Property identification of a deformable belt object from its static images toward its manipulation

A method to identify the flexural rigidity and the initial curvature of a deformable belt object from its static images is proposed. Automation of assembly of belt objects such as flexible flat cables hasn't progressed because of the difficulty in estimating their deformation. Furthermore, they have the variation which impedes successful manipulation with a uniform trajectory. In this paper, we consider a procedure to identify the properties of each cable and to generate an appropriate trajectory for it just before its manipulation. First, a method to identify the flexural rigidity of a belt object from its one static image is proposed. Next, this method is extended to identify both the flexural rigidity and the initial curvature of the object simultaneously. Finally, an example of planning adapted to the variation of the object properties is shown.