper investigates the relationship between running motions and the result of evaluating motions during jump training. To clarify the relationship, two experiments were performed using 17 subjects as follows: i) obtaining sequences of human joints during running to evaluate running motions, and ii) obtaining motions during jump training which could skill up the running motions. According to the result of those experiments, we confirmed that whether a running motion is good or not relies greatly on the number of acquired skills.

To improve running performances, each runner's skill, such as characteristics and habits, needs to be known, and feedback on the performance should be outputted according to the runner's skill level. In this paper, we propose a new coaching system for detecting the skill of a runner and a method of giving feedback using a sprint motion dataset. Our proposed method calculates an extracted feature to detect the skill using an autoencoder whose middle layer is an LSTM layer; we analyse the feature using hierarchical clustering, and we analyse the human joints that affect the skill. As a result of experiments, five clusters are obtained using hierarchical clustering. This paper clarifies how to detect the skill and to output feedback to achieve a level of performance one step higher than the current level.

The current stage of the development of academic systems for education in robotics may be considered to be at a threshold. In order to formulate academic systems for robotics education, researchers need to evaluate the effectiveness and influence of robotics education, objectively or quantitatively, in their research studies. In this study, the authors quantitatively evaluated the educational effects of a robotics classroom for elementary school children in the 5th and 6th grades using psychological measure and statistical analysis. Pre- and post-questionnaire surveys were conducted using a questionnaire that had been used and validated in other studies, and two-factor analysis of variance was applied to the results of the questionnaire. From the results, the following findings are reported. (1) The interest in making things by the group who had little experience making things increased more than the group who did have such experience. (2) The boys group generally had more interest in making things than the girls group. (3) Participants in the 5th grade had greater willingness to do something creative, and in acquiring the skill to use a tool, than did those in the 6th grade. (4) Participants in the 6th grade wanted to create something more specific than did those in the 5th grade. Overall, it was quantitatively shown that the children participating in the robotics classroom originally had a high interest in making things. Moreover, their interest in making things improved through their participation in the robotics classroom.

This paper explains a motion planning strategy that is not based on time but on "event", which can be explained as the path that the end-effector of robot tracks in our study. Arc-length parameterization is used in this paper to convert a time-based planning into an event-based planning. With our planning, it is possible to choose various time planning on the same given path so that tasks such as dynamic obstacle avoidance and performance test of motors become available. Simulation results about time planning implemented to a given path verifies the effectiveness of our event-based planning.

Project based learning (PBL) is an effective method for education to make students put their knowledge into practice or to train solving problem by themselves. PBL is also recognized to be effective and popular in robotics and manufacturing educations. However, it is not guaranteed that traditional scoring methods, like final exam, are proper to assess students' activity in PBL. One of the noteworthy methods that may be suitable to assess their activity there is the peer assessment. The benefit of carrying out peer assessment is that it has effects on building students' motivation, making feedback, and grading reliably. The author conducted peer assessment in a mobile robot development project in his own laboratory, and in a certain class where students are requested to make a paper bridge as light as possible but can support a 300[g] weight. In this paper, the data of peer assessment in those projects are analyzed to verify its effectiveness.

Recently, e-assessment has become popular, and it is now technically possible for people to take exams at home. However, because the testing environment is dramatically different from the traditional one, new technologies are required to prevent cheating. Many studies have addressed testing environments where a keyboard and mouse are used as input devices, but few have focused on an environment using tablet computers, which are widespread nowadays. The aim of our study is to develop a method of detecting impersonation using touch gestures on a tablet computer. We verified that tap gestures performed by different people on a tablet computer can be distinguished by extracting some features from finger tapping images. An experiment using 30 subjects was conducted for verification. © 2013 IEEE.

The manner of holding a pen is distinctive among people. Therefore, pen holding style is useful for person authentication. In this paper, we propose a biometric person authentication method using features extracted from images of pen holding style. Images of the pen holding style are captured by a camera, and several features are extracted from the captured images. These features are compared with a reference dataset to calculate dissimilarity scores, and these scores are combined for verification using a three-layer perceptron. Preliminary experiments were performed by using a private database. The proposed system yielded an equal error rate (EER) of 2.6%.

The progress of information technology and the popularization of the personal computer brought the growth of computer-based testing (CBT), which utilizes computer to hold an exam and rate examinees. Most of the current CBT can be regarded as an extension of the paper-and-pencil exam, and aim to assess the knowledge or the achievement of the examinees. On the other hand, very few CBT systems address assessing examinees' skill in sports or manufacturing. This paper discusses a method to evaluate examinees' physical skill from their motion data, taking golf putt swing as an example. The motion data is generally represented as a vector time series. The recurrent neural network is introduced to deal with such data. We verified the effectiveness of the proposed method by collecting actual examinees' motion data and assess them using the proposed method.

Generating motion is important for the robots over doing some kind of task. Especially, generating motion to cover certain area utilizing the size of the robot is used to realize robotic tasks, such as sweeping floors. Most of these robots generate their motion by reactive algorithm with easy sensors mounted. However, with these algorithms, the sweeping task would be done uneven to the sweepable areas. So, there are still rooms for developing a more efficient algorithm. To develop an efficient algorithm, evaluation method is in need. However, the methodology to evaluate the efficiency of state informed algorithm has not yet established. Cleaning robots do not always clean the same environment so, the difficulty of this research is that consideration of various kinds of environments is needed. The chapter develops sweeping robot simulator to analyze the existing sweeping algorithms and reveal their inefficient features to improve the existing algorithms. The algorithms that commercial robot cleaners install are analyzed, and those efficiencies are estimated. © 2006 Elsevier Ltd. All rights reserved.