Positioning systems play an important role in industrial and scientific applications, making their performance improvements and functional expansions crucial. Positioning systems generally consist of mechanisms, actuators, controllers, and sensors, with performances and functions depending on their integration. Accordingly, positioning technologies focus on these elements and their integration into the system. Actuators are fundamental and critical elements that characterize positioning systems. Electromagnetic actuators are the most widely used because of their high basic performance, and piezoelectric actuators are often used for fine movement. In addition, research on new actuators with unique characteristics is actively being pursued. In component integration, control technology plays a crucial role in effectively utilizing the characteristics of elements and addressing their drawbacks. With the recent advances in digital technology facilitating nonlinear representations and structural modifications, various control methods have been proposed and evaluated to benefit from these advances. In addition to proposing modeling methods, the effects of manufacturing processes on system performance have been evaluated.

This special issue contains 18 papers covering the design and characteristics of mechanisms and actuators, control technology, system design, and modeling, along with three review papers on positioning technology and its applications. We would like to express our sincere gratitude to the authors for their excellent contributions to the special issue. We also thank the reviewers for their incisive efforts in editing this special issue. We hope that this special issue will help readers enhance their knowledge and understanding of positioning technology and its applications, thereby contributing to its progress.

The Technical Committee for Ultra-Precision Positioning, the Japan Society for Precision Engineering has conducted a questionnaire survey on ultra-precision positioning technology comprising precision mechatronics and precision measurements every four years since 1986, analyzing the status and evolution of positioning technology and specific problems. This article presents an overview of the results of the latest survey conducted in 2022, as well as the results by equipment of interest to the respondents. Furthermore, attention is paid to the results of the survey on energy saving and the environmental impact of positioning equipment, and the efforts of the survey respondents to reduce the environmental impact are reported.

This article reviews machine tool research trends in Korea over the past decade, focusing on government-supported projects and corporate-led initiatives. Achievements in machine tool element technologies are introduced, including performance analysis of bearings, ball screws, and feed systems. Studies on motion error simulations and feed control are also highlighted. For material removal processes, government-funded research has focused on enhancing the machining performance of difficult-to-cut materials. Additionally, this review introduces ongoing research on machine tool digital twins, mobile platform-based machining systems, and hybrid machines integrating additive manufacturing technology.

Positioning technology is an interdisciplinary field of engineering that supports the development of industrial manufacturing. Well-developed positioning technology also demonstrates a certain degree of product competitiveness. It typically comprises four main components: actuator, sensor, controller, and transmission element. This review briefly introduces research related to positioning technology conducted in Taiwan, focusing on studies published mainly after 2009. The papers surveyed are categorized according to the four main components, which are discussed separately in Sections 2 through 5. Section 2 presents various types of actuators, including piezoelectric, electromagnetic, electrostatic, shape memory alloy, fluidic, and microelectromechanical-system-based actuators. Section 3 covers displacement sensors. In addition to contact-type sensors, noncontact sensors such as eddy current sensor, optical sensor, encoder, and laser interferometry system are described. Section 4 discusses controllers, introducing both classical control-based strategies and modern control schemes, such as sliding mode control, fuzzy control, repetitive control, iterative learning control, and others. Section 5 focuses on transmission elements, including flexure hinge, ball screw, linear guideway, ball bearing, fluidic bearing, magnetic bearing, and other related components. Applications of positioning technology in various fields are addressed in Section 6. Finally, based on the surveyed papers, a brief summary and concluding remarks are provided in Section 7. This review offers an overview of positioning technology and highlights the research capabilities related to this field in Taiwan.

In recent years, as semiconductors have become finer and finer, their manufacturing equipment has been required to have nanometer-scale positioning. In such systems, the main sources of positioning error are friction, thermal deformation, and strain from the roller guides. Throughput is mainly limited by the heat generated by the motor coils and thermal deformation of the stage due to friction of the roller guides. Magnetic levitation (maglev) guides can prevent these two effects that limit positioning accuracy because they can provide non-contact support for the table. In particular, compared to a stacked maglev stage, a planar-type-maglev stage has a dramatically smaller moving mass, enabling high-speed movement with high positioning accuracy. From the viewpoint of reducing heat generation in the moving unit, it is desirable to install the magnet on the levitating side and the coil on the fixed side. However, a movable magnet type planar stage requires the use of a coil switching control system. In this study, a new magnetic levitation planar stage system with the movable magnet array and the coil switching control system was prototyped and evaluated. The basic characteristics of the prototype magnetic levitation stage were evaluated. As a result, stable levitation positioning by coil switching control was confirmed.

A four-axis positioning mechanism using flexure guide and electromagnetic actuator has been developed. Additionally, a prototype of a surface topography processing and measurement system applying the developed mechanism has been produced. In a previous study, the authors constructed a three-degree-of-freedom planar positioning mechanism with nanometer resolution in the X–Y–θ-axes with a stroke of 1 mm using a flexure guide composed of 32 leaf springs and an electromagnetic actuator. To improve the motion performance in continuous-path positioning with multi-axis control, a new mechanism is constructed applying a monolithic flexure mechanism. By developing a new vertical Z-axis mechanism and integrating it with the planar positioning mechanism, a four-axis positioning mechanism with X-, Y-, θ-, and Z-axes is completed. A simple stylus probe using a thin cantilever with semiconductor strain gauges is developed and attached to the Z-stage to construct a topography processing/measurement system. Two modes of operation are defined for using this system. One is a processing mode in which the stylus is pressed against the specimen and X–Y scanning is performed while applying a sufficient contact pressure to create a scratched groove. The other is a measurement mode in which the stylus traces the surface while gentry being in contact with it to obtain topography measurements. A straight line with a length of 0.3 mm is scratched using the processing mode. The surface profile is then derived using the measurement mode. The measurement results show the actual surface shape and, thereby, demonstrate the potential of the system.

A contact-type tool setter for measurement of lengths of small-diameter tools and tools in rotation has been developed. The tool setter consists of a seesaw structure and air pressure to maintain a small contact force. A contact probe at the tip of the seesaw structure is brought into contact with the tool, and the tool length is measured by accurately measuring the movement of the seesaw structure using a laser scale. To evaluate the developed tool setter, probe pushing tests were conducted, and the tool setter was confirmed to have a resolution of 0.1 µm. Additionally, a precision machining test confirmed that the tool length adjustment using the tool setter improved the machining accuracy. Furthermore, the tool setter was shown to be capable of measuring the operating characteristics of a machine tool at the sub-micrometer level.

This paper deals with an equivalent circuit model of a stacked piezoelectric actuator. The displacement of a piezoelectric actuator is usually controlled by adjusting the applied voltage. However, the displacement shows hysteresis relative to the applied voltage. The proposed model is based on Martin’s model for a stacked piezoelectric actuator in which the mechanical compliance is represented with a Cauer circuit composed of Zener diodes and nonlinear capacitances using hyperbolic tangent functions. The parameters were identified from the initial displacement during polarization or the hysteresis loop in a steady state. The trends of the hysteresis of the displacement relative to the applied voltage coincided with the results of experiments on static displacement and frequency characteristics. Less hysteresis in the displacement by controlling the supplied charge was also represented.

This paper presents a novel wireless microactuator that utilizes a temperature-sensitive magnetic material (TSMM) for remote operation. The microactuator consists of a TSMM part, permanent magnets, and a laser light source that heats a specific area of the TSMM part. The driving principle is based on a decrease in the saturation magnetic flux density of the TSMM with increasing temperature, which enables continuous force generation and movement. Two prototype mechanisms, namely a microgripper and a rotary micromechanism, were designed and fabricated to demonstrate the capabilities of the proposed actuator. The prototype microgripper was designed such that the finger carrying the microobject and the actuator environment were separated by a thin glass. The finger and the actuator were connected by magnetic coupling such that the microobject and the actuator could be used in their respective environments. The prototype of the rotary micromechanism was designed to consist of a small number of parts with simple shapes. Magnetic field analysis showed that the microgripper generated a gripping force of 2.7 mN, and the rotary micromechanism generated a torque of more than 50 µNm. The experimental results confirmed the continuous movement of the two prototype mechanisms under laser irradiation. The motions of the actuator and the magnetically coupled finger in the microgripper were measured, and the differences in motion were confirmed. The motion of the rotary micromechanism was measured under three different laser powers, and it was confirmed that the higher the power, the faster the TSMM ring rotated.

Characteristics of the coupling system influence the characteristics of entire machines, such as ball screw feed drive systems. However, quantitatively evaluating the characteristics of coupling systems is difficult because coupling systems have frictional contact surfaces that exhibit nonlinear characteristics. Hence, effective evaluation and analysis methods have not been proposed, and factors that influence the characteristics have not been investigated. In this study, the influences of contact characteristics between shaft and borehole surfaces are investigated. The characteristics of several samples were evaluated by measuring frequency responses, and it was found that the stiffness and damping of the system varied with the fixing torque and geometry of the clamp hub. It is assumed that this phenomenon is attributable to the change of the contact surface pressure, which depends on the fixing torque and geometry of the clamp hubs. Concerning the stiffness, in this study, a simple model considering the real contact area to calculate the relationship between the contact surface pressure and contact stiffness was constructed. Results obtained by applying the model were compared with experimental results, and the proposed model was confirmed to be representative of the influences of the fixed torque and coupling clamp hub bore diameter.

High-precision motion control requires robust controllers that balance performance with design simplicity. Nominal characteristic trajectory following (NCTF) control has been effective in achieving precise motion without relying on accurate dynamic models or advanced control expertise. However, its performance is influenced by external signals, particularly reference inputs. This paper proposes an additional compensator based on the internal model principle (IMP) to simplify the design of controllers for high-precision motion control. To improve the sinusoidal tracking performance of an NCTF control system while preserving the simplicity of its control design, we implement an additional resonant (R) compensator with a sinusoidal structure, adhering to the IMP for an ideal control structure, without requiring additional parameter tuning. Additionally, a repetitive compensator, designed specifically for tracking periodic signals, and the classical cascade controller commonly used in industry are implemented for comparative performance evaluation. The effectiveness of the additional R compensator was verified through simulations and experiments using a ball-screw mechanism. The results show that incorporating the R compensator significantly reduces tracking errors for low-amplitude, high-frequency sinusoidal signals (2 µm at 5 Hz), achieving an 80% reduction in maximum tracking error. The findings demonstrate that the R compensator effectively enhances tracking precision while preserving the simplicity of NCTF control design.

In industrial applications such as numerical control machine tools and robots, high-speed positioning operations are always required to improve productivity. However, high-speed positioning operations cause residual vibration owing to the inertial force of the driven object. To suppress this vibration, researchers have proposed adjusting the time constant and/or applying filters for the positioning command; these are generally applied in the field. However, the acceleration and deceleration time increases when the conventional vibration-suppression methods are applied to suppress lower frequency vibration. The purpose of this study was to establish a design method for positioning commands that minimizes the residual vibration and positioning time. A method to eliminate the natural frequency component was applied. This method is based on the principle that when a specific natural frequency component of an external force is reduced to zero, the natural frequency component of the system will not oscillate. Although the approach can be used to design positioning commands with any positioning time, for actual systems, the torque limitation of motors should be considered because it limits the minimum positioning time. Therefore, in this study, a positioning command that can minimize the positioning time and residual vibration was designed based on the formulated relationships between positioning time and maximum acceleration of positioning command. The effectiveness of this method was verified using an experimental system with torsional beams. We confirmed that the positioning commands designed using the proposed method can minimize residual vibration and positioning time.

Nominal characteristic trajectory following (NCTF) control has been proposed to realize high-precision control systems with minimal information and expertise. NCTF control is effective at suppressing the adverse effects of friction and has demonstrated precision motion and high robustness. To improve the vibration-suppression characteristics of NCTF control systems, a method of incorporating additional compensators (which are easy to design) has been proposed and proven effective. However, the robustness of NCTF control with additional compensators has not been verified. They have also not yet been evaluated for application to a rotary mechanism with a gear reducer (commonly used in robots). This study experimentally investigated the robustness and performance improvement of NCTF control of a rotary mechanism with a gear reducer using additional compensators. To evaluate robustness, additional units were used to increase the moment of inertia and to add disturbance to the mechanism. Despite some degradation in performance owing to the added inertia and disturbance, the overall performance remained good, demonstrating the efficacy of the proposed control approach. This study demonstrated the potential of NCTF control systems augmented with additional compensators to boost servomechanism performance in automation, particularly in robots subjected to variable loads and disturbances.

In equipment manufacturing processes, pneumatic systems are commonly used because of their relatively simple design, high power-to-weight ratio, clean power source, safety, zero heat generation, and easy maintenance. Pneumatic systems contain devices such as linear actuators that convert air pressure and create straight-line motion and are widely used in industrial automation, robotics, and machinery. However, these pneumatic actuators are susceptible to frictional forces and difficult to control owing to their inherent air pressure characteristics. They therefore lack the precision and accuracy of electric actuators. Precision motion systems have become increasingly popular in manufacturing equipment, and this industrial revolution has led to the reduction of pneumatic systems adopted in precision motion systems. Many special type pneumatic servo systems have, through years of research and engineering efforts, been designed to meet special requirements such as precision and speed. However, these involve high investment and engineering time to adopt. Hence, this paper presents a novel compensator-based control system that uses a standard type linear pneumatic actuator to achieve ultra-high precision positioning control. The ultra-high-precision control feature, endowed with the inherent benefits of pneumatic systems such as safety, provides an all-around approach for future industrial automation. The research methodology involves the design of a control system with a novel compensator algorithm to address friction interference in pneumatic actuators, which is the primary causation factor of poor positioning accuracy in pneumatic servo systems. Moreover, the compensator can be easily integrated into any main control system, allowing easy adoption in any industrial field. The experimental results prove that high accuracy in the steady-state position can be achieved by this proposed algorithm. A steady-state accuracy of ±1 nm was obtained. Compared to previous studies in which accuracies of approximately 5 nm were obtained, the compensator-based control system achieves a considerably higher accuracy than the previously reported accuracies, matching that of electric actuators. The final section of this paper will present potential directions for future work.

Machine-stand vibrations often occur in industrial machines with table-positioning systems during the high acceleration/deceleration motions of the table, which deteriorate positioning performance. To enhance the performance, the mechanism and control system of the equipment should be effectively designed using a simulator that can adequately simulate machine-stand vibrations. Machine-stand vibrations exhibit strictly nonlinear characteristics, but the underlying mechanism remains unclear. Therefore, it is difficult to construct a model that can accurately reproduce the vibration characteristics. In this study, a prototype with a machine stand and support mechanisms used in actual industrial machines was modeled using multibody dynamics software while considering contact as a nonlinear element. As an element of contact in the support mechanism model, a nonlinear contact force was defined between the leveling bolt and leveling plate based on Hertz’s contact theory. Furthermore, linear spring forces between the leveling plate and floor surface were used to reproduce the contact effect between them. The frequency characteristics of the drive systems with a basic closed-loop control system were measured using sinusoidal sweeps, and the dependence of the machine-stand vibration and its characteristics on the disturbance amplitude were investigated. The model was validated by comparing the experimental and simulation results. The results revealed that it is possible to reproduce the softening-spring characteristics of machine-stand vibrations by properly incorporating the contact between the bolt and plate into the model.

Cascaded positioning stages offer a systematic design approach for achieving larger dynamic range and more degrees of freedom (DOFs). Designs such as 1-DOF coarse-fine stages and 2-DOF cascaded stages can be realized effectively. However, due to the coupling resulting from inertial forces caused by the interaction between sub-stages, the dynamic performances are usually limited, and it is desired to study the interaction between sub-stages and develop controllers to reduce the coupling. Here, a single-axis heterogeneous cascaded stage is designed and realized to serve as the platform for addressing the concerns. This novel stage integrates a rubber bearing positioning stage as the upper, and a compliant metallic positioning stage as the bottom components. Through modeling and dynamic testing, the stage dynamics are established, and controller designs based on loop transmission shaping method are implemented. The motion due to inertial coupling is then studied, and the motion can be effectively suppressed after optimizing the controller design by considering the coupling dynamics and positioning-axis error. In summary, the loop transmission shaping control scheme is successfully developed for controlling the motion and reducing the coupling of this cascaded positioning stage. In the future, the stage can be used as a test platform for developing other control schemes for further enhancing the performance of precision motion stage.

In recent years, road collapse incidents owing to underground cavities have occurred frequently. To investigate underground cavities before accidents occur, periodic cavity inspections are conducted using ground penetrating radar (GPR) equipment. When cavity inspection is performed by humans directly moving a GPR-equipped hand cart in areas inaccessible to vehicles, such as sidewalks, the area that can be inspected simultaneously is limited, often resulting in missed inspections. To address this issue, we developed a robotic system with omnidirectional mobility using a four-wheel independent steering and driving mechanism, primarily aimed at automating sidewalk cavity inspection. Cavity inspection requires path tracking control along the designated survey line; however, conventional path tracking methods are mainly designed for robots with nonholonomic constraints, such as two-wheel independent driving systems, making them unsuitable for omnidirectional mobile robots. In this study, we propose a control method called simultaneous path and orientation tracking control, which considers omnidirectional mobility. The proposed method consists of feedforward control for the target path and orientation, combined with feedback control to correct deviations. By distributing the control inputs into two degrees of freedom for the translational velocity and one degree of freedom for the angular velocity, the position and orientation of the robot can be controlled independently. Several experiments were conducted in a real-world environment to evaluate the reproducibility of the tracking control, demonstrating the effectiveness of the proposed method in terms of path and orientation tracking performance.

Continuous generation of shock waves without replacing the diaphragm between the high-pressure and low-pressure sections of the shock tube could be applied to actuators with pistons moving at high speed and high power. The realization of actuators using such shock waves is expected to play an important role in industrial equipment that requires high-speed actuation. Previous reports described performance experiments involving a shock wave generator with two shock tubes facing each other in which it was shown that it is possible to cause the drive piston to reciprocate by adjusting the initial conditions. In this study, we report the behavior of a piston in a shock wave generator that moves the piston from the initial position to the target position.

This study proposes applying adaptive incremental sliding mode control (AISMC) to a dual-axis core-type permanent magnet servo synchronous linear motor to establish a high-precision dual-axis motion control platform. First, this study uses the magnetic field-oriented theorem to convert the three-phase control current of the drive motor into the d-q axis control current and integrates it with the magnetic thrust equation and mechanical model of the linear motor to obtain the biaxial linear dynamic equations of the motor platform. However, many external disturbances will be encountered during the operation of the motor, such as friction, ripple effects, and variations in the internal parameters of the system. We collectively refer to these disturbances as system uncertainties and incorporate them into the dual-axis linear motor platform design. Considerations are made within the dynamic equation to establish a more refined dynamic equation. Then, a high-order controller can be designed based on the system dynamic equation established above. In the controller design stage, we will first design the SMC. Due to its simple structure and high robustness, it is very suitable for uncertain systems such as linear motors. In many systems, its shortcoming is the chattering phenomenon in the smooth mode. To improve this phenomenon, we designed AISMC. Its feature is that the past control input will be considered when designing the controller to suppress chattering. The vibration phenomenon and adaptive control are used to compensate for system uncertainty. Finally, experiments prove that this research successfully improves the precision of the dual-axis motion platform.

The demand for sensing in robotics and automation has increased due to the decrease in the labor force. Recent advances in computational performance have advanced the widespread use of image processing technology in various applications. This special issue aims to provide researchers with an opportunity to access the latest research and case studies on advanced image processing, computer vision, and sensing techniques for robotics and automation. The topics of interest in this special issue are as follows:

1) Theory and algorithms: Image processing, computer vision, pattern recognition, object detection, image understanding, media understanding, machine learning, deep learning, 3D measurement, simultaneous localization and mapping (SLAM), multispectral image processing, visualization, virtual reality (VR) / augmented reality (AR) / mixed reality (MR), and datasets for image processing;

2) Industrial applications: Factory automation, machine vision, visual inspection, monitoring, surveying, logistics;

3) Sensing techniques for robotics and automation: Robot vision, advanced driver-assistance systems (ADAS), autonomous driving, robotic picking, assembly, and palletizing;

4) Image processing hardware and software: Image acquisition devices, image sensors, image processing systems, sensor information processing;

5) Man machine interface: Visualization, human interface devices.

This special issue features 20 research articles that highlight the latest advancements in advanced image processing techniques for robotics and automation (Part 1: 10 articles, Part 2: 10 articles). We extend our heartfelt gratitude to all the contributors, reviewers, and editorial staff for their dedication and support in realizing this special issue.

In irradiated environments that are inaccessible to human workers, operations are often conducted via teleoperation. Consequently, robot operators must maintain continuous situational awareness of a previously unknown working environment. Visual information regarding task targets and robot manipulators is of utmost importance. The proposed method employs rail-mounted observation robots to position easily replaceable cameras capable of long-term deployment in such environments. To reduce the cognitive load on teleoperators, the automatic viewpoint selection system eliminates the need for direct control of observation robots. This research presents a method for using a single rail-mounted observation robot to gather information on an unknown environment and automatically determine an optimal viewpoint. A key contribution of this study is the viewpoint presentation system, which can adapt to occlusions caused by robots and adjust its positions accordingly. The proposed method was validated through computer simulation using a hybrid model consisting of a static environment and a dynamic robot arm, which moves within the environment and may obstruct views. Furthermore, the feasibility of the approach was demonstrated in a real-world experiment involving a robot arm performing a teleoperation task.

A method is proposed for interpolating pose information by integrating data from multiple thermal cameras when a global navigation satellite system temporarily experiences a decrease in accuracy. When temperature information obtained from thermal cameras is visualized, a two-stage temperature range restriction is applied to focus only on areas with temperature variations, making conversion into clearer images possible. To compensate for the narrow field of view of thermal cameras, multiple thermal cameras are oriented in different directions. Pose estimation is performed with each camera, and the estimation results of one camera are interpolated using those of other cameras based on reliability derived from predicted values of the camera pose. Experimental results obtained in a low-light nighttime environment demonstrate that the proposed method achieves higher pose estimation accuracy than other state-of-the-art methods.

In recent years, autonomous mobile robots have been deployed in outdoor environments, including challenging conditions such as snow. In snowy environments, stable motion control is difficult because detecting pavement edges from camera images becomes unreliable due to snow coverage. To address this limitation, we propose a novel framework for autonomous motion control in snowy environments, utilizing semantic segmentation and generative adversarial networks (GANs). In our approach, winter images captured by a camera are transformed into summer-like images using a GAN, enabling automatic detection of snow-covered pavement through semantic segmentation. However, conventional GAN-based image translation has limited accuracy because it does not account for the temporal consistency of time-series images. To overcome this issue, we improve the temporal consistency of GAN-based image translation by incorporating the sequential characteristics of images captured by a monocular camera mounted on a mobile robot. The improved GAN demonstrates high temporal consistency in real-world datasets. Furthermore, we achieve stable motion control in snow-covered environments using a novel scheme that generates optimal subgoals based on pavement coplanarity.

The objective of this study is to propose and evaluate a method for automatically constructing a training dataset specific to each closed-circuit television (CCTV) camera to identify ice-road surface images using anomaly detection. In this approach, dry-road surface images were defined as normal, while ice-road surface images were considered as anomalous. By training the anomaly detection model solely on normal images, the system can identify ice-road surface images. However, the accuracy of identification may decline owing to variations in shooting locations, individual CCTV camera installation sites, and environmental conditions. To address these challenges, we used data from the Automated Meteorological Data Acquisition System (AMeDAS) near CCTV cameras to automatically select normal images, thereby enabling the construction of camera-specific training datasets. Specifically, if precipitation and temperature data from the nearest AMeDAS station—often located several kilometers away—consistently remained below predefined thresholds for a certain duration, the corresponding CCTV images were assumed to depict dry-road surfaces. We applied the proposed method to CCTV camera images and experimentally evaluated it using the PatchCore anomaly detection technique. The results demonstrated high accuracy, achieving an area under the receiver operating characteristic score of 0.961 and 0.999 during the daytime and nighttime, respectively, validating its effectiveness in detecting ice-road surface images. To demonstrate the practical viability of the proposed method, we validated it using footage from two additional CCTV cameras. The results showed that ice-road surface images could be easily identified using the proposed approach.

Cooking robots provide life-assistance and should be capable of performing suitable tasks using simple language instructions. While previous studies have employed LLMs to develop action plans for the upper symbolic layer, there are many issues regarding its connection to the lower physical layer. For example, considering scooping, although the motion to scoop food from a container can be generated, the actual task may still fail. This is because the executability of the desired task is not examined during the trajectory generation stage. In this study, we propose the “margin of task achievement” comprising two components, the “margin for interference” and the “margin for execution,” to evaluate the degree of task accomplishment, and propose a method to determine the optimal trajectory for the task and, if necessary, the switching of tools. In an experiment in which three different tools were used to perform the task of scooping green tea powder, the average success rate was 93.3%, which was 61.6 percentage points higher than that when the margin of task achievement was not considered. These results confirm the validity of the proposed method.

We propose the construction of image datasets via data cleansing for food recognition using a convolutional neural network (CNN). A dataset was constructed by collecting food images and classes from web crawling sites that post cooking recipes. The collected images included images that cannot be effectively learned by the CNN. Examples include images of foods that look extremely similar to other foods, or images with mismatched foods and classes. Here, these images were termed “content and description discrepancy images.” The number of images was reduced using two criteria based on the food recognition results obtained using CNNs. The first criterion was a threshold for the difference in the estimated probabilities, and the second was whether the estimated class and food class matched. These criteria were applied using multiple classifiers. Based on the results, the dataset size was reduced and a new image dataset was constructed. A CNN was trained on the constructed image dataset, and the food recognition accuracy was calculated and compared using a test dataset. The results showed that the accuracy using the dataset constructed using the proposed method was 7.4% higher than that of the case using web crawling. This study demonstrates that the proposed method can efficiently construct a food image dataset, demonstrating the data-cleansing effect of the two selected criteria.

This study proposes a method for detecting scallops in seabed images using deep-learning based instance segmentation and active learning techniques. This method uses a mask region-based convolutional neural network (Mask R-CNN) combined with active learning to enable efficient annotation and adaptive learning in different seabed environments. A comparison with the transformer-based deformable detection transformer (Deformable DETR) model provides a detailed evaluation of the detection performance. The proposed method proves to be effective in detecting of object features while removing unnecessary background regions in noisy seabed environments. Active learning with margin sampling enhances the annotation process and creates an effective dataset from numerous seabed images. Experiments conducted on a large dataset of over 83,000 seabed images show that Mask R-CNN outperforms Deformable DETR, achieving an F-measure of 0.89 compared to 0.85. This study contributes to the field of fishery resource investigations by providing an approach for efficient learning using new data, which is crucial for maintaining accurate scallop detection systems over time.

Turing patterns can be used to reproduce biological skin patterns. However, the similarity between the reproduced Turing pattern and the actual skin pattern has not been evaluated quantitatively and has only been determined visually. We propose a method for quantitatively evaluating the similarity between the pattern reproduced by the Turing pattern and biological skin patterns using persistent homology, which is a topological data analysis method. Persistent homology can quantitatively extract the structural information of the pattern, allowing an analysis that focuses only on the structure of the biological skin pattern without being affected by variation in the pattern due to individual differences. The experiment tested the effectiveness of persistent homology analysis on multiple Turing patterns generated by varying the parameters of the Gray–Scott model. We used real images of Scomber japonicus as the organisms represented by Turing patterns, and the similarity between the real images and Turing patterns was calculated using persistent homology. The results confirmed the tendency for a high degree of similarity between real images of several S. japonicus and the Turing patterns generated with certain parameters. These results suggest that persistent homology can be used to quantitatively evaluate the reproducibility of Turing patterns.

Red tides are phenomena caused by the abnormal proliferation of marine phytoplankton, leading to mass fish mortality and severe economic damage to fisheries. Currently, the detection and quantification of harmful phytoplankton rely primarily on manual inspection using optical microscopes. This process is time-consuming, labor-intensive, and requires specialized expertise in species identification. In this study, we propose an automated detection system using deep learning-based object detection methods to classify various marine phytoplankton species from microscopic images and identify harmful red tide-related species. Our approach aims to enhance early detection capabilities, reduce the burden on researchers, and improve the accuracy of harmful phytoplankton monitoring.

This paper introduces a novel VGG-16-based visual navigation architecture for a differential drive robot, Social Plantroid, using a temporal vision mosaic, a novel approach which joins current and previous robot vision frames for heading estimation. As a minor contribution, it integrates a novel sunlight/shadow detection algorithm using Gabor filters. The neural network is trained with simulated data employing the artificial potential field method, which is another novelty for map-less robot navigation. Virtual and real-world experiments validate the effectiveness of this architecture in obstacle avoidance and navigation.

In the deposition-based 3D printing process, the material stacking error during printing is a major defect that affects the entire process. Therefore, a printing method that compensates for the nozzle movement path during the printing process is needed. Our research mainly aims to demonstrate the effect of the compensation control method for path changes during the printing process by vision-based real-time feedback control. In this approach, the whole area can be monitored, and region of interest (ROI) is defined to observe a specific area. Two cameras observe the area around the nozzle and avoid the effects of occlusion. Additionally, binarization and edge detection are applied to the ROI. The feedback controller acquires the distance between the center coordinates and the target path in real time and calculates a compensation motion. In this research, a printing experiment is conducted using two types of materials. In the case of the 3D printing experiment for a cylindrical structure with a radius of 50 mm, the compensation effect was verified by the convex path change of the target path. Feedback experiments using mortar materials confirmed the effect of feedback of visual information in suppressing shape errors by compensating for convex path changes. The root mean square error of the nozzle trajectory was 2.43 mm relative to the predicted trajectory. The shape error relative to the circular shape decreased with each layer. These results showed the effectiveness of the failure suppression method based on visual feedback, which is sufficiently practical for large-scale printing processes.

Recently, the manufacturing industry has digitized skills through motion capture to solve issues such as human resource development and skill transmission. However, the amount of data on body movements obtained from motion capture is enormous, and machine learning techniques are required for data mining. Elemental tasks are useful for conducting work analysis, where the unit of analysis or unit element is divided by the entire work. This study proposes an assembly movement analysis method based on work classification using motion capture and machine learning. Here, the differences between motions of experienced and inexperienced workers were classified using motion capture and deep learning software for the worker’s experience level and body part analysis.