A flexible parts assembly algorithm based on a visual sensing system

Unlike rigid parts, flexible parts can be deformed by contact force during assembly. For successful assembly, information about their deformation as well as possible misalignments between the holes and their respective mating parts are essential. Such information can be acquired from visual sensors. In the case of deformable part assembly, the corrective assembly motion to compensate for such misalignments needs to be determined from the measured informations. In order to tackle these problems, authors' previous work presented a visual sensing system for measuring parts deformation in any direction and misalignments. This paper presents a visual sensor-based error-corrective algorithm using a neural network and utilising the developed sensing system. A series of experiments to compensate for the lateral misalignment are performed. The experimental results show that the proposed sensing system and the error-corrective algorithm are effective in the assembly tasks of deformable parts, thereby dramatically increasing the rate of success in assembly operations.