STTR Phase I: Weed Control Via Terradynamically Robust Robots

Ground Control Robotics, Inc., Atlanta GA

Investigators

Abstract

This Small Business Technology Transfer (STTR) Phase I project develops a robotic platform that can provide automated weed control throughout crop development stages. In recent years, weed control costs have been growing due to the rise of herbicide-resistant weeds and the increase in costs of agricultural labor. Additionally, increased demand for fruits and vegetables leaves these specialty crop farmers struggling to find options to increase productivity while keeping expenses manageable. Several companies offer automated weed control in vegetables; however, these large platforms struggle not to damage fruit in berry orchards. This project aims to develop swarms of robotic devices that can operate underneath the plant canopy to provide mechanical weed control for berries throughout the year without impacting plant growth. This technological development will enable domestic fruit production to meet the growing consumer demand and allow for less chemical use in fruit production, reducing herbicide-associated health risks to farm workers and consumers. Long term, these devices can augment weed control strategies in other crops and perform different agricultural tasks such as fungicide treatments and plant health monitoring, with the goal of automating agriculture to be more efficient and sustainable. This Small Business Technology Transfer project aims to develop rugged, low-to-the-ground, multi-legged robots that can locomote in various agricultural fields. This technology builds off of recent works demonstrating the effectiveness of centipedes and centipede-like robots when traveling over diverse terrains. When properly coordinated, these mechanically redundant legged systems demonstrate robust locomotion in complex terrain without the need for sensory feedback. This project will leverage this platform and perform systematic robot experimentation and theoretical modeling to develop coordination schemes for various maneuvers in agricultural terrain analogues. These strategies will then be implemented on a hardened robot to reliably locomote beneath the canopy in crop fields and identify weeds using an onboard camera and computer vision techniques. This device will make use of low-cost components and principles of mobility in complex environments to deliver guaranteed locomotion in these unpredictable terrains. Eventually, swarms of these devices will be deployed on various crop fields to provide autonomous weed management throughout the growing season, decreasing the costs of production for farmers and consumers. This project will result in a robust robotic platform that can provide cheap, reliable, all-terrain locomotion and such a device can extend beyond agriculture to address other U.S. sectors such as search-and-rescue and defense. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.