Science Based Genuine Digital Manikin
The computer manikin is a software tool for assessing the compatibility of the human body with environmental/industrial products in a CAD system.
Computer manikins are currently used for design of automobiles, aircraft, production lines, and housing.
These software tools are imitating human body structure, but their accuracy has not been properly evaluated.
The main use of the software is for determining product dimensions, and it is not possible to simulate mobility/usability and feel when substituting for a human being.
Here, we aim at the next generation manikin software development that can simulate the perception of load by automatically generating movements and incorporating scientific findings on the human body that have been either already unraveled or are possible to unravel. We are working on the development of a "Science Based Genuine Digital Manikin", that reconstructs postures and body dimensions based on anthropology, anatomy, and muscle physiology, and then generates the movements that can satisfy dynamic conditions, according to one's physical constitution and the environment in which the product is used.
We are advancing development by specifying challenges which focus on housing and automobiles.
To do this, a cooperative system among private companies, universities, national institutes and ministries is essential.
Researchers at universities and national institutes can create innovative technologies.
Then software companies provide the technology as a packaged tool.
Finally, housing and automobile companies utilize the tool for product design.
The Digital Human Technology Consortium was launched in September, 2003 in order to share the technological information on digital humans and
to promote collaborative research. The following companies and universities are currently participating:
Digital Human Research Center,AIST
(As of April 30, 2007)
We launched a study group, aside from the consortium, that focuses on information exchange and education. Five private companies are participating in this study group (as of January, 2004) to work together on projects such as: data infrastructure development to learn human body structure quantitatively, techniques to measure human body motions corresponding to a digital manikin model, techniques to measure motion within automobile mock-ups, techniques to validate postures and anthropometric accuracy of a digital manikin, and development of a web database that organizes reference data on human characteristics.Data Infrastructure Development
We are taking measurements in such a way that shapes, motions, MR images, and passive joint properties are consistently measured for each subject. We are studying and incorporating into models the relationship between different body properties and unraveling how motions and loads are related to these properties.
Accuracy Validation Technique
We are proposing a validation method for functional anthropometric accuracy of computer manikins that uses an anthropometric database. Measurements of 9 members of the boundary family are imported into a computer manikin system, and computer manikins for these members. By having these manikins take postures for functional anthropometry, their body dimensions are measured within the computer manikin system, and compared with the actual measurements.
The centers of joints that correspond to a digital manikin model are estimated from the measured movements of body surface markers.Human Body Properties Database
This database covers the major references on human body measurements, passive joint properties, and body segment parameters (mass, center of gravity, moment of inertia). The database is arranged to clearly show the differences in definition. The database is open to the public.Human Motion Generating Technique
Based on measured motion data, virtual motions with various body shapes and environmental restrictions may be generated. We are also developing a system for the digital right management (DRM) of captured and synthesized motion data using a watermarking method. (motion watermarking)