Human Motion Adaptation: A Multi-Criteria Approach

The paper faces the problem of human motion simulation by providing a new approach for the adaptation of a real motion to new virtual anthropometrics and environment conditions. A multi-criteria motion adaptation algorithm has been conceptualized and implemented as an alternative to the existing solutions of human motion simulation. Having as input a real human motion, in terms of sequential postures and new virtual conditions, the algorithm generates new random human postures and evaluates them based on a number of criteria providing such a motion that would satisfy the new conditions. Real motion captured data of a car's door opening from the driver's seat position, derived from experiments, have been used in order to test and demonstrate the algorithm's efficiency in motion simulation. The results presented in this paper show that realistic and accurately adapted human motions can be generated and used in computer-aided design for ergonomics.