Abstract
This study develops a passive shopping cart robot designed to support visually impaired individuals and proposes an obstacle avoidance method based on a 3D LiDAR‒driven artificial potential field. The method fixes the attractive force in the forward direction. It combines it in real-time with repulsive forces from obstacles to generate continuous, feasible direction vectors, which are then mapped to independent braking torques for the left and right servo brakes. This enables natural and intuitive path guidance. Experiments conducted in both simulated and real supermarket environments demonstrated that participants were able to safely pass through complex aisles, confirming the effectiveness of the proposed method.
Type
Publication
第26回計測自動制御学会システムインテグレーション部門講演会, 1D3-14
