ADLs Measurement System in Home
Due to the decline in the nursing care workforce caused by the falling birthrate and aging population, the extension of healthy life expectancy is being promoted. Various exercise tests, such as the Timed Up and Go test, are conducted at local exercise testing sessions to evaluate physical functions to determine the risk of falls and diseases. However, it is difficult to evaluate the true physical functions of people because they cannot perform natural movements in unusual environments.
Therefore, this study proposes a motion measurement system that can be installed in a living space to evaluate daily physical functions such as walking, stair walking, and standing/sitting movements. For privacy protection, we do not use RGB data; we only use depth data representing the distance from the sensor to estimate joint position and angle information. We use a unique algorithm to construct a monitoring system that does not burden the occupant.
Development of Physical Function Evaluation Indexes
Knee Osteoarthritis, one of the most common musculoskeletal diseases in the elderly, is said to be difficult to recognize as symptoms progress, although early intervention is effective.
In response to this, this research aims to detect subtle changes in the body using information measured by a daily life movement measurement system. By conducting large-scale measurement sessions using the movement measurement system, we are developing a new evaluation index useful for screening for knee osteoarthritis by analyzing the presence or absence of knee osteoarthritis and other musculoskeletal disorders and the measurement data acquired.
Wellness Infill
Wellness infill refers to a system for designing the daily activities of residents through the design of living spaces, such as floor plans and furniture layouts, and the control of the living environment. The load on each part of the body during daily activities varies depending on the spatial design, such as the design of stairs, the height of steps and furniture, and the layout of the room. For example, if the load on the lower back is hefty during a particular movement, the repetition of the movement may cause back pain at some point in the future. Based on the hypothesis that hints for avoiding such an injury may be hidden in the design of living spaces, we aim to support healthy living for residents by designing appropriate spaces for their movements.
Augmented Architectural Space for Casual Connections with People
The increasing trend toward nuclear families and changes in residential patterns have led to a weakening of human ties, and social isolation has become a severe social problem in contemporary Japanese society. However, there is still a gap between face-to-face communication and telecommunication, and this gap causes physical and mental fatigue.
Therefore, we are building a system to create casual connections between people by incorporating augmented architectural space into remote communication as an element of space normally shared in face-to-face communication.
Supportive Performance of Living Space for ADLs
This is a study on movement support by space according to the health condition of residents. When a resident is discharged from the hospital with residual motor function impairment, the living environment needs to be improved to provide movement support without excess or deficiency, but currently, this is made by a trial-and-error process based on experience.
We aim to construct an evaluation system for walking while walking by acquiring information on contact with furniture, i.e., mechanical interaction and gait parameters during walking. This research will evaluate the support performance of walking, i.e., how much the space can support human motion, to realize effective and efficient environmental improvement.
Optimal Sensor Position Estimation in Home
In recent years, research on sensing technology for the purpose of evaluating motor function in living spaces has progressed from the viewpoint of preventive medicine. On the other hand, since motor function evaluation involves the measurement of daily activities using sensors, it is necessary to pay more attention to the placement of sensors and the setting of data acquisition conditions from the viewpoint of privacy protection.
In this research, we are constructing a system that proposes the optimal placement of sensors for motor function evaluation in a living space. We are attempting to propose optimal sensor placement by simulating residents' walking paths and measurable movements based on floor plan information.
Prediction of Turning
When robots and sensors are introduced into living spaces, systems that predict human movements are needed to ensure safety and comfort. Among the daily activities to be measured, changing direction while walking is challenging with high frequency and a high risk of falling.
Therefore, in this study, a prediction system was constructed for the change of direction during walking in a real-life environment. We extracted useful functional prediction parameters that reflect preliminary movements during change of direction from joint position information of the whole body acquired from sensors, improved the prediction accuracy, and investigated the difference in prediction accuracy depending on the characteristics of space and individuals.
Habits-based Space Planning System
The study reflects individual habit information in the design of a new home. Many people experience significant stress when they move, renovate their homes, or otherwise change their environment. The occupants must adapt their lifestyles to the new architectural space. This stress is considered one of the biggest problems, especially for older people, who cannot adapt to the environment. However, in actual architectural design, it isn't easy to directly incorporate the residents' lifestyles.
In this research, we are attempting to model the habitual information in a familiar house and to design architecture based on the customary information.
Posture Evaluation System for Backage Prevention
It is known that one of the causes of low back pain is the continuation of improper posture, and that worsening posture has a negative impact not only physically but also mentally. Since sitting is the most common posture in the workplace in developed countries, it is expected that prolonged poor sitting posture contributes to the occurrence of low back pain. To prevent back pain, it is important to maintain good posture. However, the definition of good posture varies from person to person due to differences in skeletal structure and muscle tone, as well as the surrounding environment, such as chair height and the presence or absence of a desk.
Therefore, this research proposes a system that estimates the load on each joint based on information obtained from sensors and provides feedback to residents on their objective posture assessment according to the individual and spatial environment.
Related research
To Publication
Concentration Estimation System
This is a study on quantifying concentration. Due to the recent surge in work-at-home and online learning due to the global situation, people of all ages are spending more and more time working and studying alone. In the past, teachers have been able to monitor learners' concentration levels and adjust educational programs, such as study content and breaks, as needed. However, online, unsupervised learning requires uniform learning content and schedule management, making independent learning difficult, especially for younger learners.
In this study, we propose a method for estimating the degree of concentration, which is useful for understanding the learner's state. This method uses only the information obtained from the camera built into the PC, which is essential for online learning, to estimate the subject's concentration without any contact and without restraining the subject.
Related research:
To Publication