2.1. Role of Robotic and AI Systems in Horticultural Pest and Disease Management

Pests and diseases pose a significant threat to horticultural crops, resulting in plant damage, diminished yields, and overall crop health concerns [22]. Insects, mites, nematodes, and rodents are among the culprits that directly harm crops by feeding on plant tissues, leading to plant weakness and a decline in overall quality [23]. Meanwhile, fungal, bacterial, and viral diseases can cause issues such as leaf spots, wilting, and fruit rot, ultimately hampering plant growth and affecting the quality of the produce [24]. It is worth noting that some pests also serve as disease vectors, accelerating the spread of ailments. These challenges not only compromise plant defenses but also create opportunities for secondary infections, affecting crops both before and after harvest. The economic implications are significant, encompassing reduced income, higher production costs associated with pest and disease management, and the potential rejection of affected produce in the market [24, 25]. In response, farmers implement a range of strategies, including IPM, crop rotation, and the cultivation of disease-resistant crop varieties. Plant pathogens, the organisms responsible for plant diseases, have the capability to disrupt essential plant functions, weaken defense mechanisms, and cause structural damage. These pathogens can transmit diseases and may reside within or outside the host plant, with transmission often occurring through specific vectors [26, 27].

Robotic and AI systems play a vital role in minimizing horticultural crop losses caused by pests and diseases. Equipped with sensors and cameras, robotic systems continuously monitor crops, identifying early signs of pest infestations or disease symptoms. This early detection allows for timely action to prevent significant damage. AI algorithms analyze the data collected by the robotic systems, enabling precise treatment application to affected areas. By utilizing real-time data, these AI-driven robotic sprayers optimize dosage, reduce chemical usage, and effectively manage pests and diseases [28]. Moreover, AI algorithms integrate pest and disease data with other factors like weather conditions and plant health indicators, facilitating proactive decision-making. This integration enables farmers to implement IPM strategies that combine various methods and minimize reliance on chemicals. Additionally, AI-powered image recognition and ML models accurately identify diseases and pests based on visual symptoms. These systems aid in correct diagnosis and recommend suitable treatments, preventing further spread. Data-driven decision-making is another advantage of AI systems in pest and disease management. AI algorithms process extensive data, including historical pest and disease records, weather patterns, and crop health information. By analyzing these data, AI systems generate predictive models that anticipate outbreaks, empowering farmers to take preventive measures [28]. Robotic systems also contribute to the reduction of crop losses by autonomously monitoring crops for pests and diseases. They patrol fields or greenhouses, detect anomalies, and trigger alerts, facilitating timely intervention and preventing extensive damage. Furthermore, AI-powered analytics assess plant health indicators, identifying early signs of stress or diseases. This proactive approach allows farmers to promptly address issues, reducing the risk of crop damage. Through the adoption of robotic and AI systems, horticulturists can optimize pest and disease management, resulting in reduced crop losses, improved yields, and efficient resource utilization [28]. These technologies facilitate proactive and sustainable practices, ensuring the health and productivity of horticultural crops.

Opara et al. [29] highlight the transformative impact of AI and ML technologies in pest and disease prediction and detection. Their study illustrates how ML algorithms like convolutional neural networks and support vector machines (SVM), utilizing features such as gray-level co-occurrence matrix, vary in efficacy across diverse crop types and farming scales. This variability underscores the importance of tailoring AI solutions to specific agricultural contexts, enhancing accuracy and applicability from small-scale farms to large commercial operations. Deeper exploration of these specific ML algorithms not only enhances understanding but also showcases their potential to revolutionize agricultural practices by enabling precise, data-driven management strategies and sustainable crop production.

Hruday, Kiran, and Reddy [30], present groundbreaking findings on the application of advanced robotics and AI technologies in horticultural pest and disease management. Their study highlights the effective utilization of ML algorithms, specifically the single shot multibox detector (SSD) with MobileNet architecture (SSD-MobileNet), and image processing techniques enabled by OpenCV to detect and classify diverse plant diseases such as late blight and leaf curl viruses. This innovative approach integrates the Raspberry Pi as a central orchestrator, coordinating operations of servo motors, motor drivers, and a camera module for precise plant monitoring. Autonomous, targeted pesticide application is facilitated by high-capacity water pumps, underscoring the system’s capability to optimize treatment effectiveness while minimizing environmental impact. The integration of sophisticated navigation systems with ultrasonic sensors ensures seamless operation in farm environments, marking a transformative leap toward sustainable agriculture and improved crop health management.

For instance, Peixoto et al. [31] employed fuzzy systems (a type of AI system that can handle imprecise or uncertain data. They are based on the concept of fuzzy logic, which is a mathematical system that allows for partial truths) to create a comprehensive model for controlling soybean aphids by considering both biotic and abiotic factors. This model significantly improved the timing of predator release, enhancing the success of IPM. Similarly, Siraj and Arbaiy [32] introduced FuzzyXpest, a fuzzy expert system offering practical insights into insect pest activities in Malaysian rice fields, moving away from traditional mathematical models. In the realm of tea cultivation, Ghosh and Samanta [33] developed “TEAPEST,” an object-oriented expert system, to address the scarcity of human expertise. Boissard, Martin, and Moisan [34] utilized image processing and knowledge-based systems to automate the rapid detection of whiteflies on rose leaves. In disease diagnosis, Kolhe et al. [35] harnessed a new fuzzy logic approach, while Karmokar et al. [36] used artificial neural networks (ANNs) to develop a tea leaf disease recognizer. Furthermore, Hahn, Lopez, and Hernandez [37] employed spectral detection and neural networks to identify Rhizopus stolonifer spores on red tomatoes, addressing significant losses in tomato production due to this pathogen. These innovative approaches exemplify the power of technology and data-driven solutions in agricultural challenges [18].

Data privacy and security are crucial in using AI for pest and disease management. Privacy involves controlling how personal information is used, shared, and stored and is protected by laws and contracts. In agriculture, major concerns include what data AI systems collect, how it is used, and who it is shared with [38]. On the other hand, farmers often worry about losing control over their data. Around 78% of farmers fear companies will share or sell their data without consent. Ownership and privacy issues arise at all stages of data handling in agricultural AI. Farmers should have the right to collect, access, and control their own data, but some companies have been criticized for not providing adequate options for data access and collection. For example, the Federal Trade Commission investigated John Deere’s data practices, highlighting privacy violations [38]. Such cases emphasize the need for AI systems that respect farmers’ data privacy and give them control over their information. Addressing these concerns is essential to build trust and ensure AI technologies benefit agriculture sustainably.

2.2. Role of AI and Robotic Technologies in Mitigating Weather-Related Crop Losses

Weather events present considerable dangers to horticultural crop production manifesting in various forms. Extremes in temperature, whether in the form of heatwaves or sudden cold snaps, can harm crops by causing wilting, sunburn, reduced photosynthesis, frost damage, or even plant death. Drought conditions, characterized by insufficient rainfall or inadequate irrigation, induce water stress in plants, resulting in diminished yields, stunted growth, and heightened vulnerability to pests and diseases. Heavy rainfall and flooding have the potential to saturate the soil, leading to waterlogging, root rot, nutrient leaching, erosion, and the propagation of waterborne diseases [39]. Hailstorms physically damage crops by causing leaf bruising, stem breakage, or fruit puncturing, resulting in significant losses and rendering harvested produce unsuitable for sale. Strong winds and storms can flatten or uproot plants, making them more susceptible to diseases, reducing yields. Frost and freezing temperatures inflict tissue damage and impede plant growth, particularly during crucial growth stages. Hazy conditions and air pollution, such as smog or haze, obstruct sunlight, hindering photosynthesis and adversely affecting overall crop productivity. Additionally, harmful pollutants deposited on leaves compromise plant health [40]. To counteract crop losses stemming from these weather events, farmers must implement protective measures, leverage suitable technologies, and adopt resilient agricultural practices.

Extreme weather and climatic conditions pose significant risks for on-farm crop losses, often beyond the control of standard farming operations. Agricultural insurance serves as a vital financial safeguard for farmers in mitigating these risks. Insurance institutions perform field evaluations to estimate yield losses following crop damage from natural disasters, focusing on identifying “homogeneous damage zones” to extrapolate localized data for broader field loss estimates [29]. ML has introduced various applications aimed at predicting and detecting crop losses caused by natural events. These applications include methods for identifying hailstorm-damaged zones using sentinel images and damage evaluation data, systems for evaluating drought risk’s impact on crop production, and models for predicting frost damage distribution. Despite advancements in ML, predicting crop loss due to natural causes remains challenging due to the limited availability of comprehensive data. Improving our understanding and predictive capabilities of natural events affecting crop production is essential for ensuring global food security.

Through the utilization of AI algorithms, robotic weather stations and sensors gather up-to-the-minute weather data, empowering farmers to make well-informed decisions concerning irrigation, pest control, and crop protection. AI-powered systems evaluate weather data, soil moisture levels, and crop water requirements to optimize irrigation scheduling, effectively reducing water stress and the likelihood of waterlogging issues [41, 42]. Drones, equipped with sensors and cameras, capture detailed crop images that are then analyzed by AI algorithms. This analysis enables the early detection of signs of stress, diseases, or nutrient deficiencies, enabling timely intervention. By processing historical weather data, AI algorithms generate predictive models that assist farmers in forecasting weather-related risks and taking preemptive measures [41, 42]. AI-controlled greenhouses integrate weather monitoring and automated climate control systems, allowing for the adjustment of growing conditions based on real-time weather data. AI also aids in crop selection by examining weather patterns and soil conditions, facilitating the identification of resilient crop varieties. Furthermore, harvesting robots equipped with AI vision systems autonomously identify and harvest ripe produce, diminishing labor requirements and minimizing losses [41]. These technological advancements bolster resilience, decrease crop losses, optimize resource allocation, and enhance overall productivity in an environmentally sustainable manner.

Hence, the adoption of new technologies has become a critical factor in ensuring the sustainability of the agricultural sector and, by extension, food security in Sri Lanka. Similar to many other nations, Sri Lanka is currently in a relatively slow phase of adopting emerging technologies, including AI. Recent records indicate only a handful of research initiatives exploring the application of AI in agriculture. These include the utilization of ANN for predicting paddy yield based on climate data, the development of pedotransfer functions tailored to Sri Lankan soils, and the creation of an intelligent chat-bot aimed at addressing farming-related issues, as evidenced by the work of Amaratunga et al. [43], Gunarathna et al. [44], and Ekanayake and Saputhanthri [45], respectively.

AI and robotic technologies are instrumental in precision agriculture, offering advanced capabilities like weather data integration and predictive analytics to address weather-related crop losses. AI-driven systems leverage weather forecasts to optimize resource allocation and decision-making, ensuring timely interventions for crop health and yield optimization even in fluctuating environmental conditions. These technologies also excel in pest and disease management, with AI-powered systems identifying insects in real-time and guiding precise pesticide applications via drones and autonomous rovers [46]. However, despite this promise, practical hurdles and environmental considerations remain. The high initial cost of high-tech equipment and the specialized knowledge required for operation can significantly limit adoption, especially for smaller farms. Additionally, the energy consumption of certain technologies and the potential for increased electronic waste raise concerns about their environmental footprint.

To overcome these challenges, we can implement government incentives to make these technologies more accessible and develop training programs to bridge the technical knowledge gap. Furthermore, by prioritizing renewable energy sources and adopting modular designs for equipment, we can minimize the environmental impact of these powerful tools. Ultimately, by addressing these practical and environmental concerns, AI and robotics can unlock their full potential for building a future of resilient and sustainable agriculture.

2.3. Leveraging AI and Robotics for Nutrient Management in Horticultural Crop Loss Minimization

Nutrient imbalances can have a significant impact on horticultural crop loss, affecting plant growth, yield, and susceptibility to diseases and pests. Nutrient deficiencies, such as nitrogen, phosphorus, potassium, or micronutrients, disrupt the availability of essential elements for plant growth and development. This leads to stunted growth, chlorosis (yellowing) of leaves, reduced flowering and fruiting, and overall decreased productivity. Conversely, excessive levels of certain nutrients can be detrimental to plants. High nitrogen levels, for example, result in excessive vegetative growth but reduced fruiting and quality. Excess phosphorus or potassium can interfere with the uptake of other nutrients, leading to imbalances and deficiencies [47, 48].

Imbalanced nutrient ratios in the soil play a crucial role in plant nutrition. An imbalance, such as a high nitrogen-to-potassium ratio, negatively impacts plant health and can make plants more susceptible to diseases, reduce tolerance to environmental stress, and decrease overall crop yield. Nutrient imbalances also weaken plants’ immune systems, making them more vulnerable to pests, diseases, and environmental stresses. Calcium deficiency, for instance, increases susceptibility to blossom end rot in tomatoes or tip burn in lettuce. Imbalances in other nutrients like magnesium, iron, or zinc can impair plant health and increase disease susceptibility [49].

Furthermore, nutrient imbalances disrupt metabolic processes and photosynthesis in plants. This disruption leads to reduced energy production, impaired nutrient uptake and utilization, and decreased carbohydrate production. As a result, plants experience poor growth, smaller fruits, and lower yields. Nutrient imbalances can also affect the uptake and transport of other essential nutrients. Excess nitrogen, for instance, can inhibit the uptake of calcium or magnesium, leading to deficiencies of these nutrients even if they are present in the soil [48]. This disruption in nutrient uptake further exacerbates imbalances and negatively impacts crop health and yield.

To mitigate the impact of nutrient imbalances on horticultural crop loss, farmers must implement proper soil fertility management practices. This includes conducting regular soil testing and nutrient analysis to assess nutrient levels and imbalances. Based on the crop’s nutrient requirements, balanced fertilizer application is essential to provide adequate nutrients while avoiding excesses or deficiencies. Regular monitoring of plant health and nutrient status allows for timely corrective measures, such as foliar sprays or targeted nutrient applications, to address deficiencies or excesses [50]. By implementing these practices, farmers can promote healthy plant growth, optimize yield, and minimize crop losses associated with nutrient imbalances.

AI and robotics have become invaluable tools in minimizing horticultural crop losses due to nutrient imbalances. These technologies offer a range of capabilities. AI-powered sensors and robotic systems can be deployed in fields or greenhouses to continuously monitor crucial parameters such as soil nutrient levels, plant nutrient uptake, and nutrient deficiencies. The real-time data collected is then analyzed by AI algorithms to detect and identify nutrient imbalances and deficiencies in plants. Robotics guided by AI algorithms enable precise and targeted application of fertilizers according to the specific nutrient requirements of crops. By integrating data from nutrient monitoring systems and AI algorithms, robotic fertilizer applicators such as automated sprayers, fertigation robots, robotic spreaders, and drone-based fertilizer applicators can deliver the appropriate amount and composition of nutrients directly to the plants or specific areas in the field. This optimized approach enhances nutrient uptake while minimizing waste [28]. AI-based decision support systems (DSS) offer farmers valuable insights for informed nutrient management decisions. These systems integrate data from diverse sources, including soil nutrient analysis, weather conditions, and plant health indicators, to provide recommendations for optimal nutrient application timing, dosage, and composition. This empowers farmers to address nutrient imbalances effectively. AI algorithms leverage extensive historical data, encompassing nutrient levels, crop performance, and environmental factors, to create predictive models. These models can forecast potential nutrient imbalances and their potential impact on crop health and productivity. By proactively identifying and addressing nutrient imbalances, farmers can prevent crop losses and optimize nutrient management practices [51]. Robotics and AI are instrumental in automating nutrient delivery systems, such as hydroponics or fertigation setups. These technologies ensure precise and controlled nutrient supply to plants, facilitating optimal nutrient uptake and minimizing imbalances. Robotics and AI systems enable remote monitoring and control of nutrient delivery systems, granting farmers the ability to adjust nutrient applications in real-time based on data-driven insights. This remote capability ensures timely and accurate nutrient management, even in large-scale horticultural operations. By harnessing the power of AI and robotics in nutrient management, horticulturists can effectively minimize crop losses arising from nutrient imbalances. These technologies enable precise nutrient monitoring, targeted fertilization, data-driven decision-making, and proactive management of nutrient levels [52]. Ultimately, they contribute to optimizing nutrient uptake, enhancing crop health, maximizing yields, and reducing the adverse impacts of nutrient imbalances on horticultural crops.

An innovative mathematical model, complemented by AI and ML, has been developed to process data obtained from the specialized soil sensor known as “chemPEGS” (chemically functionalized paper-based electrical gas sensor). This sensor serves the purpose of measuring ammonium levels in soil, a fundamental chemical that undergoes conversion into nitrites (NO²) and nitrates (NO³) due to the activity of soil bacteria. The model effectively harnesses a range of data, including information on weather conditions, timing for fertilization, soil acidity, and soil conductivity. Through this data analysis, it can accurately predict nitrogen levels for the upcoming 12 days. This prediction serves as a valuable tool in identifying the ideal timing for fertilization [53].

AI and robotics offer exciting possibilities for nutrient management and minimizing losses in horticultural crops, but economic feasibility remains a hurdle for small-scale farmers. High upfront costs for robots and sensors can be a significant burden, while the return on investment depends heavily on factors like farm size and potential yield increases. Technological limitations also exist. Existing robots may not be suitable for smaller farms with irregular fields or diverse crops, and AI algorithms trained on specific conditions might struggle to adapt to the varied environments small farms often face. Additionally, reliable Internet connectivity, crucial for data transmission in AI-powered systems, can be scarce in rural areas. Reliability concerns further complicate matters. Operating and maintaining complex robots require specialized knowledge, and system breakdowns can disrupt critical nutrient management processes. Data security is another potential concern. However, the future is not bleak. Government incentives, modular and scalable robotic designs, cloud-based AI platforms with offline capabilities, and collaborative service models can all help bridge the gap [54, 55]. By addressing these challenges and fostering collaboration, AI and robotics have the potential to revolutionize nutrient management in horticulture, empowering small farmers to optimize resources, minimize losses, and build a more sustainable agricultural future.

2.4. Harnessing AI and Robotics for Efficient Irrigation Management in Horticultural Crop Loss Minimization

Improper irrigation practices can have a significant impact on horticultural crop loss. Underirrigation or insufficient water supply can lead to water stress, resulting in reduced growth, wilting of leaves, flower abortion, and decreased fruit or vegetable production. On the other hand, overirrigation or poor drainage can cause waterlogging, leading to root damage, root rot, and nutrient imbalances. Waterlogged conditions also create favorable environments for pathogen growth, increasing the risk of disease outbreaks. Improper irrigation techniques can result in uneven water distribution across the field, causing areas of excessive water and waterlogged conditions, as well as areas of insufficient water leading to water stress. This inconsistency leads to uneven growth, reduced yields, and overall crop inconsistency [56]. Moreover, excessive or poorly timed irrigation can lead to nutrient leaching, where essential nutrients are washed away from the soil, resulting in nutrient deficiencies and reduced crop health. Improper irrigation practices, especially overhead sprinkler irrigation, can also promote the spread of diseases by creating wet foliage and high humidity. This increases the risk of diseases such as leaf spots, blights, and molds, which can damage leaves, stems, and fruits. Additionally, inadequate water supply to crops can create favorable conditions for weed growth, leading to competition for resources and reduced crop growth, yield, and quality. To mitigate the impact of improper irrigation, horticulturists should implement proper irrigation management practices. This includes regular monitoring of soil moisture levels, using appropriate irrigation methods and equipment, optimizing irrigation scheduling based on crop needs, and ensuring proper drainage. Precision irrigation technologies, such as drip irrigation or sensor-based irrigation systems, can improve water efficiency, reducing water stress, waterlogging, and nutrient leaching [57]. By adopting proper irrigation practices, horticulturists can promote healthy plant growth, optimize water and nutrient use, and minimize crop losses caused by improper irrigation.

The integration of AI and robotics has revolutionized the agricultural landscape, offering invaluable solutions to address the challenges of improper irrigation practices and minimize crop loss in horticulture. These advanced technologies offer a range of capabilities that revolutionize irrigation management. By employing AI algorithms to analyze real-time data from sensors measuring soil moisture levels, weather conditions, and plant water requirements, robotic systems can autonomously determine precise irrigation needs. This enables the delivery of the appropriate amount of water to crops, ensuring optimal irrigation and minimizing the risk of water stress or waterlogging. The integration of AI and robotics facilitates remote monitoring and control of irrigation systems. Farmers can access real-time data on soil moisture levels, irrigation flow rates, and system performance, empowering them to make informed decisions and adjust irrigation schedules or settings as necessary. This remote capability enhances efficiency and allows for timely intervention, even in large-scale horticultural operations [58]. Robotics guided by AI algorithms enable precise and targeted irrigation to specific areas or individual plants. These technologies ensure accurate water delivery to where it is most needed, minimizing water wastage and optimizing water use efficiency. By preventing overirrigation or underirrigation, precision irrigation mitigates the adverse effects of improper water supply on crop health and productivity. AI algorithms leverage historical data on irrigation practices, crop performance, soil characteristics, and weather patterns to develop predictive models. These models forecast optimal irrigation schedules, identify potential risks of water stress or waterlogging, and guide farmers in making data-driven decisions to prevent crop losses resulting from improper irrigation practices. AI-powered systems integrate data from various sources, such as soil moisture sensors, weather forecasts, and crop water requirements, to optimize water management strategies. By considering factors like evapotranspiration rates and plant growth stages, these technologies enable farmers to schedule irrigation effectively, adjust irrigation volumes, and avoid excessive water use or water stress [59]. Robotics can be employed to automate drainage systems, ensuring proper water flow and preventing waterlogging in fields or greenhouses. AI algorithms monitor drainage performance, detect potential blockages or inefficiencies, and trigger automated actions to maintain optimal drainage conditions. This contributes to improved water management and reduces the risk of crop loss due to waterlogging. With the power of AI and robotics in irrigation management, horticulturists can significantly minimize crop losses caused by improper irrigation practices. These technologies enable precise and targeted water delivery, optimize irrigation scheduling, improve water use efficiency, and provide real-time monitoring and control [58]. Ultimately, they contribute to sustainable water management, healthier plant growth, and increased productivity in horticultural crops.

Munyaradzi et al. [60] designed and tested wirelessly connected soil sensors for automated irrigation control in winter wheat fields at the University of Zimbabwe’s Research and Development station over three winter seasons (2016, 2017, and 2018). The researchers implemented a drip irrigation system with various treatments, including farmer practice (control), tensiometer-based, weather data-based, and wireless sensor-based (WCSS) methods. The system employed locally built, cost-effective soil moisture sensors connected to wireless nodes and a base station for real-time irrigation scheduling. Additionally, AI algorithms were used to analyze soil moisture data and optimize irrigation schedules, while robotic applications, such as automated valves, were employed to control water delivery precisely. The study aimed to improve irrigation water use efficiency and crop yield by comparing these treatments, demonstrating the effectiveness of the WCSS, AI, and robotic applications in saving water while maintaining or enhancing wheat yield.

Athani et al. [61] contributed to this field by designing an IoT-based soil moisture monitoring system, employing Arduino sensors to facilitate optimal soil and irrigation management for pasture growers in North Karnataka, India. This system leverages data from input sensors, managed through neural network algorithms and correction factors to monitor soil conditions. Notably, the use of affordable, readily available components has significantly reduced both the manufacturing and maintenance costs, rendering the technology accessible to small-scale farmers in rural areas.

Arif, Mizoguchi, and Setiawan [62] also contributed to irrigation management, focusing on paddy fields. They proposed an ANN model for estimating soil moisture levels in such fields, even when meteorological data are limited. Their model relies on reference values of evapotranspiration and precipitation to make accurate predictions. Furthermore, in the realm of paddy cultivation, land leveling stands out as a critical operation that enhances water coverage uniformity and, subsequently, irrigation efficiency. Si et al. [63] developed a fuzzy-based laser leveling system for precisely this purpose, employing components like a laser transmitter, receiver, controller, hydraulic system, and a bucket. The fuzzy logic system embedded in the controller enables the machine to gauge the bucket’s position, crucial for adjusting the field’s height and the leveler’s overall functioning.

De Benedetto et al. [64] implemented a time domain reflectometry (TDR)-based system for extensive and continuous monitoring of soil water content in agricultural contexts using elongate sensing elements (SEs) placed along cultivation rows. This method involves embedding SEs, such as coaxial cables, in the soil, where they interact with the electromagnetic pulses emitted by TDR instruments. The system’s effectiveness was validated through experiments in tree and open-field plant cultivations, demonstrating its ability to accurately detect variations in soil moisture content before and after irrigation. This real-time data facilitate precise irrigation management by automatically adjusting water delivery based on current soil conditions, thereby optimizing agricultural practices [64]. Integration with AI and robotic technologies could further enhance operational efficiencies by enabling autonomous decision-making processes that respond dynamically to environmental conditions, enhancing crop yield while conserving water resources. The scalable design of the system allows for flexibility in deployment, catering to diverse agricultural landscapes and cultivation requirements.

The assessment of plant water status is a key consideration in modern agriculture to create effective irrigation schedules. Valdes-Vela et al. [65] introduced a soft computing technique to construct a model capable of estimating plant stem water potential. This model uses soil water content and meteorological data, making use of fuzzy rules to improve approximation accuracy. Additionally, characterizing and assessing contaminated soils is a substantial challenge for soil scientists worldwide. Lopez et al. [66] tackled this issue by developing a classification tool based on fuzzy logic. This tool falls into the category of AI-based DSS. The application of fuzzy set formalism within this system enhances precision by reducing deviations and data variance compared to conventional computer-based models.

AI and robotics for efficient irrigation management in horticultural crop loss minimization require not only technological innovation but also effective adoption and utilization by farmers. AI algorithms and robotic systems offer promising solutions for optimizing irrigation schedules and minimizing crop losses due to water stress. However, to fully leverage these advancements, farmers need proper training and support to operate and maintain these complex systems. Training programs should focus on teaching farmers how to interpret AI-generated recommendations, operate robotic irrigation systems, and troubleshoot common issues [54]. Moreover, ongoing technical assistance and access to reliable support are crucial to ensure sustained adoption and effective use of AI and robotic technologies in agricultural practices. This approach not only enhances water use efficiency but also empowers farmers to adapt to changing environmental conditions and improve overall crop productivity.

2.5. Leveraging AI and Robotics for Effective Weed Management in Horticultural Crop Loss Minimization

Weed competition poses a significant threat to horticultural crops, leading to crop losses and reduced yields. One of the primary ways in which weeds impact crops is through resource competition. Weeds aggressively compete with crops for essential resources like water, nutrients, sunlight, and space. They can quickly establish a dense canopy, shading the crops and limiting their access to sunlight. As a result, photosynthesis is reduced, crop growth is slowed down, and plants become weaker, leading to lower yields [67]. Furthermore, weeds have the ability to deplete nutrients from the soil. They outcompete crops for available nutrients, causing nutrient deficiencies in the soil and reducing the nutrient uptake by crops. This nutrient competition hampers crop growth and development, impacting overall productivity. In situations where water availability is limited, weeds with extensive root systems can also compete with crops for water. They extract significant amounts of water from the soil, leaving the crops vulnerable to water stress. This can result in reduced crop growth, wilting, and decreased fruit or vegetable production [68].

Some weed species possess allelopathic properties, releasing chemicals into the soil that inhibit the growth of neighboring crops. This allelopathy phenomenon negatively affects crop germination, root development, and overall crop health. The presence of allelopathic weeds can lead to stunted growth, yellowing leaves, and reduced crop yields. Weeds can also interfere with important cultural practices such as tillage, weed control, and harvesting. Dense weed populations impede mechanical operations, making it challenging to effectively manage weeds or access crops for harvesting. This obstruction reduces the efficiency of weed control measures, leading to increased weed competition and subsequent crop loss [67]. Moreover, weeds can serve as hosts for pests and diseases, acting as a refuge and breeding ground for harmful organisms. They harbor pests that can migrate to crops, increasing the risk of infestation. Weeds also facilitate the spread of diseases to nearby crops, resulting in reduced crop health and potential yield losses.

To mitigate the impact of weed competition on horticultural crop loss, farmers employ a variety of weed control strategies. Cultural practices such as crop rotation, mulching, and intercropping are implemented to suppress weed growth and minimize competition. Mechanical methods like hand weeding and cultivation are used to physically remove weeds from the field. Herbicides may be judiciously applied to manage weed populations, taking care to minimize their impact on crops. Integrated weed management approaches that combine multiple strategies are often adopted to effectively control weeds and reduce their competitive effects on horticultural crops [68]. By implementing these strategies, farmers can minimize crop losses and optimize the productivity of their horticultural crops.

AI and robotics provide efficient remedies for reducing crop loss in horticulture resulting from weed competition. These technologies provide a range of capabilities that enhance weed management practices. Firstly, AI algorithms coupled with image recognition and ML enable real-time weed detection and classification. Robotic systems equipped with cameras and sensors autonomously navigate fields, scanning and identifying weeds. These data empower farmers with accurate information on weed presence, density, and distribution, enabling targeted weed management strategies. Secondly, robotics guided by AI algorithms facilitate precision weed control by selectively applying herbicides, mechanical methods, or other weed management techniques directly to the weeds while minimizing the impact on surrounding crops. This targeted approach allows farmers to effectively manage weed populations and reduce competition without harming their crops. Thirdly, robotics can autonomously perform weeding tasks, such as mechanical removal or the application of targeted herbicides [69]. AI algorithms guide the robots to efficiently identify and remove weeds, reducing the need for manual labor and improving the overall effectiveness of weed control. This automation enhances efficiency, saves time, and ensures consistent weed management practices. Additionally, AI-powered DSS integrate data from various sources, including weed monitoring, soil conditions, crop growth stages, and weather forecasts. These systems provide real-time insights and recommendations for optimal weed control strategies. By considering multiple factors, such as weed species, growth stage, and the effectiveness of control methods, farmers can make informed decisions and implement the most suitable weed management practices. Furthermore, AI algorithms analyze extensive historical and real-time data to develop predictive models for weed growth and spread. These models help farmers anticipate weed outbreaks, plan weed control measures in advance, and optimize timing and effectiveness of weed management interventions. By proactively managing weeds, farmers can reduce crop losses and ensure timely weed control actions. In conclusion, AI and robotics play a vital role in weed management, enabling precise weed detection, targeted control, automated weeding, data-driven decision-making, and proactive weed management strategies [70]. These technologies empower farmers to effectively suppress weed growth, reduce competition with crops, and maximize the productivity and quality of horticultural crops.

Tobal and Mokthar [71] pioneered an AI-assisted image processing method for the successful identification of weeds in agricultural fields. Their system introduced an evolutionary ANN to the domain. By optimizing the neural parameters using a genetic algorithm, their system efficiently reduced classification training time and operational errors. Over the past two decades, numerous similar AI-assisted weed identification and classification systems have emerged, with the majority utilizing ANN technology [72, 73]. Unmanned aerial vehicles (UAVs) offer significant potential for weed monitoring. Perez-Ortiz et al. [74] introduced an innovative approach to weed management, combining UAVs, image processing, and ML. UAVs provide exceptionally high spatial resolution, a notable contrast to the limited temporal and spatial resolutions of conventional platforms like piloted planes and satellites. Consequently, this technology holds great promise, particularly in the realm of early postemergence weed control.

AI and robotics offer promising solutions for effective weed management in horticultural crop loss minimization, but integrating these technologies into existing farming practices poses significant challenges. One major hurdle is the initial cost of adopting AI and robotic systems, which includes procurement, installation, and maintenance expenses. These technologies often require specialized infrastructure and skilled personnel for operation and upkeep, adding to the financial burden for farmers. Furthermore, integrating AI algorithms with robotic platforms to accurately identify and selectively manage weeds in varying field conditions demands robust technical expertise and ongoing software updates. Despite these challenges, the potential benefits of reduced herbicide use, labor efficiency, and improved crop yield underscore the importance of overcoming these barriers to achieve sustainable and effective weed management strategies in agriculture [75]. Despite these challenges, the potential long-term benefits in terms of efficiency, sustainability, and yield optimization underscore the importance of overcoming these barriers to realize a more efficient and environmentally friendly future for weed management in agriculture.

2.6. Harnessing AI and Robotics for Enhanced Planning and Monitoring in Horticultural Crop Loss Minimization

Insufficient planning and monitoring practices can result in substantial losses of horticultural crops. Without proper planning, farmers may choose crops that are not well-suited to the local climate, soil conditions, or market demand. This can result in poor crop performance, lower yields, and difficulty in selling the produce, leading to financial losses. Insufficient planning and monitoring can lead to improper planting practices, such as incorrect spacing, inadequate soil preparation, or improper depth of planting. These factors can negatively impact crop establishment, root development, and overall plant growth, resulting in reduced yields and poor crop quality. Without proactive planning and monitoring, farmers may fail to implement effective pest and disease management strategies. This can result in uncontrolled pest outbreaks, leading to crop damage and reduced yields. Lack of monitoring also means that early signs of pests or diseases may go unnoticed, allowing them to escalate and cause more significant losses. Without proper planning and monitoring, farmers may struggle to optimize irrigation and nutrient management practices. This can result in under- or overirrigation, leading to water stress or waterlogging, and nutrient deficiencies or excesses, negatively impacting crop growth, yield, and quality [76]. Inadequate planning and monitoring can result in poor weed control practices. Failure to implement timely and effective weed management strategies can lead to increased weed competition, resource competition, and reduced crop yields. Without proper planning and monitoring, farmers may miss the optimal harvest window, resulting in overripe or underripe produce. Delayed harvesting can lead to quality deterioration, reduced market value, and potential crop losses. To mitigate the impact of the lack of planning and monitoring on horticultural crop loss, it is essential for farmers to adopt systematic planning and monitoring practices. This includes proper crop selection based on local conditions and market demand, implementation of good agricultural practices, regular monitoring of crop health, timely pest and disease management, efficient irrigation and nutrient management, effective weed control strategies, and timely harvesting. By incorporating these practices, farmers can minimize crop losses, optimize yields, and ensure the overall success and profitability of horticultural crops [77].

AI and robotics play a crucial role in minimizing horticultural crop loss caused by a lack of planning and monitoring. AI algorithms analyze various data sources, including climate data, soil conditions, and market trends, to provide farmers with insights for informed decision-making. By integrating these data, farmers can make better choices regarding crop selection, planting practices, irrigation schedules, pest and disease management, and harvesting timing. AI-powered DSS assist farmers in optimizing their planning and monitoring processes, reducing the risks associated with inadequate planning. Robotics equipped with remote sensing technologies, such as drones or satellite imagery, can monitor crops and collect data on plant health, growth patterns, and stress indicators. AI algorithms analyze these data, providing real-time information on crop performance and alerting farmers to potential issues. Remote monitoring enables farmers to detect problems early, such as nutrient deficiencies, pest outbreaks, or water stress, allowing for timely intervention and adjustment of management practices. Robotics guided by AI algorithms can automate various tasks in horticultural operations, ensuring precision and efficiency [78]. These technologies reduce human error and ensure optimal resource allocation, mitigating the negative effects of a lack of planning and monitoring.

ML applications have transformed various aspects of horticultural production and retail operations. ML models such as support vector regression, linear regression, and random forest are effectively used to predict crop yields under varying environmental conditions. Studies by various researchers, including Opara et al. [29], demonstrate how these models consider factors like weather patterns and soil properties to optimize harvest timing and minimize storage losses. Additionally, ML technologies play a crucial role in assessing crop quality, employing algorithms such as SVM, linear discriminant analysis, and ensemble methods to classify produce based on attributes like color and texture, ensuring adherence to market standards and reducing postharvest wastage. In postharvest handling, ML systems automate sorting and grading processes, improving efficiency and accuracy compared to traditional methods. Research in retail operations highlights ML’s role in predictive analytics for inventory management and consumer behavior, enhancing decision-making to reduce waste and maximize profitability. Overall, the integration of ML in horticulture and retail offers promising advancements in operational efficiency, sustainability, and economic performance throughout the supply chain.

Elaborating on the accuracy and reliability of predictive models used for planning and monitoring in horticultural crop loss minimization underscores their critical role in modern agriculture. AI algorithms are adept at analyzing extensive datasets encompassing weather patterns, soil conditions, and crop health indicators. These models deliver highly precise forecasts and actionable insights, facilitating optimized resource allocation, anticipation of market demands, and effective risk mitigation. Feedback from farmers who have integrated AI-driven systems into their operations provides practical insights [79]. AI predictive models empower farmers to make informed decisions about planting schedules and crop management practices based on real-time data analysis, thereby enhancing crop yields and minimizing losses by predicting optimal storage conditions and shelf life.

Additionally, the integration of AI-driven robotics, such as autonomous harvesters and precision spraying systems, enhances operational efficiency and reduces labor costs. Farmers value the precision and reliability of these technologies, ensuring tasks are performed with minimal environmental impact and maximum productivity [79]. These applications underscore AI and robotics’ transformative potential in modern agriculture, optimizing yields and promoting sustainable farming practices crucial for global food security.

2.7. Leveraging AI and Robotics for Enhanced Postharvest Management in Horticultural Crop Loss Minimization

Postharvest losses have a significant impact on horticultural crop loss. Horticultural crops, including fruits, vegetables, and ornamental plants, are highly perishable and require careful handling and storage to maintain their quality and value. However, inadequate postharvest management practices can lead to substantial losses throughout the supply chain, from the farm to the consumer [80]. Quality deterioration is one of the key consequences of postharvest losses. Improper handling, storage conditions, and delays in postharvest processes such as cleaning, grading, and packaging can result in the deterioration of horticultural crops’ quality [77]. This can include physical damage, decay, wilting, discoloration, and loss of flavor and nutritional value. Such quality losses not only reduce the market value of the crops but can also render them unsuitable for sale or consumption.

Quantity losses are another significant effect of inadequate postharvest management practices. For instance, poor harvesting techniques, such as improper timing or incorrect methods, can lead to damage to the crops and loss of yield. Similarly, ineffective storage facilities or transportation methods can result in increased spoilage, reducing the quantity of marketable produce [77]. The economic impact of postharvest losses is substantial, affecting both farmers and the entire supply chain. Farmers bear the direct costs of lost production and wasted inputs such as seeds, fertilizers, and labor. Additionally, reduced crop quality and quantity can result in lower prices received for the produce, impacting the income and livelihoods of farmers [28]. The losses also affect traders, processors, and retailers, leading to financial losses throughout the horticultural value chain.

Postharvest losses also undermine food security and nutrition. Horticultural crops play a significant role in global food security and nutrition as they provide essential vitamins, minerals, and dietary diversity. However, when crops are lost or wasted due to inadequate postharvest management, it not only reduces the availability of nutritious food but also wastes the resources invested in their production, such as land, water, and energy [81]. Furthermore, postharvest losses have environmental consequences. When horticultural crops are lost or wasted, the resources used in their production, such as water, energy, and fertilizers, are also wasted. Improper disposal of spoiled or unsold produce can further contribute to environmental pollution if not managed properly [28, 81].

Addressing postharvest losses requires the implementation of various measures. These include improving infrastructure, providing access to appropriate storage facilities, developing better transportation systems, and promoting the adoption of good agricultural and postharvest practices. Reducing postharvest losses enhances food security, improves farmers’ incomes, reduces waste, and promotes sustainable agricultural systems that benefit both the economy and the environment.

AI and robotics have become instrumental in mitigating horticultural crop losses resulting from postharvest challenges. These technologies offer a range of solutions that can revolutionize the horticultural industry and improve its efficiency [28, 82]. One significant way in which AI and robotics help is through quality monitoring and sorting. With the development of computer vision systems, AI can analyze images of horticultural crops to assess their quality and identify any signs of damage or decay. By automating the sorting and grading process based on quality parameters, AI ensures that only high-quality produce reaches the market [82]. This early identification and separation of damaged or low-quality crops prevent the mixing of good and poor-quality produce, thereby reducing overall postharvest losses.

Predictive analytics is another valuable application of AI. By analyzing various data inputs, such as weather patterns, temperature, humidity, and crop conditions, AI algorithms can make predictions about the likelihood of postharvest losses. By utilizing historical data and real-time monitoring, these systems provide valuable insights into when and where losses are likely to occur. Armed with this information, farmers and stakeholders can take proactive measures to minimize losses, such as adjusting storage conditions, modifying transportation routes, or implementing timely interventions [28].

AI and robotics also contribute to intelligent storage and preservation. AI-powered systems can monitor and control storage conditions, including temperature, humidity, and ventilation, in storage facilities. This ensures optimal conditions for each type of horticultural crop, helping to maintain their freshness and quality. Robotics can assist in automated handling and stacking of produce, reducing physical damage and contamination during storage. By maintaining ideal storage environments and minimizing physical damage, these technologies help prevent spoilage and extend the shelf life of crops, ultimately reducing postharvest losses [28, 83].

Precision harvesting is another area where robotics shine. Advanced robotic systems equipped with computer vision and ML algorithms can identify ripe fruits or vegetables and perform precise harvesting actions. By minimizing damage to the crop and ensuring accurate harvesting, these robots reduce losses caused by improper harvesting techniques and maximize the yield of marketable produce [83].

AI’s ability to optimize the entire supply chain is a game-changer. By analyzing data from various sources, including production, transportation, and market demand, AI algorithms can provide insights into demand forecasting, inventory management, and efficient route planning. This enables timely delivery of produce while minimizing waste and losses. By optimizing the supply chain, AI reduces delays, spoilage, and quality deterioration, leading to lower postharvest losses [6].

AI and robotics offer innovative solutions to minimize horticultural crop losses caused by postharvest losses. These technologies enable proactive decision-making, early detection of quality issues, precise handling, and optimized storage and transportation [6, 82]. By harnessing the power of AI and robotics, the horticultural industry can significantly improve its postharvest management practices, enhance efficiency, and ultimately reduce losses.

AI leverages extensive data processing from sources such as weather forecasts, satellite imagery, soil sensors, and historical records, offering real-time insights, pattern recognition, and actionable recommendations for informed decision-making. In contrast, conventional methods rely on experiential knowledge, often lacking thorough data analysis. AI-driven systems excel in early detection by identifying subtle shifts in plant health, pest activity, and environmental conditions, enabling swift action and preempting potential issues [84, 85]. Conversely, conventional methods depend on manual farmer inspections, which might be less frequent and accurate. AI empowers precision interventions, enabling exact adjustments in irrigation, fertilization, and localized pest control, minimizing resource waste and enhancing efficiency. Conventional approaches often rely on general treatments, risking overuse or insufficient control. AI’s predictive modeling capabilities utilize historical and real-time data to anticipate disease outbreaks, pest infestations, and other threats, allowing preventive measures, whereas conventional methods tend to be reactive [28]. Real-time monitoring through AI technologies like sensors, drones, and cameras ensures prompt identification of changes or anomalies, surpassing the limitations of human observation in conventional methods. AI’s climate adaptation aids in adjusting cultivation practices to changing conditions, a challenge for conventional practices. AI optimizes resource usage by recommending precise resource amounts based on real-time data, minimizing waste, while conventional methods might rely on manual estimation. AI accumulates knowledge, enabling cross-farmer and regional knowledge sharing through data-driven insights, whereas conventional methods are more localized. In summary, AI offers data-driven insights, early detection, precision interventions, predictive capabilities, and adaptability, although its effectiveness relies on technology, training, and local integration [80, 82].

ML has demonstrated significant potential in various facets of horticultural production, including predicting yields, detecting diseases, and evaluating crop quality. These applications highlight its ability to quantify and forecast postharvest wastage across the food value chain, essential for planning production and shaping policies. Despite its usefulness in these areas, there remains a noticeable gap in research focusing on using ML to physically measure postharvest losses in fresh produce. Current studies, like those examining socioeconomic factors in food loss classification, primarily concentrate on grains rather than direct quantification. Effectively addressing postharvest losses through ML requires integrating various data types such as environmental conditions, storage methods, and logistical information. However, incorporating physically measured data, like through weighing or counting, is crucial for accurate ML-based assessment of postharvest losses [29]. Closing this research gap is crucial for making informed policy decisions and enhancing sustainability in horticultural supply chains.

Kaul, Hill, and Walthall [86] introduced an AI system employing ANN for predicting corn and soybean yields. They developed yield prediction models using historical data from multiple locations, and reports indicate that these ANN models yielded more accurate predictions compared to regression models. Similarly, Ji et al. [87] crafted a system for rice yield prediction using ANN technology, specifically for the Fujian province in China, where unpredictable weather conditions, such as typhoons, floods, and droughts, pose threats to rice production. This model relied on location-specific historical weather data and, like its corn and soybean counterparts, proved to deliver more accurate yield predictions in Fujian.

Numerous other ANN models have been designed for yield prediction across various crops, including cotton (Gossypium hirsutum L.) [88], wheat [89], jute (Corchorus capsularis L. [90]), and tomatoes [91]. Beyond direct yield forecasting, ANN models demonstrate their ability to capture other variables that impact agricultural production. For instance, Nabavi-Pelesaraei, Abdi, and Rafiee [92] developed an ANN model capable of predicting energy usage and greenhouse gas emissions in watermelon (Citrullus lanatus L.) production systems. These predictions offer valuable insights for future farmers and are poised to enhance food availability for consumers in the future.

Pathmanaban et al. [93] explores several innovative postharvest technologies and their contextual effectiveness. AI-driven thermal imaging has proven highly effective in assessing fruit quality by detecting temperature variations associated with ripeness, damage, and disease, thereby enabling precise sorting and grading. Robotic systems equipped with AI enhance handling efficiency and reduce fruit damage during sorting and packaging processes. These technologies offer significant benefits such as improved quality control, reduced wastage, and enhanced operational efficiency in diverse fruit handling contexts. However, regulatory challenges related to data privacy, standardization of AI algorithms, and compliance with food safety regulations pose potential barriers to widespread deployment. Addressing these regulatory concerns will be crucial to realizing the full potential of AI and robotic technologies in transforming postharvest fruit management practices globally.

2.8. Comparative Analysis: Effectiveness of Traditional Horticultural Practices vs. Advanced Technologies

Traditional horticultural practices encompass a range of techniques rooted in local wisdom and sustainable principles, beneficial across various horticultural contexts. These methods, often labor-intensive and holistic, include crop diversification, companion planting, and organic pest management. They promote soil health, biodiversity, and resilience against pests and diseases, making them particularly effective in small-scale operations or regions where sustainable farming is prioritized [94]. However, their reliance on manual labor and natural inputs may limit scalability and efficiency compared to technologically advanced approaches.

Advanced technologies like AI and robotics offer transformative capabilities in horticulture, enhancing productivity, precision, and sustainability. AI-driven systems analyze data from sensors, satellites, and weather forecasts to optimize irrigation, pest control, and crop monitoring. Robotics automate tasks such as planting, harvesting, and pesticide application with precision, reducing labor costs and increasing efficiency [95]. These technologies are particularly suited to large-scale operations where economies of scale justify initial investment costs and where labor shortages are prevalent.

Implementing AI technologies like robotics, sensors, and data analytics platforms can be expensive, making them inaccessible for small- and medium-sized farms. This can exacerbate existing inequalities in the agricultural sector. Automation through AI could also lead to job losses in agricultural labor, impacting rural communities and livelihoods. Additionally, large corporations often own and control agricultural data, potentially creating monopolies and disadvantaging smaller players [95]. Farmers with limited access to technology and digital literacy skills risk being left behind in the AI revolution, creating a digital divide. AI algorithms trained on biased data can perpetuate existing inequalities, for example, by discriminating against certain types of farms or crops. Overreliance on AI could lead to the erosion of traditional farming knowledge and practices accumulated over generations, posing a loss of cultural heritage.

AI-driven intensification of agriculture could have negative environmental consequences, such as increased use of pesticides and fertilizers, leading to soil degradation and biodiversity loss [95]. The complexity of AI algorithms can make it difficult to understand how decisions are made, raising concerns about transparency and accountability in the food system. Integrating both traditional knowledge and advanced technologies tailored to local conditions holds promise for achieving sustainable, resilient horticultural systems that balance productivity with environmental stewardship. However, addressing economic, social, and ethical considerations is crucial to ensure that AI benefits all stakeholders in the food system and contributes to a sustainable and equitable future for agriculture.