Airborne ultrasound has been widely employed in various industrial applications and generating high-power airborne ultrasound is important for these performances. For this purpose, we propose an air-type double-parabolic-reflectors acoustic waveguide (Air DPLUS) to realize high power ultrasound. Air DPLUS reflects the incident waves twice and the sound pressure is focused in a small radius for high power ultrasound. FEM simulation was carried out to confirm the proposed principle. Based on this simulation result, a prototype was fabricated with 3D printing and short-burst excitation was applied. In experiments, the acoustic pressure distribution was measured and the output acoustic pressure was confirmed to be magnified by four times with Air DPLUS.

Piezoelectric actuators have played a critical role in the field of precise positioning and driving. However, existing actuators cannot achieve the performance of high speed and positioning accuracy at the same time. Therefore, in this study, a piezoelectric actuator with triangular mechanisms that enables high-speed operation and high-precision positioning was proposed and fabricated. Experimental results indicated that the actuator can work in a wide range of driving frequencies. Under the resonance frequency of 14 kHz, the minimum voltage that could drive the actuator was 6 Vpp, and the maximum velocity of 24.6 mm/s was achieved at a voltage of 36 Vpp. Moreover, the proposed actuator could realize precise positioning under the non-resonance frequency. The positioning resolution of 144 nm was obtained, and the measured repeatability was ±25, ±35, ±40, and ±50 nm for four strokes of 325, 425, 625, and 725 nm, respectively.

The displacement magnifying mechanism dealt with in this paper is to magnify small displacement generated by the multilayer piezoelectric actuator. Equivalent circuit representation is indispensable for estimating transient phenomena of this system. The equivalent circuit representation also enables calculation of the force applied to the multilayer piezoelectric actuator. The theory is based on the fact that the degrees of freedom of a two-port circuit is three. Therefore, it has generality and can be applied to all displacement magnifying mechanisms.

Strong holding torque is one of the advantages of ultrasonic motors. However, the previous report shows the rotors turn backward at the start-up of the stator vibration if large load torque is applied. This slope start phenomenon prevents the utilization of the holding torque of the ultrasonic motors in some applications, such as robot end effectors. In this research, the rotation response at start-up was simulated with the equivalent circuit model and the friction contact model. The simulation shows that the backward rotation can be mitigated by accelerating the start-up time of stator vibration.

The relationship between worker concentration and work performance was examined by measuring the force applied by the operator and the instantaneous heart rate simultaneously during drilling. A correlation was found between the instantaneous heart rate changes and the degree of work proficiency. In tasks that one has become familiar with through repeated training, the heart rate is high at the beginning of the task but decreases as the task progresses. In tasks that the worker self-assessed as difficult, the heart rate tended to keep a high value throughout the drilling and the variation tended to be small. We calculated the Mahalanobis distance in the space of bivariate variables of mean heart rate and variation during drilling, where the data in operations that the worker self-assessed as difficult were used as reference data. Mahalanobis distance was correlated with task difficulty. This suggests that the relationship between worker concentration and work performance can be evaluated from the Mahalanobis distance.

It is known that the recognition performance of stereo cameras is affected by rainfall, but it is difficult to validate the recognition performance under actual rainfall with good reproducibility. Therefore, we validated the recognition performance of a stereo camera in a rainfall environment using the stereo camera VR validation system that has been verified so far, by drawing rainfall by simulation, projecting it on a display, and capturing it with a stereo camera.

In 3D scan data, point cloud and mesh are very important data formats to represent geometric shapes. In most cases, point clouds are often arranged irregularly, so it is common to use generalized method. However, in line laser scanning data, we found that the data was arranged with certain rules. We often take many scans to get the whole objective rather than a single scan, so we took advantage of this feature and carried out line segmentation and scanning segmentation on the data. Then averaged distances between different scans are evaluated, and this is used in the overall data process and resample. Finally, we obtain relatively improved high-quality 3D mesh data.

Image averaging with a convolution kernel consisting of brick-wall cutoff along salient edges is useful for robust estimates of an image manifold and its color statistics, but their fast computation is non-trivial. Although some simplification of such averaging by using half-box kernels has been proposed, the conventional methods require multiple filtering processes. In this research, we propose a fast algorithm for image averaging using half-box convolution kernels. Our algorithm is based on summing the total colors within adjacent rectangular regions, where these total colors are efficiently obtained by applying a half-radius box filter only once via the moving average method. Our numerical experiments show that the proposed method is faster than conventional algorithms.

Fused Deposition Modeling (FDM) allows the creation of specific shapes by stacking materials, which makes it difficult to check the quality of the products. Using an X-ray CT scanner to inspect voids inside products that affect mechanical properties is an effective method. However, traditional void analysis method is difficult to extract voids from the CT volume. In this research, the toolpath, which contains the ideal shape of the product, is used to improve the void extraction. Besides, since the toolpath represents the FDM process, it can be used to infer the formation mechanism of each void. Based on the toolpath, a novel void classification is proposed that can help the user rapidly determine the cause of each void and find ways to avoid them. Products with different infill densities are inspected, and the results are shown by slices.

Wave-dissipating blocks made of large concrete slabs are essential components of the armor layer of breakwaters that protect their core from direct wave attacks. However, the blocks will gradually sink or ablate as the corrosion. Therefore, regular block stacking is required to maintain the height or shape of the breakwater. This study aims to develop a system that can simulate the replenishment works utilizing a physics engine and deep learning, which can predict the additional block amounts and their stacking poses and provide pre-visualization of their stacking operations. Deep learning was used to estimate the additional block poses that better fit the currently stacked blocks. The simulation was applied to actual block stacking operation in a local port at Hokkaido. The accuracy and practicality of the simulation were verified with the final construction results.