An Open-source Software-hardware Integration Scheme for Embodied Human Perception in Service Robotics

摘要

Perception of human beings is one of the basic capabilities of service robots and is the prerequisite for interaction between robots and humans. Although enabling hardware and software technologies have made great strides, there are not many open-source solutions that organically integrate the two. To address this shortfall, this paper introduces an open-source scheme of hardware and software integration for robotic embodied human perception. The embodied entity includes a robot chassis, a computing unit based on ARM architecture, a 3D lidar, a 2D lidar and four RGB-D cameras for robot exterior perception, a display panel for human-robot interaction, a set of LED lights to show the robot’s status and a sonar strip for low-level obstacle avoidance. The perception software is fully based on the Robot Operating System (ROS) which allows high modularity, fully deployed to the embodied entity and running at a rate of 30 Hz. The entire integration solution is very portable and publicly available at https://github.com/ Nedzhaken/human_aware_navigation.