Development of Back-Mounted Payloads for Additional Capability (PAC) for Unitree Go1 Quadruped

摘要

Popularity in Quadrupedal robotics has grown in recent years due to their increased availability and improved capability. These robots are able to perform a variety of tasks unmanned and autonomously such as transport and surveillance. The majority of this innovation, however, requires costly hardware or external hardware hindering the robot’s mobility. This research enhances the Unitree Go1 quadruped’s capabilities by integrating the RPLIDAR sensor and YOLO (You Only Look Once) object detection into modular Payloads for additional Capacity (PACs.). These auxiliary sensory payloads provided a low-cost and open-source mechanism to augment the Go1’s sensory capacity. Two back-mounted PACs were developed with one Raspberry Pi single-board computers integrated into each. Field testing indicated that the PACs added the desired functionality to the Go1 in both indoor and outdoor environments.