Hybrid Soft Sensing in Robotic Systems

Abstract

The desire to operate robots in unstructured environments, side-by-side with humans, has created a demand for safe and robust sensing skins. Largely inspired by human skin, the ultimate goal of electronic skins is to measure diverse sensory information, conform to surfaces, and avoid interfering with the natural mechanics of the host or user. These demands require an alternative approach to conventional sensing electronics, which are typically planar and rigid. Soft sensors fill this gap by employing materials and substrates that mimic the mechanical properties of living tissues. Their natural compliance makes soft sensors well-suited for a variety of robotic applications including object manipulation, tactile sensing, and human-robot interaction.

However, stretchable sensors face unique challenges that require the development of novel fabrication techniques and methods to handle the non-linear characteristics of soft materials. Designing larger robotic systems for robust interaction is additionally constrained by the limited availability of sensing modes. For these reasons, soft sensors have been largely limited to pressure and strain sensing and not often integrated into larger systems. These challenges not only restrict their functionality, but prevent further potential work towards soft sensors for safe and intelligent interaction. This work introduces three soft sensor designs that expand available soft sensing modes for hybrid robotic systems development. Each approach balances simple fabrication and signal interpretation to explores their use in the context of robotic gripping.

First, a biohybrid pneumatic gripper was designed to take advantage of the inherent material compatibility between synthetic biology and soft robotics. Chemical signals in the environment are expressed by the development of fluorescent proteins, which are detected by on-board opto-electronics in the gripper's strain-limiting layer. This implementation supports the integration of biological cells and soft robotic materials for advanced chemical sensing. Second, I scale up a technique to fabricate a soft and stretchable electronic skin with commercial IC chips. The combination of stretchable substrate, liquid-metal traces, and IC chips achieves soft sensing skins with on-board processing, communication, and MEMS sensors for the first time. Lastly, I introduce a novel magnetic elastomer composite for deformation sensing. We show practical applications for this composite as a continuous tactile surface and environmental marker for sub-mm object localization. Taking inspiration from many fields, this thesis work contributes to the diversity of soft sensing technologies to advance the development of complex, robust, and soft interfaces and systems.