The objective of this study is to achieve the target soil weight in excavation of incline terrain. We propose an autonomous excavation method using a dynamical model of excavation that can estimate the soil weight from the excavation work. In this method, the penetrating length is adjusted to control the weight of the soil, and the penetrating depth is also adjusted to prevent bucket sticking due to overload. The trajectory modification is performed by estimating the height of the soil surface and by estimating the slope angle based on the model. The effectiveness of the proposed method is verified through experiments using an excavation robot.

One of the problems with teleoperated hydraulic excavators is communication latency, which can worsen the operator’s sense of agency and narrow the effective field of vision. This study investigated the effect of communication delay on the operator’s sense of agency and effective field of vision in teleoperated hydraulic excavators. The results of a visual field task showed that there was a positive correlation between the sense of agency and the accuracy rate at 20 ° left and right. This suggests that achieving a high sense of agency is effective in preventing narrowing of the effective visual field.Furthermore, a significant tendency was confirmed in the difference in the rate of correct answers to the tasks between the latency conditions. The above suggests that increased latency and the resulting decrease in a Sense of Agency may narrow the effective field of view.

In recent years, the shortage of human resources at earthwork sites has been accelerating due to the declining birthrate and aging population. Additionally, industrial accidents remain a serious problem. To overcome these problems, we are developing a Cyber-Physical System platform named ”ROS2-TMS for Construction” as a system that simultaneously improves the efficiency and safety at earthworks. As part of this development, we recently conducted an autonomous construction experiment focusing on the process in which soil discharged into a pit during shield-tunnel drilling is excavated by a backhoe, loaded onto a dump truck, and then transported and disposed of outside the construction site. This paper provides an overview of the experiment, which was carried out using ROS2-TMS for Construction.

This study proposes a system to derive in real time an excavation trajectory suitable for actual soil properties. The proposed system consists of a database construction process, a soil identification process, and an optimal excavation trajectory application process. In the database construction process, a database of optimal excavation trajectories for 1000 soil types is constructed in advance. In the soil identification process, the soil type at the site is estimated based on excavation data from the actual excavation site. In the process of applying the optimum excavation trajectory, the optimum excavation trajectory for the estimated soil type is obtained from the optimum excavation trajectory database and applied to the actual machine. Finally, the effectiveness of the proposed method is demonstrated by comparing the actual excavation by hand using construction equipment and the excavation using the excavation trajectory derived by the proposed method on the simulator.

Illuminance measurement is essential in building facility construction to ensure that illuminance conditions comply with JIS regulations before completion. In conventional methods, illuminance measurement is performed manually, requiring long hours of work at night. To address this issue, we have been developing a swarm robot system for autonomous illuminance measurement. This paper discusses improvements in communication stability for the swarm robot system and presents evaluation experiments conducted with six robots. Additionally, we report on the implementation of a step detection and avoidance function using ToF sensors attached to the robots. Experimental results confirm that the six robots can perform illuminance measurement stably and that the step detection and avoidance function operates effectively．

Hydraulic excavators play a crucial role in construction operations, particularly in hazardous or complex environments. However, their operation presents significant risks, and the shortage of skilled operators further exacerbates logistical and economic challenges. Remote operation and unmanned construction technologies have been explored as solutions, yet efficiency remains a major concern, as remote operation efficiency is reported to be below 50% of direct operation. This paper proposes a semi-autonomous excavation system that integrates remote control with autonomous trajectory generation based on nonlinear dynamical systems. By introducing trajectory attractors, the system adapts to ex-cavation environments and autonomously generates excavation paths, reducing the operator’s workload. Additionally, the system allows flexible human intervention. Experimental validation demonstrates the system’s effectiveness in enabling continuous excavation while permitting intuitive operator adjustments, contributing to practical advancements in remote excavation technology.

Concrete 3D printing (3DCP) is gaining attention as a technology for rapidly constructing buildings and other structures. However, conventional 3DCP systems have large-scale layering mechanisms, making deployment and transportation difficult in confined spaces. To address this issue, we propose a novel 3DCP technique in which a robot moves over the deposited layers, spreading and layering mortar pellets transported by drones or other means. Currently, no existing robot is capable of both moving over the deposited layers and simultaneously spreading and layering the transported mortar. To overcome this limitation, we have developed a screw-type spreading and layering robot and successfully demonstrated its ability to perform these tasks. These results indicate the potential of a new 3DCP system. This paper presents a new layering method in which transported mortar is processed by a screw-type spreading robot, offering the possibility of constructing large-scale structures even in confined spaces.

Excavation with hydraulic shovels presents challenges such as operator danger in hazardous environments and decreased efficiency due to fatigue. While remote control and autonomous systems have been proposed, human intervention remains necessary in emergencies. A semi-autonomous system using trajectory attractors was introduced to integrate human input without conflicts, but previous methods restricted excavation depth. To overcome this, we extend the index value to two dimensions in a toroidal space, allowing independent control of both excavation position and depth. This enables smooth transitions between excavation locations while adjusting depth dynamically. By implementing a cylindrical attractor with a multi-helical structure, the system ensures stable operation and seamless human intervention. The proposed method is applied to a leader-follower excavation system to verify its effectiveness, demonstrating improved adaptability and efficiency in excavation tasks.

Hydraulic systems are highly power-dense and widely used in construction machinery such as excavators. However, hydraulic actuators, including cylinder-piston systems, exhibit load-dependent piston velocity responses, which degrade controllability for human operators. To address this issue, some hydraulic systems in excavators incorporate pressure compensation mechanisms, but they may induce vibrations in the piston velocity transient response. Therefore, theoretical analysis-based methods for suppressing hunting are required. In this paper, we propose a mathematical model that describes the behavior of a hydraulic system in construction machinery equipped with a load-sensitive inter-cylinder pressure compensation mechanism. Additionally, we develop simulation models of the system and analyze its behavior based on the results. Furthermore, we propose design directions for hydraulic system components to suppress hunting using the simulation results.Through this method, we showed these mathematical models are effective in the dynamic analysis of hydraulic systems with such hydraulic circuits.

To scoop up a massive rock efficiently with a wheel loader, we have proposed a method in which a massive rock is pushed in with the tip of the bucket. In this scooping method, the vertical component of the bucket reaction force is upward, even though the bucket is in contact with the lower part of the massive rock. This is believed to be due to the moment of rotation of the column caused by the passive earth pressure generated when the column is pushed into the soil. The relationship between the bucket reaction force and the passive earth pressure during the scooping operation will be investigated to construct a reaction force model and to elucidate the mechanical phenomena acting on a massive rock.

Bridges often have beams, making it difficult for conventional bridge inspection mobile robots with wide wheel spacing to traverse. In this study, we developed a flexible wheel diameter mechanism that expands the wheel diameter to reduce the wheel spacing when encountering uneven surfaces and contracts the wheel diameter during wall climbing to improve energy efficiency and stability. After determining the necessary wheel size for overcoming beams, we designed a wheel mechanism incorporating a radially extendable pantograph mechanism and a Hoberman mechanism. It also enables expansion under conditions with propeller-generated thrust. We conducted experiments on wheel expansion under expected thrust conditions and measured power consumption differences based on wheel diameter variations. As a result, the wheel expansion could be traversal of uneven surfaces, while contraction improved energy efficiency during movement.

Cut and fill earthworks are the preliminary work of the construction projects in remote areas. Haulage costs typically accounts for around 30% of the total projects costs. The temporary haul road constructed on-site is one major factor influencing haulage costs. Planning efficient earth allocation, temporary roads, and dump trucks’ routes remains an empirical design problem at present. This study aims to optimize not only the placement of temporary roads but also the construction timing of temporary roads, considering the possibility that temporary roads may disappear due to changes in the terrain. The optimization approaches are applied in a real-world project scenario and proven to reduce the cost of the projects.

In this study, we develop Pile-type Sensor Pods for monitoring conditions at civil engineering sites. The Pile-type Sensor Pod consists of a pile, two multi-core microcomputers, two 220° cameras, a GNSS receiver, and a vibration sensor. In this report, we have redesigned the housing for waterproofing. In addition, the accuracy of the data acquired by the vibration sensor was verified.

In this paper, we explore the development of wall-traversing robots equipped with rebar detection devices for concrete structures and the challenges associated with this process. The aging of concrete infrastructure is a significant concern, increasing the need for quality inspections. Often, inspection sites include hazardous locations such as high altitudes, necessitating the implementation of extensive scaffolding and safety measures. This underscores the importance of deploying wall-traversing robots. However, these robots tend to be large and heavy, posing a considerable risk in the event of a fall. Our research has been directed towards simplifying the structure of these robots, thereby facilitating the creation of a more compact prototype. This advancement has enabled the development of lighter wall-traversing robots compared to previous models.

This study proposes a method for estimating the underground soil state during excavation using reaction forces measured by an excavator’s internal sensors. A dataset is constructed using Discrete Element Method (DEM) simulations, where soil distributions and reaction forces are recorded during excavation. A diffusion model is trained on this dataset to generate local cross-sectional soil distribution images around the excavator bucket’s tip from reaction force and end-effector velocity data. To estimate the overall soil distribution, local images generated at different time steps are sequentially integrated using a log-odds-based occupancy grid framework. This approach allows the estimation of soil displacement during excavation without relying on external sensors.

This research incorporates the idea of embodiment into a remote control of a construction robot and develops a method in which motions of a visual camera and manipulator are linked to ones of the operator's head and arm. This paper evaluates the effects of the motion scale between the camera and manipulator movements on operability. The evaluation results suggest that operability was improved when the motion scale of the manipulator was equal to or slightly higher than that of the camera. This is thought to be due to the fact that the head motion range and arm motion range interfere with each other when the manipulator motion scale is low, resulting in a decrease in operability.

We modeled the cylinder system of a hydraulic excavator from the lever input (spool opening command) to the piston speed. After creating a simulator from the obtained differential equations, the transfer function of the model was obtained through system identification. Using the transfer function, we designed a two-degree-of-freedom control system that combines feedback control and feed-forward control. In addition, a compensator was designed to multiplicatively guarantee the gain reduction due to external force. Experiments were conducted on a simulator and an actual machine to demonstrate the usefulness of the proposed method, and both systems were able to track the target speed.

During disaster scenarios, rescue operations require rapid data collection in environments too dangerous for human entry. Here, we present SNAKE, a slim, horizontally deployable multi-joint robot with minimal actuators and passively deformable joints, designed to resist structural buckling under compression while remaining flexible upon contact with obstacles. Experiments using a vertical plane as a model obstacle highlight the necessity of joint stiffness to maintain smooth propulsion and avoid local bending or locking. Furthermore, we integrated SNAKE with a legged robot platform to extend mobility, verifying its feasibility in narrow-space exploration. Future work will further focus on enabling multi-axis passive deformation and evaluating performance in simulated environments, thereby increasing safe and efficient search-and-rescue operations.

In recent years, the construction industry has been facing various challenges, such as a declining workforce and an aging population. These issues have been further exacerbated by increasing demand for construction due to aging infrastructure. To address these challenges, we are developing ROS2-TMS for Construction, a cyber-physical system (CPS) for earthwork sites. As part of this research, we are developing OperaSimVR, a visualization system for cyberspace within ROS2-TMS for Construction. In addition to the existing playback mode, we have newly developed a control mode and a preview mode. In control mode, users can remotely operate construction machinery via OperaSimVR, while in preview mode, the simulator predicts the state of the actual machine or earthwork site and visually presents it to users. This paper describes each function in detail and presents the verification experiments conducted to evaluate the control mode and preview mode using actual OPERA-compatible construction machinery.

Robots will not be installed unless they are cost-effective. To improve their effectiveness, robots are needed to cooperate with building equipment such as elevators, automatic doors, and security. To solve this problem, Robot Friendly Asset Promotion Association (RFA) of Japan have published the standards on communication interfaces between robots and building equipment. This paper reports the project achievement led by Octa Robotics, Inc. including the implementation of RFA standards, the new issues to standardize (including an A-frame sign application of RFA Shared Markers and the shared notification system) and their conformance test programs.

This study reports on the development of teaching materials for autonoums driving education using a small EV, utilizing the technology and knowledge gained from the development of an autonomous mobile robot. The development of elemental technologies for use as teaching materials for automated driving education was conducted in cooperation with regional companies. As a first step in autonomous driving, an environment recognition function using 3DLiDAR was implemented and evaluated.

With the advancement of motion capture technology for measuring body movements, its application in teaching vertical articulated robots is expected. Recently, models have been proposed to estimate 3D human posture from 2D keypoints of the skeletal structure captured by a monocular camera, drawing attention as a motion capture technology for low-cost and easily accessible robot teaching systems. This study aims to develop a 3D motion capture system using a monocular camera, assuming the utilization of such 3D human posture estimation models. In this paper, we report on a 2D keypoint observation system focusing on hand movements during penmanship.

Robotics education plays a crucial role in fostering problem-solving and independent thinking skills. As technological advancements continue to reshape society, the development of science and technology innovation professionals has become a key priority. Since its inception, the Department of Electronic Control Engineering at National Institite of Technology, Numazu College has implemented a project-based learning (PBL) curriculum focusing on the development of small autonomous mobile robots, known as MIRS (Micro Intelligent Robot System). This program emphasizes not only hands-on manufacturing but also system development processes and real-world implementation experiments. To address challenges in system design, the fifth generation of MIRS education introduced agile methodologies, Proof of Concept (PoC), and Minimum Viable Product (MVP) approaches. The curriculum was restructured to include standard model development, technology exploration, system proposal and design, field implementation, and video documentation. This paper reports on the first-year implementation of the fifth-generation MIRS program, comparing it with previous iterations and discussing future prospects.

In this paper, we have developed an Autonomous Mobile Robot (AMR) system to support independent and comfortable living for elderly individuals requiring mild to moderate care (Support Levels 1 and 2, and Care Level 1). The AMR integrates LiDAR-based environmental recognition technology and a ROS2-based control system, enabling autonomous navigation and obstacle avoidance. Additionally, it features voice command recognition and manual control via a game controller for more intuitive operation. A battery management system and a graphical user interface (GUI) have also been developed to enhance operational efficiency. Experimental validation was conducted in the ”Sanpo no Sumai” model room at the Pharma Valley Center, confirming the effectiveness of voice-commanded task execution. However, unstable behavior was observed in narrow or complex pathways, indicating the need for further refinement. At the CareTEX Tokyo exhibition, industry professionals provided valuable feed-back, highlighting key areas for improvement, including enhanced environmental recognition, improved voice command accuracy, simplified operation, and extended battery life. Future research will focus on optimizing power management, ensuring a safe and user-friendly design, and improving long-term operational stability. These enhancements are expected to make the AMR a more practical and reliable system, contributing to elderly independence and reducing caregiver workload.

Automated journalism is emerging as an efficient means of delivering news in our increasingly AI-driven world, yet current systems remain limited by their reliance on single-modal inputs and the need for extensive human oversight. This often results in articles that lack the rich contextual detail and accuracy expected in today’s dynamic media landscape. To address this shortcoming, our research proposes a novel approach that harnesses both visual and auditory data from raw video inputs to generate comprehensive and engaging news narratives. By integrating multiple data modalities in a unified framework, our method aims to capture a fuller picture of events, thereby enhancing factual accuracy and overall quality while reducing the necessity for manual intervention. This work represents a significant step toward scalable, robust, and fully automated news production systems that meet the evolving demands of modern journalism.

In recent years, smart agriculture has been gaining attention, and active research is being conducted on automation through robotics. Generally, farms involve vast areas for cultivation regardless of the type of crops, so it is more efficient for multiple harvest robots and container transport robots to collaborate on tasks rather than a single robot performing both harvesting and transportation tasks. While research on agricultural harvest robots is reported, there are few reports on the automatic transport of containers after harvesting or the collaborative operations between harvest robots and container transport robots. This research focuses on the development of harvest robots and container transport robots, and examines their collaborative operations and communication methods between the robots, which is reported here.

In this report, we propose a system integration (SI) for a robot system that can be used by users without knowledge of industrial robots by using LLM/VLM as a teaching method for industrial robots. As a specific task, we assume the pick-and-place task of objects in a variety of environments. The performance of the proposed robot system is clarified through a demonstration of the pick-and-place task of fine wires less than 1 mm in size.

In this study, we attempt to design a control system using an adaptive optics simulator for the Tweeter system, which is responsible for measuring and compensating for the high-frequency vibrational components of wavefront disturbances. This system is a part of the adaptive optics system planned to be installed in SEICA (Second-generation Exoplanet Imager with Coronagraphic Adaptive Optics), an observation instrument intended for the direct observation of Jupiter-like exoplanets with the Seimei Telescope.

To establish a robot-friendly environment, shared markers were specified as robot-oriented signs in RFA B 0005. However, the standard color scheme of the shared markers stands out against the design and furnishings of the facility. This study examined a color scheme for the shared markers specified that harmony with the environment while maintaining detectability. The foreground and background colors of the markers, wallpaper color, illuminance, and camera-marker positioning were varied, and the number of detected frames within a fixed time was recorded. Results showed that the standard color scheme ensured high detectability regardless of illuminance. When the wallpaper was brown, using the same color for the marker’s foreground maintained high detectability while achieving visual harmony.

Activities related to the social implementation of robotics technology inherently require a process of accumulating complex and integrated technologies. This process, which can be a barrier, can be appropriately promoted by an integrated prototyping system consisting of a platform for robot development. With this in mind, we are developing integrated prototyping systems with a robot arm mounted on a mobile robot platform. As a specific subject, mobile manipulation, which is expected to spread rapidly in factories, was taken up. This is a joint project of National Institute of Technology (KOSEN) colleges within the framework of social implementation. In this paper, we describe an example of system development of a 4-DOF single-arm or 14-DOF cooperative dual-arm robot mounted on a differential two-wheeled or mechanum wheeled mobile mechanism, respectively, for mobile manipulation prototype.

This study proposes a structured control education approach utilizing MATLAB/Simulink and introduces an experimental method incorporating agile prototyping. At National Institute of Technology, Kitakyushu College, a curriculum was developed to systematically teach control engineering from the second to the fifth year, integrating simulations and practical control experiments. Additionally, a low-cost rotary inverted pendulum was designed and implemented to enhance hands-on learning. Using Simulink, students conducted parameter identification, PID control, and trajectory tracking, experiencing an iterative process of design, evaluation, and refinement. This approach allowed students to develop flexible design thinking and gain practical control engineering skills. The results demonstrate that incorporating agile prototyping in control education effectively bridges the gap between theoretical learning and real-world applications, fostering adaptability and problem-solving abilities essential for engineering practice.

For the automated disassembly of commonly found mating parts in large home appliances by robots, autonomous force control remains challenging, necessitating a system for intuitive human demonstration. This study proposes a system that presents affordances in a virtual environment to facilitate efficient human demonstrations. By visualizing disassembly directions based on mating part constraints and suitable grasping poses, the system simplifies the process. Additionally, impedance control is introduced to prevent excessive disassembly trajectories during affordance-guided demonstrations. Experiments validate the effectiveness of the proposed affordance-based disassembly method by comparing it with a conventional approach.

Against the backdrop of rapidly increasing demand for resources in line with the growth of the global economy and the worsening of environmental problems such as global warming, a shift to a recycling-oriented economy is urgently needed. In the global effort to build a circular economy, Japan is pioneering urban mines that reuse industrial waste as raw materials. However, domestic scrap metal recycling is currently hampered by manual sorting, which has led to a decline in recycling, especially of less profitable items such as home appliances. Therefore, we are developing an autonomous sorting system for wastes small home appliances in order to effectively utilize wastes small home appliances, which are considered urban mines, and to realize a recycling economy. In this study, we report on an automated sorting system for products with lithium-ion batteries, which are the main cause of fires.

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.

Amount of e-waste is rising globally, and contains valuable metals like copper. However, disassembly is still mostly done manually because of the wide variety and small volumes of discarded electronics. This research proposes an automated cable cutting method that uses instance segmentation and key-point detection. By identifying individual cables and detecting cut positions near their end points in a sequential process, the method reduces missed detections and improves recycling efficiency. The evaluation results of cut position detection for outdoor unit cables demonstrate higher accuracy compared to parallel approaches and ensure reliable endpoint identification. Future work includes recognition of various types of cables and connector detection to foster sustainable resource recovery.

To achieve the SDGs, technologies for automating the disassembly of high-mix,low-volume home appliances and for improving material separation have been developed. Their disassembly automations rely on repetitive programmed motions for specific products, but it is difficult to adapt to variations in products for disassembly. While many disassembly processes can use CAD data, the process of cutting wires cannot. In this process, cutting motions must be generated in accordance with wire situations of the products. Furthermore, to improve material separation, it is desirable to cut a wire near its terminal point. However, interference from components near the cutting points makes the process even more difficult. In this study, we propose a motion planning algorithm for the process of cutting wires using a manipulator. This method integrates motion planning for interference avoidance and cutting sequence determination. We conducted validation experiments to evaluate its effectiveness.

This paper investigates the applicability of quantum computing technology in robotics from a functional safety perspective. Quantum computing has the potential to significantly enhance robot systems in tasks such as path optimization, machine learning, and quantum sensing. However, critical challenges exist regarding the inherently probabilistic nature of quantum algorithms, physical errors such as decoherence and crosstalk among qubits, and difficulties in verifying the validity of quantum-generated results. Considering these issues, three approaches for applying quantum computing are proposed: use in non-safety-related systems, cautious integration into safety-related systems, and application for human decision support. It concludes that direct application to safety-related systems necessitates classical verification and careful hybrid design

The use of drones for plant inspections has been increasing due to their efficiency and ability to access hazardous areas. However, ensuring consistent service quality and operational safety remains a challenge. This study develops an operational guide map for drone-based plant inspections, based on JIS Y 1011:2024, which defines process requirements for drone service providers. The guide map organizes key considerations and response measures for each requirement, referencing existing manuals and best practices. Additionally, it includes supplementary explanations on safety measures and plant-specific inspection procedures. By following this guide map, drone service providers can enhance service quality, improve safety management, and comply with industry standards. This research contributes to the establishment of standardized drone operations for plant inspections, supporting safer and more efficient industrial applications.

The most common problem in heavy snowfall areas is human casualties caused by snow damage, the most common of which are accidents during snow removal work such as removing snow from roofs. This study aims to develop an extendable and retractable convex tape telescopic mechanism for snow removal to prevent accidents during the removal. We have improved the mechanism to prevent unwinding and succeeded in extending and retracting the convex tape over its entire 17m length. For practical application, In order to increase the strength of the telescopic mechanism, the tape will be folded like quasi-pipe form as half, about 8.5m length. It is thought that it can be used for clearing thin piles of snow without any problem in terms of power.

This research focuses on developing an omnidirectional mobile robot with independent steering and drive mechanisms for each wheel. Unlike our previous designs that relied on complex power transmission systems, this robot simplifies its structure by mounting the drive motor directly below the steering motor. To effectively realize this structure, an in-wheel motor, in which the tire and motor are integrated, is used to make the mechanism more compact and improve track following performance. Wireless communication is used to increase the flexibility of mechanism placement and the efficiency of communication connections. Although there is still room for improvement in tuning the control system for the in-wheel motor and steering motor, it was confirmed that overall basic operation is possible.

We present the development of a cord-like robot with redundant degrees of freedom, utilizing differential modules, designed to operate not only in hard-to-access Satoyama regions but also in diverse environments such as caves and animal burrows. This type of robot offers excellent adaptability and mobility in various terrains. To fully leverage these advantages, it must achieve high torque and robustness to withstand physical damage in harsh conditions. The robot’s main functions include "three-dimensional movement, monitoring, and measurement." Since direct visual observation of the robot is not always possible, it is controlled via a remote interface.

Vulcanized fiber, a cellulose-based material, has attracted attention as an environmentally friendly alternative to plastics. This study focuses on optimizing the cutting conditions for side and hole machining of vulcanized fiber using Random Forest regression and the TPE algorithm. The goal was to minimize machining time while maintaining target machining accuracy. The results successfully identified optimal conditions that balanced precision and efficiency. Additionally, a biodegradability comparison was conducted, highlighting the material's environmental advantages. A prototype of a robot part was also fabricated using vulcanized fiber, demonstrating its potential for industrial applications. These findings suggest that vulcanized fiber is a viable material for sustainable manufacturing, offering both machinability and biodegradability.

When considering the design of underactuated systems, the concept can be extended beyond internal body coordination to include Trinity Design—a framework that accounts for the physical interactions among the body, the object, and the environment. This idea can further evolve into Triadic Interaction Design, which emphasizes mutual influences among these three domains, or even into a Unified Agent-Environment Design, where the object and environment are collectively treated as the external counterpart to the robot. In other words, the focus shifts from merely optimizing the number, position, and orientation of components within the robot body to also designing systems of element exchange and transformation across the following three domains: intra-body (within the robot), body-object, and body-environment interactions. In this paper, we introduce this novel design philosophy and illustrate its utility through several concrete examples.

The development of sustainable mobility technologies aimed at addressing global warming and reducing dependence on fossil fuels is highly anticipated. Systems have been proposed that use hydrogen supplied from commercially available portable hydrogen storage alloy containers to power electric scooters and electric bicycles through a fuel cell stack. In this report we study the feasibility of powering small four-wheeled mobility vehicles with such portable hydrogen storage alloy containers and hydrogen fuel cell stack.

We hypothesized that by intentionally making the mechanical structure more complex, it would be possible to increase the capabilities of each module, thereby improving the robot's efficiency and energy-saving potential. We designed a mechanism that uses eight motors installed on the same plane as the power source. By connecting each motor to a helical gear and using a bidirectional helical gear with teeth cut in both directions as the output, we were able to obtain both rotational and linear motion from a single module. Additionally, by using two bevel gears as inputs, we created a mechanism that provides yaw and pitch rotational movements at the output bevel gear. By combining these two mechanisms, we constructed an 8-axis manipulator capable of generating changes in posture or rotating the output shaft. The experiment will verify whether posture changes can be achieved and whether multiple forces can be concentrated on the output axis.

Push-off plays an important role for stable and smooth gaits. Its shortage results in relatively slow walking and a high risk of falling. To overcome this problem and enhance the push-off moment in terminal stance of gait, we proposed an unpowered push-off support orthosis. This device is composed of a spiral spring and its cover, a pendulum, shoes, a frame and a cuff. The spiral spring is compressed during stance phase and provide support moment at push-off. On the other hand, it doesn’t provide any support moment in the swing phase. The results of gait experiment show imperfect lock/unlock motions. To confirm its mechanism, dynamics of the pendulum were simulated.

Most commercially available prosthetic legs are equipped only with plantar flexion and dorsiflexion functions. As a result, many prosthetic users experience difficulties when walking or standing on uneven terrain. Therefore, this study aims to develop a prosthetic shoe sole with inversion and eversion functions for prosthetic users. By combining this sole with an existing prosthetic foot that features plantar flexion and dorsiflexion capabilities, the prosthesis users can stably stand on uneven terrain at a low cost. This article proposes its basic structure and design method for the fluid system whose internal fluid is controlled by two tubes and a non-return valve. Based on the control objectives, the inner diameter of Tube 2 was determined to be 1.0 mm, with a length of 1.2 m. The developed prosthetic shoe sole showed no water leakage in both neutral and pseudo-eversion situations.

The number of elderly individuals living alone has been increasing annually. This trend raises concerns regarding the growing population of frail older adults and the prevalence of dementia. To address these issues, we propose the development of the Selfish Pretty Robot (SPRO), a socially assistive robot designed to encourage elderly individuals living alone to go outside. Unlike conventional robotic companions, SPRO cannot be recharged within the user’s residence, thereby promoting outdoor activities. SPRO functions as a substitute for companion animals, supporting users in maintaining an active lifestyle and extending their healthy lifespan. SPRO operates in conjunction with ECSPOT, a dedicated charging station placed outside the user’s residence. ECSPOT serves multiple roles, including monitoring the user and facilitating interactions. By requiring users to leave their homes to recharge SPRO, the system naturally encourages mobility and social engagement. This paper presents the basic structure and function of the first prototype.

In this study, the evaluation of assistive devices using a bipedal robot was considered effective for the development of assistive devices, focusing on the reproducibility and the possibility of objective evaluation. Therefore, we investigated the evaluation of a walker using a model of a small bipedal robot, which can be used at a low cost; a model of a bipedal robot and a model of a walker were constructed by simulation using MATLAB, and a walking simulation was conducted. The reaction forces on the legs and the torques on the knees were measured to check whether the model of the small bipedal robot could be used for the evaluation of the walker.

In recent years, force-display technology has been increasingly anticipated to be widely used. Force-display technology enables the reproduction of reaction forces when interacting with virtual objects. Force-display devices can be classified into two categories based on the method of force generation: active and passive type. Active type devices use actuators such as motors and pneumatic artificial muscles, and are capable of active force presentation. While there are safety issues with motor-based actuators,the balloon actuator used in this study is safe. On the other side, passive devices cannot present the force sensation acting during stoppage. The aim of this study is to develop a force-display glove as a new rehabilitation method for patients with hemiplegia of the hand. This device, which is expected to synergise the advantages of active and passive types, is called a hybrid force-display glove. This paper describes the results of an improved internal pressure control model for a balloon actuator and a force control evaluation.