EMSOR: Embodied Multi-Modal Sensing in Self-Healing Soft Robots

Soft robots, a next generation of robots, ensure safe interactions with humans due to their inherent
softness. But their softness comes at a cost, making them susceptible to damage. Self-healing (SH)
soft robots address this vulnerability by recovering from macroscopic damages. Stretchable sensors
are embedded in their soft bodies to track deformation, touch, and damage. However, embedded
sensors face limitations, as their discrete integration results in areas where interactions remain
undetectable. EMSOR moves away from discretely embedded sensing and aims ambitiously to create
full-bodied multi-modal sensing in SH soft robots. To achieve this, new SH composites will be
researched that combine SH and piezoresistivity, along with the ability to 3D print airtight structures
with isotropic mechanical and electrical properties. This material innovation will enable the
exploration of two fundamentally new methods for embodied sensing in 3D printed SH soft robots,
leveraging pressure sensing and electrical impedance tomography (EIT). These allow the robots to
localize deformation, touch, and damage across every part of their soft bodies and monitor SH. For
both methods, EMSOR is challenged with researching strategies to distinguish between deformation,
touch, and damage, as well as to (re)calibrate via model-based approaches and machine learning.
Ultimately, the strengths of these methods will synergize in a multi-modal sensing system that will
be embodied in a single soft robot.