Digital Theses Archive


Tesi etd-05072021-142856

Type of thesis
Scientific disciplinary sector
Istituto di Biorobotica - BIOROBOTICS
relatore Prof.ssa LASCHI, CECILIA
  • material intelligence
  • mechanical intelligence
  • mechanochromism
  • octopus
  • optical fiber
  • physical intelligence
  • proprioception
  • soft robotics
  • spiropyran
  • suction cups
  • surgical robotics
Exam session start date
“Mechanical” and “material intelligence” in soft robots toward smart electronics-free solutions for proprioceptive machines, is the title of the Goffredo Giordano’s Ph.D. thesis. This manuscript shows a fascinating connection between the biological world (i.e. cephalopods’ capabilities of light-manipulation and morphological adaptation), and bioinspired industrial and technological systems developed for the emergent field of soft robotics. Particular attention will be posed to bidirectional transducers acting as both actuator and sensor, exploiting the “mechanical intelligence” (namely conformability and morphological adaptability), and the “material intelligence” (here considered as smart mechano-responsive properties). A universal tendon-driven octopus-inspired robotic arm developed for energetic industrial scenario, will be presented as a multi-level platform to retrieve object from harsh and constrained environment, exploiting its “mechanical intelligence”. This platform will be considered along the manuscript as the most likely system on which we will integrate our developed smart solutions or soft sensor systems. The main scope is to obtain innovative solutions to provide soft robots with mechanical proprioception, at the same time, minimizing the adoption of standard electronics that hamper the intrinsic softness of the robot body (i.e. wirings, encapsulation, batteries, etc.). To achieve a step toward this challenge, and without interfering with the “mechanical intelligence” during actuation, we designed sensing systems considering the mechanical force has potential for productive use (i.e. activation of smart embedded properties). The mechanical proprioception challenge is here entrusted to recognize the spatial shape of a continuum-like arm, or the achieved grasping, or a pressure-mediated task. For the former, we will present a tendon-driven robotic mock-up, on which has been obtained the reconstruction of a one- and large-bending motion, fusing a plastic optic fiber as curvature sensor, a simplified steady-state mathematical model, and an adaptive filtering technique. Contrary, to detect grasping or pressure-mediated task, we will present a smart chemical functionalization of typical constitutive elastomeric matrices of end-effectors, by covalently bonding a mechanochromic spiropyran molecule into the chain-like polymeric system (i.e. molecular mechanochromism). We will provide a grounded optical, mechanical, and thermal mechanochromic material characterization. The latter was useful to the development of a proof-of-concept of a mechanochromic embodiment in a surgical tool in a minimally invasive soft robotic surgery application. To the best of our knowledge, this is the first electronics-free smart embodiment furnishing the operator with visual cues for incipient potential organic damage, and for safe tissue handling. Nonetheless, we will present some preliminary results regarding the integration of spiropyran chemical species in end-effector suction cups to provide grasping proprioception in the aforementioned octopus-like robot. A brief analysis on structural mechanochromism, replacing stretchable photonic crystals, will be presented as electronics-free adds-on to detect strain or damage thresholds with optical feedback. We envisage a fusion between soft sensor systems and smart molecular and structural mechanochromic properties, which could advance the proprioception in our developed soft platforms. The “mechanical” and “material intelligence” can open to adaptive displays and dynamic camouflage skins in soft robots, e.g. to obtain mechano-imaging outcomes, which can enhance the coexistence of soft machines in open-world realms. In conclusion, our thesis is that the mechanical proprioception in soft robots could come by means of chromogenic devices (“material intelligence” as e.g. structural and molecular mechanochromism) and innovative soft photonic strategies, respecting the inherent “mechanical intelligence” of the end-effectors.