Dataset for magnetic soft robots and materials

Abstract

(1) Dataset for magnetic soft robots and materials This dataset is a result of the "Data-driven design of shape-programmable magnetic soft materials" paper. Karacakol, A.C., Alapan, Y., Demir, S.O., Sitti M. Data-driven design of shape-programmable magnetic soft materials. Nat Commun 16, 2946 (2025). https://doi.org/10.1038/s41467-025-58091-z It involves a raw dataset of 53 learning runs (resulting in ~1.5e7 simulation evaluations) conducted for the demonstrations and algorithm benchmarking cases. Each data provides the position information of the robot in cartesian coordinates along its body at the final steady-state shape for a given design and actuation signal. The dataset is shared to enable further research in stimuli-responsive soft robotics and soft materials fields. (2) Github link for the project Public code repository for the "Data-driven design of shape-programmable magnetic soft materials" paper could be found in; https://github.com/AlpKaracakol/data_driven_magnetic_soft_material_design