ARS Engineering Innovations Protect U.S. Cotton Quality

The United States is one of the world’s largest cotton producers, yielding millions of bales annually, according to the USDA’s Economic Research Service. And behind every bale is a story of science, innovation, and problem-solving. 

For over 60 years, engineers at the USDA’s Agricultural Research Service (ARS) have innovated and developed new technologies and methods to help the cotton industry reduce cotton contamination and minimize fiber waste, enabling U.S. cotton producers to maintain clean, high-quality, and profitable production.  

“Contamination in cotton lint significantly reduces its quality, posing a critical issue for ginners,” said John Wanjura, a research agricultural engineer at the ARS Cotton Production and Processing Research Unit in Lubbock, TX. “Contamination can occur at various stages throughout the process, from harvesting to gin processing. However, the primary source has been identified as the plastic wrap used with cotton modules from specialized harvesters. The  work from scientists and engineers at the ARS locations in Lubbock, TX, Stoneville, MS, and Las Cruces, NM provide cotton producers and ginners with much-needed technology and training that can help them minimize plastic contamination, thereby maintaining high quality cotton.”

Let’s dive into years of innovative ARS-engineered projects:

The Cotton Gin Stand Machine-Vision Inspection and Removal System for Plastic Contamination uses cost-effective color cameras to detect and remove plastic from cotton in real-time. This system automates the process, ensuring minimal plastic contamination from the module wraps. It captures multiple images of the cotton as it travels down the feeder apron just before entering the gin stand. When foreign materials are detected, an automatic air jet ejects them. 

Over the years, researchers have enhanced this technology by incorporating auto-calibration algorithms and developing a more cost-effective version known as Visual Inspection Single Node (VISN) for gins that do not require or wish to invest heavily in the full separation system.

VISN systems offer an efficient way to monitor and manage plastic contamination, ensuring the production of high-quality cotton. They have been successfully implemented in several gins across the U.S., proving to be invaluable in maintaining cotton cleanliness during processing.

The USDA Module Feeder Inspection System [for plastic contamination] employs a set of color video cameras to provide ginners with periodic views of the backside of the module feeder, where the cylinders disperse the cotton. The system’s software analyzes the video stream data and alerts workers when plastic contamination is detected, allowing them to take corrective action before the contamination proceeds into the ginning process.

The system includes network cameras to allow gins to monitor both the upper and lower cylinders and an updated algorithm that facilitates the automatic detection of plastic on the dispersing cylinders. This extension has been implemented in many gins across the U.S. as a training tool. It helps employees evaluate the effectiveness of their practices in cutting and unwrapping plastic from the modules. The video footage allows them to observe the impact of these practices on the likelihood of plastic contamination at the feeder. This training tool is used in the industry to teach best practices for properly cutting and unwrapping plastic from the modules before feeding them into the gin.

The Loader Attachment for Managing Round Modules is a tool designed for articulated wheel-loaders or telehandlers that frequently unload modules from transport trucks. Incorrect module positioning and improper wrap removal are major causes of contamination. This system uses sensors to detect the rotation position of each cylindrical module, allowing workers to adjust the modules and minimize the risk of plastic contaminating the cotton when the wrap is cut.

Recently, ARS researchers have been working on Artificial Intelligence (AI) Vision Models that can detect plastic contamination autonomously with minimal training. These systems distinguish actual plastic from irrelevant images like bugs, birds, or human arms and identify valid cotton images to improve performance as the cotton's appearance changes. 

These technologies have garnered recognition, including the VIPR Systems receiving various Federal Laboratory Consortium (FLC) Awards for Excellence in Technology Transfer, and many are now commercially manufactured. Ensuring commercial viability has been a key focus in the development and implementation of these innovations by engineers at ARS. – By Maribel Alonso, Office of Communications.