Design and Implementation of a Voice-Controlled Human-Following Mobile Toolbox

When repairing equipment, going back and forth to find the maintenance tools we need exactly is so difficult and will waste too much time. To solve the problem, we designed a voice-controlled human-following mobile toolbox. The layer and grid structure allowed it to hold tools of different types and sizes. The wheeled structure made it mobile. With specific voice commands, the layers where corresponding tools were located would rotate to the front of the toolbox to facilitate users to take. We added the target following and obstacles avoidance functions for the toolbox to make it avoid obstacles automatically while following the red target. We also used the filtering and anomaly detection algorithms to make signals of the visual sensor more stable. To test the toolbox’s humanfollowing and obstacles avoidance performance, we tied a red marker around the shank, walked around in front of the toolbox freely and placed obstacles randomly. The results showed that the toolbox could avoid obstacles timely and realize the voice-controlled and human-following functions.