DESIGN AND TESTING OF AN AUTONOMOUS GROUND ROBOT FOR AGRICULTURAL APPLICATIONS

Project Abstract

<p>&nbsp;                 <b>ABSTRACT&nbsp;</b></p><p>This senior project discusses the design and testing of an autonomous ground robot for agricultural applications such as strawberries. The vehicle will feature a robotic arm that will be programmed to perform various tasks, such as collecting soil and leaf samples of the crop or measuring soil moisture and salinity. Various components were chosen to be implemented on the vehicle due its power requirements and operating environment. Finite Element Analysis testing was done on the frame of the vehicle to ensure the adequacy of the design. <br></p>

Project Overview

<p>1.1 INTRODUCTION&nbsp;</p><p>Agriculture is a significant part of our past as human beings, and it will need to be a significant part of our future, especially with the growing world population. Technology and Agriculture will continuously be associated with each other, as both will have to work harder and smarter to feed our world. In more ways than one, California is representative of a growing world that will always rely on agriculture, but will also become increasingly reliant on technology due to the difficulties it can alleviate. California is known for being “America’s Salad Bowl” due to its large agricultural presence, as well as home to several of the largest technology companies in the world. That being said, it was inevitable that technology would become increasingly incorporated in agriculture here in California. Today, technologies such as drones, robots, GPS, and thermal and 3-D imaging are widely used in agriculture to gather and analyze data, with the goal of decreasing labor and increasing production. Specifically, the use of ground robots is becoming more widely used in agriculture to gather data across a large field area, sometimes autonomously. In this growing field, the greatest challenge is to “develop smarter machines that are intelligent enough to work in an unmodified or semi natural environment” (Design and development of the architecture of an agricultural mobile robot). In other words, these robots are usually not designed for a typical indoor, controlled environment, which poses exciting new challenges for engineers and designers. In Time-optimal guidance control for an agricultural robot with orientation constraints, the author says this well by explaining that, “Differently from the well-structured environment, the working environment of agricultural robots imposes varied constraints on the movements of the vehicles due to contact surface of loose soil and the specialties of crop cultivation features.” Robots such as these can be used to gather and analyze soil samples, use thermal or 3D imaging, apply nutrients or pesticides, as well as many other different tasks associated with agricultural data. However, these robots are sometimes limited to only one of these tasks, and for a hefty price tag, a farmer or field manager might only get one aspect of data out of it. These robots also sometimes require user operation, as some of them are not autonomous. This senior project is to bring all of these aspects together into one robot. A robot will be designed and built so that multiple sensors, cameras, and robotic arms can be added or removed from the robot, depending on the specific operation. This would make it possible to accomplish several different tasks that collect many kinds of data that a farmer or field manager would find useful. Designers that have created similar robots in the past have had to deal with the challenges associated with designing a robot that could simulate the intricacies and complexities in dealing with agricultural products. This fact sets apart agricultural robots from similar robots in different applications, as fruits, vegetables, and nuts are more delicate in handling and most processes still require a human eye to determine the state of the product. In Evaluation of a Strawberry Harvesting Robot in a Field Test, this is reinforced when the author states that “it is necessary to design an intelligent robot with human-like perceptive capabilities; for instance, the machine would need to calculate fruit position, assess maturity level and pick the fruit without damaging the pericarp.” This will likely displace some of the labor associated with managing a farm, either by replacing certain workers, or by simply adding to the work force by working at off-hours, because the robot can work 24/7. In addition to this, the robot would also be designed to be autonomous, making it more convenient for the user, without adding any unnecessary labor. Such a robot would prove useful in agriculture, as it would be able to give a good overall representation of different kinds of data in a field. The goal of this project is to design a prototype that utilizes the current technologies seen in similar agricultural robots in use today, while experimenting with new ideas in an attempt to keep the project innovative and original. <br></p>