Graphical Abstract Development of clothing in consideration of the shape and body function of a person with spinal cord injury is an important task. Then, a dummy robot with a deformation mechanism was developed in this study for evaluating the comfortable level of clothings. Specifically, a trunk joint mechanism and an abdominal mechanism that can realize various deformations of the abdominal area and various trunk poses were developed. The trunk joint mechanism was implemented in order to simulate the seated posture of persons with spinal cord injury. The abdominal deformation mechanism was implemented using linear actuators and rotating servomotors in order to simulate abdominal obesity of persons with spinal cord injury. Further, a tactile sensor system was developed for measuring the clothing pressure on the abdominal area and evaluating the comfort or discomfort of clothing.

Computer vision plays an important role in various applications. However, the accuracy of computer vision is often degraded by specular reflection. To solve this problem, a system utilizing a high-speed camera and a strobe for removing specular reflection has been shown to be effective. This method utilizes a principle of estimation using images with luminances varied by the flickering of a strobe. However, the method generates noise due to the difference in position of specular reflection among images in video imaging. In this paper, the authors propose a method combining optical flow with the above-mentioned method. The proposed method enables the removal of specular reflection for video imaging without generating noise. Experimental results and comparison with other methods of specular reflection removal show the effectiveness of the proposed method.

Prevention of pressure ulcers, which result from persistent mechanical loading of soft tissues, is a common issue among wheelchair users. In the case of deep tissue injury (DTI), it is difficult to predict where tissue breakdown will occur, as it initiates from tissues under the skin. In this paper, a new methodology to assess the risk of DTI using a compliant buttock dummy is proposed. The dummy is composed of pelvis model with six-axis force sensors and silicone rubber covering it to imitate the anatomical form of a human buttock. Six-axis force sensors attached to the pelvis model enable the measurement of external forces and moments during the sitting posture and estimate points of action on bony prominences such as ischial tuberosities by an algorithm. Experiments have been performed to test the accuracy of position estimation by placing point load on the bony prominences model. Next, the relationship between external forces and internal pressure has been confirmed by adhering pressure sensors on the surface of rigid models covered by soft materials. The results suggested the potential of the proposed dummy for the assessment tool of DTI risk.

This paper describes the development of an ankle training robot force visualization for eccentric contraction training. In the mechanism of walking, the eccentric contraction of the tibialis anterior muscle plays an important part in the phase between heel-strike and foot-flat as a shock absorber. However, the eccentric contraction has not been reported in previous research on ankle rehabilitation robots, despite its importance. In addition, traditional training methods both unaided and with tubing have difficulty making the training load constant. Hence, this study proposes a system that uses visual feedback of the forces to help the patient keep the training load constant. We present an estimation method for the lower extremity posture and torques. We also experimentally subject compared output force with and without the use of visual feedback.The results show that the system is able to make the training load constant. Furthermore, frequency analysis using the fast Fourier transfer (FFT) of the force information measured during training and passive motion shows that the robot is capable of detecting unwanted vibrations.

Robotic devices for nursing care are expected to help caregivers work with the elderly. Some robotic devices assist in the physical transfer of the elderly, and these robots come in contact with large surfaces of the human body. The regions of the buttock and the back may be uncomfortable due to the surface material or motions of the robotic devices. Therefore, sensing devices simulating a human body were developed to quantify and evaluate the load of a human body objectively. This simulated buttock consists of simulated bone and soft tissues, which include muscle, fat and skin. These regions have multi-axis force sensors to enable the quantification of the load due to the robotic devices used for nursing care. On measuring the soft exterior, it was found that the stiffness of the buttock dummy was similar to the human buttock. The comfort of a robotic bed was measured using the simulated buttock, and it was found that the shear force increased due to the deformation of the robotic bed. Thus, it was proven that the simulated buttock was capable of measuring the load of the human body when being used with robotic devices for nursing care. GRAPHICAL ABSTRACT

Accurate 6-DoF tracking technology is indispensable in robotics research, particularly for assessing the accuracy of 3D localization. We have developed an innovative 6-DoF tracking system for mobile objects that is cost-effective, passive, portable, scalable, and accurate. This proposed system employs high-accuracy markers, which have been enhanced from commonly used planar fiducial markers. All the operator needs to do is place these markers in the environment and on the measurement targets, and capture them with cameras. Subsequent processing, including marker measurements, establishing the world coordinate system, determining camera pose, outlier removal, calculating relative marker poses, handling occlusions, and generating 6-DoF tracking results, is automatically executed by the analysis software we have developed. We conducted experiments to measure the 6-DoF travel path of an electric wheelchair in a $\mathbf{6}\times \mathbf{9}\ \mathbf{m}$ indoor space, demonstrating that highly accurate measurements with a position error of less than 10 mm and an orientation error of less than 0.5° were achievable.