Real-time 3D Robotic Arm Tracking in Indoor Environment by RF Nonlinear Backscattering

Xiaonan Hui,Yunfei Ma,Edwin C. Kan
DOI: https://doi.org/10.1145/2987354.2987358
2016-01-01
Abstract:We demonstrate accurate real-time 3D robotic arm tracking in indoor environment via broadband RF nonlinear backscattering from passive tags, which is immune to mild visual obstruction from clothes and interior walls as well as static magnetic field distortion from the metallic structures and motors. The marker tag on the robotic arm has nonlinear backscattering built on an open-source RFID platform in order to resolve the self-jamming and multi-path problems in conventional RFID system. The sparsely selected frequencies in a broad bandwidth and the antenna spatial diversity enable efficient and reliable phase-based 3D localization against the ghost images from multi-path interference. We utilized one Tx and four Rx antennas with five frequencies from 829MHz to 1080MHz. The measured 20 Hz sampling had centimeter-level accuracy against an estimated 3D ground truth and millimeter-level deviation in many repeated measurements.
What problem does this paper attempt to address?