دانلود کتاب اینترنت اشیاء (IoT) و هوش مصنوعی (AI) در کشاورزی: ​​راهکارهایی برای توسعه فضای کشاورزی هوشمند جهت تولید محصول در مزرعه و محیط‌های سرپوشیده

The Tsukuba Conference (TC) and Tsukuba Global Science Week (TGSW) provide global platforms to address future policy challenges, develop solutions, and cultivate future leaders. This book is the outcome of our yearly efforts, stemming from the Tsukuba Conference 2025, aimed at addressing agricultural challenges, new developments to mitigate climate change, and increasing food production to meet global demands. This year, we collaborated on a special session, “AI and IoT to Integrate Automation in Farm Management: A Path to Sustainable Agriculture,” with the University of Tsukuba and the National Agricultural Research Organization (NARO). Investing in the agricultural sector can address not only hunger and malnutrition but also other challenges, including poverty, water and energy use, climate change, and unsustainable production and consumption. In Chap. 1, high-throughput plant phenotyping (HTPP) is discussed that was initially developed for model plants in controlled environments and that has recently transitioned to the more challenging context of field crop phenotyping. Despite the challenges posed by unstable environmental conditions, technological advancements have made efficient and precise field phenotyping feasible, enabling the assessment of crops in their actual growing environments using platforms such as satellites, unmanned aerial vehicles (UAVs), and ground-based vehicles (UGVs). High-throughput field phenotyping (HTFP) is vital for diverse and structurally complex horticultural crops. These crops are crucial for nutritional security but have received less attention in phenotyping research, resulting in a significant knowledge gap. This chapter reviews key HTFP technologies applied to selected horticultural crops, including advanced imaging sensors deployed on RGB, thermal, and hyperspectral platforms. In Chap. 2, the development of microcontroller-based semi-automated primary root and shoot length measurement system for tomato (Solanum lycopersicum) is discussed. Accurate quantification of root and shoot growth is essential for assessing plant development, but manual measurement methods are laborious and fail to capture the actual geometry of plant structures. This study presents a novel semi-automated system for nondestructive measurement of tomato (Solanum lycopersicum) root and shoot growth in an aeroponic setup. The low-cost system integrates a Raspberry Pi computer for automated image acquisition, custom software for user-guided annotation of root and shoot paths, and algorithms to extract length measurements from the annotated images. Chapter 3 discusses on the global demand for sustainable agriculture in driving innovations in precision farming technologies. Grapes are a high-value crop, and their cultivation is widespread worldwide. Vineyard irrigation and disease control should be improved to meet the growing demand for grape production in response to an increasing population. This research presents the development and implementation of a smart drip irrigation and fungicide application system designed for grapevine management, integrating Internet of Things (IoT) technologies, cloud-based monitoring, and automated control mechanisms. Chapter 4 discusses on mini-plant factories that are controlled by Environmentally Friendly Agriculture (EFA) systems, which are reported for small-scale indoor crop production. This chapter will critically review the research on using mini-plant factories for growing homegrown vegetables, herbs, and spices, covering the challenges of design, implementation, and technology integration, as well as the benefits. The chapter will also focus on developing a conceptual model to improve the boundary conditions and achieve more benefits with suitable crops well-suited to mini-plant factory cultivation. Future trends in mini-plant factory design, as well as the application of modern technologies, including AI, the Internet of Things (IoT), and automatic control, are reviewed in the chapter. These systems typically utilize hydroponics or aeroponics to cultivate plants without soil, often incorporating artificial lighting, microclimate controllers, and nutrient management systems. In Chap. 5, computational intelligence in climate-adaptive agriculture is highlighted with an extensive review approach. Intensifying climate shocks, ranging from unpredictable rainfall to extreme heat, are heightening the vulnerabilities of smallholder farmers, who dominate agricultural landscapes across the globe. Relying on traditional, labor-intensive practices, they are increasingly under-equipped to adapt to the evolving realities of climate change. Small-scale mechanization offers a promising solution, expanding beyond basic tools to include affordable, context-specific equipment integrated with digital technologies such as sensors, IoT, and solar energy systems. In Chap. 6, the agronomic performance and physiological responses of inbred rice using Drone Seeding Technology are discussed, and a study is presented. A Randomized Complete Blocked Design (RCBD) with a two-factorial structure was employed. The study aimed to evaluate the agronomic and physiological response of inbred rice under different drone seeding parameters through two experiments. Chapter 7 reported short strategic notes on deep learning strategies for smart agricultural spaces. This note highlights the strategic convergence of deep learning technologies across three critical domains: multimodal plant health diagnostics, autonomous precision field management, and intelligent product quality optimization. In Chap. 8, feature selection techniques for model development are highlighted. One of the most critical issues in developing the model to analyze real-world problems is the input variable selection problem. Determining the number of input variables the model requires to predict the target variable and gain insight into the research question is one of the most important considerations when developing a model, especially for an Edge-AI model. Chapter 9 outlined the development of remote autonomy of multi-cluster agricultural machinery. Considering the shortage of the agricultural workforce and the consequent rise in the need for integrated machinery, this chapter presents a study that endeavors to construct a remote autonomous system for multi-cluster agricultural machinery by implementing an IoT-based and AI-based system using the WebSocket protocol and the YOLO deep learning algorithm. Considering the difficulties that contemporary machinery encounters, including substantial operational expenses and arduous learning curves, this research aims to enhance operational efficiency to support labor shortages and increase productivity, with the ultimate goal of achieving autonomy and intelligent machinery. Chapter 10 discusses the development of a navigation system for the crawler tractor, designed for use on rural roads, to enhance productivity. Labor shortages in the agricultural sector have become a global concern, especially in countries with an aging population. The use of AI technology in agriculture not only addresses the workforce shortage but also increases efficiency by monitoring plant growth and diseases, as well as enabling autonomous transportation. In agriculture, autonomous tractors have been increasingly used, requiring advanced navigation systems to support various agricultural operations. Chapter 11 reported the development of an orchard autonomous system using Zigbee Communication and Visual SLAM Technology. The orchard management system encounters skilled labor shortages globally. Additional care, such as spraying at the tree location and continuous monitoring, is also required for the site-specific plantation in the orchard. Automation systems can be utilized in the agricultural sector to address labor shortages. Chapter 12 discusses deep learning-based instance segmentation algorithms for autonomous navigation in orchards. Autonomous navigation in orchards presents unique challenges due to the dynamic environmental conditions and various farming tasks, including monitoring, weeding, spraying, harvesting, and transporting goods. In field operations, GNSS-based navigation systems often experience signal interference caused by dense tree canopies, making them unreliable for precise local environmental perception. Chapter 13 highlights the intelligent recognition of orchard environment by instance segmentation of trees and roads. Orchard target detection is a key prerequisite for realizing agricultural automation. With the continuous development of agricultural science and technology, agricultural production is gradually moving toward intelligence and automation. As a vital part of the agricultural sector, the management and production of orchards must also utilize advanced technology to enhance efficiency and quality. Deep learning models utilizing instance segmentation, such as Mask R-CNN and YOLACT, can not only detect trees in the orchard but also segment the safe range for vehicles to operate, thereby realizing accurate identification and localization of the orchard environment, as well as guaranteeing vehicle safety during operation.