Recipient Organization
SOUTHERN ILLINOIS UNIV
(N/A)
CARBONDALE,IL 62901
Performing Department
(N/A)
Non Technical Summary
In the coming decades, soybean farming will confront substantial challenges due to anticipated impacts of climate change. Ongoing shifts in temperatures, precipitation patterns, and extreme events necessitate the implementation of adaptive strategies and innovative agricultural practices by soybean farmers. This proactive approach is imperative to strengthen the resilience and sustainability of soybean production. In the United States, particularly in the Midwest region, drought has emerged as a pivotal stress factor in soybean cultivation. Under severe drought conditions, soybean crops may experience yield losses of up to 40%. Therefore, the precise and prompt identification of drought characteristics in soybean plants holds paramount importance for the development of drought-resistant soybean varieties.Up to date, measuring soybean drought characteristics on a large-scale farm is challenging. First, large-scale soybean farms exhibit notable spatial variability in soil types, topography, and micro-climates. This diversity makes it hard to comprehensively understand drought conditions across the entire farm. Second, implementing accurate and extensive drought measurements requires substantial resources, including monitoring equipment and skilled personnel. Third, as drought conditions evolve over time, large-scale farms may experience temporal variations in drought intensity and duration across different sections. Fourth, the financial cost associated with deploying and maintaining a large-scale drought monitoring system can be prohibitive. Balancing the need for comprehensive data collection with the economic constraints of large farms adds an extra layer of difficulty. Last, analyzing and interpreting the vast amount of data collected from a large-scale soybean farm requires advanced computational tools and expertise. The fast extraction of meaningful insights from datasets is a bottleneck.This project aims to address the soybean drought challenge by developing advanced artificial intelligence (AI) technologies that can help farmers manage and mitigate the impacts of drought. This is important not only for the agricultural community but also for the broader society, as it supports food security, economic stability, and sustainable farming practices in the face of climate change. To achieve this, we will utilize unmanned aerial vehicles (UAVs) equipped with cameras to monitor soybean fields. The data collected will be processed using AI models to predict drought conditions and recommend optimal irrigation strategies. These models will be designed to be efficient and user-friendly, ensuring they can be easily adopted by farmers. We will also conduct educational seminars and workshops to train farmers and agricultural students on how to use these new technologies effectively. Through these educational activities, this research enables algricultural students and farmers to understand new AI technologies and encourage them to wholeheartedly embrace these technology innovations. The ultimate goal of this project is to enhance the resilience of soybean farming to drought, thereby increasing crop yields and economic returns for farmers. By providing timely and accurate information on drought conditions, we hope to empower farmers to make informed decisions that will benefit their crops and livelihoods.This project holds the potential to deliver significant societal benefits, including improved food security, better resource management, and enhanced sustainability in agriculture. By implementing this initiative, we can address key challenges such as climate change, which poses risks of crop failures and food shortages, thereby enhancing food security in the long term. The project focuses on promoting sustainable farming practices, which not only protect against the impacts of climate change but also lead to the development of climate-resilient crop varieties suitable for changing climate conditions. Furthermore, this project aims to achieve more efficient use of resources such as water, land, and energy. For instance, the adoption of precision irrigation systems can significantly reduce water wastage. These resource-efficient practices not only benefit the environment but also contribute to improved agricultural productivity and economic gains for farmers. Moreover, beyond the environmental advantages, this project has significant socioeconomic benefits. By increasing agricultural productivity, it enables farmers to generate higher incomes, thus helping to alleviate poverty and enhance living standards. Additionally, the project creates employment opportunities in the agricultural sector, particularly benefiting rural populations and contributing to overall economic growth and poverty reduction.
Animal Health Component
50%
Research Effort Categories
Basic
20%
Applied
50%
Developmental
30%
Goals / Objectives
Climate change is poised to elevate temperatures and the frequency of severe heat waves throughout US, posing a significant threat to soybean farming. Soybean plants, inherently susceptible to drought stress, face productivity reductions, thereby impacting the income of farmers. The intensification of drought stress necessitates more active cooling methods, such as water irrigation, yet these practices consume substantial amounts of electricity and water resources. To effectively mitigate heat stress, early detection of drought signs and characteristics is paramount to prevent long-term impacts. Traditionally, farmers rely on visual cues to identify drought stress in soybean plants, observing symptoms like leaf rolling, wilting, leaf cupping, and changes in plant color. However, this approach is less viable in large-scale farms where soybean areas are vast per farmer. Therefore, the overarching goal of this project is to enhance the resilience and sustainability of soybean farming against climate change through the development and application of cutting-edge AI technologies. This initiative aims to empower farmers with timely information and insights, allowing them to make informed decisions to effectively manage drought stress and improve crop yields. This research will disseminate research findings through educational seminars and workshops, equipping agricultural students and soybean farmers with the skills needed for effective drought stress management.Objective 1: AI-Driven Cost-Effective Drought Detection and Classification:Establish a regular unmanned aerial vehicle (UAV) monitoring system for soybean farms.Gather data from visible or multispectral cameras mounted on UAVs.Create an advanced AI model to process collected data into a holistic predictive framework to assess drought stress.Investigate AI model compression to reduce memory and computational demands for real-time solutions.Objbective 2: Timely Alerts and Recommendations:Develop an AI model to inform irrigation decisions using regular UAV images and weather forecasts.Provide instant notifications to farmers about the health status and irrigation needs of soybean farms.Offer agricultural guidance, including precision irrigation plans, moisture-retaining agents, and drought-resistant varieties.Objective 3: Educational Outreach and Engagement:Conduct educational seminars and workshops to disseminate research findings.Equip agricultural students and farmers with skills for effective drought stress management.Maintain continuous communication and seek feedback from farmers to refine and adapt the AI model for practical use.By achieving these goals and objectives, the project aims to introduce efficient, intelligent, and sustainable solutions for soybean agricultural production, ultimately safeguarding crop health and yield against climate-related challenges.
Project Methods
The project will employ a multifaceted approach, integrating various scientific methods and innovative techniques to achieve its objectives. The methods include:Data Collection:Weather Data and UAV Imagery: Weather forecast data and high-resolution images of soybean fields will be collected using UAVs. The images will capture drought severity and other relevant conditions.Manual Classification: Soybean plant images will be manually classified by soybean experts on a scale of 0 to 4 for wilt severity, serving as a reference for AI model training.AI Model Development:Baseline AI Models: Review and selection of top-performing AI models from existing literature to establish baseline models. These will be used for developing efficient AI models for edge computing devices.AI Model Compression: Techniques like weight quantization, quantization-aware training, and automated network pruning will be applied to create compact models suitable for deployment on edge devicesLSTM Models for Smart Irrigation: Development of LSTM models to optimize irrigation strategies based on drought severity assessments and weather forecasts. These models will predict soil moisture levels and recommend appropriate irrigation schedulesField Testing:Practical Trials: Field tests will be conducted at selected farms to evaluate the performance of AI models in real-world conditions. Data from these trials will inform further refinements and optimizationsData analysis will be conducted through the following steps:Preprocessing: Collected data, including weather records and UAV imagery, will undergo cleaning and normalization to ensure quality.Deep Learning Applications:Feature Extraction: Deep learning techniques will extract features and insights from the data.Time-Series Analysis: Trends in soil moisture levels and climatic conditions will be modeled using time-series analysis.Computer Vision: Convolutional neural networks and image segmentation will assess drought severity on soybean imagesSynthesis and Interpretation: Results will be synthesized and interpreted within the context of research objectives, guiding decisions on irrigation scheduling and other farming practices.