Kyushu Institute of Technology
Department of Mechanical and Control Engineering
Komura Lab.
Scientific Investigation of Human Tactile Perception Toward Next-Generation Engineering Technologies.
Development of a Three-Axis Tactile Sensor for Robotic Applications
Abstract
If the manual skills of highly skilled experts can be recorded and reproduced, their applications can be greatly expanded. Achieving this goal requires three key technologies: (1) extraction of tactile information from skilled operators, (2) high-precision recording of that information, and (3) robotic reproduction of the recorded tactile skills. In this study, we develop a three-axis tactile sensor for robotic applications as a fundamental technology enabling these capabilities. A three-axis tactile sensor can measure not only the normal pressure distribution at the contact surface but also the shear force distribution simultaneously. The proposed sensor consists of an acrylic core shaped to mimic a human fingertip, covered with a thin rubber sheet that envelops the fingertip and finger pad as a unified structure. The contact state between the rubber sheet and the acrylic core can be observed optically from a region corresponding to the fingernail. By analyzing the brightness distribution of the contact area and the displacement of its centroid through image processing, the three-axis contact forces (normal force and two orthogonal shear forces) can be estimated. In addition, the material properties of the object being touched can also be inferred. This technology aims to serve as a foundation for implementing human-level manual dexterity in robotic systems.
Research on Kinesthetic Illusions Toward Achieving Human Supportive Technologies
Abstract
Kinesthetic illusion refers to a perceptual phenomenon where appropriate vibratory stimulation of tendons creates a sensation as if the muscle has moved in the direction of elongation. Unlike the tonic vibration reflex that moves in the direction of contraction or the antagonist vibratory response where the antagonistic muscles relax, kinesthetic illusion is characterized by the sensation of movement without actual movement. This phenomenon occurs throughout the body, involving muscles and tendons. Discovered in the 1970s, precise experimental results with controlled stimuli remain scarce, varying significantly among researchers.
Novelty and Originality of This Study
In this research, we utilize a high-precision vibration generator to conduct experiments, controlling parameters such as compressive force, waveform, frequency, and amplitude to clarify the conditions under which kinesthetic illusions occur. To achieve this, we have developed the device depicted in the diagram below. Using this apparatus, vibratory stimuli are applied to the flexor carpi radialis tendon of the right wrist, enabling the manifestation of kinesthetic illusions that can be quantified in real-time using the left hand.
Applications
This study envisions devices that induce the sensation of hand movement in stroke patients, potentially enhancing recovery when used alongside other rehabilitation methods.
Development of a Sensor for Recording Human Tactile Information
Abstract
In today's aging society, the development of technology for recording and reproducing human tactile information is highly anticipated for skill transmission and remote medical applications. Our research focuses on a sensor that estimates three-axis forces (normal and shear forces) acting on the fingertip based on changes in the nail color of the index finger. Various attempts have been made to record human tactile sensations, including methods that capture "material perception" based on skin vibrations and techniques that estimate "force" using nail color changes. However, factors such as fingertip stiffness, nail color variation due to ambient temperature, finger posture, and capillary blood flow fluctuations can significantly affect these measurements, reducing the accuracy of force and material perception estimation when environmental conditions differ from those in which the sensor was calibrated.
In this study, we aim to develop a high-precision technology that estimates "material perception" and "force perception" by real-time tracking of changes in the mechanical properties of the finger and blood flow. This approach will allow for accurate estimation of the interaction between the finger and external objects.
Novelty and Originality of This Study
Although previous research has utilized the physical deformation of the finger during touch as a principle, it has not sufficiently accounted for variations in finger properties under different conditions. For instance, fingers become stiffer in low temperatures, and bending the finger alters nail color. Since tactile perception is fundamentally based on skin deformation caused by external stimuli, estimating the condition of the skin is essential. The novelty of this study lies in the detailed estimation of the finger’s state before analyzing tactile perception. Furthermore, the originality of this research is in integrating the estimated finger state with tactile perception to analyze the characteristics of skilled techniques.
Applications
This research not only enables the recording of highly skilled techniques but also captures the unique characteristics of the fingers of experts. By doing so, it aims to reveal the underlying factors that contribute to exceptional skills and dexterity.
Development of Sports Motion Assistance Technology Using Pneumatic Actuators
Abstract
We are developing an assistive device aimed at enabling the efficient and accurate acquisition of sports movements. Pneumatic joint-fixation actuators and contraction-type actuators are mounted on the body, and their optimal placement and actuation timing are investigated from a biomechanical perspective. The accompanying video presents an experimental demonstration of batting motion assistance, in which an actuator is attached to the left elbow.
Development of a Measurement Shoe for Caregiving Posture Estimation
Abstract
With the rapid aging of society, the physical burden on caregivers in nursing care settings has become a serious social issue. In particular, improper postures during patient transfer tasks can lead to lower back pain and other work-related musculoskeletal disorders. Therefore, there is a growing need for systems that can detect hazardous postures and provide appropriate warnings.
In this study, we develop a measurement shoe capable of capturing caregivers’ postures during transfer assistance tasks. Based on the acquired sensor data, we construct a posture estimation system using deep learning techniques.
Fundamental Study on the Velvet Hand Illusion
Abstract
The engineering design and creation of tactile sensations is a crucial research topic that can drive new innovations. In our previous studies, we have focused on the Velvet Hand Illusion (VHI), a tactile illusion, as a method for enhancing tactile perception and have conducted fundamental research on its underlying mechanisms. VHI is a phenomenon in which rubbing a wire mesh between both hands generates the sensation of a smooth surface. Our findings indicate that VHI does not occur with a single wire but only emerges when the wires form an enclosed structure. This suggests that the illusion is related to the concept of Gestalt perception—an integrated whole that cannot be explained by individual components—originally developed in the context of visual perception.
We hypothesize that tactile information is grouped according to the Gestalt principle of Prägnanz, leading to the emergence of a unified tactile experience. By elucidating the process through which tactile information is integrated into a Gestalt and contributes to the perception of surface texture, we aim to advance the understanding of tactile perception mechanisms. This study seeks to clarify how the brain integrates multiple tactile stimuli into a cohesive object surface representation and explore applications in haptic devices.
Novelty and Originality of This Study
This study is not only utilizing a tactile illusion but also striving to reveal its underlying mechanisms, which is a novel approach. We attempt to explain tactile illusions through the established Gestalt framework, which has traditionally been used to describe visual perception. To achieve this, we conduct both psychophysical experiments and neurofunctional measurements, providing a comprehensive investigation of the phenomenon.
Applications
The ultimate goal of this research is to leverage tactile illusions to create and present diverse tactile experiences, opening new possibilities for haptic technology.
