Relation-Based Posture Modeling for DHMs

Abstract

Posture modeling for DHMs has significant effect on their usage in evaluation of product ergonomics. Direct manipulation schemes, such as joint level maneuvering for changing posture, are tedious and need more user intervention; complex scenarios are hardly simulated. This paper presents a high-level, relations based, description scheme for human postures and demonstrations for executing these descriptions using a digital human model (DHM). Here, posture is viewed as a pattern of relations of body segments among themselves and with the environment. These relations are then used as the criteria for the novel description based control. A few basic postures have been derived using the conventional principal planes. The basic postures and the composition rules enable description of complex postures in an easy and unambiguous way. We discussed the issues involved in the execution of descriptions and developed methods to resolve the conflicts due to link fixations. This scheme is effective for both lower and higher level control. Illustrative examples from the implementation of the concepts in our native DHM ‘MayaManav’ are included.