Robotic Sensing and Manipulation of Deformable Linear Objects with Learning-based methods

Nowadays robotic applications are widespread and most of the manipulation tasks are efficiently solved. However, Deformable-Objects (DOs) still represent a huge limitation for robots. The main difficulty in DOs manipulation is dealing with the shape and dynamics uncertainties, which prevents the use of model-based approaches (since they are excessively computationally complex) and makes sensory data difficult to interpret. This thesis reports the research activities aimed to address some applications in robotic manipulation and sensing of Deformable-Linear-Objects (DLOs), with particular focus to electric wires. In all the works, a significant effort was made in the study of an effective strategy for analyzing sensory signals with various machine learning algorithms. In the former part of the document, the main focus concerns the wire terminals, i.e. detection, grasping, and insertion. First, a pipeline that integrates vision and tactile sensing is developed, then further improvements are proposed for each module. A novel procedure is proposed to gather and label massive amounts of training images for object detection with minimal human intervention. Together with this strategy, we extend a generic object detector based on Convolutional-Neural-Networks for orientation prediction. The insertion task is also extended by developing a closed-loop control capable to guide the insertion of a longer and curved segment of wire through a hole, where the contact forces are estimated by means of a Recurrent-Neural-Network. In the latter part of the thesis, the interest shifts to the DLO shape. Robotic reshaping of a DLO is addressed by means of a sequence of pick-and-place primitives, while a decision making process driven by visual data learns the optimal grasping locations exploiting Deep Q-learning and finds the best releasing point. The success of the solution leverages on a reliable interpretation of the DLO shape. For this reason, further developments are made on the visual segmentation.

Advanced electrospun nanofibrous mats for hindering delamination, improving damping and for sensing of composite laminates

Carbon Fiber Reinforced Polymers (CFRPs) display high specific mechanical properties, allowing the creation of lightweight components and products by metals replacement. To reach outstanding mechanical performances, the use of stiff thermoset matrices, like epoxy, is preferred. Laminated composites are commonly used for their ease of manipulation during object manufacturing. However, the natural anisotropic structure of laminates makes them vulnerable toward delamination. Moreover, epoxy-based CFRPs are very stiff materials, thus showing low damping capacity, which results in unwanted vibrations and structure-borne noise that may contribute to delamination triggering. Hence, searching for systems able to limit these drawbacks is of primary importance for safety reasons, as well as for economic ones. In this experimental thesis, the production and integration of innovative rubbery nanofibrous mats into CFRP laminates are presented. A smart approach, based on single-needle electrospinning of rubber-containing blends, is proposed for producing dimensionally stable rubbery nanofibers without the need for rubber crosslinking. Nano-modified laminates aim at obtaining structural composites with improved delamination resistance and enhanced damping capacity, without significantly lowering other relevant mechanical properties. The possibility of producing nanofibers nano-reinforced with graphene to be applied for reinforcing composite laminates is also investigated. Moreover, the use of piezoelectric nanofibrous mats in hybrid composite laminates for achieving self-sensing capability is presented too as a different approach to prevent the catastrophic consequences of possible structural laminate failure. Finally, an accurate, systematic, and critical study concerning tensile testing of nonwovens, using electrospun Nylon 66 random nanofibrous mats as a case study, is proposed. Nanofibers diameter and specimen geometry were investigated to thoroughly describe the nanomat tensile behaviour, also considering the polymer thermal properties, and the number of nanofibers crossings as a function of the nanofibers diameter. Stress-strain data were also analysed using a phenomenological data fitting model to interpret the tensile behaviour better.