Session: 02-01-02: Product and Process Design

Paper Number: 145669

145669 - A Prescription Control Model for an Adaptive Sprayer System Implement 

With a growing imperative for increased sustainability and efficiency in farming practices, Precision Agriculture has emerged as a solution to this need. Precision Agriculture utilizes information technology, automation, drones, sensors and an IoT approach to reduce resource consumption while maintaining or increasing yield. Within Precision Agriculture there is a sub-realm of precision sprayer technologies. Current spray irrigation, and pesticide/insecticide techniques, albeit effective, are inefficient; it is susceptible to wastage and overspraying. This can potentially lead to environmental impacts and higher operating costs. Santa Clara University’s (SCU) Robotics System Laboratory (RSL) has created a novel precision spraying system which employs an adaptive prescription-based control scheme. This system is our first attachment for AG-Bot we have developed for our broader Ag-tech initiative where we have multiple field rovers capable of autonomous navigation. The AG-Bot with the sprayer attachment makes two significant contributions. Firstly, it can serve to remedy labor shortages in the agricultural industry. As the general population of the world increases, crop production needs increase, which in turn requires a larger workforce. Secondly, it can provide spraying capabilities in tough and difficult terrain and climates, alleviating the need for human-labor in harsh working conditions.

 The precision sprayer system consists of one reservoir, two pumps, four nozzles, four valves, and four flow sensors. A nozzle unit consists of a valve, flow sensor and the nozzle. A total of four nozzle units constitute the sprayer system. The system is segregated into two zones: an independently controlled left and right zone, two nozzles in each zone, which can administer the fluid as a volume per time rate or a total desired volume. The prescription-based control scheme relies on flow rate data as well as GPS data to administer varying chemical prescriptions across a given area; PID control with pre-mapped feedforward controller is employed. The underlying control scheme relies on a handshake between the sprayer system and the rover to query/control the rover velocities, target waypoints, locations and so on. Similar to the physical modular nature of the attachment, the control scheme is also designed to be a modular ROS based framework to accomplish the desired spraying needs in a field.

The system is capable of administering  between 80 - 400 liters/hour within 10% accuracy of the desired flow rate. The dispensing data is collected via flow meter sensors mounted in-line with the nozzles. The dispensing data and accuracy is verified via fluid collection and manual measurement in the field during several runs lasting a few hundred meters to study the effects of uneven terrains and other unknown field parameters.

Presenting Author: Katie Ott Santa Clara University

Presenting Author Biography: Masters Student in the dept. of Mechanical Engineering

Authors:

Largim Zhuta Santa Clara University
Katie Ott Santa Clara University
Manoj Sharma Santa Clara University
Christopher Kitts Santa Clara University