Search

To accommodate various applications in optical access systems, the Software-Defined Network (SDN) with edge computing is the key allowing telecom operators to achieve flexible and cost-effective deployment. To control the various network functions needed to meet the variety of requests, softwarization of all access system functions, including transmission functions, and open Application Programming Interfaces (APIs) have been studied. As well as function changes, we aim at maximizing the merits of SDN by adaptively configuring network function parameters to meet individual application requirements. This paper proposes a novel access system architecture with configuration optimizer that adaptively generates parameter configuration values from application requirements, and adaptive configuration values for image transmission to enhance computation efficiency. Our proposal includes Forward Error Correction (FEC) with error correction control in decoding to suit image quality requirements. Our proposal is implemented on a general-purpose server and we demonstrate its performance using the softwarized 10G-EPON Physical Coding Sublayer (PCS) functionality on an optical system. The demonstration results show that the proposed algorithm adjusts the error correction capability of FEC, and that the received image quality approaches the required value. Evaluations of FEC computation load show that the processing time of FEC is reduced compared to the standard functions of 10G-EPON depending on the optical received power and the required image quality by adjusting the error correction capability.

Tracked vehicles are expected to be used for factory inspection such as in petrochemical refinery, in which the robot needs to navigate in narrow spaces. During the narrow space navigation, the robot needs to adjust its position in lateral direction. However, it is a difficult problem to realize lateral motion of tracked vehicles for factory inspection, because mechanical complexity is increased if the additional actuator is installed. In this paper, the authors propose a lateral skidding motion of tracked vehicles using wall reaction force. The proposed method convert the direction of driving force by actively colliding the wall with passive wheels, and realize the lateral motion without attaching the additional actuator. Experimental results show that the lateral skidding motion is realized on two types of floor material. In addition, the time for lateral positioning with manual operation was 20 % reduced when the proposed method was used. The use of external force is one solution to change the motion direction in narrow spaces.

In order to prevent potential accidents between pedestrians and vehicles, it will be effective to introduce a vehicle-to-pedestrian warning system. The system detects potential accidents between pedestrians and vehicles in advance and sends warning to pedestrians and vehicles ahead of time so that potential accidents can be avoided. The time ahead of the potential accidents is named prior notification time and it is set in advance in the system. However, there is a difference between the prior notification time and the actual notification time due to system delay and inaccuracy of GPS localization. This difference is named warning error. The location estimation error and the speed of pedestrians and vehicles will affect the appropriate estimation of prior notification time. In this paper, how warning error is changed with vehicle speed is investigated through simulation. The simulation results show that the warning error increases as the vehicle speed increases. The maximum warning error is 2 seconds regardless of the vehicle speed. Thus, notification time delay can be reduced by estimating the maximum value of the warning error and increasing the prior notification time in advance.

In the high-frequency Wireless Power Transfer(WPT) system, the system size of the transformer and power converter depends on the transfer frequency. This paper clarifies the effects of transformer and power converter on system loss and volume, moreover establish the guideline for determining the transfer frequency in a small-capacity, high-frequency WPT system.

This paper presents a single-layer ring microstrip antenna (MSA) fed by an L-probe with a variable reactance circuit for a beam adjustable microstrip antenna arrays (BA-MSAA). The BA-MSAA consists of a central patch element fed by a feeding line and two parasitic patches with a variable reactance circuit (VRC) adjacent to the center patch. Radiation patterns of the BA-MSAA are controlled electronically by reactance values of the VRCs. In order to enhance the beam-shaping performance of the BA-MSAA, it is necessary to feed the all patches and to control the radiation phases of the fed patches independently. For this purpose, a single-layer ring MSA fed by an L-probe with a VRC is presented in this paper. The prototype MSA with the VRC is designed at 5 GHz. The radiation phase of the prototype MSA can be controlled in the range of about 80 degrees when the normalized reactance of the VRC is varied from –2.0 to 2.0.

Backhoe loads sediment onto the bed of dump trucks during earthmoving work. The prediction of backhoe loading time is essential for ensuring safe cooperation between the backhoe and dump trucks. However, it is difficult to predict the instant at which the backhoe is ready to load sediment, because of the similarity in motions observed during gathering sediment. Moreover, since operators have different skill levels, the prediction requires a unique model for each operator. In this study, we attempt to predict the instant at which the backhoe is ready to load sediment into the dump truck. For this purpose, the beta-process hidden Markov model (BP-HMM) is employed to build a backhoe motion model for a specific operator. Time series data of backhoe loading motions for crushed rocks and wood chips, which were measured using 6-axis inertial measurement unit (IMU) sensors equipped at the cab, boom, and arm of the backhoe, were used for modeling with the BP-HMM. Several primitive motions of the backhoe, which occur at the completion of preparation before the loading process begins, were discovered as a result of the motion modeling based on the BP-HMM. We developed the prediction of the instant using three primitive motions. At best, the proposed method could predict the instant with a probability of 67% and 100%, at 6.0 s and 0.7 s before the loading motions began, respectively. This phased prediction can be used to reduce the idle time and risk for dump trucks during earthmoving work with the backhoe.

This paper proposes a drive control method for a linear compressor driven by a linear motor. Generally, the efficiency of a linear motor itself is relatively lower than that of a rotational motor. However, higher efficiency should be achievable by adding assist springs and driving at the resonant frequency. We have developed a novel resonant frequency estimation method that calculates the phase difference between a supplied voltage and a mover position. Since the phase difference changes according to the drive frequency with respect to the resonant frequency, our method calculates the phase difference by using the arctangent of a Fourier coefficient, which controls the drive frequency to keep the phase difference at 90°. Experimental results show that the proposed method can successfully track to the resonant frequency dynamically even if the mechanical condition is changed.

Contracture is assessed by the palpation performed by a physician or physical therapist based on the range of motion and slidability of a joint. However, the diagnosis of contracture depends on the experience and skills of the therapist and is subjective. Quantification could improve the efficiency and accuracy of the diagnosis. Therefore, we aim to quantify contracture palpation using a wearable tactile sensor by focusing on the frictional vibrations that are generated by the disturbance in the slidability during palpation. In a previous study, the pulse density that was measured by detecting a pulse from the sensor output at a threshold exhibited a good relationship with the subjective rating of the frictional vibration. However, as this sensor measures the propagational vibrations of the skin, it is influenced by the characteristics of individual therapist’s fingers. Thus, in this paper, three types of pulse detection threshold setting methods (constant, adjustment with a sine wave, and adjustment with white noise) are proposed. While conducting experiments, the same frictional vibration was applied on several participants by reproducing the frictional vibrations with a vibrator as a tactile display. The results demonstrated that the threshold adjusting method with white noise, which overlapped with the frequency band of the frictional vibrations, significantly reduced the variance in the finger characteristics.

Higher efficiency for a linear motor should be achievable by adding assist springs and driving the motor at the resonant frequency, even if the efficiency of the motor itself is relatively lower than that of a rotational motor. When using a linear motor with assist springs in a sealed chamber such as a compressor, it is difficult to install a position sensor in such an environment; thus, position sensor-less resonant frequency control is essential. This paper proposes a drive control method that realizes resonant frequency driving without a position sensor. The resonant frequency is estimated on the basis of the detected motor phase current, and the position is estimated using only the fundamental component of the current after removing the offset of the detected current with a high-pass filter. Experimental results show that the proposed method can successfully track the resonant frequency dynamically without a position sensor.

Contracture is diagnosed by doctors or physical therapists through palpation on slidability of patient’s joints. However, the diagnosis depends on tactile perception of each therapist and subjective. Quantifying contracture palpation can increase diagnosis accuracy and improve its efficiency. This paper proposes a contracture diagnosis system with a wearable skin vibration sensor. The sensor is worn on the finger pad of the therapist, allowing the therapist to touch patient’s joint for the contracture palpation with his/her bare fingertip. The sensor detects frictional vibration induced by sliding disturbance of the soft tissue around the joint when the patient’s arm or leg is moved. The system comprises the sensor, an audio interface, and a PC. The sensor output based on the frictional vibration is intermittent pulse waveform, and mean pulse density in the sensor output during the palpation is proposed for the evaluation. Experimental results on knee, shoulder and waist by a well-trained physical therapist show that the obtained pulse density has a good relation with subjective ratings and is consistent with the transition of the friction vibration according to increasing the repeated number of the palpation. These results indicate the availability and validity of the sensor system.

Sensing position and orientation of construction vehicle is an important issue for automation of construction process. We aim to develop sensing and visualization technologies for construction vehicles. Our target construction vehicle is backhoes. Construction vehicles are usually rented in construction fields. However, construction vehicles that can be rented do not have the sensing function. It is too hard to obtain backhoe position and manipulator pose without sensing information. This paper proposes an attachable sensor box that can measure backhoe position and orientation and the manipulator pose. The sensor boxes can be attached on backhoe metal surface by magnetic force without any additional manufacturing on the construction vehicle surface. By using WiFi communication and mobile battery the sensor box can be easily attached on large-size backhoe without wiring. After the work is done, it is easily detached by changing magnetic force power. At loading and scooping motion, backhoe arm and boom have large force exceeding 16 G. To attach the sensor boxes to the arm and boom, we designed an additional magnetic frame that can generate force of 1960 N. The sensor boxes were firmly attached to each joint and prevented to drop by any force from backhoes movement. We also measured the behavior of the backhoe in loading works in hot conditions over 30 °C and visualized backhoe pose and tip manipulator position.

Network requirements in optical access networks continue to expand, and softwarization of transmission function has been studied for complete virtualization of access systems. In order to meet the requirements posed by Internet of Things (IoT) applications, SDN/NFV-based access network is an excellent candidate for the edge computing platform needed. However, using a common transmission method is inefficient and will increase the computational loads placed on edge servers. We are aiming to maximize the benefits of SDN/NFV by realizing transmission that can respond to individual IoT application requests. This paper proposes novel application-oriented optical transmission control to enhance computational efficiency, especially for image processing applications. Our proposal is implemented on a general-purpose server and we demonstrate its performance on a 10-Gbps optical transmission system. The results show that the proposal does achieve transmission control to suit application requests.

Real-time motion control of Internet of Things (IoT) devices over the network has been studied with the emergence of advanced applications such as drones and industrial robots. However, delay fluctuation is caused by the network, which leads to the deterioration of control performance. This study introduces the concept of access edge that performs real-time motion control on an edge server located in the central office. To reduce the effect of delay in the access network on control performance, we propose and evaluate the control method that uses the delay information. In the proposed method, the delay information is actively obtained and sent to the controller for delay compensation. We construct an experimental system that simulates peer to peer communication in the access network, and evaluate the real-time control performance. We confirm that the settling time is improved by the proposed method in motor and drone experiments.

Mobility as a Service (MaaS) is the integration of various forms of transportation services for smart cities. The accuracy of estimating travel time is very important for MaaS. We focused on passenger travel time between the door of the train and the taxi stand in a railway station. Travel time depends on factors such as the distance, width of the platform or steps, number of people, etc. We proposed a behavioral model of Japanese passengers by observing passengers at the station. We improved our model. In the near future, we will implemented the passenger model on the microscopic simulator and also will show the results of the evaluation our model.

This paper assesses application-level QoS in the case where medical data is transferred with the block acknowledgment (ACK) mechanism of the IEEE 802.15.6 CSMA/CA. In the IEEE 802.15.6 medium access control (MAC), a node can transmit multiple MAC protocol data units (MPDUs) successively as a data block with the block ACK. By simulation, we first examine the effect of the block ACK mechanism on the application-level QoS in the case where medical data and background data are transferred from nodes to the hub. Simulation results show that the block ACK mechanism can improve the average media unit (MU) delay for medical data. We also reveal that the average MU delay for medical data decreases as the MU rate becomes larger, while the throughput for background data deteriorates under heavy traffic conditions. Furthermore, we examine the influence of medical data transmission with the block ACK on emergency data transmission. Numerical results show that the block ACK mechanism can deteriorate the average MU delay for emergency data if a large number of medical MPDUs are transferred successively as a data block.

This paper summarizes our studies on the software implementation of passive optical network (PON) physical-layer processing to maximize the flexibility of optical access systems. The key technologies to softwarerize optical line terminal (OLT) functions are described.

We propose a method for deciding an avoidable stop position for a vehicle to not disturb the surrounding traffic. A vehicle needs to pull over and make an autonomous emergency stop when a driver has a seizure and loses consciousness. To develop such an autonomous stop system on a narrow road, considering the surrounding vehicles and not hindering traffic are imperative to rescue a patient. We present an approach based on predicting the paths of other vehicles to avoid the ego-vehicle, a load index of a following vehicle driver for predicted paths, and an index that shows the interference between the paths of the following and oncoming vehicle. The path prediction is computed by applying a lattice path planner comprising the vehicle's position, posture, and path curvature. The load index of the following vehicle is calculated with respect to the cumulative curvature of the predicted path. The interference index of the following and oncoming vehicles is estimated by evaluating the area that both vehicles pass to avoid the ego-vehicle. The simulation results showed the ego-vehicle had to pull over by 1.5 m over the road shoulder and 1.0 m ahead of a parked vehicle on the opposite lane.

This paper presents a performance analysis of the IEEE 802.15.6 CSMA/CA by an approximate analytical technique called equilibrium point analysis (EPA). We develop an approximate Markov chain model of a wireless body area network (WBAN) with the CSMA/CA protocol taking into account data and ACK traffic. We then calculate throughput and average response time by the technique of EPA. The accuracy of analytic results obtained from the model is examined through simulation. Furthermore, we evaluate the stability behavior of the CSMA/CA network for varying the number of nodes, minimum number of contention window (CW), and maximum number of CW. Numerical results from both analysis and simulation show maximum number of CW strongly affects maximum number of nodes under constraint that the network is stable.

Disaster response robots are important for the investigation of nuclear power and chemical plants. To investigate an area, robots are required to climb straight or spiral staircases because equipment components are installed on different-level floors, which are connected by stairs. Installed stairs are narrow and steep because the location of equipment components is given priority in the environment inside plants. It is considerably difficult for a tracked vehicle, which has high mobility on stairs or rough terrain, to climb spiral staircases. We propose a method for a tracked vehicle to climb spiral staircases. A characteristic of the method is that the tracked vehicle uses the reaction force from a safety wall, which is installed to prevent objects from dropping down and damaging the equipment in plants. It is shown that the climbing motion is easy to perform using the wall without complicated control, as compared to when the wall is not used. The climbing motion of a robot on a spiral staircase using a wall is analyzed by employing a dynamic model. Experimental results show that it is possible to climb without slippage and at a high speed by maintaining contact along the wall. These results show that a tracked vehicle can access areas that could not be reached earlier through contact with the environment. Furthermore, it is considered that the climbing motion using contact with the environment can be applied to an autonomous mobile robot for automating inspection because complicated control is not required.

Pulmonary aspiration (mis-swallowing) causes more than 90% of pneumonia in elderly people. Ingested food unconsciously remaining in the piriform sinus occasionally Hows into the trachea leading to aspiration. If food remaining in the piriform sinus is noninvasively detected from outside the body, the risk of pulmonary aspiration can be evaluated We propose a method using near-infrared fluorescence emitted from fluorescence food, and this study reports the results of preliminary experiments using meat phantoms. The results showed that fluorescence from the fluorescence food embedded in thick meat was detectable and indicated that the method could be potentially applicable for evaluating the risk of pulmonary aspiration.

In this study, we aim to achieve path-planning for firefighter robots in large petrochemical complexes. In large environments, path-planning (e.g., Hybrid A*) requires a large computation memory and a long execution time. These constrains are not feasible for firefighter robots. In order to overcome these two challenges, we propose a two-stage hybrid A* path-planning. For the first stage we use a global path-planner that makes a path using a low-resolution grid map of 2 m. The global path-planner generates a path for an area of approx. 500 m ×1000 m in 10 seconds. In the second stage, we refine the path by using a local-planner that uses a local-map of 100 m ×100 m size around the robot with a high resolution grid of 1 m. The local planner receives its sub-goal from the global planner and recalculates a local path at a high speed of a few hundred milliseconds. Therefore, the local-planner can react to changes of the map due to obstacles in real-time. We evaluated our proposed method by comparing with conventional hybrid A* in simulated as well as real experimental data of petrochemical complexes. By employing the local-planner our method could drastically reduce the used memory and execution time for the re-planning. For a trajectory of 600 m, our method reduces the execution time by 99.2% for real data and by 94.34% for simulated data. The memory usage was likewise drastically reduced by 97.45% for real data and by 97.91% for simulated data.

Virtualization of optical access systems is being pursued to speedup service deployment and reduce CAPEX/OPEX. Several passive optical network (PON) specifications have been standardized and the applicability of virtualization to access networks is being considered. We have proposed a programmable optical line terminal (OLT) that performs OLT functions by software processing on general-purpose hardware.Unfortunately, general-purpose hardware struggles to achieve the 10-Gbps throughput demanded of access systems. Our work solves this problem by introducing software implementation techniques for PHY processing, a key component of computation complexity. This paper proposes a 2-stage parallel pipeline based on data-parallelism, and details its implementation with the use of look up table (LUT) for the Galois-field multiplier for implementing the standard OLT functions of forward error correction (FEC) and encryption. We implement our proposals on a general-purpose server with optical transmitter. Evaluation results show that proposals drastically accelerate algorithm performance and attain the throughputs of 10 Gbps with RS(255,239), 5.1 Gbps with RS(255,223), 2.5 Gbps with CTR-AES-128 and 1.2 Gbps with GCM-AES-128.

In large industrial plants, the inspection of production lines is a heavy and costly task that puts human inspectors at high risk. In order to overcome these challenges, we have developed an autonomous plant inspection system using a mobile tracked vehicle. In this paper, we propose an autonomous navigation method for tracked vehicles based on a contact force model that enables the robot to compensate for collisions with obstacles. The model considers the influence of the contact force on the linear and angular motion of the robot. Using the model, the controllable velocity range is derived during collisions. The experimental results show that the robot is safely controlled by complying with velocity constraints. In addition, our method can generate motions such as leaving wall, L-shaped curve and crosswise locomotion in straight passage while navigation alongside the walls. The method allows the robot to smoothly follow a target path, despite colliding with obstacles.

Structure tensor analysis on epipolar plane images (EPIs) is a successful approach to estimate disparity from a light field, i.e. a dense set of multi-view images. However, the disparity range allowable for the light field is limited, because the estimation becomes less accurate as the range of disparities become larger. To overcome this limitation, we propose a new method called sheared EPI analysis, where EPIs are sheared before the structure tensor analysis. The results of analysis obtained with different shear values are integrated into a final disparity map. As verified by extensive evaluations on 12 datasets with large disparity ranges, our method is comparably accurate to and much faster than a multi-view stereo method.

Network Function Virtualization (NFV) and Software Defined Network (SDN) are attracting attention with the goal being enhanced networks efficiency. To enhance flexibility and to meet user requests and conditions, software implementation of communications equipment is pursued as our approach. It becomes possible to implement various optical line terminal (OLT) functions on the common hardware by realizing the functions as software. The proposed approach prevents kinds of systems from increasing and simplifies maintenance. Throughput on lower-performance software comparing with dedicated circuits is a problem. This paper focuses on a cipher algorithm for the access network as typical of the more demanding OLT functions and proposes parallel implementation on CPU/GPU resources that offers real-time processing. This paper targets the CTR and GCM which are utilized on PON systems. For software implementation, this paper proposes a data-parallelization-based algorithm architecture. Evaluation results gathered from a many-core CPU simulator and GPU show that throughput is sufficient for the 5.37-Gbps CTR-AES128 process and the 914 Mbps GCM-AES128 process. The proposed parallel algorithm on a GPU is 141 times faster than serial processing. It is also found that the cipher circuit of 1-Gbps-class PON systems utilizing CTR-AES128 can be replaced with GPU.

CuInSe2 (CIS) thin films prepared on quartz substrates by the spin-coating method using the nanoparticles of CIS have been investigated by X-ray diffraction, scanning electron microscope, energy dispersive X-ray analysis. The chemical composition of CIS thin films sintered in N2 atmosphere was Se/metal(Cu+In) = 0.55 or 0.58, indicating a deficiency of selenium. However, CIS thin films sintered in atmosphere of N2 including Se vapor were improved Se/metal molar ratio. It can be seen that the improvement of crystallinity of CIS thin films using the nanoparticles of CIS were relatively effective for the sintering time and the sintering Se vapor.

We have developed a method to visualize search and rescue (SAR) dogs' activities from sensor data recorded by the SAR dogs' sensor vests. This paper proposes two methods for detecting continuous barking actions of SAR dogs, which locate victims by smell and then bark continuously to tell handlers where victims are located. Continuous barking action is detected from audio information and a dog's body motions. This detection method is based on dynamic time warping (DTW), which has been used successfully to analyze human audio information. Cyclic body motion was observed during dogs' barking motions. This cyclic motion can be detected by an inertial measurement unit (IMU) attached to the vest. A fast Fourier transform (FFT) is used to analyze a dog's barking motion. The proposed detection methods were evaluated using audio and IMU data recorded during actual SAR dog training sessions. The F-scores of the audio and motion-based barking detection methods were 0.95 and 0.90, respectively. As a trial, we marked victim locations on a map based on the body motion.

By the diffusion of Hospital Information Systems, many medical documents have been computerized. In addition, most of paper documents before computerization have been also scanned and archived as document images. These were usually converted to text data by using document analysis techniques and Optical Character Reader (OCR) and archived for medical document retrieval. However, the resolutions of some documents are not sufficient for character recognition because of storage spaces, scanning regulations and so on. Therefore, we cannot search desired keywords in the documents, as a result, these documents are not still used effectively in medical document retrieval systems. In this study, we discuss a keyword detection and extraction methods for these document images. As the first step of this study, this paper proposes a method to detect and extract desired words from these documents by using weighted dissimilarity and character sequence. Evaluation experiments using actual medical documents are conducted to discuss the effectiveness of the proposed method.

A new sensorless control method that uses the neutral point voltage from stator windings is proposed. Each inductance of the three phase coils of the permanent magnet synchronous motor (PMSM) is slightly affected by the rotor magnet flux. As a result, the neutral point voltage also changes depending on the rotor position. The rotor position can be estimated by detecting the neutral point voltage even in the zero-speed regions based on this principle. In addition to this, this method can be applied to nonsaliency PMSMs. This paper presents the principle and the experimental results when using a small-capacity PMSM without saliency.

A walking assistance apparatus which can promote gait training for motor palsy patients and outdoor walking for the elderly was developed. Our prototype uses a novel spatial parallel link mechanism and restraint free type of equipment. The apparatus supports complete leg alignment, including the soles of the feet from under the soles with flat steps, and assists walking behavior at the ankle, knee, and hip joints simultaneously without the use of Velcro tape. We proposed the control method according to the purpose of the user. For example, to promote gait training, this apparatus can select the impedance control method. User's legs are assisted along the trajectory with the ideal gait as well as walking. To promote outdoor walking for the elderly, this apparatus can select the torque control method. The apparatus calculates the necessary torques of each joint according to the posture and motion of the apparatus and user. In the case of the impedance control method, by adjusting the natural frequency of the ideal dynamic equation for the flat steps under the soles, the user felt it was easier to walk by frequency entrainment. On the other hand, to promote outdoor walking of the user, this apparatus has to be used not only on the flat ground but also on stairs or slopes with freedom behaviors, and it has to make the user feel comfortable. To address this problem, the torque control method proposed to use real-time acceleration data instead of using the template trajectory. By using this method, the subject could walk on the slope and stairs continuously. Finally, %MVC of the user's EMG while walking equipped with this apparatus was measured, and the influence of the difference using between the impedance and torque control method was confirmed by interviewing the subjects. From the result, the impedance control method assisted the motion of kicking the ground, and the torque control method assisted the motion of raising-up the leg. Therefore, the impedance control method is good for gait training, and the torque control method is good for walk promotion.

A new position sensorless method is proposed for surface permanent magnet synchronous motors in all speed regions. In this method, two sensorless methods are switched in accordance with the rotating speed. In low speed regions, an Induced Voltage caused by a Magnetic Saturation (IVMS) based method is used. This method is switched to a back EMF based sinusoidal current method at a medium or high speed. Some control blocks are commonly used for both methods. Therefore, the methods can be switched smoothly and rapidly. The basic idea and the experimental results are described.

Potential of a cloud system combining a smart device and cloud servers is increasing. One of the representative examples is "Siri" which offers a friendly web knowledge navigator based on natural language user interface. Since latest portable devices equip not only a microphone but also a high resolution camera, this kind of cloud based framework is also promising to create various kinds of video based recognition systems. There are two essential components for it: a smart device with a high-resolution camera which is responsible for detecting feature from input video and cloud servers which execute recognition or data search using big data as shown in Fig. 1.

We developed a calorimetric device for measuring thermal diffusivity non-destructively. The calorimetric device was based on the principle that temperature phase delay between a heater and a resistance temperature detector (RTD) is affected by thermal diffusivity of the region between them. The device consisted of an Au wire, as oscillating heat source and a piezoresistor as an RTD. We exerted the experiment, in which air pressure was changed, and observed that the device measured thermal diffusivity of the air.

This paper presents a new system for the online and real-time recognition of handwritten mathematical formulas. Mathematical formulas are inputted into the system as hand drawings on a computer screen using a data tablet. The system analyzes the inputted data stroke by stroke in order to recognize the structure of the formula based on the relationship of bonding boxes that include components of the formula. A recognized result is obtained as a source code from the "LATEX" type serving system. Our system can recognize arbitrary combinations of superscripts, subscripts, square roots, overlines, underlines, and fractions. The input of the mathematical formula is independent of stroke order. In several experiments, the system proved to be very effective for inputting various mathematical formulas in real-time.

Keypoint extraction has lately attracted attention in computer vision. Particularly, Scale-Invariant Feature Transform (SIFT) is one of them and invariant for scale, rotation and illumination change. In addition, the recent advance of machine learning becomes possible to recognize a lot of objects by learning large amount of keypoints. Recently, cloud system starts to be utilized to maintain large-scale database which includes learning keypoint. Some network devices have to access these systems and match keypoints. Kepoint extraction algorithm utilizes only spatial information. Thus, many unnecessary keypoints for recognition are detected. If only Keypoints of Interest (KOI) are extracted from input images, it achieves reduction of descriptor data and high-precision recognition. This paper proposes the keypoint selection algorithm from many keypoints including unnecessary ones based on spatio-temporal feature and Markov Random Field (MRF). This algorithm calculats weight on each keypoint using 3 kinds of features (intensity gradient, optical flow and previous state) and reduces noise by comparing with states of surrounding keypoints. The state of keypoints is connected by using the distance of keypoints. Evaluation results show that the 90% reduction of keypoints comparing conventional keypoint extraction by adding small computational complexity.

In this paper, we propose a producing method of softness sensation using device vibration. The basic idea of the softness producing method is to provide a softness illusion of stimulating cutaneous mechanoreceptor by vibration. A prototype device is developed to verify effect of parameters of the vibration: frequency, amplitude and wave shape, on softness sensation. We create a database which indicates relationship between softness sensation and the vibration parameters to control the prototype device. Then subjective experiment is conducted for validation of the softness producing method. Experimental result shows that the proposing method can be used for softness sensation producing with valid resolution.

In this paper, we discuss a position sensorless drive for an axial structure motor without rare-earth metals in a low-speed region. Since this motor has no saliency, a conventional position sensorless drive using harmonic injection cannot be applied. Therefore, we apply a position sensorless drive based on Induced Voltage caused by Magnetic Saturation (IVMS) to this motor. As IVMS becomes very small for motors without saliency, we introduce a differential amplifier and adjust its gain appropriately. We implemented this method by using a microcomputer and an analog circuit. As a result, excellent drive performance in a low-speed region was obtained in experiments for the motor (200V/11kW).

This paper proposes a start-up method for reducing electrical energy at starting periods. Minimum current startup is achieved by combining two position sensorless control methods that require no harmonic injection. Position sensorless control for low-speed regions utilizes the induced voltage caused by the magnetic saturation of a permanent magnet synchronous motor with conventional block commutation drive. At acceleration, the position sensorless control is changed to the second method based on back-EMF with sinusoidal commutation drive. This paper also addresses when to change the control, as unconsidered switching may result in start-up failure when used with applications that have periodic load properties. A simple yet powerful method to identify the most suitable timing considering load variation is proposed. The proposed methods are validated using a reciprocating compressor with a large periodic load torque.