Biomechanics and Ergonomics

The project main objective is to make a realistic prediction of the position of the human hand grip for everyday objects. The work is centred in the positions of maximum opening and final gripping posture for pinch grip, one of the most widely used in human manipulation. This is the first step towards the generalization to any grip. Neural networks are useful as a predictive tool for  human grip postures as they are able to learn the underlying rules that are adopted during this process. In the project there are three phases. The first involves identifying the parameters that most affect the reached grip, through experimentation. The second is the development of a prediction module for grip posture based on neural networks for pinch grasp. Finally, the third module will involve the integration of the prediction module in the biomechanical hand model developed by the research group. This integration will automate the production of grip position for a given object that is biomechanically correct and representative of normal human grip for that object.