Bioinspired Robotic Cutaneous Sensor System for Softness Detection

Softness perception is a unique feature of the human skin, which is very useful during object manipulation tasks. However, this distinctive function relies on the skin deformation under pressure, and is hard to mimic in traditional robotic technology.<br/>Although some research groups have developed several compliance sensing modules based on flexible or stretchable electronic technologies, the interaction between a soft artificial device and a target object, in terms of Young's modulus, is rarely considered. Specifically, when the target material is stiffer than the sensor, a large deformation occurs in the device while there is negligible deformation on the contacted object. Conversely, a softer object may deform more than the soft sensing module itself. As a result, even under the same external pressure, the deformation of the sensor depends on the difference in Young's moduli of the materials forming the target object and the sensor. Such variation could cause significant errors in the measurement and thus, hinder the correct characterization of the compliance.<br/>Inspired from the human tactile system, we report on liquid metal-based soft sensor modules for softness detection. Each sensing unit comprises a strain sensor and a pressure sensor to measure deformation and pressure simultaneously. We propose an analytical mechanical model to describe the interaction between the soft sensor and the soft target, and simulate the system using finite element modeling. Following a calibration, we successfully distinguished engineering materials such as rubber and hydrogel and some biological tissues including chicken chest and heart. Next we prepared a 4x4 sensor matrix and successfully quantified the compliance and spatial distribution of several objects. This softness sensing technology may be implemented in applications such as robotic palpation or precise teleoperation.