ETD

• artificial vision

This work focuses on the task of palletizing and depalletizing using robotic manipulators in the context of Industry 4.0 and transition to automation. More precisely, the process of unpacking a pallet requires an ample variety of jobs, namely the detection of the goods, their localization in the space, and finally the picking, so that they are ready to be placed in the desired position on a second pallet. The objective of this thesis is to find a way to determine the said pose in an automatic manner, perceiving the objects in the space using the information generated by 3D visual sensors, in particular from RGB-D cameras, also determining how they are located in the space, i.e. their pose, and finally to utilize this result in the subsequent motion control. The work is subdivided into many steps, so the first concerns the implementation of a state-of-the-art convolutional neural network for object recognition. This is exploited for the classification of a particular object belonging to a previously created database, supplying it with 3D grids generated starting from depth data so that the type of the recognized object can be obtained in output. Then an extension of the network is required to detect many instances of the object, and to find how they are placed in more complex scenarios and then devise the strategy to pick them up; in more detail, the algorithm performs point clouds registration and then computes the position and the orientation of the objects, in order to give them in input to the inverse kinematics part. The validity and effectiveness of the method is shown both through simulation results and hardware experiments carried out using cameras to test the neural network and the whole algorithm with the robot.