Quick Buzz Feed
East Carolina University researchers are using artificial intelligence to transform data collected in Hyde County canals and fields to provide real-time, usable information for managing water levels, salinity intrusion and other environmental conditions.
East Carolina University (ECU) researchers are pioneering the use of artificial intelligence (AI) and the Internet of Things (IoT) to provide actionable, real-time data to farmers in Hyde County, North Carolina. This initiative aims to assist in critical environmental management tasks such as maintaining optimal water levels and combating salinity intrusion in agricultural fields and canals. The project introduces the "Piton" platform, an innovative system designed to deploy low-cost sensors that collect vast amounts of environmental data. Initially, the analysis of this data was conducted manually by the research team, leading to delays in relaying vital information to farmers. This highlighted the need for a more efficient and automated system to deliver timely insights. The overall objective is to bridge the gap between complex scientific data and practical, on-farm decision-making, ensuring that farmers can react quickly to changing conditions and optimize their operations. This foundational effort is setting the stage for a new era of data-driven farming, moving academic research into direct, beneficial application for the agricultural community.
A pivotal advancement in the project came with the involvement of Colby Sawyer, an ECU graduate student. Sawyer, with his background in technology and a personal connection to farming as a grandson of a retired farmer, applied his expertise in large language models (LLMs) — the same type of AI used in advanced programs like ChatGPT and Google's Gemini — to the agricultural data. He custom-trained these LLMs to not only process and analyze the extensive data stream from the field sensors but also to interpret and present this information in a farmer-friendly format. This groundbreaking work transformed raw sensor data into practical, understandable insights, enabling the AI agent to communicate directly with farmers. Sawyer's contribution was crucial in automating the data analysis process, significantly reducing the lag between data collection and the delivery of useful advice, thereby making the technology truly impactful for daily farm management. His innovative application of AI is a testament to how academic research can be tailored to solve specific real-world problems.
One of the most significant real-world applications of this AI-driven system is in water level management for a Hyde County farmer. Traditionally, this farmer had to physically check canal water levels multiple times a day using a measuring stick, a time-consuming and often inefficient process. The ECU team installed affordable IoT sensors, costing approximately $150 each, directly into the canals. These sensors continuously transmit data to the AI agent. This intelligent system then analyzes the data and provides real-time updates and recommendations to the farmer via a mobile application. For instance, if water levels become excessively high, the farmer receives a text alert. The AI can then respond to queries, providing context-aware advice, such as confirming if the current level is problematic or if action can be delayed based on historical data and weather forecasts. This innovative approach has saved the farmer considerable time and effort, eliminating the need for frequent manual checks and enabling proactive water management, thereby preventing potential flooding and optimizing resource use. The ability to monitor critical parameters remotely offers immense flexibility and peace of mind to farmers.
The success in water level management spurred the expansion of AI applications to address other critical agricultural challenges. Inspired by direct feedback from their farm partner in Hyde County about the detrimental effects of saltwater intrusion on crops, the researchers began developing solutions to monitor this issue. They are now deploying sensors to measure salinity levels in both canal water and soil, aiming to understand the extent of saltwater leaching and its impact, also considering environmental factors like temperature and weather. This data, processed by the AI, will provide farmers with insights to mitigate damage from salinity. Furthermore, another project involves equipping grain bins with sensors to accurately measure stored grain volumes and, crucially, monitor internal temperatures. The goal is to eliminate the hazardous manual climb to check bin levels and prevent spontaneous combustion by detecting abnormal temperature rises, which can lead to significant losses. These diverse applications demonstrate the versatility of AI and IoT in tackling a range of practical problems in modern agriculture, moving beyond initial water management to comprehensive farm monitoring.
The ECU team is actively collaborating with RIoT (Regional Internet of Things), an incubator dedicated to rural entrepreneurship. This partnership focuses on integrating technology and AI into various farming and business ventures, exemplified by a facility in Wilson that supports experimental crop cultivation. Popoviciu and Sawyer are supplying this facility with sensors and instrumentation to monitor crucial environmental parameters like temperature, humidity, and irrigation in greenhouses and hoop houses. This monitoring helps growers assess the feasibility of new crops and alerts staff to critical changes, such as a recent incident where propane tanks unexpectedly ran out of fuel during a cold snap. The long-term vision extends beyond immediate problem-solving; the researchers hope to inspire the next generation of farmers, including the son of their Hyde County farm partner, to embrace technology in agriculture. By demonstrating how technology can streamline operations and solve complex challenges, they aim to bridge the generational gap and attract tech-savvy youth back to the family farm, ensuring the future sustainability and innovation of rural farming communities.
Despite significant progress and positive impact, the project faces financial constraints, with Popoviciu and Sawyer investing substantial personal time alongside their initial grant funding. The future funding model for the system remains uncertain, with possibilities ranging from county funding for community services to transitioning into an external, university-affiliated entity. Crucially, the system has been developed with cost-effectiveness in mind, making it relatively inexpensive to implement. This grassroots approach contrasts with other projects in wealthier regions that have received millions in grants but, according to Popoviciu, haven't achieved the same practical impact. The core mission is to make this advanced technological "revenue" accessible and easy to adopt for small and medium-sized farms, which typically lack the resources for high-cost solutions. This commitment to affordability and practical utility ensures that the benefits of smart agriculture can reach a broader spectrum of the farming community, democratizing access to cutting-edge AI and IoT solutions.