Condition monitoring of rotor problems such as demagnetization and eccentricity in permanent-magnet synchronous motors (PMSMs) is essential for guaranteeing high motor performance, efficiency, and reliability. However, there are many limitations to the offline and online methods currently used for PMSM rotor quality assessment. In this paper, an inverter-embedded technique for automated detection and classification of PMSM rotor faults is proposed as an alternative. The main concept is to use the inverter to perform a test whenever the motor is stopped and to detect rotor faults independent of operating conditions or load torque oscillations, which is not possible with motor current signature analysis (MCSA). The d-axis is excited with a direct-current+alternating-current signal, and the variation in the inductance pattern due to the change in the degree of magnetic saturation caused by demagnetization or eccentricity is observed for fault detection. An experimental study on a 7.5-kW PMSM verifies that demagnetization and eccentricity can be detected and classified independent of the load with high sensitivity.

Individual tooth segmentation from cone beam computed tomography (CBCT) images is an essential prerequisite for an anatomical understanding of orthodontic structures in several applications, such as tooth reformation planning and implant guide simulations. However, the presence of severe metal artifacts in CBCT images hinders the accurate segmentation of each individual tooth. In this study, we propose a neural network for pixel-wise labeling to exploit an instance segmentation framework that is robust to metal artifacts. Our method comprises of three steps: 1) image cropping and realignment by pose regressions, 2) metal-robust individual tooth detection, and 3) segmentation. We first extract the alignment information of the patient by pose regression neural networks to attain a volume-of-interest (VOI) region and realign the input image, which reduces the interoverlapping area between tooth bounding boxes. Then, individual tooth regions are localized within a VOI realigned image using a convolutional detector. We improved the accuracy of the detector by employing non-maximum suppression and multiclass classification metrics in the region proposal network. Finally, we apply a convolutional neural network (CNN) to perform individual tooth segmentation by converting the pixel-wise labeling task to a distance regression task. Metal-intensive image augmentation is also employed for a robust segmentation of metal artifacts. The result shows that our proposed method outperforms other state-of-the-art methods, especially for teeth with metal artifacts. Our method demonstrated 5.68% and 30.30% better accuracy in the F1 score and aggregated Jaccard index, respectively, when compared to the best performing state-of-the-art algorithms. The primary significance of the proposed method is two-fold: 1) an introduction of pose-aware VOI realignment followed by a robust tooth detection and 2) a metal-robust CNN framework for accurate tooth segmentation.Index Terms-Cone beam computed tomography image segmentation, pose-aware tooth detection, pose regression neural network, tooth instance segmentation.

Recently, fire accidents in buildings have become bigger around the world, and it has become necessary to build an efficient building disaster management system suitable for fires in a Smart City. As building fires increase the number of casualties and property damage, it is necessary to take appropriate action accordingly. There has been an increasing effort to develop such disaster management systems worldwide by applying information communication technology (ICT), and many studies have been conducted in practice. In this paper, an augmented reality (AR)-based Smart Building and Town Disaster Management System is suggested in order to acquire visibility and to grasp occupants in case of fire disasters in buildings. This system provides visualization information and optimal guide for quick initial response by utilizing smart element AR-based disaster management service through linkage of physical virtual domain in the building. Additionally, we show a scenario flow chart of the fire extinguishment process according to the time from the ignition stage to the extinguishment stage in the building. Finally, we introduce the related sensors, the actuators, and a small test-bed for AR-based disaster management service. This test-bed was designed for interlocking and interoperability test of the system between the sensors and the actuators. It is expected that the proposed system can provide a quick and safe rescue guideline to the occupants and rescuers in the building where fire is generated and in regions of poor visibility.

We report on an Er-doped fiber oscillator that produces 146 fs pulses with 156 mW average power at a repetition rate of 49.9 MHz. The pulse energy reaches 3.13 nJ, surpassing the conventional power limit in the dispersion-managed soliton regime. Such high pulse power is obtained by devising a hybrid mode-locking scheme that combines saturable absorption with nonlinear polarization evolution. The oscillator also offers excellent temporal purity in the generated pulses with high power, providing a robust fiber-based frequency comb well suited for industrial uses.

No abstract