Design, Development and Characterization of Tactile Sensors Solutions in Multi-Modal Sensing Systems for Robot Self-Aware Control

This doctoral research addresses the development of scalable, printable, and mechanically compliant capacitive tactile sensing solutions for robot self-aware control and large-area robotic skin integration. As robots increasingly operate in close proximity to humans, tactile sensing becomes a fundamental technology for safe physical interaction, compliant manipulation, and whole-body perception. However, significant challenges remain in terms of fabrication scalability, sensitivity enhancement, communication bandwidth, and integration on curved and articulated robot geometries. The thesis proposes and experimentally validates three complementary technological contributions. First, multilayer capacitive tactile sensors based on vertically stacked architectures are introduced to enhance sensitivity and dynamic range without increasing the in-plane dimensions of the sensing element. Analytical modeling and finite element simulations demonstrate that vertical stacking systematically increases capacitance variation under applied pressure. Experimental results obtained through inkjet and screen-printing fabrication confirm a mono- tonic sensitivity improvement up to five stacked layers, validating the scalability of the approach while preserving compact taxel geometry. Second, CySkin+ is presented as an evolution of an established distributed tactile architecture. The work introduces substrate optimization, improved fabrication strategies based on lamination and printing techniques, and the integration of an EtherCAT-based communication backbone. Experimental characterization confirms compliant sensitivity, repeatability, and robustness compared to the original CySkin design. The adoption of industrial real-time communication enables deterministic, synchronized, and scalable distributed acquisition, strengthening compatibility with industrial robotic control frameworks. Third, PeSkin is introduced as a large-area tactile sensing solution based on screen-printed air-gap capacitive technology. The laminated PET–air-gap architecture demonstrates high low-force sensitivity (up to 5870 f F/N | 57 f F/kPa), limited hysteresis (3.11 %), high repeatability ( 1.77 %), fast dynamic response ( 4.17 ms), and an effective dynamic range of approximately 15 bits. A cylindrical integration demonstrator composed of 12 patches and 1728 taxels validates mechanical conformability, modular acquisition scalability, and real-time distributed tactile visualization over curved surfaces. Overall, this thesis contributes a complete pipeline—from sensor architecture modeling and material selection to electronic readout design and system-level robotic integration—toward scalable and printable tactile skins. By exploiting vertical stacking, industrial communication protocols, and additive manufacturing techniques, the proposed solutions demonstrate that high-performance capacitive tactile sensing can be effectively integrated into distributed robotic platforms.