Demo Configuration Identification for a Freeform Modular Self-reconfigurable Robot - FreeSN

Authors: Yuxiao Tu, Tin Lun Lam Corresponding author: Tin Lun Lam (Email: tllam@; Website: ​) Published in: Published in: IEEE Transactions on Robotics (T-RO) Paper: Freeform Robotics: Title: Configuration Identification for a Freeform Modular Self-reconfigurable Robot - FreeSN Abstract: Modular self-reconfigurable robotic systems are more adaptive than conventional systems. This article proposes a novel freeform and truss-structured modular self-reconfigurable robot called FreeSN, containing node and strut modules. A strut module contains two freeform connectors, which can connect to any position of the node module. This article presents a novel configuration identification system for FreeSN, including connection point magnetic localization, module identification, module orientation fusion, and system configuration fusion. A magnetic sensor array is integrated into the node module. A graph convolutional network-based magnetic localization algorithm is proposed, which can efficiently locate a variable number of magnet arrays under ferromagnetic material distortion. The module relative orientation is estimated by fusing the magnetic localization result with the inertia moment unit and wheel odometry. Finally, the system configuration can be estimated, including the connection topology graph and the poses of modules. The configuration identification system is validated by accuracy evaluation experiments and two closed-loop library-based automation demonstrations.