Re-Sizable Quadrupedal Robot Platform: HARQ
This paper presents the development of kinematically adjustable quadrupedal robot platform, HARQ (Human Assistive and Robust Quadruped). Since 2019, ART (Assistive Robot Team) in University of Hartford has designed and built HARQ platform. The main objective of HARQ is to assist human labors in dangerous work environments such as radio-active areas which are destroyed by disasters.
To conduct a wide variety of tasks in such environments, robots first should be able to locomote over rough terrains and arrive the target area safely. However, the disaster areas are often not easily traversable due to high amount of obstacles and debris which are scattered by explosions. To reach task spaces, robots should pass through small cracks between obstacles and walk over scattered debris on the road. Considering the fact that such obstacles and debris have not-patternized shapes and sizes, fixed kinematic structure is not efficient design for robots which are expected to work in disasters. The net results is that robots should be able to optimize its kinematic configuration depending on the given road condition.
In previous studies, ART developed the full-sized humanoid, HART, which can adjust its limb size for driving different types of ground vehicles. Kinematic adaptability and low-cost manufacturing were main technical design requirements. In this study, they are continuously adopted to design and build the quadrupedal platform, HARQ, to enable the robot to navigate various working environments safely and adaptively.
First, the mechanical design of HARQ is described in this paper. As described above, since debris and obstacles in disasters do not have uniform shapes and sizes, the robot with the fixed structure cannot handle them efficiently. As such, kinematic adaptability is highly recommended to design component of quadrupeds for safe and efficient locomotion. To achieve the goal, each limb of HARQ’s is designed to be manually adjustable. The kinematic feature enable the robot to optimize its structure depending on the given road condition. Modular components are mainly adopted for most parts of in this study. It resulted in easy maintenance and streamlined part-list for low cost manufacturing.
Next, based on the mechanical specification of the built platform, kinematic analysis of HARQ is also implemented. The whole-body motions which enable HARQ to locomote over rough terrains are designed based on the kinematic equations. Then, the designed motions are executed thorough the actuator of the whole-body’s each joint in synchronous way using the unified real-time controller. The low-level controller was connected to the main computing system which provides the high-level computing, user-interface and sensor data communication.
Last, the built robot is tested and evaluated in human-centered environments which include both indoor and outdoor task spaces. Through the experimentation, HARQ demonstrated its kinematic adaptability to handle various types of terrains efficiently.
Re-Sizable Quadrupedal Robot Platform: HARQ
Category
Technical Paper Publication
Description
Session: 07-04-01 Design and Control of Robots, Mechanisms and Structures I
ASME Paper Number: IMECE2020-23105
Session Start Time: November 18, 2020, 12:05 PM
Presenting Author: Kiwon Sohn
Presenting Author Bio: Kiwon Sohn received B.S. and M.S. degree of electrical engineering from Kyungpook National University (Daegu, South Korea) and University of Pennsylvania (PA, USA) in 2005 and 2007. He received Ph.D. degree of mechanical engineering from Drexel University (PA, USA) in 2014. He served as a Chief of Engineering in team DRCHUBO@UNLV (the finalist of DARPA Robotics Challenge Finals) and DASL (Drones and Autonomous Systems Lab) in University of Nevada, Las Vegas, between 2014 and 2016. Currently, he is working in University of Hartford as an assistant professor. Prof. Sohn is a member of IEEE Robotics and Automation society and ASME.
Authors: Kiwon Sohn University of Hartford
Salman Hussain University of Hartford
Bradnan Matthew University of Hartford
Owen May University of Hartford