Soft sensors made of highly deformable materials are one of the enabling technologies to various soft robotic systems. However, major drawbacks of soft sensors compared with traditional sensors are their nonlinearity and hysteresis in response, which are prominent especially in microfluidic soft sensors. In this research, we investigate the above issues of soft sensors by taking advantage of deep learning. We implemented a hierarchical recurrent sensing network to the calibration of soft sensors for estimating both magnitude and location of a contact pressure simultaneously. We collaborate with Prof. Yong-Lae Park’s group at Seoul National University.
Related publications
1. S Han, T Kim, D Kim, Y Park, S Jo, Use of Deep Learning for Characterization of Microfluidic Soft Sensors, IEEE RA-L, 3(2): 873-880, 2018 (to be presented in ICRA 2018).[LINK] [PDF][CODE]
