Configuration Identification for a Freeform Modular Self-Reconfigurable Robot - FreeSN
Yuxiao Tu, Tin Lun Lam
Abstract
Modular self-reconfigurable robotic systems are potentially more robust and 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 node module contains a low-carbon steel spherical shell. A strut module contains two magnetic-based freeform connectors, which can connect to any position of the node module and provide spherical motions. Accurate configuration identification is essential for the automation of modular robot systems. 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 then 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 a series of accuracy evaluation experiments and two library-based automation demonstrations based on closed-loop control.