Complex systems are known to pose significant risks due to unexpected defects arising from interactions between their subsystems and components, even when all components are in safe conditions. Two safety analysis methods have been proposed for such complicated system: the Functional Resonance Analysis Method (FRAM) and the System-Theoretic Process Analysis (STPA). FRAM takes a relatively macroscopic view of the system and considers resonance between functions with six aspects: input, output, precondition, resource, control, and time. These aspects enables the description of complex dependencies among functions. On the other hand, STPA considers the interactions between components inside the functions using a control structure diagram to identify the loss scenarios following any unsafe control actions. This study applies an integrated safety analysis combining FRAM and STPA to ensure resilient operation in the proposed smart antenna system for astronomical observations using balloon-borne very-long-baseline interferometry (VLBI). First, the integrated safety analysis uses FRAM to model the system from a macro perspective. Then, STPA is applied to model the control structure and identify the loss scenarios within the FRAM functional models. This method makes it possible to trace loss scenarios in accordance with operational scenarios focused on safety by the FRAM functional model. Consequently, this method easily reveals new loss scenarios that were not easily accounted for in our previous research, which only used STPA analysis.

In recent years, the transformation of manufacturing industries toward "product-as-a-service" has changed the market environment, resulting in placing the entire customer experience at the center of value creation. Consequently, word-of-mouth (WOM) and post-purchase satisfaction have become critical drivers of product diffusion. In the phases of product diffusion, characteristics of target customers shift from one to another as the number of customers increases. Product designers can no longer rely on a linear, one-way design process until launch. Instead, they must continuously analyze customer feedback and iteratively refine designs based on real-world evaluations. This study aims to provide a novel framework for rapid and precise design decision-making by constructing a product diffusion simulation model. As the situation involves emergent mechanisms among various factors and their complicated interactions, the framework integrates system dynamics as a macro-level model with multi-agent simulation (MAS) as a micro-level model to capture how individual uncertainties aggregate into macro-level diffusion patterns. By employing MAS, the model reproduces the distribution of adoption attitudes and the emergent effects of WOM at an individual level. The simulation effectively serves as a design framework by analyzing dominant factors and critical transition points that dictate diffusion outcomes. Through a virtual case study, this paper demonstrates that specific corporate investment balances significantly trigger the branching of diffusion patterns, providing actionable insights for strategic product design.

This study investigated localized obstacle avoidance motions for AMRs within warehouses using vortex fields. For obstacle detection, 2D-LiDAR was employed to measure distances and acquire point cloud data. By generating vortex fields at the coordinates of these point clouds and superimposing them, a final vector was calculated. The direction of this vector was then used to determine the rotational speed. Translational speed was controlled by setting a maximum speed and decelerating based on the direction of travel and distance to obstacles. When the vortex field generated had only one rotation direction, adding a virtual wall’s vortex field enabled more efficient avoidance. Since this system uses vortex fields instead of repulsive forces, stagnation does not occur as long as there is sufficient width for the AMR to pass through. Particularly when obstacles appear in the presence of walls, the system avoids stagnation and collisions with walls, largely due to the vortex field effect. When an obstacle exists in the target direction, the vortex field boundary primarily selects the direction closer to the target direction on either side of the obstacle. Therefore, even when a moving obstacle appears, avoidance can be performed without interfering with the obstacle's progress. As a result, AMRs can pass each other without issue. This method generates vortex fields on both sides, causing the AMR to travel along the center of the path. This maximizes the distance from walls, enabling safer navigation. Furthermore, normalizing the strength of the left and right vortex fields reduces the impact of obstacle conditions on the vortex fields, allowing for more stable navigation.

In thermal network modeling, accurate parameter identification, particularly of thermal conductance, is essential for reliable temperature prediction. Sequential data assimilation techniques, which integrate measurement data with numerical models, have attracted increasing attention for parameter identification. In sequential data assimilation, model parameters are estimated at each measurement step, updating the numerical simulation and enabling robust thermal predictions. Accordingly, this approach has been widely applied to thermal network models. However, treating thermal conductance as a model parameter generally yields a nonlinear state-transition equation, often requiring nonlinear sequential Bayesian filters, such as the unscented Kalman filter or the ensemble Kalman filter. These nonlinear filters often entail high computational costs, limiting their applicability to large-scale thermal network models and real-time state monitoring under constrained computational resources, such as onboard systems. This study proposes a fast sequential data assimilation method for thermal network models. Variations in heat transfer induced by changes in thermal conductance are represented as correction heat quantities and introduced as model parameters. With this formulation, the state-transition equation becomes linear. Consequently, sequential data assimilation can be performed using a linear Kalman filter, which can reduce computational cost. The thermal network model incorporating the correction heat quantities and its associated data assimilation procedure are presented. The proposed method is validated experimentally using measured temperature data from a simulated motor system. Thermal conductance estimation results are compared with those obtained using an unscented Kalman filter and an extended Kalman filter. The results show that, in this experiment, the proposed method achieves comparable estimation accuracy while reducing computation time. These results indicate that the proposed approach is suitable for real-time thermal state monitoring and parameter identification in thermal models with limited computational resources. The present validation is limited to steady-state thermal test conditions, and verification under continuous large temperature variations remains future work.

This paper presents a 3D-printed sit-to-stand assist device that operates without actuators and evaluates its assistive performance. The purpose of this study was to experimentally verify whether the assistive performance assumed based on the device structure can be achieved in the actual prototype and to identify design improvements through evaluation experiments. Experiments were conducted with healthy participants, focusing on the buttock lift-off phase during the sit-to-stand task. Assistive performance was evaluated using three indices: (i) the proportion of body weight supported by the assisted limb, (ii) comparison of muscle activity between the assisted and non-assisted limbs, and (iii) the input–output torque ratio of the device. In particular, the validity of the device design was examined by comparing the measured and estimated input–output torque ratios. Although the measured assistive load accounted for only part of body weight because of soft tissue deformation and voluntary muscle activity, relatively large ground reaction forces were observed in the device-worn limb even when muscle activity was intentionally suppressed. These findings suggest that the proposed device has the potential to support part of the body weight borne by the paretic lower limb in individuals with post-stroke hemiplegia. In addition, the device increased the ground reaction force of the assisted limb relative to the non-assisted limb, even when sufficient muscle activity was not achieved during postural maintenance. These findings indicate that the device may increase weight-bearing on the assisted limb while reducing the physiological burden on the user. On the other hand, the measured input–output torque ratio was lower than the estimated value. Although this discrepancy may have been partly influenced by soft tissue deformation, structural improvements are necessary to enhance the sit-to-stand assistance performance.

This study aims to develop a power output control mechanism for the Stirling Engine. In response to global climate and environmental challenges, there is a growing demand to establish new energy systems. The Stirling Engine is recognized for its clean and eco-friendly characteristics; however, there are several obstacles to broader industrial application. One of the most significant difficulties is the lack of a power output control mechanism. Effective power output control is crucial for the reliable operation of the engine. This study connected the third cylinder and piston to the Stirling Engine and then successfully developed a control mechanism, which could adjust the rotational speed and power output as desired. Experimenting with this mechanism also clarified the relationship between internal pressures and output power within the engine. This research confirmed that this mechanism operates as a variable compression ratio engine. This research contributes to harnessing the Stirling Engine as a clean and sustainable energy solution. Furthermore, this mechanism indicates the potential application as the internal combustion engine.

Systems engineering has been applied to the development of complex systems that require high safety, such as aircraft and space systems. Recently, with the increasing sophistication and complexity of systems, such as automobiles, medical systems, and other general systems, there is a growing need to apply systems engineering to the development of these systems. However, these complicated systems require sufficient knowledge and experiences to perform safety analysis. This study proposes a new framework to integrate safety system design with systems engineering called system and safety design matrix (SSDM). SSDM connects system operation, functional requirements, and safety analysis to the system architecture, which is the relationship between functional requirements and components. This allows for a comprehensive overview of the entire system from multiple perspectives. SSDM makes it possible to efficiently progress and manage the system design together with safety design. This efficient process is valuable not only during the initial design, but also for preventing oversights during design changes.

Springs are widely used in various industries, with high-hardness springs being in particularly strong demand in the automotive sector. From a manufacturing perspective, it is essential to establish a method for predicting elastoplastic deformation behavior, including cyclic plastic deformation induced by processes such as the shot peening, in spring steels. In this study, the extended subloading surface model is applied to accurately describe the behavior of high-hardness spring steel under uniaxial tensile loading and tensile-compressive cyclic loading. To determine appropriate material parameters for the extended subloading surface model, the particle swarm optimization (PSO) method is employed, because the model involves a relatively large number of parameters, making manual calibration difficult. Numerical simulation using material parameters obtained by the PSO method show good agreement with virtual test data in terms of stress-strain and internal state variables-strain relationships. In addition, experiments on the spring steel were also conducted under uniaxial pulsating and reversed cyclic loading conditions. The results exhibited mechanical ratcheting and cyclic softening, both of which were well reproduced by the extended subloading surface model with PSO-identified material parameters. These results demonstrate that the PSO method is an effective approach for identifying material parameters in sophisticated constitutive models without repeated manual parameter tuning.

The total acid number (TAN) serves as a critical indicator of lubricant degradation; however, conventional measurement techniques are time-consuming and require the use of organic solvents. In this study, a partial least squares (PLS) regression approach was employed to predict TAN from spectral data obtained via Fourier transform infrared (FT-IR) spectroscopy and ultraviolet–visible–near-infrared (UV-Vis/NIR) spectrophotometry. Oil samples were intentionally oxidized under controlled conditions using various base oils and additive formulations, prepared through the rotating pressure vessel oxidation test (RPVOT). The acquired spectral data were preprocessed using smoothing, standard normal variate (SNV) correction, and Savitzky–Golay filtering prior to model construction. A PLS regression model was then developed to predict TAN, and its performance was evaluated using the root-mean-square error (RMSE) as a metric for comparing predictive performance. The results demonstrated that the type of base oil exerted a pronounced influence on model performance. Both FT-IR and UV-Vis/NIR spectra enabled reliable TAN prediction when models were developed for individual oil types. Among these, FT-IR spectroscopy exhibited high predictive capability even when spectra from all oil types were combined into a unified model. Further optimization of spectral resolution and calibration ranges yielded additional improvements. In UV-Vis/NIR, lower spectral resolution improved prediction performance for individual oil models, whereas in FT-IR, restricting the wavenumber range to 1800–1650 cm^-1 was effective for improving prediction performance. Notably, when the base oil type was held constant, TAN could be predicted using only the FT-IR spectral region of 1800–1650 cm^-1, even for samples excluded from the calibration data set. In addition, a prediction model based on synthetic oil showed potential applicability to lubricants with different base oil types.

In Engineering-to-Order (ETO) products, high customization demand from customers leads to a wide variety of product variations. This results in increased design effort, longer lead times, and strong reliance on experience-based design processes. To address these challenges, this paper proposes a design configurator based on a three-layered hierarchical design thinking approach: product structure definition, module selection, and determination of layout and geometric dimensions. The proposed method adopts a hybrid approach that integrates rule-based logic using decision tables with similarity-based retrieval utilizing a Domain Mapping Matrix (DMM). The weights in the DMM are derived from expert knowledge, ensuring high reliability in recommendations. This hierarchical structure prevents the black boxing of complex design logic and ensures high maintainability. The developed configurator was applied to the layout design of high-speed elevators in two large-scale building projects. The evaluation focused on design effort, measured in person-hours, including both layout design and subsequent drawing generation. As a result, the required design effort was reduced to approximately 85–90% of that of conventional design without the configurator. In addition, design knowledge such as decision tables, weight matrices, and dimension tables is managed as external master data. Operational rule changes were handled only by updating these master data, without modifying CAD models or program code, and such updates were performed by engineers themselves. These features confirm that the proposed configurator provides not only a reduction in design effort but also high maintainability and operational sustainability, which are essential for ETO product development environments. These results demonstrate that the proposed approach is effective for supporting systematic and sustainable early-stage layout design in ETO products.

This paper investigates the feasibility of in-process flank wear monitoring using motor torque command values obtained from the internal sensors of an NC lathe without external sensors. Tool wear degrades production stability and is one of the primary factors determining tool life. However, the increasing use of multitasking machines has made tool wear behavior more complex, and the declining workforce has increased the need for automated tool wear management independent of skilled workers’ experience and knowledge. Therefore, an automated tool wear monitoring method is strongly required. In this study, cutting forces were estimated from motor torque command values acquired through the internal sensors of an NC lathe. Furthermore, the relationship between the estimated cutting force components and flank wear width was evaluated to investigate the feasibility of tool wear monitoring. The results showed that the thrust force is strongly affected by the resistance generated during the reversal of the feed mechanism. Therefore, when the X-axis is positioned in the same direction as the thrust force during measurement, the estimated thrust force shows good agreement with that measured using a cutting dynamometer. In contrast, the cutting force and feed force could be estimated by subtracting the average torque command value immediately before cutting from that during cutting. Furthermore, a strong correlation was observed between the estimated cutting force and flank wear width.

Computer-aided wear prediction is necessary for exploring the design solutions of artificial joints and reducing the burden of long-term wear testing on them. However, since a contact model between the ultra-high molecular weight polyethylene (UHMWPE) and metal (Co-Cr-Mo alloy, 316L stainless steel) components has not been established yet, calculating the contact pressure between them remains a major obstacle in wear prediction. The purpose of this study was to calculate the contact pressure between the UHMWPE and metal components using a computational method, in which the deformation of the contact surfaces was divided into indentation and sinking under the assumption of elastic contact. This method was developed based on the displacement equation for a normal force acting on the contact surface in the elastic half-space theory. The experimental conditions were determined by considering the viscoelasticity and elastic–plastic characteristics of UHMWPE identified through instrumented indentation testing. The calculated results were compared with the measured values obtained using pressure-sensitive films and a dye solution and both values were in good agreement for each contact condition. Under the ball-on-disc contact condition, no appreciable difference was observed between the results with and without sinking deformation. Furthermore, under the ball-on-cup contact condition, the calculated values were close to the measured ones, even without including the sinking deformation. In contrast, calculations that included sinking deformation were necessary to determine the contact pressure distribution under the pin-on-disc contact condition. The indentation modulus of the UHMWPE was nearly constant for plastic strains below 0.17 %.

Head checks are a persistent maintenance problem on curved tracks of urban railways, including Japanese Linear Metro lines. Current Japanese guidelines limit rail grinding to within 5 mm below the top surface of the rail, often leaving deeper cracks untreated and necessitating costly rail replacement. Expanding the grinding allowance could significantly reduce such replacements. However, to implement this change, the impact of ground rail profiles on running safety must be evaluated. This study investigates the relationship between expanded grinding profiles and running safety for the Linear Metro system. Three profiles were assessed: two single-angle profiles ground to depths of 10 mm and 14 mm, and one double-angle profile at 10 mm. The 10 mm depth was determined considering the flange angle of the Linear Metro wheel profile. Wheel-rail static contact analysis evaluated wheel-climb margin using the critical contact-angle-reduction wheel raise limit. Dynamic behavior was simulated using a validated multi-body dynamics model of a Linear Metro vehicle, incorporating initial wheel load imbalance, a 100 m radius curve, and track irregularity. The evaluated 10 mm grinding profiles showed no significant deterioration in static or dynamic safety indices; derailment coefficients and maximum wheel raise remained comparable to those of the unmodified rail. In contrast, the 14 mm profile reduced static climb margin and yielded a negative residual allowance, suggesting increased risk. These findings indicate that, under the vehicle and track conditions examined in this study, extending the allowable gauge-corner grinding depth to 10 mm is unlikely to compromise running safety.

This study proposes a multi-stage computer aided engineering (CAE) surrogate modeling framework for tactile design support of rubber dome keyboard switches. Because rubber dome deformation involves snap-through buckling, material nonlinearity, and contact nonlinearity, direct nonlinear CAE is costly for design exploration. To improve design efficiency, surrogate models were constructed from nonlinear CAE results obtained for 5000 random combinations of 10 design variables using an axisymmetric finite element model. First, design feasibility was classified from the force-stroke response. In the subsequent stages, only the true feasible designs (Status = 2) identified from the CAE results were used for regression modeling. The click ratio was predicted by regression. The force-stroke curve was also predicted as a functional output by applying principal component analysis (PCA) to normalized curve data and regressing the principal component coefficients. In addition, displacement and stress fields at the force maximum and minimum points were reduced by proper orthogonal decomposition (POD), and the corresponding modal coefficients were regressed for field reconstruction. The results showed that binary classification was effective for feasibility judgment and that deep neural networks achieved high accuracy in both feasibility classification and click-ratio prediction. The PCA-based curve model and the POD-based field model also provided practically useful accuracy. These results demonstrate that the proposed framework is effective for efficient tactile design support of keyboard switches.

Linkage structures are primitive mechanical systems that are useful in industrial applications. However, their sparse characteristics make it difficult to apply topological optimization techniques, except in the field of statics. We proposed a continuous feature field for the linkage structure via a generative neural network, which allows for topology optimization in the dynamic kinematics field. In order to examine this, a conditional generative adversarial network (cGAN) was employed, and it was connected with a differentiable multi-body simulator. It was demonstrated that the rigid-body multi-pendulum systems transitioned in a stepwise manner, in accordance with the gradient of the objective function. Consequently, the results showed that the number of links, their connecting topology, and the scale of the linkage bar could be updated along the optimization objective. This approach expands the benefits of differentiable simulators, such as policy-gradient-based behavior optimization and parameter tuning, to include topology optimization.

The axis of value provision is shifting from products to services, and further, to the orchestration of experiences. Such a movement is closely tied to broader transformations in society and daily life. At its core lies a narrative that traces a chain of interactions between humans and artifacts, and from which innovative product and service concepts can emerge. Issues surrounding such narratives have long been the subject of conceptual design, but the scope must now expand to what might be called ideational design, a practice that includes defining values themselves. However, conventional methods or frameworks for conceptual design focus on products. Their scope is gradually expanding to services, but experiences are out of focus in the ordinary sense. This study proposes an ideational design framework to comprehensively describe such narratives through a diagrammatic method called the Story Diagram (SD). The SD serves fundamentally as a mechanism that begins with a value graph, leads to the functional structure of individual products and services, explores various candidates for functional components, and morphologically integrates them into promising alternatives of the entire shape of transformation. This paper demonstrates the practical use of the SD as a tool for divergent thinking in designing the integration of products, services, and experiences, through a retrospective case study on the history of music listening devices. The study validates the applicability of the SD to the ideational design of transformation.

To prevent traffic accidents caused by human error, it is important not only to support immediate responses to hazards but also to maintain and restore drivers’ attention allocation to safe driving. However, the effects of state-dependent visual warnings presented during reduced attention to driving have not been sufficiently examined, particularly in terms of warning color and flashing frequency. This study investigated the effects of visual warning color and flashing frequency on attention guidance during driving under cognitive workload, using eye-movement-based anomaly scores and subjective evaluations. Twelve students holding Japanese driver's licenses (6 males and 6 females) performed a driving task using a driving simulator. To induce a state in which attention allocation to driving was reduced, a 2-back task was imposed during driving. Seven warning conditions, consisting of three colors (blue, yellow, and red), two frequencies (1 Hz and 5 Hz), and a no-warning condition, were tested. Visual warnings were presented only when the anomaly score exceeded a predetermined threshold, indicating reduced attention to driving. Subjective evaluations revealed that flashing frequency significantly affected perceived driving interference, with the 5 Hz condition showing higher interference than the 1 Hz condition. In contrast, no significant effects of color or flashing frequency were observed for perceived urgency. Temporal analysis of anomaly scores indicated that the blue 1 Hz condition showed relatively stable decreases in anomaly scores after warning presentation, whereas the 5 Hz conditions did not consistently improve the driving state. These results suggest that warnings perceived as more interfering are not necessarily more effective for guiding attention back to driving. Overall, visual warnings for attention guidance should attract attention while avoiding excessive interference. Because luminance and background contrast were not strictly controlled across color conditions, the findings should be interpreted as effects of the LED stimulus conditions rather than pure color effects.

Serious games are designed to address social challenges, such as education, by combining learning and entertainment to enhance motivation. However, their effectiveness varies among individuals, and optimal design strategies remain underexplored. This study investigated whether tailoring game designs to players’ personality traits could influence six dimensions of user experience (UX): effectiveness, efficiency, satisfaction, immersion, motivation, and emotion. Participants’ personalities were assessed using the Big Five traits and a player trait model comprising aesthetic, narrative, goal, social, and challenge orientations. A user study was conducted using five English vocabulary quiz-based game designs: (1) a basic quiz, (2) a quiz with scoring elements, (3) a quiz with adventure elements, (4) a quiz with action elements, and (5) a multiplayer quiz with action elements. Each participant was assigned to one design and played under controlled online or offline conditions, followed by tests and questionnaires evaluating UX. The results suggested that certain personality–game design combinations were associated with more favorable UX outcomes in quiz-based serious games. Participants high in openness or aesthetic orientation showed improved UX in games with adventure elements. Participants with a strong social orientation experienced UX improvements in games with multiplayer elements. In contrast, action elements appeared to be more influenced by usability and player proficiency than by personality traits. These findings suggest that both player personality and game mechanics should be considered in the design of quiz-based serious games. This study provides preliminary design implications for personalized quiz-based serious games and calls for validation across broader game content and player populations.