Design and Analysis of Booms for Wheeled Mobile Platform for Crop Phenotyping

Abstract

Crop phenotyping is frequently used by breeders and crop scientists to monitor the growth of plants and to relate them to genotypes of plants. Seemingly, this contributes to better crop growth and results in higher yield in solving food insecurity from growing world population. Instead of traditional crop monitoring, which is labor intensive, high-throughput phenotyping (HTP) using ground-based vehicle has several advantages over manual methods. Equipped with advanced sensors, the high-throughput phenotyping platforms quickly, accurately, and automatically, measure and record plant traits, such as appearance, height, and temperature. Although there have been many studies on plant phenotyping, there is still needs for ground-based HTP platform to perform accurate phenotyping on targeted crops (e.g. canola and wheat). Previous studies using ground-based HTP platforms focus primarily on leafy plants rather than densely cultivated crops. Besides, the previous platforms are designed for specific vehicles or sensors, and they are inappropriate for canola or wheat, which are targeted crops of this study. In this research, the main objective is to develop appropriate mechanical structures that are attached to different wheeled mobile platforms for HTP study. Using sensors attached to these mechanical booms, data are collected automatically for several traits such as height, temperature, greenness, and photos. These collected data are compared with manual measured data to evaluate the performance of the system, including suitability of mechanical structure. Three generations of the HTP platform are developed. The 1st and 2nd generation booms with simple structures use C-channel as the key component. While developing these booms, the stress, deformation, and vibration, are assessed with the finite element analysis (FEA). Meanwhile, it is necessary to understand the actual vibration pattern of these relatively long cantilever beams when attached to moving vehicles; however, previous research have little or limited investigation on vibrations influence on long booms in a farm setting. Thus, part of this research investigates how different factors, such as vehicle selection, vehicle speed, sensor locations, and road conditions, influence the boom attached to a farm machine, its vibration, and its effects on sensors performance for phenotyping. Then, an ideal operating conditions for HTP were obtained. The measurements from sensors confirm that the proposed mechanical structures and their ideal operating conditions are fulfilling the requirements for accurate sensor measurements. Finally, the 3rd generation boom/robotic arm featured of a hybrid structure is proposed and analyzed for its kinematics and dynamics suitability. Through the calculation and simulation, it shows that this robotic arm meets the requirements, including long-reach and high-payload capability, while maintaining a lightweight and relatively compact size after folding. Moreover, comparing results from path planning routines between Newton-Euler iterative method and simulations, it illustrates that they correlate well.