An electrical impedance tomography based artificial soft skin pressure sensor: Characterisation and force modelling

Abstract

Using electrical impedance tomography (EIT) to drive a pressure mapping device shows great potential, due to the customisability of the sensing domain and the non-invasive nature of the boundary electrodes. A pressure mapping system has been developed in this work that uses a carbon black silicone rubber (CBSR) nanoparticle sensing domain, giving the sensing domain a comparable softness to human skin tissue. To take this technology into a commercial application the performance of such an EIT-based sensor must be quantifiable and repeatable. In this work a series of experiments were repeated for various load locations, strains, and carbon black percentages. Capturing this data gave insight into the how the sensing domain performs over time and captured the transient events limiting the sensor. Metrics were determined to quantify the sensor’s spatial resolution. Load localisation could be determined with error values as low as 0.67 mm. A series of randomised test loads gave similar spatial performance results to the more structured experiments. A quasi-static conductance-force model of the material was developed with an accuracy of ± 0.78 N. One important metric is temporal resolution, as it is the least quantified performance metric in literature, however can be the most important for some applications. For the sensor domains tested, average settling times of between 19.0 – 44.5 s and 22.5 – 36.0 s were determined for 8 and 9 wt% CBSR samples. This sensor platform shows promise for future soft surface pressure mapping applications. Further use of the developed performance metrics will allow for a variety of sensor applications to be validated.