Keywords:
Abstract: According to the ASEAN's guidelines for sustainable agriculture, there is an urgent need for reduction in resource use and more climate resilient methods in the agricultural process. High precision techniques practiced by control and instrumentation engineers can directly contribute to sustainability in food production and agriculture. Robotics and automation technology can reduce labor and eliminate the use of toxic chemicals and pesticides, thus allowing environmentally-friendly and cost effective solutions for farmers throughout the life cycle of a crop. However, there are challenges in how to make autonomous systems operate reliably in uncontrolled outdoor environments and to be able to interact with living things.

Keywords: Guidance and Flight Control, Autonomous Decentralized Systems, Intelligent Control
Abstract: This paper proposes the concept of demand capacity balancing (DCB) for urban air mobility (UAM) using multi-agent reinforcement learning (MARL). It is crucial to manage the demand capacity balance for UAM with limited flight time prior to flight in order to ensure safe and efficient flight. Urban air traffic management (UATM) that manage UAM operations are expected to be decentralized systems in terms of scalability since the number of UAMs in operation is expected to increase. In DCB for existing aircraft, the buffer that determines the ground delay time is not optimized and is an centralized system, so new DCB system is needed. The concept of DCB proposed in this paper is expected to resolve the problems of efficiency and scalability of conventional DCB system. The DCB algorithm based on the concept is developed using MARL and simulation was conducted with a case study. The result of simulation shows that the developed DCB algrithm work effectively and fairly. The concept and algorithm proposed in this paper will serve as a basis for the development of future DCB systems for UAM.

Keywords: Guidance and Flight Control, Transportation Systems
Abstract: The use of unmanned aerial vehicles (UAVs) for logistics has received growing attention due to labor shortages and other societal challenges. However, UAV operations are constrained by inherent limitations such as low battery capacity, short flight range, and high sensitivity to environmental disturbances, particularly wind. Moreover, effective route planning becomes increasingly complex in dynamic environments with time-varying wind conditions. This paper proposes a route planning algorithm for logistics drones that simultaneously considers multiple transportation cost factors—namely, inter-station distance, wind direction, and the number of power supply stops—under time-varying conditions. A multi-objective Dijkstra algorithm is applied iteratively to compute routes that minimize the aggregated transportation cost. The structure of the proposed method is presented, and its effectiveness is demonstrated through numerical simulations.

Keywords: Robust Control, Guidance and Flight Control, Mechanical Systems Control
Abstract: This paper gives a comparative study of Sliding Mode Control (SMC) with smoothing factors and Linear State-Feedback Control (LSFC) with internal saturation for satellite attitude control. It is well-known that SMC has a strong robustness against so-called ``matched uncertainty'', which is characterized as the uncertainties belonging to the image of the control input matrix of plant system. Thanks to this property, visual-feedback based satellite attitude control using SMC is discussed for Hayabusa2 in a flyby imaging mission near an asteroid ``2001 CC21'' on the way for its ``Extended mission'' to the asteroid ``1998 KY26.'' The feasibility study of capturing the asteroid image of ``2001 CC21'' using SMC with smoothing factors during the flyby has been conducted, and the usefulness of SMC with smoothing factors compared to PD (Proportional and Derivative) controller has been demonstrated. We show the result of applying LSFC combined with internal saturation to the same mission, i.e.~capturing the asteroid image of ``2001 CC21'' during the flyby, and also point out issues for further control performance improvement.

Keywords: Guidance and Flight Control, Mechanical Systems Control, Robotic and Automation Systems
Abstract: Tilt-rotor aircraft has both the advantages of fixed-wing aircraft and rotor-wing aircraft. However, tilt-rotor aircraft decreases its stability in the transitional mode according to the change of the rotors’ tilt-angle. This study proposes a new type of tilt-rotor aircraft with arm-mechanisms, which can keep the flight stability in the transitional mode. First, the advantages of the aircraft are explained. Then, from the governing equations of the aircraft, its equilibrium state is explained, and the control strategy is proposed based on the equilibrium state. Finally, simple experimental results of a tilt-rotor aircraft are shown to verify the stability in the beginning of the translational motion.

Keywords: Guidance and Flight Control, Discrete Event Systems, Autonomous Decentralized Systems
Abstract: By configuring a laser interferometer among three spacecraft in orbit, their relative positions can be measured and controlled with high precision. This enables us to construct a large-scale gravitational wave telescope that meets advanced requirements, which are impossible to achieve with a ground-based telescope. This study focuses on formation flight control to achieve laser link acquisition for establishing a laser interferometer among three spacecraft. Specifically, we propose an event-triggered controller of each spacecraft, and derive stability conditions to ensure successful laser link acquisition. A numerical example illustrates that the proposed controller effectively reduces the communication among spacecraft while maintaining the laser link acquisition capability.

Keywords: Transportation Systems, Guidance and Flight Control
Abstract: At Wakkanai Airport, the northernmost passenger jet airport in Japan, northerly winds blowing from Soya Bay create crosswinds on the runway and are the main cause of the low utilization rate. Various measures have been implemented to improve the service rate, but have yet to have a noticeable effect. As a new attempt, a feasibility study of a hybrid concept of carbon neutrality and crosswind protection has been initiated, utilizing the effects of wind turbines and windbreaks along the coastline. This paper reports some of the results of simulations on the effects of localized weakening of crosswinds on airport runways on flight characteristics during landing.

Keywords: Mechatronics Systems, System Engineering, Standard of Measurement
Abstract: The increasing demand for energy efficiency and sustainable transportation has led to the integration of regenerative braking systems (RBS) in electric vehicles, including electric bicycles (e-bikes). This paper investigates the design and performance of a RBS tailored for electric bikes, focusing on capturing and storing kinetic energy during deceleration. The system utilizes a battery energy storage unit to absorb recovered energy, thereby enhancing overall energy efficiency and extending the operational range of the bicycle. The results demonstrate that implementing an RBS can significantly improve energy utilization, reduce dependency on frequent battery recharging, and contribute to environmentally friendly transport solutions. This paper provides both theoretical analysis and experimental design considerations for developing practical regenerative systems in small-scale electric vehicles.

Keywords: Mechanical Measurement, Mechanical Systems Control, Modeling, System Identification and Estimation
Abstract: The objective of this study is to analyze the dynamic behavior of a bicycle under rear-load conditions as a preliminary step toward designing a steering-assist method to improve stability. To this end, a mathematical model of the bicycle is first derived. Simulations using the model are conducted to evaluate the effects of variations in front-wheel geometry and center of mass location on the bicycle's dynamic response. Subsequently, real-world riding experiments are performed using an actual experimental bicycle. The collected data are analyzed to identify unintended behaviors induced by the rear load. As a result, both simulations and experimental results reveal oscillatory behavior in the steering response. These findings confirm that the dynamic characteristics of the bicycle under rear-load conditions have been successfully captured and analyzed.

Keywords: Modeling, System Identification and Estimation, Mechanical Systems Control
Abstract: The objective of this research is to realize a posture stabilization control system that suppresses bicycle wobble, thereby reducing accidents caused by rider-induced unbalance. To achieve this goal, we first derive a dynamic model that accurately represents bicycle behavior. Based on this model, a speed-assist control system is designed to enable the bicycle to stably travel along a specified turning radius. The effectiveness of the proposed control system is then verified through experiments using the developed bicycle hardware.

Keywords: Modeling, System Identification and Estimation, Mechatronics Systems, Intelligent Systems
Abstract: Instead of mathematical or matrix-based observers applied in recovery systems, this paper proposes an alternative based on a series of recurrent neural network based models that estimate the position signal of each joint in a six-degree-of-freedom manipulator robot. These estimations could serve as a part of a recovery system for fault detection and correction. Two types of faults, bias and noise are induced into the sensor signals as a benchmark to be detected in real-time to evaluate the robustness of this proposal. Therefore, the whole system considers two types of recurrent neural networks, where their input is the torque applied to each joint of the 6-degree-of-freedom manipulator robot. The proposed models manage to estimate the position sensor signal with a mean squared error lower than 1x10−4.

Keywords: Robotic and Automation Systems, Robust Control, Adaptive and Optimal Control
Abstract: This paper proposes a controller design approach that applies the repetitive control (RC) technique to inverse dynamics control to ensure the stability of path-tracking control for robotic manipulators. The RC technique is utilized to compensate for model uncertainties that may arise in controllers based on conventional inverse dynamics control methods. By leveraging the advantage of RC in stabilizing periodic input signals, the proposed approach enhances robustness against model uncertainties in manipulators that track periodic trajectories. Furthermore, an optimal controller is designed for the applied RC technique, developing a control framework that guarantees robust optimality. Finally, simulation verification is conducted on a multi-axis manipulator system.

Keywords: Mechanical Systems Control, Nonlinear Control
Abstract: This paper proposes a new admittance controller that combines independently designable task- and joint-space dynamics by solving a prioritized minimization problem. The controller can be used with redundant manipulators and is capable of tolerating singular configurations. The end-effector follows the task-space proxy, and the behavior in the nullspace is determined by the joint-space dynamics. Task- and joint-space dynamics are combined through inverse kinematics formulated as a lexicographic minimization problem. The controller allows for the independent design of task- and joint-space dynamics and is capable of realizing nonlinear dynamic characteristics. The controller was tested with a seven-DOF collaborative robot, Kinova Gen3, which is equipped with joint torque sensors.

Keywords: Mechatronics Systems, Robotic and Automation Systems, Mechanical Systems Control
Abstract: This paper proposes a workspace position control method for a parallel-link robot. In general, forward kinematics, which is the calculation of motion information in the workspace from joint space, is computationally expensive because it is difficult to solve directly. To address this, the paper applies an extended Kalman filter (EKF) to estimate the forward kinematics of the parallel-link robot. Additionally, a workspace position controller is designed based on the values estimated by the EKF. The effectiveness of the proposed method is demonstrated through simulation results.

Keywords: Safety, Environment and Eco-Systems, Robotic and Automation Systems, Nonlinear Control
Abstract: This study examines the effectiveness of safety assist control in the torque control of robot manipulators. Because robot manipulators are subject to input constraints, we designed a control barrier function using a viability kernel. Furthermore, to enhance computational efficiency, the control barrier function is improved based on hybrid systems theory, and its effectiveness is validated through simulation.

Keywords: Robotic and Automation Systems, Factory Automation, Mechanical Systems Control
Abstract: In this study, a 3-D stable grasping using a surface contact model of a 2-D fingered hand is investigated. Conditions that only gravity is considered and the normal vector of the grasping surface is in the horizontal direction, developed in the past, are no longer required, and grasping stability that allows grasping in any direction is discussed, taking inertial forces into account. Simulation and actual machine verification were conducted.

Keywords: Robotic and Automation Systems, Mechanical Systems Control, Modeling, System Identification and Estimation
Abstract: Small unmanned surface vehicles (USVs) are increasingly being used for applications such as environmental monitoring and coastal research. However, lightweight and compact USVs weighing less than 10 kg are highly susceptible to disturbances such as wind and waves, and their control systems are limited in complexity owing to limited onboard computational resources. In this study, a practical control allocator is implemented by developing an efficient and power saving mechanism for a small catamaran USV. Station-keeping was implemented using this mechanism with proportional derivative control, and experiments were conducted in a real environment. As a result, stable position keeping and azimuth control were achieved. Furthermore, the data obtained during station-keeping suggested that a simplified Fossen model might be applicable.

Keywords: Mechatronics Systems, Guidance and Flight Control, Modeling, System Identification and Estimation
Abstract: We developed a pacing robot for long-distance running in athletics. This robot detects the lane markings on an athletics track using a camera to recognize the running lane. In the recognition process, tracking windows are placed at regular intervals along the lane markings, and an approximate straight line is derived from the centroid positions within the tracking windows.

The recognized running lane is used to calculate the steering angle based on the Pure Pursuit method, which takes into account the relative angle between the target point and the robot, as well as the lateral deviation of the robot from the running lane.

Furthermore, the robot is equipped with a 3D LiDAR sensor, allowing it to estimate its position while driving. By combining the camera-based and 3D LiDAR-based navigation, the system functions as a backup in case of camera detection failures.

During 3D LiDAR-based driving, the amplitude of the steering angle during motion is larger compared to camera-based driving. From this observation, we hypothesized that time delay might be related to the system's stability and conducted a stability analysis considering the time delay.

Keywords: Robotic and Automation Systems, Transportation Systems, Mechanical Systems Control
Abstract: A robot system that can autonomously navigate between indoor and outdoor environments is developed by using the same algorithm. A series of operations from map data construction to autonomous driving can be performed in a flat indoor environment by self-localization with LiDAR sensing and odometry. However, in outdoor environments, the accuracy of odometry is reduced due to the characteristics of the road surface, which limited the range of autonomous driving to a small area and revealed challenges related to self-localization.

Keywords: Robotic and Automation Systems
Abstract: This study aims to automate an outdoor mobile robot in an environment with restricted areas. The proposed method using a stereo camera to detect traffic cones and generate restricted zones while mapping and coverage path planning to navigate autonomously. The system was evaluated through real-world experiments, demonstrating the effectiveness of the approach in detecting traffic cones, generating restricted areas for cleaning tasks.

Keywords: Robotic and Automation Systems, Networked Sensor System, Transportation Systems
Abstract: In this paper, we propose an autonomous mobile robot system that uses an ultra-wideband (UWB) positioning system in an overhead trellis environment composed of pears and grapes. The environment makes it difficult to use satellite positioning systems due to the presence of branches and wires. To realize autonomous driving by mobile robots in an overhead trellis environment, we propose a mobile robot system that uses UWB. In this system, the UWB positioning area is split into multiple cells, each of which has its own coordinate system. To facilitate seemless navigation between these cells, we propose a cell switching method that enables robots to transition across different coordinate systems. The experimental results validate the effectiveness of the proposed method for autonomous robot navigation along a predefined multi-cell path.

Keywords: Robotic and Automation Systems, Mechatronics Systems
Abstract: This paper is concerned with an advanced object recognition approach for realizing an innovative core-setting robot in foundry. The accuracy of keypoint matching in object recognition can be improved by detecting a greater number of correct correspondences. However, extracting correct correspondences from the 3D reference model and the scene point cloud is challenging due to occlusion and differences in data accuracy. In particular, the proportion of incorrect correspondences tends to be increased by the mismatch between the 3D reference model representing all surfaces and the scene point cloud representing a partial view of the target object. To address this issue, we propose the keypoint detection method that reduces the number of incorrect correspondences by removing hidden surfaces from the 3D reference model prior to keypoint detection. These hidden surfaces refer to regions that cannot be observed from any viewpoint. By excluding them, the number of incorrect correspondences can be reduced in advance. Experimental results show that the proposed method improves the precision of keypoint matching compared to conventional approaches using full 3D models.

Keywords: Biological and Physiological Engineering, Analytical Measurement, Integrative Biophysical Engineering and Informatics
Abstract: This study investigates the dynamics of autonomic nervous system regulation during the Eight Stages of Kyudo shooting by comparing expert and novice practitioners. Heart rate variability (HRV) was recorded using Polar H10 sensors during standardized four-shot trials. Frequency domain analysis using Welch’s method quantified low frequency (LF), high frequency (HF) components, and their ratio (LF/HF), across shooting phases. Experts exhibited stable LF/HF ratios (2.0–3.5) with minimal phase-to-phase variation (<15%), whereas novices showed pronounced fluctuations (LF/HF >5.0), particularly during transitions such as kai and hanare . Pre-shooting LF/HF biofeedback training reduced peak LF/HF variability by approximately 30%. These findings reveal distinct psychophysiological patterns related to skill level and suggest the efficacy of biofeedback in enhancing autonomic regulation learning in Kyudo. 　The LF/HF ratio, an index of autonomic nervous balance, was used to assess sympathetic and parasympathetic activity. Implications for training optimization and future research are discussed.

Keywords: Biological and Physiological Engineering, Standard of Measurement, Safety, Environment and Eco-Systems
Abstract: This study presents a method for evaluating electromagnetic field (EMF) exposure from multiple sources operating simultaneously across a wide range of frequencies, based on the guidelines defined by the International Commission on Non-Ionizing Radiation Protection (ICNIRP) in its 2020 publication. The proposed approach adopts the normalized exposure ratio, which combines the exposure values from all relevant sources and frequency bands. A key issue addressed in this study is the lack of standardized measurement methods for absorbed power density (Sab) in high-frequency bands above 10 GHz, despite the existence of defined exposure limits. A comparison reveals that the difference between the limits for Sab and incident power density (Sinc) for the general public is relatively small when compared to the gap between the general public and occupational exposure limits. Therefore, the use of Sinc is proposed as a temporary surrogate for Sab in frequency ranges where direct measurement is not yet feasible. This approach enables comprehensive exposure assessment from multiple sources under both near-field and far-field conditions. It also serves as a practical alternative while Sab measurement techniques for high-frequency applications such as 5G, WiGig, and next-generation wireless technologies are still under development.

Keywords: Biological and Physiological Engineering, Integrative Biophysical Engineering and Informatics
Abstract: Microneurography remains the only method for measuring peripheral sympathetic nerve activity, but the percutaneous insertion of microneedles imposes a significant burden on patients. Therefore, a non-invasive alternative is needed. This study investigates the feasibility of using fingertip peripheral arterial stiffness and knee-lead electro-sympathetic nerve activity (ESNA) as non-invasive indices of peripheral sympathetic nerve activity. Seventeen healthy male participants underwent a cold pressor test (CPT) while ESNA (chest and knee leads) and fingertip peripheral arterial stiffness were recorded simultaneously. Additionally, in a separate invasive experiment with one male participant, muscle sympathetic nerve activity (MSNA) was recorded using microneurography, alongside the aforementioned non-invasive measures, to validate their potential as proxies for MSNA. The results showed significant increases in peripheral arterial stiffness (p < 0.01), chest-lead ESNA (p<0.01), and knee-lead ESNA (p<0.01) during cold water stimulation. Furthermore, in the invasive experiment, knee-lead ESNA exhibited a downward trend, similar to fibular MSNA, during cold water stimulation. These findings suggest that knee-lead ESNA and fingertip arterial stiffness reflect peripheral sympathetic responses and may serve as non-invasive alternatives to microneurography. Further studies with larger sample sizes are needed to validate their clinical utility.

Keywords: Biological and Physiological Engineering, Human Interfaces, Medical and Welfare Systems
Abstract: Mechanical vibration delivered to muscle or tendon can evoke the tonic vibration reflex (TVR), a spinal response that generates involuntary muscle contraction without external actuation. Leveraging this mechanism, vibration-based stimulation offers a lightweight alternative for motor rehabilitation, particularly in patients who lack sufficient voluntary control. This study examines how joint-specific vibration frequencies—i.e., heterogeneous frequency combinations across muscles—affect multi-degree-of-freedom (multi-DOF) upper-limb motion driven solely by TVR. Vibration was applied to four muscles spanning the shoulder and elbow joints: posterior deltoid and pectoralis major (controlling shoulder horizontal adduction/abduction) and biceps and triceps brachii (controlling elbow flexion/extension). Experimental results show that differential stimulation frequencies significantly modulate joint-specific activation strength and inter-joint coordination, revealing frequency-dependent tuning characteristics of TVR in a multi-joint context. These findings provide a foundation for adaptive vibration-based rehabilitation systems that align assistance with individual neuromuscular needs while avoiding bulky robotic actuation.

Keywords: Temperature Measurement, Signal and/or Image Processing, Biological and Physiological Engineering
Abstract: Infrared thermography has gained interest as a tool for non-contact measurement of blood circulation and skin blood flow due to cardiac activity. Particularly, blood vessels on the surface, such as on the back of the hand, are suited for visualization. However, standardized methodologies have not yet been established for areas such as the face and neck, where many blood vessels lie deeper beneath the surface, and external stimulation for measurement could be harmful. Here, we propose Synchro-Thermography for stable monitoring of facial temperature changes associated with heart rate variability. We conducted experiments with eight subjects and measured minute temperature changes with an amplitude of about 10 mK on the forehead and chin. The proposed method improves the temperature resolution by a factor of 2 or more, and can stably measure skin temperature changes caused by blood flow. This skin temperature change could be applied to physiological sensing, such as blood flow changes due to injury or disease, or as an indicator of stress.

Keywords: Signal and/or Image Processing, Biological and Physiological Engineering, Intelligent Systems
Abstract: In conventional esophageal cancer diagnosis, the presence or absence of a lesion is confirmed by observing a virtual slide (WSI: Whole Slide Image) [1] during a biopsy, and the lesion is then removed by endoscopic surgery (ESD), after which the stage of the cancer is determined. ESD requires advanced skills from doctors, and also places a heavy burden on the treated patient, who must be hospitalized for about a week. In particular, advanced patients must undergo treatment surgery in addition to ESD. This is a current challenge in the diagnosis of esophageal cancer. In this study, we aim to reduce the burden on advanced patients by skipping ESD by analyzing virtual slides using AI to estimate the stage of cancer at the biopsy stage.

Keywords: Adaptive and Optimal Control, Intelligent Control, Robotic and Automation Systems
Abstract: This paper propose a path planning method for the automatic parking with naturally generating a switching point by formulating a nonlinear MPC problem where the forward motion from an initial point and the backward motion from a target point are simultaneously considered. The forward and backward vehicle motions are composed into a whole kinematic model and the condition for the switching point that the forward and backward trajectories coincide with each other is integrated as the termination one into the evaluation function of the MPC problem formulation. The simulation result is performed to show that the whole vehicle trajectory for the automatic parking is efficiently generated without assigning the switching point in advance.

Keywords: Robotic and Automation Systems, Transportation Systems, Modeling, System Identification and Estimation
Abstract: In recent years, high-precision self-position estimation has become essential for autonomous driving. Various position estimation methods are widely used for autonomous driving, including Scan Matching, RTK-GNSS/IMU, and DR. However, it is known that these position estimation methods suffer from degraded estimation accuracy depending on the environment. Therefore, previous research proposed an integrated positioning approach. By pre-evaluating the position errors of each method, this approach achieves position estimation that adapts to different environments. However, this approach cannot adapt to real-time environmental changes and cannot consider coordinate system offsets. Therefore, this study aims to achieve reliable self-position estimation in diverse environments and to contribute to expanding the operational range of autonomous driving. To resolve the above problem, we propose an integrated position estimation method that considers environment-dependent uncertainty and the coordinate system offset. In the evaluation tests, 98.8% of the two-dimensional errors were within 0.1 m, and accuracy improved by 22.1% compared to the method without coordinate system offset correction. Our proposed method provides highly accurate position estimation at the 10 cm level and contribute to expanding the operational range of automated driving systems.

Keywords: Adaptive and Optimal Control
Abstract: This research proposes a test pattern methodology that ensures comprehensive ACC validation by achieving 100% coverage of the area of interest. The experiment involves two vehicles: an ACC-equipped following car and a preceding car that generates the test pattern. The preceding vehicle follows a reference trajectory designed using an ellipsoidal reference and controlled by Model Predictive Control (MPC), which accounts for the ACC system’s behavior and constraints. By adjusting the reference parameters, the ellipsoidal graph’s size can be modified, enabling complete scenario coverage. This approach ensures a realistic and safe testing environment, effectively validating ACC performance under diverse conditions.

Keywords: Adaptive and Optimal Control, Mechanical Systems Control, Nonlinear Control
Abstract: This paper presents an application of model predictive control (MPC) with constraints using control barrier functions (CBF) to adaptive cruise control (ACC) with passenger safety to maintain a predefined distance between two consecutive vehicles on a line. The paper discusses applications of the CBF-based MPC for ACC to vehicles on the plain.

Keywords: Signal and/or Image Processing, Identification and Estimation, Analytical Measurement
Abstract: Many organizations have tried to improve communication through various efforts, but most focused on settings like meetings or phone calls. Past studies mainly analyzed verbal interactions, with few examining nonverbal cues such as facial expressions and gestures. This study investigates how nonverbal communication in everyday lab conversations relates to productivity. Using Raspberry Pi devices to record and analyze interactions, the results show that serious expressions during discovery or learning promote task progress. Stable emotions made participants more open to advice, while emotional shifts often led to helpful support for completing tasks.

Keywords: Identification and Estimation, Signal and/or Image Processing
Abstract: Assessing avocado ripeness accurately is essential for post-harvest management and quality control. This study uses Near-infrared (NIR) spectroscopy and machine learning to classify avocados as Raw or Harvestable. To improve existing categorization methods, a non-invasive, cost-effective, and efficient solution is needed. Chiang Mai's Royal Project Gardens yielded 120 kg of Buccaneer avocados. An affordable NIR sensor collected spectral data at 18 wavelengths. Multiplicative Scatter Correction (MSC) and outlier elimination improved spectral quality. Outlier reduction and MSC improved spectral quality. Spectral characteristics enabled a Random Forest model to classify 99% with an area under the curve (AUC) of 0.99. This study introduces a reliable, automated ripeness classification system to improve agricultural precision. Next, the AI model will be optimized for mobile edge devices to make it more accessible to small farmers.

Keywords: Signal and/or Image Processing
Abstract: Image denoising is a fundamental task in image processing, where deep learning models have demonstrated remarkable effectiveness. While much attention has focused on architectural advancements, the selection of loss function remains a critical yet often underexplored factor influencing denoising quality. Many studies default to Mean Squared Error (MSE) loss, which, although widely used, tends to produce over-smoothed outputs with diminished fine details. Additionally, alternative loss functions are frequently adopted without systematic justification. This study presents a comprehensive evaluation of an Attention U-Net trained using eight loss functions: MSE, MAE, SSIM, Huber, Charbonnier, and three hybrid combinations of MSE and SSIM. Experiments were conducted on 8-bit RGB facial images corrupted with random Gaussian noise, random salt-and-pepper noise, and a combination of both. Results show that MSE consistently yields the highest PSNR and IEF under Gaussian and mixed noise, while the hybrid loss (0.8×MSE + 0.2×SSIM) achieves the best SSIM scores and overall performance in salt-and-pepper noise scenarios. These findings underscore the importance of loss function selection in optimizing denoising outcomes for diverse noise conditions.

Keywords: Networked Sensor System, Remote Sensing, Signal and/or Image Processing
Abstract: In recent years, with the rise of information-based societies, smart cities have been proposed, and Wireless Sensor Network (WSN) have attracted attention as a means of collecting information. In addition, the location information of each node is essential for the operation of WSN, so it is necessary to consider the localization system to be installed. A localization system for a WSN must be usable both indoors and outdoors, and it should be compact and power-efficient. In this study, we aimed to improve the localization accuracy over a wide range by applying the Apollonius circular trajectory theory to RSSI (Received Signal Strength Indicator) from three or more localization anchor nodes and performing maximum likelihood estimation for the likelihood function. As a result, in simulation verification, the localization error was reduced by approximately 79.7% compared to conventional methods. However, increased distance between the nodes leads to a larger discrepancy between the true and estimated RSSI distance ratio values, which in turn degrades localization accuracy. In addition, since multipath fading and antenna directivity characteristics are not taken into account, further consideration, including experiments, is required.

Keywords: Robotic and Automation Systems, Signal and/or Image Processing, Mechatronics Systems
Abstract: With the aging and shrinking workforce responsible for maintaining power equipment, there is an increasing need for research and development of robotic applications in substations to enhance patrol and inspection efficiency, as well as to enable rapid response to equipment failures. In this study, we developed an autonomous push-button operation system using the quadruped robot Spot, which is equipped with a manipulator arm designed for remote operation tasks. Experimental results confirmed that a series of operations—from detecting a 2D tag located beneath the button to executing the button press—can be performed autonomously.

Keywords: Robotic and Automation Systems, Signal and/or Image Processing, Mechanical Systems Control
Abstract: As global resource demand and environmental issues escalate, transitioning to a circular economy is essential. This study, part of a NEDO project, develops an autonomous sorting system for electronic waste in home appliance scrap yards. Traditional localization methods like SLAM and GNSS struggle in dynamic environments. To address this, we propose an artificial landmark-based approach. The system uses vertically stacked geometric landmarks, enabling robust detection and self-localization within a 30 m range. Experiments confirm that localization accuracy improves when landmarks are spaced further apart. Future work will focus on reducing detection errors and stabilizing performance in scrap yards. Additionally, we will implement a switching mechanism that uses either camera images alone or a combination of LiDAR and images based on landmark distance.

Keywords: Modeling, System Identification and Estimation, Intelligent Control, Network System Integration
Abstract: We address the problem of estimating both the self-location and target positions of multi-agent systems while controlling agents to improve performance under noise with variance dependent on distance. We define a likelihood function of the estimation probability, which takes the observed relative positions as input, to obtain the estimates by maximizing it and formulate a movement strategy to sharpen its shape through minimization the likelihood. Furthermore, we conduct the estimation and movement processes alternately using Gradient Descent/Ascent (GDA) and demonstrate the effectiveness of the proposed method through simulations.

Keywords: Nonlinear Control, Robust Control, Computer Aided Design
Abstract: This paper presents a new framework for static output feedback stabilization of uncertain polynomial systems with bounded actuators. The design procedure is divided into two steps. In the first step, parameter-dependent state-feedback controller is obtained to stabilize the given system with a guaranteed bound on the input magnitude. Then, this controller is used as input to the second step to synthesize a robust static output-feedback gain which is independent of the uncertain parameters. In both steps, the synthesis conditions are given in terms of parameter-dependent linear matrix inequalities which are solely convex in the decision variables and hence do not require any iteration. For the both steps, a parameter-dependent Lyapunov function is employed to reduce the conservativeness of the design conditions. The proposed approach leads to enhanced static output feedback design with more computationally tractable formulation than the existing iterative approaches. Effectiveness of the proposed approach is demonstrated by numerical experiments.

Keywords: Process Control, Computer Aided Design, Modeling, System Identification and Estimation
Abstract: This paper considers the combination of disturbance observer (DOB) design with the conventional PID controllers to improve the ability to compensate effects of valve stiction. In the proposed method, the process delay is firstly approximated to a high-order transfer function using Pade approximation. Then an additional low-pass filter is designed to cope with the nonminimum-phase characteristic of the approximated process. Therefore, a DOB with a specific structured is constructed to estimate the dynamic of valve stiction on the system output. A numerical example is provided to demonstrate the benefit of the proposed approach.

Keywords: Robust Control
Abstract: This paper proposes an H-infinity controller design method for two-degree-of-freedom (2-DOF) control systems with lower sensitivity characteristics in the low-frequency range similar to that achieved by disturbance observers. The slopes of the magnitude plots of the sensitivity functions achieved by disturbance observers are 40 [dB/dec], while the systems are servo systems for step signals. Previous research demonstrated that incorporating integral weighting for input disturbances could achieve such low-sensitivity characteristics. However, with that method, the dynamics of the plant affect the output sensitivity function in the evaluated transfer function. Therefore, it is difficult to shape the output sensitivity function precisely. This paper emphasizes the importance of directly shaping the output sensitivity functions and proposes to move the integral weight from the input side of the plant to the output side. This approach allows for precisely shaping sensitivity functions while maintaining servo characteristics. Numerical examples demonstrate the effectiveness of the proposed method.

Keywords: Multivariable Control, Robust Control, Mechanical Systems Control
Abstract: This paper discusses reference tracking control for a four-wheeled vehicle using a robust model predictive control (MPC) method. For a linear parameter varying (LPV) model of the vehicle, the paper applies an LPV-MPC method. The LPV-MPC method can impose constraints on the state and input and treat the variation of the model.Numerical examples illustrate the effectiveness of the tracking performance.

Keywords: Adaptive and Optimal Control, Computer Aided Design, Multivariable Control
Abstract: Infinite-horizon sparse optimal control is considered with an objective function composed of the quadratic form and the 1-norm of the input and the state. In particular, its value function (or cost-to-go function) is investigated and its lower and upper bounds as well as evaluation of its decrease with time are given. These results can be used for estimating the time after which an optimal input becomes constantly equal to zero, combined with past results of the present authors. Estimation of this time is expected to be an important step toward solving an infinite-horizon sparse optimal control problem by noticing its essential finite part.

Keywords: Modeling, System Identification and Estimation, Adaptive and Optimal Control, Guidance and Flight Control
Abstract: This paper considers output feedback model predictive control (MPC) based on a realigned model of quadrotor attitude dynamics. In a setting where only the attitude angles are measurable, the variations of the angles are neces- sary to apply a conventional state feedback MPC method. The realigned model can represent the quadrotor without the variations. On the other hand, the state of the realigned model includes the current and past output and the past input sequences, meaning that MPC based on the realigned model does not require variations of the angles. Nevertheless, the computational burden is almost identical to the conventional state feedback MPC method. Numerical examples illustrate how the realigned MPC for quadrotors works.

Keywords: Modeling, System Identification and Estimation, Power Systems Control, Guidance and Flight Control
Abstract: This paper proposes a dynamic current estimation system to estimate the battery state of charge (SOC) in multicopters without using current sensors. A dynamic current estimation method is developed using a brushed DC motor that approximates the behavior of a multicopter's brushless DC motor. The model parameters are identified using the least-squares method based on measured motor speed, PWM command values, and current values. The system estimates the in situ current utilizing the time-series PWM signals. The battery state is estimated through the estimated current integration. The effectiveness of the proposed method is shown by comparing it with a previous static estimation method, utilizing a realistic simulation scenario for a specific multicopter.

Keywords: Robotic and Automation Systems, Nonlinear Control, Robust Control
Abstract: This paper introduces a novel robust visual servoing approach for tilt-rotor quadrotor systems designed to create airflow-free zones beneath the vehicle for precision agricultural spraying. We propose modeling the tilt-rotor dynamics as a thrust loss problem, where fixed rotor tilt angles directly reduce effective vertical thrust while complex aerodynamic couplings are treated as composite disturbances. Our integrated control framework combines a virtual camera-based robust observer with a sliding mode controller and adaptive thrust compensation mechanism to ensure stable visual servoing across various tilt angles. Theoretical stability analysis proves global asymptotic convergence of the closed-loop system. Numerical simulations demonstrate effective control at tilt angles up to 40° with an optimal adaptation gain of (gamma_b)=0.2, while ROS Gazebo simulations validate the thrust-loss modeling approach in realistic environments. This work establishes a foundation for developing practical tilt-rotor UAVs that can minimize pesticide drift through deliberate airflow manipulation, addressing a critical limitation in current agricultural drone technology.

Keywords: Robotic and Automation Systems, Mechatronics Systems, Nonlinear Control
Abstract: With the rapid aging of social infrastructure, the inspection and maintenance of bridges has become an urgent issue. As the number of bridges increases, the demand for new technologies such as multicopters is growing due to manpower shortages and reduced work efficiency. However, conventional multicopters struggle to maintain stable flight in confined spaces. To address this challenge, we are developing a multicopter capable of horizontal movement with minimal changes in attitude, equipped with multiple transformation mechanisms to adapt to infrastructure inspection tasks. In this paper, we present simulation results using nonlinear model predictive control (NMPC) and demonstrate that the proposed aircraft can maintain stable flight during transformation.

Keywords: Mechatronics Systems, Safety, Environment and Eco-Systems, Robotic and Automation Systems
Abstract: Amid global efforts to conserve resources and transition toward a circular economy, Japan is focusing on “urban mines,” where discarded small home appliances are treated as sources of valuable materials. However, domestic recycling has stagnated due to the high cost and inefficiency of manual sorting—especially for low-profit items like small appliances. To address this challenge, we are currently developing an automated sorting system designed to improve the efficiency of resource recovery and reduce recycling costs. This paper reports on the development and evaluation of the Auto-Sort system for lithium-ion battery equipped products, which are a major cause of fires and explosions in stockyards.

Keywords: Intelligent Systems, Robotic and Automation Systems, Innovative Systems Approach for Realizing Smarter World
Abstract: Large Language Models (LLMs) show potential for enhancing robotic path planning. This paper assesses visual input's utility for multimodal LLMs in such tasks via a comprehensive benchmark. We evaluated 15 multimodal LLMs on generating valid and optimal paths in 2D grid environments, simulating simplified robotic planning, comparing text-only versus text-plus-visual inputs across varying model sizes and grid complexities. Our results indicate moderate success rates on simpler small grids, where visual input or few-shot text prompting offered some benefits. However, performance significantly degraded on larger grids, highlighting a scalability challenge. While larger models generally achieved higher average success, the visual modality was not universally dominant over well-structured text for these multimodal systems, and successful paths on simpler grids were generally of high quality. These results indicate current limitations in robust spatial reasoning, constraint adherence, and scalable multimodal integration, identifying areas for future LLM development in robotic path planning.

Keywords: Safety, Environment and Eco-Systems, Robotic and Automation Systems, Intelligent Systems
Abstract: One type of attack on autonomous vehicles is GPS spoofing, which involves altering GPS signals to deviate from their true values. If this attack succeeds, the autonomous vehicle can be directed to arbitrary locations. The authors have demonstrated that adding a constant signal as a method of shifting the true value is effective in the sense of likelihood ratio testing. In this study, we experimentally evaluate the effectiveness of this attack and identify the differences between theoretical predictions and experimental results. In the experiment, we construct a system consisting of a robot subject to attack and a monitoring system that utilizes likelihood ratio testing. The attacker’s objective is to guide the robot into an unsafe region. The experiment examines whether the robot can move into the unsafe region while deceiving the monitoring system.

Keywords: Robotic and Automation Systems
Abstract: A key challenge for multi-robot navigation systems is the lack of global information in an unknown environment that impedes efficient path-finding, especially when subjected to local optima. To address this problem, a twin-delayed deep deterministic policy gradient (TD3) algorithm with a modified reward function is proposed for performance improvements in navigation tasks. The reward function incentivizes each robot to move towards a goal position, while penalizing any collision with the environment or other robots. It also incorporates a negative reward mechanism when the robot remains stationary for an extended period of time or fails to make meaningful progress. This encourages the robots to avoid local optima and seek alternative pathways toward their goals. The proposed method is evaluated using three different environments and shown to achieve high success rates in dynamic situations, while avoiding stagnation and collision. This validates the impact of the modified reward function and the effectiveness of deep reinforcement learning for autonomous multi-robot navigation systems.

Keywords: Robotic and Automation Systems, Network System Integration, Safety, Environment and Eco-Systems
Abstract: As societal challenges become more complex, the use of robots for problem-solving is increasingly desired. Instead of applying highly functional robots, it is more effective to network a diverse range of robots and sensors with the necessary functions for each task using the common communication network protocol RSNP. This creates a robot network platform that can address regional challenges. We have developed a control system that seamlessly switches between remote operation and autonomous navigation of multiple connected robots using this platform. Additionally, we have expanded the application of robot services by adding object detection capabilities locally and globally, to reduce human burden, and including drone for rapid and extensive movement and patrols. This paper describes these advancements.

Keywords: Human Interfaces, Virtual Reality Systems, Guidance and Flight Control
Abstract: In this paper, we propose an intuitive control framework for a multi-UAV system, enabling simultaneous control of UAVs and their onboard five-degree-of-freedom manipulators using virtual reality (VR) controllers. Unlike conventional methods that depend on complex button mappings or predefined gestures, our approach leverages natural and direct user motions to control both the UAVs and their manipulators seamlessly. By integrating motion tracking and real-time mapping of the VR controllers, the UAVs smoothly respond to positional and orientational changes, while the manipulators accurately replicate dexterous hand movements to perform precise grasping tasks. This intuitive interface significantly lowers the operational learning curve, allowing even non-expert users to effectively operate the aerial robotic system. Experimental results validate the effectiveness of the proposed method across various grasping tasks, demonstrating its potential for telemanipulation, search and rescue, and industrial automation applications.

Keywords: Modeling, System Identification and Estimation, Process Automation, Process Control
Abstract: Based on commonly captured sensor measurements like temperature or pressure, a “soft sensor” enables the estimation of quality values that are usually difficult to be measured in real-time. It has taken large efforts to build soft sensors by trial and errors hitherto, accordingly, “comprehensive soft sensor design tool” was developed based on the collaboration of academia and industry. The developed tool standardized and automatized the soft sensor building process and largely accelerates implementing high accuracy soft sensors with small efforts.

Keywords: Sensors and Transducers, Intelligent Systems
Abstract: We introduce soft somatosensory—a sensing paradigm in which soft-bodied robots perceive external stimuli via internal receptors embedded in deformable structures. Unlike rigid robots, soft robots allow environmental forces to deform the body, enabling internal receptors to detect contact, texture, or friction. Through dynamic touch experiments and learning-based slip detection, we show that internal receptors can function contextually as environmental sensors. This approach reduces sensor complexity and utilizes body compliance for perception, offering a new framework for embodied intelligence where the body itself becomes an active part of sensing.

Keywords: Sensors and Transducers, Micro and Nano Devices, Mechatronics Systems
Abstract: Recent advances in flexible electronics have led to the development of numerous sensors and systems designed to accommodate various applications. Among these, wearable devices represent a particularly promising field due to their potential for seamless integration with the human body. A key requirement for skin-adhesive wearable devices is stretchability, which allows them to conform to the natural movements of the skin without compromising performance. In this context, liquid metals have attracted considerable attention as stretchable conductive materials, owing to their remarkable mechanical robustness under bending and stretching. Integrating liquid metal wiring into stretchable devices can therefore provide enhanced reliability and stability, even under large deformations.　In this study, we introduce both physical and chemical sensors that utilize Ga-based liquid metal, and demonstrate their implementation in controlled stretchable devices. By harnessing the inherent flexibility and durability of liquid metal, our approach paves the way for robust, high-performance systems suitable for next-generation wearable applications.

Keywords: Signal and/or Image Processing, Process Automation, Factory Automation
Abstract: A new method for measuring the weld displacement of steel pipes using AR markers and a smartphone camera was developed. Increasingly strict inspection specifications are now applied to welded steel pipes, but since the efficiency of manual inspection is still low, new automatic inspection methods are required. Therefore, we proposed a new method to measure weld misalignment from images of the end faces of steel pipes, and developed a smartphone application to realize automatic misalignment measurement of welded pipes simultaneously with saving the photographs as inspection records. As a result of experiments, the inspection time was shortened from 90 seconds to 25 seconds, and the measurement error when measuring actual pipes with wall thicknesses over 9.5 mm was less than 0.5 mm.

Keywords: Sensors and Transducers, Human Interfaces
Abstract: In this paper, we show that a plane plate with a periodic comb-like structure on its surface functions as a surface waveguide for ultrasonic waves. In the waveguide, ultrasonic waves propagate in the tangential direction of the plane but are localized in the normal direction. The waveguide function is confirmed by numerical simulations.

Keywords: Signal and/or Image Processing, Remote Sensing, Sensors and Transducers
Abstract: This study aims to establish a quantitative evaluation technique for surface roughness using aerial ultrasonic waves, in order to construct a diagnosis method for the degree of deterioration of metal surfaces. First, experimental and theoretical verifications are conducted to acquire the relationship between surface roughness and reflectance ratio derived from the amplitude of reflected ultrasonic wave from the test specimen, in an environment where the influence of directivity characteristic of ultrasonic transducers is suppressed. Then, it is demonstrated that there is a linear decreasing trend in the reflectance ratio with increasing surface roughness. Furthermore, by estimating the directivity characteristic of the ultrasonic transducers, the attenuation correction based on its characteristic is applied. A sensitivity of the quantitative evaluation technique for surface roughness considering the directivity characteristic has been improved approximately 6.3 times compared to the results of previous studies.

Keywords: Medical and Welfare Systems, Integrative Biophysical Engineering and Informatics, Analytical Measurement
Abstract: The knee abduction moment (KAM) is a key biomechanical metric linked to medial knee osteoarthritis (OA) and ACL injury risk. Conventional KAM measurement methods using 3D motion capture and force plates are accurate but impractical for real-world use due to their bulk and cost. To address this, we propose a portable KAM prediction system using a smart shoe embedded with six force-sensitive resistors (FSRs) and an inertial measurement unit (IMU), combined with a machine learning model. Sensor data from the wearable, including force and motion signals at 100 Hz, were collected alongside ground-truth KAM data from a 3D tracking system. Two datasets from two subjects were used, totaling 97 gait cycles. Missing values were handled via linear interpolation and K-Nearest Neighbor (KNN) imputation. A Histogram-Based Gradient Boosting Regressor (HGBR) was trained to predict KAM, and model performance was evaluated using R², RMSE, and percentage error. Results show that individual-specific models with linear interpolation achieved the highest accuracy (R² = 0.986), while the generalized model showed strong cross-subject performance. KNN imputation significantly reduced accuracy. This research validates the use of wearable sensors with machine learning for reliable KAM monitoring, offering practical, scalable solutions for injury prevention and rehabilitation outside laboratory settings.

Keywords: Medical and Welfare Systems, Modeling, System Identification and Estimation, Flow Measurement and Control
Abstract: Cardiopulmonary bypass (CPB) is the temporary replacement of the heart and lungs during heart surgery. Its operation varies significantly depending on the surgical procedure and requires the operator to manage several parameters. One such parameter is the control of an occluder, which has been investigated for automated regulation. Previous studies, however, have not accounted for variations in the speed of the centrifugal pump and the height difference between the venous drainage site and the reservoir, both of which vary during surgery. In this study, we propose a model for estimating the blood flow rate based on the occluder opening ratio, incorporating the effects of pump speed and height difference. Perfusion experiments with a glycerol solution were performed to characterize the pressure-flow rate and pressure-occluder opening ratio relationships. These characteristics were modelled using fluid dynamics theory for sudden contraction and expansion in tubing and showed strong agreement with experimental measurements (R² = 0.9975 and 0.9969; p < 0.0001). By integrating the identified characteristics, the proposed model accurately estimated blood flow rate, with Bland–Altman analysis confirming high agreement with measured values and minimal systematic bias.

Keywords: Human-Machine Systems, Medical and Welfare Systems, Sensors and Transducers
Abstract: This paper proposes a visual-assistance system for a stationary gait trainer, which employs Neuro-Developmental Treatment (NDT) for patients with stroke. NDT is an effective technique that can help the recovery of walking ability and motor coordination for stroke patients through repetitive practice of correct motor patterns. However, traditional NDT training methods require the presence of therapists, limiting patients' training time during the critical rehabilitation period. Therefore, we built an NDT trainer to repeat the therapists' intervention techniques and stimulate the subjects' consecutive movements by cuing the subjects' anterior-superior-iliac spine. This paper proposes visual stimulation to enhance the effects of rehabilitation. First, we develop a visual assistance system using the Unity game engine, which can provide real-time visual stimulations to help stroke patients correct their gait patterns more efficiently. Finally, we invite stroke patients to conduct clinical NDT experiments. Based on the results, the visual-assistance system effectively improves the subjects' gait symmetry, pelvic rotation, walking speed, and step length.

Keywords: Sensors and Transducers, Medical and Welfare Systems, Ubiquitous Healthcare
Abstract: Respiratory diseases are health issues that require early detection for effective treatment. Conventional pulmonary function tests require measurement devices available only at medical institutions. To address these accessibility challenges, we propose a simple, non-invasive diagnostic system for pulmonary function assessment that can be used at home by analyzing respiratory characteristics from an image sequence of breathing chest motions.

Keywords: Medical and Welfare Systems, Electrophysiology and Kinesiology, Mechanical Systems Control
Abstract: This paper considers the safety control using Control Barrier Function(CBF) of the tele-rehabilitation system with electrical stimulation. The main contribution is to improve the safety of our tele-rehabilitation system by using the CBF based quadratic problems. Then, even if errors occur in the control input to the patient given by the physical therapist, the control input can be shaped so that the patient's knee joint angle is within a safe range. First, the dynamical model of the patient's knee joint is considered. Next, we propose constraint conditions in a quadratic programming method using a CBF. Finally, the proposed safety control system using the CBF is applied to our tele-rehabilitation system. It is verified that control inputs are shaped into safe control inputs.

Keywords: Information Management Systems, Integrative Biophysical Engineering and Informatics, Networked Sensor System
Abstract: This paper presents a hybrid sensing architecture for real-time acquisition, fusion, and analysis of environmental and physiological data to support AI-driven monitoring in smart indoor healthcare environments. The system integrates fixed environmental sensors and wearable physiological devices to continuously track key parameters, including carbon dioxide (CO₂), volatile organic compounds (VOCs), particulate matter (PM), temperature, humidity, acoustic activity, heart rate, and blood oxygen saturation (SpO₂). A microcontroller-based edge gateway, built on the Infineon PSoC™ 62S2 platform, performs local data processing and lightweight machine learning inference to enable anomaly detection without relying on continuous cloud connectivity. Physiological data is acquired via a proxy Android device connected to a commercial smartwatch, while all sensor data is synchronized, formatted in JSON, and transmitted over MQTT with SSL/TLS to cloud services for extended analytics and visualization. The system was validated through controlled experiments simulating normal and adverse conditions, confirming its ability to provide timely alerts, reliable performance, and scalable deployment in residential and clinical settings. This architecture offers a robust and flexible platform for real-time health and environmental monitoring with integrated edge intelligence and secure cloud interoperability.

Keywords: Robotic and Automation Systems, Intelligent Systems, Signal and/or Image Processing
Abstract: Autonomous mobile robots frequently utilize 3D LiDAR for environmental perception; however, blind spots near the robot's base pose significant safety risks. These blind spots, especially in the forward-moving areas close to the robot's feet, can obscure small steps and potholes, complicating safe navigation. To tackle this issue, we present a novel 3D point cloud measurement system that integrates a small, lightweight multi-zone distance sensor with a side-to-side rotating mechanism. This combination effectively addresses the inherent limitations of the multi-zone distance sensor, such as lower point counts, narrow measurement angles, and restricted distance detection. The side-to-side rotating mechanism enhances the system's measurement capabilities, enabling a broader range of data collection and generating a more detailed point cloud that accurately detects ground surface irregularities, such as small steps and/or potholes. Preliminary experimental tests have shown that our proposed measurement system can successfully identify irregularities in the ground surface within the blind spots surrounding the mobile robot. This innovation significantly enhances navigation accuracy and overall safety for autonomous robots.

Keywords: Robotic and Automation Systems, Signal and/or Image Processing, Transportation Systems
Abstract: In this paper, we present the development of a specific closed-road signboard detection algorithm designed for the "Tsukuba Challenge" task. A 3D LiDAR mounted on a mobile robot is utilized to simultaneously capture both 3D point clouds and the reflected intensities of the surrounding environment. In the "Tsukuba Challenge", the target closed-road signboards used to indicate road closures are positioned symmetrically at both sides of pedestrian paths in the park. By leveraging the spatial arrangement, along with the predetermined characteristic shapes and reflection patterns of the target closed-road signboards, the proposed algorithm achieves accurate target identification. The effectiveness of the real-time detection algorithm is validated through real-world experiments.

Keywords: Robotic and Automation Systems, Signal and/or Image Processing, Intelligent Systems
Abstract: This paper presents the development of a LoRa-assisted RTK-GNSS system designed for accurate outdoor navigation of mobile robots. The proposed system leverages LoRa, a long-range wireless communication function, to enhance the reliability and coverage of RTK-GNSS corrections in challenging outdoor environments without requiring an LTE or Wi-Fi network connection. By efficiently utilizing the narrow bandwidth of LoRa modules, the system employs an Arduino-based microcontroller to compress and transmit RTCM3 correction data from the base GNSS receiver to the rover GNSS receiver on the mobile robot. As a result, the system achieves real-time position correction while maintaining efficient bandwidth usage, making it well-suited for accurate outdoor mobile robot navigation.

Keywords: Signal and/or Image Processing, Factory Automation, Identification and Estimation
Abstract: Electroluminescence (EL) inspection is frequently utilized in the manufacturing process of solar panels to identify defects such as cell cracks. To achieve this goal, defect classification models from an EL image using supervised learning (SL) methods have been proposed. Although these models demonstrated high prediction accuracy in discerning known faults, defects different from the images used for model training may go undetected. In this study, we propose an EL image inspection system that combines an unsupervised learning (UL) and SL-based models. The proposed system has a function to continuously update the models by re-training on the false negative and positive images by the system. This update process was semi-automated to minimize expert involvement. The proposed system was already deployed in a production line of solar panels at the Toyooka plant of KANEKA Solar Tech Co., Ltd. After two rounds of re-training, the classification ability has kept higher than 94% for five months.

Keywords: Signal and/or Image Processing, Mechanical Measurement, Computational Intelligence
Abstract: Currently, the ultrasonic pulse-echo method is widely used for thickness measurement. To enhance the performance of this technique, this study aims to develop a method to predict the internal geometry of pipes with periodic flaws typically caused by flow-accelerated corrosion. Ultrasonic wave spectra reflected from periodic surfaces show pattern specific characteristics, which we utilized for estimating flaw parameters. We conducted simulations that calculated the propagation of ultrasonic waves reflected on different periodic rough surfaces. These simulations were performed under the same conditions as experiments, which replicated pipe thickness measurement using the pulse-echo method. The simulated pulsed wave amplitude spectra showed the same characteristics as experimental waves when reflected from periodic rough surfaces. For parameter estimation, we trained and tuned a neural network model using only simulation data. To evaluate performance under real-world conditions, we tested the model using experimental data. The results showed relatively high accuracy in estimating flaws depth. While improvements are required, further flaw pitch estimation exhibited potential.

Keywords: Computational Intelligence, Signal and/or Image Processing, Virtual Reality Systems
Abstract: This paper addresses challenges in 3D reconstruction from degraded video data in nuclear reactor environments, focusing on post-Fukushima Daiichi decommissioning tasks. High radiation levels and image degradation due to white noise and particulate matter are critical obstacles for remote inspection using ROVs. To evaluate robust reconstruction methods, the study utilizes video footage from the ROV-A2 used by the Tokyo Electric Power Company, with preprocessing involving frame extraction, text removal, and COLMAP-based parameter estimation for SeaThru-NeRF and 3D-GS. Results comparing these approaches under degraded conditions highlight the advantage DUSt3R has in reconstructing structures of interest with high perceptual realism. Metrics such as PSNR, SSIM, and LPIPS quantify accuracy in reproducing detailed texture visuals from set positions, on which the point-based 3D Gaussian splatting scores highest. These results highlight the trade-off between perceptual realism, favoring continuous volumetric reconstructions, and metric accuracy, favoring discrete point cloud representations with explicit texture encoding.

Keywords: Identification and Estimation, Transportation Systems, Safety, Environment and Eco-Systems
Abstract: With the development of vehicular network technology, the prediction of vehicle power demand has become significant in intelligent transportation systems and energy consumption optimization. However, due to the complexity of the traffic environment, prediction of short-term vehicle power demand with great accuracy faces challenge. In order to improve the prediction accuracy, this work proposes a LSTM Encoder-Decoder prediction model. The model extracts key features from input data through the encoder, and with the features, uses the decoder to generate future predictions of vehicle power demand. To validate the effectiveness of the model, experiments are conducted using a real-world dataset, and comparisons are conducted to the LSTM model. Experimental results show that the proposed prediction model has obvious advantages in both prediction accuracy and stability. Moreover, the work proposes a systematically evaluation and analysis on the impact of different feature sets on prediction performance.

Keywords: Manufacturing Systems, Micro and Nano Devices, Innovative Systems Approach for Realizing Smarter World
Abstract: This study presents a novel optical digital reconstruction framework for layered opto-electronic devices, exemplified by Organic Light-Emitting Diodes, utilizing Conditional Variational Autoencoders (CVAE) to address the limitations of traditional deterministic neural networks in multi-objective optimization. Unlike conventional approaches that rely on bidirectional mappings and deterministic inference, the proposed CVAE framework incorporates latent space modeling, physical consistency constraints, and probabilistic sampling to generate diversity, physically reliable structural design conditioned on target optical properties. By integrating feedback from a pretrained forward model, dozens of candidate structural designs are produced and filtered by additional objectives such as color gamut coverage and angular luminance decay to identify optimal designs. Experimental results validate that the proposed framework preserves target optical performance while enhancing design diversity through latent space clustering. Post-processing optimization further demonstrates its engineering applicability in balancing color gamut and angular luminance decay, offering a scalable solution for complex optoelectronic device design.

Keywords: Flow Measurement and Control, Micro and Nano Devices, Intelligent Control
Abstract: A control strategy based on the combination of Linear Active Disturbance Rejection Control (LADRC) and Particle Swarm Optimization (PSO) algorithm is proposed to address the problem of precise control of micro-Newton cold gas thruster，providing effective technical support for their applications in space gravitational wave detection and high-precision measurement of the Earth's gravity field. The mathematical models of the piezoelectric drive system and Laval nozzle are obtained by parameter identification of the system through the PSO algorithm. Compared with the traditional PID control, the Linear Active Disturbance Rejection Control (LADRC) exhibits stronger anti-disturbance capability and faster response speed. Simulation results show that the LADRC controller exhibits obvious advantages under different operating conditions: the rise time is shortened by about 20%, the regulation time is reduced by about 25%, the overshoot is significantly reduced, and the steady-state error is close to zero.

Keywords: Robotic and Automation Systems, Mechanical Systems Control, Nonlinear Control
Abstract: In this paper, the formation control problem for multiple wheeled robots is addressed. Firstly, each wheeled robot subject to nonholonomic constraints on the lateral movement of the wheel axle is modeled under the Port-Hamiltonian (PH) framework. The formation control problem is then transformed into an optimization problem. Through solving this problem, a distributed controller is designed for each nonholonomic wheeled robot. Specifically, each robot only exchanges the estimated average position with its neighbors, thereby avoiding the risk of position privacy leakage due to direct position exchange. In addition, based on the proposed controller, each wheeled robot system still maintains the PH structure, which simplifies the stability analysis. Finally, numerical simulations are conducted to verify the effectiveness of the proposed controller.

Keywords: Intelligent Control, Nonlinear Control
Abstract: The development of the vehicle-to-everything (V2X) technology brings new opportunities for automtoive control. This report investigates a velocity planning problem for road vehicles if the traffic light information is prior known, to obtain a satisfactory fuel economy performance and lower travelling time loss.The model predictive control algorithm is adopted to solve the multi-objective optimization control problem, and the reinforcement learning algorithm based on the Actor-Critic framework is used for online adjustment of the parameters in MPC. Experiments show that, on the premise of ensuring the passage of green waves, compared with the rule-based energy management strategy, the strategy proposed in this paper saves 7.4% in electricity consumption and 2.8% in fuel consumption.

Keywords: Mechanical Systems Control, Transportation Systems, Adaptive and Optimal Control
Abstract: This paper presents a cooperative load transportation system using multiple multirotor UAVs, with a focus on collision avoidance between the UAVs. Each aerial vehicle modeled as a fully-actuated UAV is connected to a payload via cables. In such a system, there is a risk of the UAVs’ colliding with each other. To address this issue, the present method introduces collision avoidance constraints based on a conic control barrier function, applied individually in the direction of each cable. The control inputs for each UAV are then obtained by solving a constrained optimization problem. The effectiveness of the proposed method is demonstrated through simulations.

Keywords: Autonomous Decentralized Systems, Adaptive and Optimal Control, Safety, Environment and Eco-Systems
Abstract: This paper proposes an adaptive cruise control (ACC) method for quadrotor UAVs that considers both the preceding and following quadrotors. The control method is an optimal control approach. It utilizes a quadratic programming (QP) with a collision avoidance constraint formulated using a parameter-adaptive control barrier function. By considering both the preceding and following quadrotors, cooperative distance control between quadrotors becomes possible, and the feasibility of solving the QP is expected to be improved. In addition, by solving the optimization problem locally in each quadrotor, distributed control is achieved. The present method is applied to ACC in linear quadrotor platooning, and its effectiveness is demonstrated through a simulation.

Keywords: Robotic and Automation Systems, Autonomous Decentralized Systems, Networked Sensor System
Abstract: We propose a distributed mobile charging system for environmental monitoring, based on control barrier functions with capability-aware coordination. The system enables safe and autonomous recharging by coordinating monitoring robots with a charging-enabled robot, while accounting for their heterogeneous movement capabilities. A centralized control scheme is first developed and then decentralized. Simulations show that the distributed approach achieves performance comparable to the centralized one, demonstrating its effectiveness for persistent monitoring tasks.

Keywords: Robotic and Automation Systems, Guidance and Flight Control, Network System Integration
Abstract: This paper proposes a bio-inspired control framework for the flight control of a swarm of quadcopters with flocking dynamics, integrating a novel combination of dynamic goal-shifting and sinusoidal altitude modulation for adaptive 3D trajectory tracking. The swarm coordination leverages a graph-theoretic model of multi-agent systems (MAS) for decentralized communication, enabling consensus-based position synchronization. A key contribution of this work is the introduction of a goal update mechanism that adapts the reference position dynamically when a subset of agents reach proximity, ensuring seamless progression along the desired trajectory. Meanwhile, altitude is modulated using a sinusoidal profile to mimic natural undulating flight patterns and introduce vertical adaptability. By unifying Reynolds’ flocking rules (separation, cohesion, and alignment) with artificial potential field (APF)–based obstacle avoidance, the swarm achieves robust formation stability and collision-free motion in cluttered 3D environments. Simulations validate the proposed control framework in circular trajectory scenarios, demonstrating inter-agent cohesion, obstacle avoidance, and dynamic altitude synchronization. The results show promise for real-world applications such as environmental monitoring in uneven terrain or search and rescue missions in vertical or constrained spaces.

Keywords: Autonomous Decentralized Systems, Mechatronics Systems, Robotic and Automation Systems
Abstract: This paper introduces SARNet, a hybrid routing protocol for Search-and-Rescue (SAR) Flying Ad-hoc Networks that integrates event-driven OLSR broadcasts with SINR-triggered link adaptation and ant-inspired unicast reinforcement. It aim to achieve robust, low-latency communication for UAV swarms by combining deterministic broadcast with probabilistic multipath unicast. The protocol is lightweight and suited to dynamic, infrastructure-less environments. Simulations benchmark SARNet against AODV under varying mobility to validate delay stability and delivery performance.

Keywords: Multivariable Control, Transportation Systems, Autonomous Decentralized Systems
Abstract: In this study, we propose a decentralized control method that simultaneously achieves stable formation and obstacle avoidance in cooperative transport by multiple unmanned aerial vehicles. In the past, geometric methods, artificial potential methods, and path planning methods have been widely used for obstacle avoidance control in cooperative transport, but recently, a method based on the Control Barrier Function (CBF) has been proposed and is attracting attention as an approach to realize transport while guaranteeing safety. This research is novel in that N agents determine the optimal control input in a decentralized manner. This enables us to simultaneously realize robustness, in which the remaining agent can complement the system even if one of the agents fails, and high system scalability, which is independent of the number of agents.

Keywords: Modeling, System Identification and Estimation, Guidance and Flight Control, Mechatronics Systems
Abstract: This study focuses on modelling a satellite prototype designed to be manufactured with CubeSat standards and simulating attitude control to implement fault-tolerant control strategies. As part of the passive fault tolerance approach, the system is configured using a pyramid configuration with four reaction wheels. The ability to maintain attitude control using only three operational wheels is also analyzed. To increase reliability in case of sensor failure, an additional sensor is included with the physical integration of two redundant IMUs. A Two-Stage Kalman Filter is developed to detect sensor bias errors and increase the accuracy of system state estimation. In addition, a simple flip-flop mechanism is used to isolate faulty sensors and provides effective fault detection and switching. The Two-Stage Kalman Filter has shown strong performance in generating residuals indicating sensor bias errors.

Keywords: Modeling, System Identification and Estimation, Computer Aided Design
Abstract: This paper proposes a landing leg design for lunar landers using genetic programming (GP) to simultaneously achieve parameter and structure optimization. GP is an algorithm that searches for the optimal solution by simulating the process of biological evolution using a population with a tree structure. This study applies GP to optimize a landing leg structure by representing it as a tree. The multi-objective optimization is performed using two objective functions: the shock absorption performance and the number of springs and dampers to be used. The results show that the proposed method enables simultaneous search for design parameters and structure.

Keywords: Robust Control, Adaptive and Optimal Control
Abstract: This paper considers hierarchical and cooperative control of distributed electric propulsion aircraft. For the dynamics of the aircraft body, a one-degree-of-freedom model is employed. The model includes a single rotational motion affected by thrust generated by multiple distributed electric motors. The dynamics of thrust are approximated by a first order lag system. The overall model is first represented as a system consisting a several decoupled identical subsystems. Then, hierarchical optimal control is applied to the system. The obtained optimal control law is composed of an individual feedback loop regarding the regulation of each thrust and a cooperative feedback loop regarding the regulation of rotational motion. Parameter uncertainties of the thrust dynamics are also taken into account by using H∞ control with a hierarchical structure. The effectiveness of the proposed control laws are confirmed by numerical simulations.

Keywords: Adaptive and Optimal Control, Guidance and Flight Control, Nonlinear Control
Abstract: This study investigates the effect of a singularity avoidance term in the cost function of the controller on the control performance in spacecraft attitude maneuvers using a control moment gyro system with model predictive control. Monte Carlo model predictive control simulations were conducted to examine whether explicitly incorporating a singularity avoidance term is necessary for achieving high control performance. The results indicate that if the prediction horizon is sufficiently long relative to the characteristics of the system, the controller can achieve superior performance compared to cases where singularity avoidance is explicitly handled.

Keywords: Guidance and Flight Control, Mechatronics Systems, Robotic and Automation Systems
Abstract: Flapping-wing robots have attracted significant attention due to their ability to replicate natural flight mechanisms observed in birds and insects, offering advantages in maneuverability, energy efficiency, and safer interaction with the environment compared to conventional aerial vehicles. However, achieving stable and autonomous flight remains a substantial challenge because of the highly nonlinear and time-varying nature of flapping-wing aerodynamics. This study proposes an integrated Proportional-Integral-Derivative (PID) control system for a flapping-wing robot to enhance stability and responsiveness during free flight. Building on previous work that employed independent PD controllers, the present system unifies pitch, roll, and altitude control under a single activation scheme, with all control loops operating simultaneously based on real-time feedback. Controller gains were empirically tuned through iterative flight testing to address the complex, unsteady aerodynamic behavior intrinsic to flapping-wing locomotion. Flight experiments conducted in a gymnasium environment evaluated the robot's ability to autonomously stabilize its attitude and altitude during random, untethered flight. The results demonstrate that the integrated PID system achieves rapid disturbance rejection, maintains altitude within approximately +-0.1-0.2 meters of the target, and restores roll and pitch angles to near-level within approximately 5.8 seconds after external perturbations.

Keywords: Intelligent Control
Abstract: In this paper, we propose a model-free method for state-flipped control of probabilistic Boolean networks(PBNs). A PBN is known as a mathematical model of gene regulatory networks, industrial machines, smart grid devices, and so on. The state in a PBN and its time evolution are represented as a binary variable vector and a set of Boolean functions, respectively. State-flipped control is a control method where some elements of a binary variable vector representing the state are flipped. By the proposed reward design for Deep Q Network, a controller that achieves a smaller number of flipped nodes can be derived.

Keywords: Safety, Environment and Eco-Systems, Components and Devices
Abstract: There are a few cases in which medium to low temperature geothermal resources have been utilized for power generation. Therefore, we focused on binary power generation using hot spring water as a heat source. In this study, we developed a small binary power generator (10-15 kW) that can use medium- to low-temperature heat sources and confirmed its operation.

Keywords: Entertainment Systems
Abstract: In the process of garment production, creating sewing patterns which serve as design blueprints for clothing, from fashion sketches is a highly complex task that requires extensive experience and knowledge from garment makers. In this paper, we propose a method for generating sewing patterns from clothing images by constructing a dataset based on sewing patterns used in actual dressmaking and training an image generation model using Generative Adversarial Network (GAN).

Keywords: Intelligent Systems, Autonomous Decentralized Systems, Intelligent Control
Abstract: Current research indicates a great application potential of multi-agent reinforcement learning (MARL) on realworld network systems, such as power systems, traffic networks, and multi-unmanned aerial vehicle systems. There is a challenge for MARL is efficiently achieve global cooperative performance while in a decentralized learning style that is applicable to networked systems. To this end, we propose novel decentralized multi-agent actor-critic algorithms inspired by the ideas of parameter consensus learning and counterfactual baseline of multi-agent. Specifically, we reasonably consider that the reward functions of the agents are different and are available only to the corresponding agent. A consensus update is designed to approximate the joint reward function via communication over the network to ensure global goal consistency. Based on this, in the critic step, a truncated representation of joint rewards is used for respective value function learning that reduces variance. For the actor step, the truncated-based counterfactual advantage is accordingly computed to enable an efficient credit assignment for each agent’s policy improvement that ensures effective joint policy learning. Our algorithms possess provable convergence when the approximation functions are within the class of linear functions and are general for both discrete and continuous spaces of tasks. Experimental results with both linear and nonlinear function approximations show the effectiveness of the propos

Keywords: Intelligent Systems, Safety, Environment and Eco-Systems, Innovative Systems Approach for Realizing Smarter World
Abstract: The extensive adoption of SMS for communication has resulted in an increase in phishing and misleading messages that jeopardize user security. This study introduces a classification system to sort SMS messages into three risk categories: safe, potentially harmful, and clearly harmful. The dataset, made up of 10,000 user-submitted SMS messages gathered through an online form, was labeled and analyzed manually. A refined transformer model (LLaMA 3.2) was used for multi-class classification. Performance evaluation included precision, recall, F1-score, and ROC-AUC, resulting in an overall accuracy of 81%. Tools like word clouds and confusion matrices were utilized for result interpretation. The results show that transformer models can accurately identify SMS risk categories. The model demonstrates potential, but it has limitations, such as missing contextual and behavioral metadata. Future work will incorporate explainable AI (XAI) methods to improve clarity and foster user confidence.

Keywords: Intelligent Control
Abstract: This study outlines the creation and execution of an IoT system for monitoring and analyzing environmental conditions in hydroponic green-oak cultivation. Optimize plant growth using automated control and real-time data analytics. The system uses sensors to measure temperature, humidity, and pH, transmitting data through WiFi to a cloud-hosted MySQL database. A Grafana dashboard handles visualization and user interaction, and KNIME Analytics Platform is used for data analysis and model training. The research took place in a regulated hydroponic setting, focusing on green oak as the target crop. Data on growth metrics such as leaf width, height, root length, and biomass weight were gathered from 30 samples. We applied multiple linear regression and random forest models to evaluate how environmental variables affect crop yield. The Random Forest model showed strong predictive accuracy, achieving an R² of 0.86, an MAE of 9.23 g, and an RMSE of 12.37 g. Results indicate that pH significantly impacts growth, whereas high humidity adversely affects yield. This study presents a scalable framework for smart farming utilizing affordable embedded systems. The results lay the groundwork for intelligent, data-driven agricultural control systems, even without a baseline comparison to non-IoT environments.

Keywords: Sensors and Transducers, Modeling, System Identification and Estimation, Innovative Systems Approach for Realizing Smarter World
Abstract: This paper introduces an IoT-based waste monitoring system that integrates ultrasonic sensors with an ESP32 microcontroller to detect and display waste levels in trash bins. The system is designed to show whether a bin is full or empty, providing real-time waste level updates through an LCD panel on the bin and pop-up notifications via a web application. Our experimental setup uses five ultrasonic sensors, each emitting ultrasonic pulses at 40 kHz, to determine the distance between the trash can lid and the closest rubbish. The sensor also detects the reflected pulse when an object blocks the signal. After getting data from all sensors, the ESP32, programmed in C, processes these data and determines whether the bin is "FULL" or "EMPTY." The system integrates an IEEE1888 gateway for standardized data exchange and a Pocket-WiFi module for simplified network connectivity. This system is invented under the Zero-Piece concept, aiming to eliminate litter on streets and roads. Given the challenges of overflowing trash in Thailand’s urban areas, we propose deploying these smart bins along city roads to improve waste management and maintain cleaner public spaces.

Keywords: Sensors and Transducers
Abstract: The terahertz photoacoustic effect enables non-contact generation of plane-wave ultrasound in aqueous media. However, for applications such as biomedical imaging, wavefront engineering is crucial for enhancing spatial resolution and signal-to-noise ratio. In this study, we explore the feasibility of non-contact implementation of ultrasound wavefront engineering through a geometric approach. We demonstrate that sub-terahertz waves impinging onto a spherical water boundary formed within an acrylic spherical tank can generate focused ultrasound in water owing to the geometric effect. We experimentally observe a focused wavefront by translating a hydrophone. This capability of non-contact generation of focused ultrasound has significant potential for sensing and imaging.

Keywords: Sensors and Transducers, Human Interfaces
Abstract: Remote work often leads to social isolation and reduced well-being for workers. To address this, we aim to share workers' concentration states to enhance their sense of connection without being monitored. However, most of the works overlook the privacy aspects of this issue. Therefore, this study proposes a privacy-preserving approach that uses keyboard and mouse logs, wrist and chair movement data, and a distance sensor to estimate concentration during desk work. To validate the privacy-preserving method to estimate concentration, we conduct controlled experiments under concentrated and unconcentrated conditions, with tasks classified by input intensity. Principal component analysis reveals that specific features of wrist and chair accelerometers are associated with concentration levels. These findings suggest that workers' concentration can be estimated without compromising privacy, providing a foundation for privacy-aware presence-sharing systems.

Keywords: Modeling, System Identification and Estimation, Robotic and Automation Systems, Sensors and Transducers
Abstract: The objective of this study is to develop a GNSS/IMU simulator that can generate realistic errors consistent with real-world environments, thereby enabling pre-validation of position estimation accuracy. Initially, for GNSS, positioning solutions are predicted based on the number of satellites obtained through a LOS/NLOS simulation using 3D building data. Subsequently, for the IMU—especially the gyro—we improve a conventional gyro model, originally based on the Allan variance in static environments, by incorporating error factors that reflect dynamic environmental influences to account for performance degradation under dynamic conditions. Finally, the accuracy of position estimation was validated using both real vehicle data and simulator-generated sensor data. The similar trends observed in position estimation results confirmed the simulator’s potential for effective pre-validation.

Keywords: Signal and/or Image Processing, Sensors and Transducers, Opto-Electronic Measurement
Abstract: Event cameras asynchronously detect changes in light intensity with microsecond resolution, offering rich motion information under dynamic scenes. However, the sparse and discrete nature of event streams hinders effective feature extraction and reconstruction for mainstream vision tasks. To address this, we propose a correlation image simulator that generates compact, structured representations of motion. Our method simulates the output of a three-phase correlation image sensor (CIS), capturing temporal correlations between the incident light and an external reference signal. By iteratively accumulating events over the exposure period, we reconstruct a relative intensity signal and compute its correlation with the reference. Experimental results on both synthetic and real dynamic scenes show that our approach enables efficient and informative event-to-image transformation, facilitating better motion characterization

Keywords: Signal and/or Image Processing, Remote Sensing, Sensors and Transducers
Abstract: Flooding remains one of the most damaging natural disasters, especially in rural or under-instrumented regions where river level monitoring is limited or nonexistent. Traditional systems rely on contact-based sensors or manual gauges, which require infrastructure, maintenance, and expert interpretation. In contrast, local communities often assess flood risk using intuitive visual cues, such as the proximity of water to tree roots or bridge posts. This study presents a lightweight, solar-powered water monitoring system that integrates deep learning with visual referencing to estimate water levels from images. The system uses a Raspberry Pi with a camera to autonomously capture riverbank images and transmit them to a remote server. A hybrid AI pipeline performs water surface segmentation using deep CNNs and detects the waterline relative to predefined pixel landmarks. A confidence-based fallback mechanism enhances robustness under suboptimal conditions. Field deployments demonstrate an average estimation error of about 6 percent, validating the system’s accuracy. This low-cost, image-based approach offers a scalable solution for flood monitoring in resource-constrained environments, bridging the gap between local intuition and automated sensing.

Keywords: Medical and Welfare Systems, Robotic and Automation Systems, Mechanical Measurement
Abstract: We have developed a CT-fluoroscopy-guided needle-puncture robot named Zerobot to assist in interventional radiology (IR) procedures. Currently, Zerobot is remotely operated by the surgeon using a controller. The next goal is to achieve automated needle insertion with Zerobot, which requires the integration of several techniques. In this study, we focus on estimating the bending moment applied to a puncture needle using CT images. In previous work, a method for automatically detecting needle puncture from a series of CT images was proposed. To obtain information on needle deflection and the forces acting on it, the curvature of candidate needle points in 3D space must be determined. We fit a polynomial curve through the thinned set of candidate points representing the needle path. Information derived from the fitted polynomial is then used to compute both the signed curvature and the corresponding bending moment.

Keywords: Medical and Welfare Systems
Abstract: This study advances routine monitoring methods for individuals with spinal cord injuries by examining the potential of daily wheelchair movement data to reflect changes in their physical condition. These individuals risk unnoticed deterioration in their physical condition owing to sensory disturbances, highlighting the need for routine assessments. We hypothesized that daily wheelchair movement could reflect physical health and aimed to identify measurable characteristics of this activity through routine, non-invasive methods. Four subjects with cervical spinal cord injuries who used manual wheelchairs were asked to engage in their daily activities for four weeks. We analyzed continuous acceleration data from the accelerometers mounted on their wheelchair frames to identify instances of wheelchair drive and estimated the probability distribution of aggregated daily drive durations. A log-normal distribution effectively characterized wheelchair drive, allowing us to summarize the daily activities into a limited set of parameters. The distribution parameters varied depending on the subject's physical condition. These findings suggest that daily wheelchair activity patterns may offer useful indicators of physical condition in this population.

Keywords: Mechatronics Systems, Human Interfaces, Medical and Welfare Systems
Abstract: This paper proposes a small and lightweight wearable robotic finger system for daily life support, especially in eating support. The proposed device has two fingers facing each other, which is driven by a single servo motor. The finger mechanism is based on a multiple four-bar linkages in order to achieve stable grasping of a cup and glass. A custom-made attachment is designed to install chopsticks and alter the finger motion from grasping mode to eating support mode. The control system employs a command generator based on the real-time food detection using YOLOv5 and an EMG-based control using brachioradialis. The proposed robotic finger system is verified through experiments.

Keywords: Medical and Welfare Systems, Human Interfaces
Abstract: Individuals with total hearing loss face difficulties in obtaining information from sounds, making it necessary to provide them with support for communication and environmental awareness. Particularly due to their inability to perceive environmental sounds, they have difficulty recognizing warning sounds, which can pose life-threatening risks. The purpose of this study was to propose and develop a system designed to assist individuals with hearing loss recognize environmental sounds through tactile feedback with reducing cognitive load. Through the basic experiment, we verified the system’s ability to recognize environmental sounds and provide appropriate tactile feedback, confirming that users can recognize sounds based on the provided feedback. The system analyzed environmental sounds using spectral flatness and the singular spectrum transformation method and recognizes these sounds through zero-shot inference. The recognized sounds are then conveyed to users through tactile feedback via a smartwatch. To confirm the feasibility of the developed system for individuals with total hearing loss, we conducted an experiment with four hearing individuals simulating individuals with total hearing loss. As a result, participants were able to recognize environmental sounds using the developed system. Therefore, we confirmed the feasibility of the developed system in supporting environmental sound recognition for individuals with total hearing loss.

Keywords: Human Interfaces, Medical and Welfare Systems
Abstract: Stuttering is a speech disorder that makes fluent speech difficult. It is believed to be a worldwide phenomenon. There is currently no established treatment for people who stutter, so new methods of providing stuttering support are needed. Therefore, we focused on the distraction effect. The distraction effect is the phenomenon whereby stuttering symptoms are alleviated by diverting a stutterer's attention away from speech. We believe that if this effect can be induced by presenting vibratory stimulation, it will be possible to develop a new stuttering support device. Additionally, previous studies have examined the effects of presenting vibration stimulation to people who stutter, but the effects of presenting vibration stimulation to multiple sites simultaneously have not been confirmed. Therefore, to develop a speech aid for people who stutter, this study created a device that delivers vibration stimulation while the person is speaking. The device was then used to deliver vibratory stimulation to one or more body parts while stuttering individuals read aloud to investigate its effect on core symptoms of stuttering.

Keywords: Medical and Welfare Systems, Human Interfaces, Safety, Environment and Eco-Systems
Abstract: With advances in science and technology, much physical labor has been automated; however, physically demanding work that is difficult to mechanize still exists, and low back pain has become a serious global issue. The authors have developed and promoted the Muscle Suit as a countermeasure against back pain, but users are demanding further miniaturization and weight reduction. In this study, we reviewed the assistive torque and developed a new model targeting a total weight of 3 kg. This paper reports on the details of that development.

Keywords: Intelligent Control, Adaptive and Optimal Control, Power Systems Control
Abstract: This study proposes an intelligent hybrid control system for EV shuttle mini bus charging, integrating photovoltaic (PV) generation, grid electricity, and battery storage. The primary aim is to minimize dependency on conventional grid power, enhance charging efficiency, and promote sustainable energy use within a university campus. Initial performance validation utilized PVsyst simulation software, followed by the development and experimental testing of a prototype system comprising a 2.88 kW bifacial PV array and a 48V, 50Ah lithium-ion battery at Rajamangala University of Technology Lanna, Chiang Mai. Experimental results demonstrated significant reductions in grid dependence, achieving stable PV outputs of approximately 530–600 W during peak hours (08:00–13:00), thereby lowering grid power usage by roughly 45%. Additionally, the integrated system provided adaptive energy management, ensuring continuous, reliable power supply across varying conditions. These findings confirm that the developed hybrid control system effectively meets energy sustainability goals, particularly suited to campus environments, with potential scalability for broader implementation.

Keywords: Transportation Systems, Social Systems, System Engineering
Abstract: This study deals with ride-sharing systems, which are expected to alleviate traffic congestion and reduce carbon dioxide emissions. Although ride-sharing systems have been defined variously in the literature, this study assumes that individual passengers share a vehicle for transportation and split the cost of transportation, such as gasoline, tolls, and parking fees, with other passengers who have similar transportation needs. In this study, the combination of multiple passengers is determined by matching using hypergraph. Then, the amount paid by passengers is determined as the Nash bargaining solution to design a price that takes into account passengers’ preferences regarding time and fare. Through simulation verification, it is confirmed that the combination of multiple passengers reduces carbon dioxide emissions and drivers, and that the design of the payment amount using the Nash bargaining solution guarantees the gain for the passengers.

Keywords: Nonlinear Control, Social Systems, Robust Control
Abstract: For mitigating the spread of infectious diseases which cannot be eliminated completely, this paper proposes a person-to-person contact regulation control law to stabilize and robustify the endemic equilibrium of SIQRS model subject to rate uncertainty of immunity, transmission, transportation, recovery, mortality, and inflows. Immunity waning creates a loop, which can induce undesired oscillations and make nonlinear stability analysis not only harder but also prone to fatal conservativeness. To avoid such conservativeness, this paper designs feedback contact regulation adding asymmetric damping instead of cancelling nonlinearities. It renders the closed loop asymmetrically dissipative in terms of a logarithmic storage function serving as a control Lyapunov function. The function conforms to variables positivity and allows the regulation control to be combined with vaccination and isolation to achieve input-to-state stability.

Keywords: Social Systems, Modeling, System Identification and Estimation
Abstract: This paper constructs a GM-BiLSTM hybrid model, which utilizes the characteristics of two-way long and short-term memory network (BiLSTM) that can accurately capture the long-term dependence and short-term fluctuation of data, and combines the error correction function of the gray model (GM(1,1)), to analyze the development trend of the population in Zhejiang Province based on the relevant data of 11 cities and municipalities in Zhejiang Province from 2008 to 2023, and to forecast the development trend of new quality productivity. Forecasting study, and at the same time construct the new quality productivity index to analyze the development level and trend of new quality productivity in Zhejiang Province. The study shows that the hybrid model has a small mean square error and good prediction accuracy and generalization ability. The model predicts that the population size of Zhejiang Province will reach 54.764 million in 2033, and the spatial distribution shows the pattern of “Hangzhou and Ningbo” dual-core drive and the gradual rise of sub-centers, but the regional disparity is gradually appearing. It is suggested to optimize regional resource allocation and strengthen cross-regional synergistic development to balance the layout and provide a decision-making basis for the demographic dividend transition.

Keywords: Autonomous Decentralized Systems, Social Systems, System Engineering
Abstract: In this paper, Choreography concept is introduced to implement the lower layer system in Cyber Physical Value System which arbitrates various social systems, Health care, Logistics, Transportation, Power, Water, etc., to meet the value required by users. Choreography we proposed provides a small amount computation cost and flexibility for the upper layer by absorbing changes in environment and system configuration occurred in the lower layer.

Keywords: