Design and development of a low-cost hybrid wheeled-leg for an agricultural robot : a thesis presented in partial fulfilment of the requirements for the degree of Master of Engineering in Mechatronics at Massey University, Manawatū Campus, Palmerson North, New Zealand

Thumbnail Image
Open Access Location
Journal Title
Journal ISSN
Volume Title
Massey University
The Author
Currently, New Zealand is financially dependent on its agricultural industry quite heavily. However, the agricultural sector faces several problems such as labour shortages, environmental issues and increasing costs. In other industries, robotics and automation have been used to combat these issues successfully. Yet, in agriculture, robotics and automation have only been adopted in horticulture but not in pastoral farming (dairy, sheep, and cattle). This is because the tasks and terrain in horticultural are well defined and structured, whereas, in pastoral farming, the terrain and tasks are unstructured and dynamic. The locomotion used by current horticulture robots is either not capable of operating in unstructured terrain or are inefficient. Therefore, pastoral farming will need to adopt new forms of locomotion in automation platforms. In this thesis, it is proposed that hybrid wheel-leg locomotion will enable robots to operate in unstructured and dynamic environments. With this in mind, a low-cost prototype hybrid wheeled leg has been designed and built. The leg has been designed to specifications which were developed based on the tasks that a multipurpose horticultural and pastoral farming robot is expected to do. A joint actuator is extremely influential towards the performance of any robotic leg. Due to the unstructured terrain, in which the leg will operate, it was concluded, that a mechanically compliant actuator is required. Because of the prohibitive cost of commercially available actuators, a prototype high torque, low-cost mechanically compliant actuator was designed and built to meet the specified torque requirements. This was in addition to the design and fabrication of the leg itself. Once the leg was assembled, the sensors, actuators and the motor were interfaced with ROS™ (Robot Operating System). ROS makes it easy to coherently control each leg's DOF (Degrees of Freedom) and makes it easy to combine and control multiple legs into a robot. Testing of the leg produced very encouraging results, but there were two issues with the performance of the actuator. The first issue is due to the poor implementation of the position control algorithm that came standard with the actuator motor driver. The problem can be resolved through software or the purchase of a different motor driver. The second issue is that the actuator only outputs 23 Nm of torque, but the motor used is rated at 50 Nm. This is due to the cheap drill motor used which is from an unknown brand; it is hoped that a more powerful drill motor from a well known reputable brand will be able to output its rated torque.
The following Figures are re-used with the publishers' permission: 9a, 11c, 13b, 14a, 16a, 19. These Figures are re-used with permission from IEEE: 10a ©2005 IEEE; 10b ©2008 IEEE; 11b ©2011 IEEE; 12a ©2010 IEEE; 13a ©2015 IEEE; 13c ©2010 IEEE; 14b ©2013 IEEE; 14c ©2010 IEEE; 15 & 22 ©2016 IEEE; 16b ©2017 IEEE; 18a, b &c ©2005 IEEE; 20a & b ©2011 IEEE; 21 ©2009 IEEE; 23 ©2016 IEEE. Other Figures are either in the public domain, or re-used under a Creative Commons license.