Bioscience Evidence 2024, Vol.14, No.6, 260-269 http://bioscipublisher.com/index.php/be 264 Figure 3 Example of WorldView-3 (WV-3) images acquired over the Coles corn field on 2 July 2018 (Adopted from Skakun et al., 2021) Image caption: True color composite red-green-blue (a); false color composite near infrared (NIR)-red-green (b); corn yields (c). ©2021 DigitalGlobe, Inc., a Maxar company, NextView License (Adopted from Skakun et al., 2021) 4.4 Yield improvements and economic benefits The adoption of precision agriculture techniques has resulted in significant yield improvements and economic benefits for the farm. Sensor-based nutrient and irrigation management have enhanced physiological performance and water productivity, leading to higher grain yields. For instance, the integration of sprinkler irrigation with precision nutrient management recorded a grain yield increase of up to 35.4% compared to conventional practices (Sachin et al., 2023a). Additionally, the use of remote sensing and machine learning for yield prediction has optimized resource allocation, further boosting economic returns (Eugenio et al., 2020; Ren et al., 2023). 4.5 Challenges and lessons learned Despite the benefits, the farm faced several challenges in implementing precision agriculture. These included the high initial costs of technology adoption, the need for technical expertise, and data management complexities. However, the long-term benefits, such as reduced temporal yield variation and increased yield stability, have outweighed these challenges. The farm has learned the importance of continuous monitoring and adaptation of precision agriculture practices to local conditions for sustained success (Yost et al., 2017; Monzon et al., 2018). In summary, the case study of the Midwest soybean farm demonstrates the transformative potential of precision agriculture in enhancing yield and economic viability while addressing environmental sustainability. 5 Data Analytics and Decision Support Systems 5.1 Role of big data in precision agriculture Big data plays a crucial role in precision agriculture by enabling the collection, analysis, and interpretation of vast amounts of data from various sources such as remote sensing, UAVs, and ground sensors. This data can be used to monitor crop health, predict yields, and optimize resource use. For instance, UAV-based multimodal data fusion using RGB, multispectral, and thermal sensors has been shown to improve soybean yield prediction accuracy significantly, demonstrating the adaptability of big data to spatial variations and its potential for high-throughput phenotyping and crop field management (Maimaitijiang et al., 2020; Ren et al., 2023). 5.2 Machine learning and AI for yield prediction Machine learning (ML) and artificial intelligence (AI) are pivotal in enhancing yield prediction models. Various ML techniques, such as Random Forest (RF), Support Vector Regression (SVR), and Deep Neural Networks (DNN), have been employed to predict soybean yields with high accuracy. For example, DNN-based models using UAV data have achieved an R² of 0.720 and a relative RMSE of 15.9%, indicating their robustness and adaptability across different soybean genotypes. Additionally, combining genotype information with UAV-based multi-sensor data using ML methods like Gaussian Process Regression (GPR) has further improved yield estimation accuracy (Maimaitijiang et al., 2020; Ren et al., 2023).
RkJQdWJsaXNoZXIy MjQ4ODYzMg==