We present a data-driven approach to simulate the BioTac tactile fingertip sensor within physics engines. The behavior of the sensor is first captured in an experimental setup that records positions and external forces of contacts as well as the sensor output. This data is then used to fit a non-linear model that maps force-annotated mesh collisions of a simulator to sensor responses. We discuss two deep network architectures that reproduce the BioTac data with high accuracy and demonstrate the simulation of simple grasps with the Shadow Dexterous Hand and five BioTac sensors. We present an open source plug-in for the simulator Gazebo and release the captured dataset alongside this paper.
CITATION STYLE
Ruppel, P., Jonetzko, Y., Görner, M., Hendrich, N., & Zhang, J. (2019). Simulation of the syntouch biotac sensor. In Advances in Intelligent Systems and Computing (Vol. 867, pp. 374–387). Springer Verlag. https://doi.org/10.1007/978-3-030-01370-7_30
Mendeley helps you to discover research relevant for your work.