The automation of production processes requires an increasingly accurate and sensitive control of product manipulation. This ensures that the finished product remains undamaged and that machines can adaptively handle process variations. To achieve the ability to manipulate an object with the same sensitivity as a human hand, it is necessary to integrate various technologies that enable the identification of the object and the gripping point in space, as well as the positioning of robots and manipulators, and the manipulation of parts with a controlled level of force.
The project aims to create a system for the robotic manipulation of delicate products without requiring programming operations. To teach the system the gripping of a new component, it will be sufficient for the operator to perform the manipulation of the pieces once, wearing a sensorized glove capable of reading the movement of the hand and the pressure exerted on the part. An AI algorithm will determine the ideal grip and pressure to apply by cross-referencing the shape and orientation data of the object from the vision system and the glove worn by the operator. Once trained, the system will be able to replicate the operation automatically, adapting to process variability.