Digital Transformation Drivers, Technologies, and Pathways in Agricultural Product Supply Chains: A Comprehensive Literature Review

Abstract

The digital transformation of agricultural product supply chains has emerged as a strategic direction that cannot be overlooked in the global modernization of agriculture. This paper adopts a narrative review framework based on the “Technology–Collaboration–Sustainability” perspective in the digital transformation of agricultural product supply chains, summarizing the drivers of digital transformation, the application of digital technologies, multi-stakeholder collaborative mechanisms, and pathways for sustainable development within these supply chains. The study finds that the core drivers promoting the digital transformation of agricultural product supply chains include external environmental factors (such as population growth, dietary shifts, and food waste) and internal demand drivers (such as industrial upgrading and increased corporate competition). The application of digital technologies such as the Internet of Things (IoT), blockchain, and artificial intelligence (AI) has significantly improved the efficiency, transparency, and resilience of the supply chains. Furthermore, various models of multi-stakeholder collaborative mechanisms have optimized resource allocation and enhanced supply chain stability. Finally, the paper proposes a pathway for the sustainable development of agricultural product supply chains based on digital transformation, providing directions for future research and practice.

1. Introduction

Agricultural product supply chains, as a core link in global food security, are facing multiple challenges such as climate change, population growth, and resource shortages [1,2]. Notably, climate change impacts these supply chains in a dualistic manner: on one hand, the agriculture and food systems themselves contribute 19–29% of global greenhouse gas emissions [3], significantly exacerbating the climate crisis; on the other hand, the resultant increase in extreme weather events has led to reduced crop yields [4] and diminished productivity. Consequently, the agricultural product supply chain is both a contributing factor to climate issues and a direct victim thereof. Particularly under the impact of the COVID-19 pandemic on traditional supply chains, existing research has predominantly focused on individual segments (such as production or logistics), lacking a systematic analysis of the chain’s collaborative and sustainable aspects. Moreover, as a pivotal element in ensuring food security and promoting rural development, the agricultural product supply chain urgently requires a comprehensive transformation to address the dual pressures of productivity improvement and sustainable development. Therefore, development strategies for these supply chains should not only consider the transformation of agricultural production methods but also incorporate sustainable objectives: to minimize harmful economic, social, and environmental impacts while harnessing positive outcomes [5].

Digital transformation offers a systematic solution for the sustainable development of agri-food supply chains. Evidence shows that the Internet of Things (IoT), blockchain, and artificial intelligence (AI) enhance resilience, raise process efficiency [6], lower operating costs and advance the United Nations Sustainable Development Goals [7]. Coordinated strategies are equally critical for mitigating the social tensions and resource depletion confronting agriculture [8,9]. However, prevailing research and practice often exhibit a “technological determinism” bias, focusing narrowly on the application effects of isolated digital technologies while neglecting the fact that technology-enabled value creation depends on complementary organizational collaboration models and must ultimately be evaluated against sustainability objectives. To address this research gap, this paper innovatively adopts an integrated “Technology–Collaboration–Sustainability” (TCS) perspective as the overarching framework for the review. The internal logic of this framework is as follows:

(1)

Technology as the enabler. The IoT, blockchain, and AI supply the data-capture, transparency and analytical capabilities required to reduce information asymmetry, raise operational efficiency and strengthen traceability [10,11].

(2)

Multi-agent collaboration as the value nexus. Digital tools generate value only when embedded in institutional arrangements that coordinate governments, firms, farmers and civil society. Ineffective governance produces data silos and entrenches power imbalances, preventing chain-wide value creation.

(3)

Sustainability as the ultimate metric. Success is judged not merely by productivity gains but by the simultaneous delivery of environmental integrity and social inclusivity.

These three elements constitute an evolving, co-dependent system: technology empowers collaboration, collaboration delivers sustainability outcomes, and sustainability objectives subsequently reshape both technological priorities and governance models. The research framework followed in the literature review section is illustrated in Figure 1.

This study does not aspire to craft a new grand theory; instead, its contribution resides in contextualizing and synthetically leveraging existing paradigms to open the black box between digital technology and sustainability in agri-food supply chains. We offer three mutually reinforcing advances.

(1)

Cross-theoretical Integration and the Construction of Mediating Mechanisms

Rather than privileging a single lens, we orchestrate Technology Acceptance Model (TAM), Institutional Theory, Stakeholder Theory, and Collaborative Governance Theory into an integrative framework. The core proposition is a mediating mechanism chain—Digital affordances (T)→Relational coordination (C)→Environmental–economic–social tripartite sustainability (S), where “C” captures both the stock (existence of governance arrangements) and the flow (dynamic trust repair, power re-balancing) of collaboration.

The frame posits that digital artifacts do not affect sustainability directly, but operate through the reconfiguration of multi-stakeholder collaborative relationships. By shifting the analytical focus from tools to relations, the specification constitutes a profound theoretical advance in research on the digital transformation of agri-food supply chains.

(2)

Socio-Technical Systems Perspective

The framework conceptualizes the digital transformation of agri-food supply chains as the evolutionary reconfiguration of a socio-technical system rather than a mere technological upgrade. It emphasizes that technology must be embedded in specific social structures, power relations and institutional arrangements in order to function. The study’s systematic treatment of social variables—including power asymmetries, trust deficits and inequitable benefit distribution—evinces a sophisticated grasp of the social dimension of digital transformation and shields the analysis from the superficiality of purely technological accounts.

(3)

Dynamic and Feedback Perspective

As shown in Figure 1, authors view the system as a dynamically evolving, mutually shaping organism, rather than nonlinear, unidirectional causality. Recursive feedback loops allow sustainability outcomes to re-shape governance arrangements, which in turn re-calibrate technological affordances (S→C→T).

Grounded in this framework, this paper reviews the drivers, technologies, and pathways of digital transformation in agricultural supply chains, aiming to construct a strategic framework of “technology empowerment–collaborative innovation–sustainable value addition”. The goal is to provide policymakers and supply chain actors with both theoretical foundations and feasible practical insights, thereby supporting decision-making in the global transformation of agricultural supply chains.

2. Literature Collection, Screening, and Organized Analysis

2.1. Preliminary Screening

This review is positioned as a structured narrative review, distinct from a systematic review or meta-analysis. While the latter adhere to rigid PRISMA protocols and emphasize quantitative synthesis, our objective is to deliver a comprehensive, interpretive synthesis of a broad and heterogeneous literature landscape concerning digital transformation in agricultural supply chains.

We justify this methodological choice for three interrelated reasons:

(1)

Interdisciplinary breadth: The field spans technology adoption, governance mechanisms, sustainability performance, and multi-stakeholder collaboration. A purely systematic approach would risk overlooking nuanced, context-specific insights that are critical for understanding socio-technical transitions.

(2)

Conceptual integration: Our aim is to develop and propose a conceptual framework (TCS) that organizes fragmented knowledge and identifies emerging research gaps, rather than to test a narrowly defined hypothesis or intervention effect.

(3)

Inclusion of practice-based evidence: We deliberately incorporate the gray literature, policy reports, and seminal case studies that would typically be excluded by rigid systematic review protocols, yet are essential for understanding real-world implementation barriers and contextual complexities.

We performed a structured-non-systematic search in Web of Science, Google Scholar and AGRIS, supplemented by snowballing, yielding 253 relevant articles for narrative synthesis.

The search strategy was built around key terms related to the core concepts of the study: “agricultural product supply chain” OR “agri-food supply chain” OR “food supply chain” AND “digital transformation” OR “Industry 4.0” OR “smart agriculture” AND “technology” OR “IoT” OR “blockchain” OR “artificial intelligence” AND “sustainability” OR “collaboration” OR “governance”. The search was confined to the title, abstract, and keywords of publications and irrelevant field classifications were excluded. The publication time frame was set from January 2000 to February 2025 to capture the evolution of digital technologies in this domain. Only articles published in English were considered.

2.2. Inclusion and Exclusion Criteria

Studies were included if they met the following criteria:

(I1) Focused on agricultural product/food supply chains.

(I2) Investigated the application of digital technologies (e.g., IoT, blockchain, AI, big data).

(I3) Discussed aspects of drivers, collaboration models, sustainability outcomes, or implementation pathways.

(I4) Were peer-reviewed journal articles, conference proceedings, or review articles.

Studies were excluded based on the following criteria:

(E1) Focused on industrial supply chains unrelated to agriculture.

(E2) Purely technical papers without discussion of management, collaboration, or sustainability implications.

(E3) Studies focused solely on agricultural production technology without a supply chain perspective.

(E4) Books, book chapters, editorials, and non-English publications.

(E5) Not accessible and No full text.

2.3. Study Selection and Quality Assessment

The study selection process involved a two-phase screening, as illustrated in the flow diagram (Figure 2). Initially, all identified records were imported into EndNote, and duplicates were removed automatically and manually. In the first phase, two authors independently screened the titles and abstracts of all retrieved articles against the inclusion and exclusion criteria. In the second phase, the full texts of the remaining articles were assessed for eligibility. Any disagreements between the reviewers were resolved through discussion or by consultation with a third author.

To appraise the methodological quality and relevance of the included full-text articles, we applied the following evaluation criteria:

C(1): Clarity of research objectives.

C(2): Appropriateness of research methods.

C(3): The credibility of the results and analysis.

C(4): The correlation between the conclusion and the focus of the review.

This evaluation helps ensure the robustness and relevance of the comprehensive literature.

2.4. Descriptive Statistics

It can be observed that starting from a single inaugural paper in 2011, the number of publications gradually increased, reaching 47 in both 2022 and 2024. This indicates that over the past decade or so, research on the digital transformation of the agricultural product supply chain has generally shown an upward trend, with a growing scholarly interest in the subject as shown in Figure 3.

A statistical analysis of the source journals revealed that Sustainabilityprovided the highest number of articles meeting this study’s selection criteria, with a total of 11 articles (As shown in Figure 4).

3. Literature Review

3.1. Drivers of the Digital Transformation in Agricultural Supply Chains

Based on Boulding’s push–pull theory, the driving forces behind individual or organizational behavioral decisions can be divided into pull factors—external opportunities or pressures (such as market demand and technological trends)—and push factors—internal needs or problems (such as inefficiencies and the demand for reform) [12]. Originally employed to analyze economic activities and the adoption of innovations, this theory has since been widely applied in areas such as technology diffusion, supply chain management, and digital transformation [13]. Accordingly, this section conducts an in-depth analysis of the drivers behind the digital transformation of agricultural supply chains based on the push–pull theory.

3.1.1. External Environmental Drivers

(1) Market and Environment

First, in terms of population growth and the looming food crisis, the 2017 report “The Future of Food and Agriculture: Trends and Challenges” published by the United Nations Food and Agriculture Organization (FAO) predicts that by 2050 the global population will increase by 33%, reaching an estimated 9.7 billion [14]. With this population growth, agricultural and food production will need to increase by 60% by 2050, and in developing countries, production might need to double in order to meet the growing demand for food [15,16]. This scenario has prompted the agricultural supply chain to seek avenues for transformation.

Second, regarding dietary transitions and nutritional needs, increasingly high consumer expectations are shifting agricultural consumption towards a more balanced, safe, and health-conscious diet [17]. Traditional agricultural production techniques are perceived as raising concerns over quality, safety, and toxicity, leading the food manufacturing sector to shift towards more cost-effective, nutritionally stable, and environmentally sustainable practices [18]. This transition not only has a significant positive impact on public health and welfare [19] but also benefits the environment. The transformation of national food systems and a widespread shift towards diets with higher nutritional standards are driving increased agricultural demand [20,21]. To achieve a balanced diet, agricultural productivity will need to improve by nearly 50% from 2012 levels by 2050 [22], further propelling the transformation of the agricultural supply chain towards greater efficiency.

Additionally, concerning food loss and severe waste, fresh agricultural products are characterized by perishability, high prices, volatility in demand, and a higher likelihood of food safety issues [23], thereby contributing to a more complex supply chain. Data released by the FAO indicates that more than 1.94 billion tons of fresh fruits and vegetables are produced globally each year over an area exceeding 122 million hectares; however, due to losses, decay, and waste, only 50% of these fruits and vegetables actually reach the consumption stage, with approximately 30% of food being lost or wasted at some point in the supply chain [24]. Food loss and waste have detrimental impacts on society, the environment, and the economy [25].

Finally, regarding the demand for information processing, there is an explosive growth in the generation of agricultural data worldwide. The value of data lies not only in its collection and storage but primarily in its reuse. Faced with such an enormous volume of available data, it becomes crucial to develop solutions capable of organizing, analyzing, and visualizing this data to provide decision-making support, thereby guiding farmers, scientists, or policymakers toward better decisions and driving actionable changes in agriculture [26]. As data volume continues to increase yet the capacity of individuals or organizations to process and utilize it remains limited, there is an urgent need for a system or tool that addresses data coordination and standardization challenges and bridges the gap between science and end users [27].

(2) Accelerated Technological Iteration

The application of emerging technologies in agricultural supply chains is becoming increasingly widespread, driving the transformation toward a data-driven, digital supply chain environment [28]. Industry 4.0 combines cyber-physical systems, IoT technology, artificial intelligence and machine learning, big data analytics, and cloud computing with agricultural machinery, marking a fundamental shift in agricultural practices [29]. Among these, IoT, blockchain, and big data technologies serve as potential catalysts for sustainable development within agricultural supply chains [30]. In certain irregular application scenarios, such as pattern recognition under uncertain or changing environmental conditions, digital technologies have demonstrated performance that exceeds human capabilities. The advancement of digital technologies provides a robust technical foundation for the digital transformation of agricultural supply chains, meeting technical requirements such as data integration and user-friendly interfaces for precision and smart agriculture [31]. Moreover, progress in blockchain, big data, and related technologies enables real-time information sharing and precise transmission, which helps resolve issues of sluggish information flow in agri-food supply chains by enhancing transparency and collaborative efficiency [32]. These technologies are poised to usher in the next generation of agricultural innovations by simultaneously boosting productivity, efficiency, and reducing risks and adverse impacts [33].

3.1.2. Internal Demand Drivers

(1) Industry and Sectoral Demands

First, the state of the agricultural products industry is challenging. In recent years, extreme weather conditions have caused widespread economic damage globally [34], and agriculture is now confronted with issues such as climate change [35], water scarcity [36,37], and limited arable land [38,39]. The agri-food sector is continually striving to navigate various market challenges, including the unpredictability of commodity prices, shifting consumer preferences, and the complexities of global trade regulations [40]. Against the backdrop of increased food waste and unequal distribution of income and food, there is a heightened need for greater agricultural output. These challenges render traditional agricultural supply chain models inadequate for adapting to market changes and meeting demand. For any enterprise pursuing growth, expansion, quality, and sustainability, digital transformation is inevitable [41]. Enhancing supply chain flexibility and resilience through digital transformation is essential for improving resource utilization efficiency and addressing environmental challenges.

Second, the imperative for sustainable supply chain development is increasingly evident. Economic, environmental, and social sustainability are receiving growing attention within academic circles and national economic policies alike [30]. As the core of the value chain, agricultural activities must both feed a growing global population and safeguard resources to the highest degree possible to ensure sustainability [42]. Establishing a sustainable agricultural food system involves the large-scale coordination of agricultural operations, value chain activities, resource utilization, and food distribution—all of which require extensive digital devices and infrastructural support to maintain a comprehensive data and digital information network [43]. The digital transformation of agricultural supply chains guarantees greater flexibility, efficiency, and effectiveness by accelerating decision-making [42]. This transformation aids traditional supply chains in realizing more efficient and sustainable practices across food production, distribution, and consumption [43]. This industrial agricultural ecosystem with real-time farm management, high automation, and data-driven intelligent decision-making will significantly enhance the efficiency of the agricultural product supply chain and the utilization rate of natural resources [44]. It is an essential condition for developing robust circular agricultural food systems and plays a critical role in balancing economic, environmental, and social demands toward achieving sustainable development goals in the agricultural supply chain.

(2) Corporate Demands

First, enterprises must take proactive measures to counteract risk and uncertainty. In today’s agricultural landscape, climate change, limited and costly resources, and market instability have collectively heightened uncertainty in decision-making among farmers and policymakers, while simultaneously increasing business risks. In 2020, the pandemic triggered a demand among supply chain stakeholders for novel decision-making frameworks, sharply intensifying the need for lean, agile, resilient, and sustainable supply chains under the severe test of COVID-19 [45,46]. The digital transformation seen in digital agriculture or Agriculture 4.0 has provided new opportunities to mitigate these risks and uncertainties [11]. In the long term, enterprises must develop data-sharing capabilities and prepare action plans to implement supply chain digitalization [47].

Second, issues related to data security and trust remain to be addressed. For agri-food enterprises undergoing digital transformation, data protection is paramount, particularly in an era of heightened geopolitical tensions and escalating cyber threats. Robust cybersecurity measures can substantially protect sensitive data and critical infrastructure from these threats [48]. Furthermore, establishing a digital trust mechanism can enhance stakeholder confidence and effectively eliminate development barriers arising from low trust levels, thereby continuously fostering high-level collaboration [49].

Third, companies pursue competitive advantage and enhanced market position. Digital transformation has a transformative impact on the global economy and society, making corporate digital transformation an inevitable trend. This process can help companies secure potential benefits such as increased sales, improved productivity, and a heightened customer-centric focus, thereby driving value creation and introducing innovative forms of interaction [45]. In doing so, they can respond to the rising and spreading disruptive threats of the digital technology era, swiftly adapt to market changes, make timely and informed decisions, and maintain competitiveness in today’s challenging markets by offering personalized products and services to their customers [50]. Therefore, for companies aiming for growth, expansion, quality, and sustainability, digital transformation is an essential development approach [41].

This section details the internal and external pressures driving the digital transformation of agricultural supply chains (e.g., efficiency enhancement, food safety, and sustainability requirements). While digital technologies themselves serve as tools to address these challenges, the ultimate realization and distribution of their value are not automatically determined by the technology. Instead, institutional arrangements define the roles, rules, modes of interaction, and mechanisms for benefit distribution among the participating actors.

3.2. Technological Applications in the Digital Transformation of the Agricultural Product Supply Chain

In the digital transformation of the agricultural product supply chain, technologies are diverse and extremely critical and are typically categorized into efficiency technologies, connectivity technologies, trust-based disintermediation technologies, and automation technologies [51]. This paper defines digital agriculture (DA) as an integrated system encompassing the following four domains: sensor technology (e.g., soil moisture probes, unmanned aerial vehicle remote sensing); data management (cloud computing, edge computing); control technology (intelligent irrigation, automated harvesting); and data modeling and communication (AI/ML, 5G networks). The extensive use of these technologies in agriculture has significantly driven transformations in agricultural production and supply chain management [52].

3.2.1. Internet of Things Technology

The advancement of smart agriculture depends on access to both real-time and historical data, enabling the monitoring and management of farms and crops, as well as forecasting factors such as crop yield and productivity [53,54]. With continuous improvements in wireless sensing technology, the application of IoT in smart agriculture has become increasingly widespread [55].

IoT devices primarily use various low-power sensors for data monitoring and collection, generating vast amounts of streaming data known as “big data” [56], assist farmers and agricultural researchers in determining the optimal timing for planting, fertilization, and harvesting, as well as in identifying issues promptly and taking appropriate actions—thus enabling real-time monitoring of production processes. Sensors and cloud computing components are mainly used to monitor soil parameters [57,58] and weather conditions to achieve intelligent irrigation, pest control, and rational pesticide usage [42,59]. Some scholars have put forward an intelligent agriculture system based on IoT and cloud computing technologies. Through big data analysis on the intelligent agriculture platform, farmers gain deeper insights into the factors affecting crop growth, allowing them to make more informed decisions [60,61] to increase crop yields and reduce losses caused by quality disputes [62].

Furthermore, through message queue telemetry transport (MQTT) protocols, IoT systems can share and exchange information with blockchain, smart contracts, and various system participants [63], providing decision-making reference information for activities across primary production, processing, and retail. Nozari et al. (2021) introduced a “four-dimensional decision framework” for the fast-moving consumer goods industry to optimize product quality during processing [64]. Li et al. (2017) developed an effective and cost-efficient real-time tracking and traceability management platform for the pre-packaged food supply chain based on IoT technology, ultimately ensuring a secure food consumption environment [65].

In terms of risk assessment, many scholars have offered new insights for optimizing the integration of agricultural production supply chains. Dai et al. (2020) used big data to analyze potential risks at the interface of agricultural supermarkets of large retail enterprises and their supply chains, highlight risks associated with farmer selection, management of production material suppliers, and distribution information flows [66]. They provided corresponding fuzzy evaluation grades and risk assessment standards. Han et al. (2024) proposed a data-driven multi-criteria decision-making framework based on cost, benefit, and feasibility to support the development and balancing of optimal risk management strategies in volatile supply chains [67]. Yadav et al. (2020) further contributed by developing an IoT-based performance measurement framework for the agricultural supply chain, providing essential tools for enhancing the competitiveness and sustainability of the supply chain [68].

From an economic standpoint, studies indicate that IoT systems generally achieve a return on investment (ROI) within one to three years through optimized resource utilization (e.g., water and fertilizer savings), reduced crop losses, and decreased labor expenditures. For example, the IoT system evaluated by Wu (2022) increased profit per mu by 18.7%, demonstrating strong financial performance [62]. Nevertheless, key barriers such as upfront hardware investments (sensors and gateways), ongoing communication costs, and cloud platform fees remain insufficiently addressed in the existing literature.

3.2.2. Blockchain Technology

Blockchain technology encapsulates transaction identifiers—including time, amount, and date—into cryptographically linked digital blocks that form a sequential, immutable chain [69]. Each verified transaction is stored within a block marked with a unique hash, creating a decentralized and tamper-resistant ledger system that ensures transparency and permanence [70].

Within agricultural supply chain management, blockchain significantly improves transactional transparency and operational visibility [71]. By providing open access to information throughout production, distribution, and trade processes, it enhances the reliability of cross-regional agricultural transactions [72]. Technologically, this is implemented through various innovative frameworks. For instance, Salah et al. (2019) developed an Ethereum-based system for soybeans that utilizes smart contracts to automate verification processes, granting consumers access to full product lifecycle data [71]. Similarly, Lin et al. (2020) introduced an integrated blockchain-IoT framework for food traceability that employed zero-knowledge proofs to maintain transparency without compromising privacy [73]. As blockchain continues to consolidate more information, the traceability of product quality and safety improves, thereby further enhancing consumer trust. As blockchain repositories grow, they facilitate more precise quality and safety tracking, strengthening consumer confidence [74,75,76].

The traceability and transparency enabled by blockchain directly contribute to key objectives of sustainable agriculture. By offering an auditable record of a product’s journey, blockchain effectively addresses inefficiencies and behavioral challenges—such as fraud—within supply chains [77]. It is particularly effective in reducing fraudulent activities in cross-border food trade [51] and establishing reliable trust mechanisms among stakeholders. The literature confirms that these capabilities are foundational to sustainability. In a case study on Pakistan’s mango exports, after introducing zero-knowledge proof (ZKP) technology to design a hierarchical blockchain model, the incidence of fraud in the targeted scenario (e.g., agricultural supply chain traceability) decreased from 15% to 3% [78]. Lin et al. (2020) note that the ability to trace product origins and enhance visibility is crucial for mitigating risks and promoting sustainable supply chain development [73]. Beyond traceability, blockchain’s impact extends to broader economic and operational dimensions. Using an evolutionary game model, Su & Cao (2023) showed that blockchain adoption can influence the lending strategies of financial institutions and shape the strategic decisions of agricultural enterprises, potentially directing capital towards more sustainable practices [79]. Furthermore, Pranto et al. (2021) and Gandhi Maniam et al. (2024) highlight how blockchain, especially when integrated with IoT and smart contracts, fosters sustainability [63,80]. It automates data collection and coordination [63], which enhances environmental compliance, reduces waste, and enables more efficient resource use through automation [80].

Although blockchain technology holds significant potential for enhancing transparency and trust in theory [71,73], the generalizability of these outcomes across diverse agricultural products and geopolitical contexts remains inadequately validated. Practical implementation faces further challenges, including high energy consumption, substantial computational requirements, and sensitivities related to cross-enterprise data sharing. Existing studies have largely overlooked the compatibility of these technological and economic demands with the realities of small- and medium-sized enterprises (SMEs), particularly smallholder farmers. Although case studies [78] provide compelling evidence regarding technical feasibility and fraud mitigation capabilities, there remains a notable lack of large-scale empirical validation using real-world financial and operational data. Consequently, the broader economic impact and tangible sustainability benefits of blockchain technology within agricultural supply chains remain predominantly exploratory and insufficiently substantiated.

3.2.3. Artificial Intelligence and Machine Learning Technologies

Artificial intelligence, which enables computational systems to simulate human cognitive functions, has emerged as a transformative tool in addressing complex challenges throughout the agricultural supply chain and facilitating its transition toward intelligent operations [81,82]. The integration of AI technologies offers significant advantages to the agri-food sector, including reduced training costs, enhanced response capabilities, improved decision-making accuracy, and decreased reliance on manual labor [83].

In crop production management, AI is extensively employed to analyze soil and environmental data, thereby assisting farmers with weather forecasting, disease diagnosis, and weed identification [82,84]. As a critical subset of AI, machine learning (ML) plays an essential role in predictive analytics for production planning, procurement strategy development, and inventory management [85]. Commonly utilized algorithms—such as decision trees, random forests, and Bayesian networks—leverage sensor-derived data to monitor environmental parameters effectively; diagnose livestock illnesses; and optimize breeding programs. This significantly enhances overall supply chain efficiency [85]. Continuous monitoring of crop health and growth status further contributes to the formulation of sustainable strategies aimed at addressing global food security concerns [86]. Notably, contemporary ML techniques—including support vector machines (SVMs) and artificial neural networks (ANNs)—have demonstrated superior performance compared to traditional time-series models (e.g., ARIMA) in predicting agricultural commodity prices while generating more reliable market forecasts [87]. Representative applications include ensemble ML-based recommendation systems for assessing crop suitability [88], automated seed classification frameworks [89,90], and ML-driven plant disease detection models tailored for low-resource farming environments [91]. Additionally, decision tree algorithms have been effectively utilized in monitoring soil moisture [92] and nitrogen content [93], thereby enhancing crop productivity. Deep learning platforms such as Plantix, developed by PEAT GmbH, facilitate real-time identification of crops and pathologies through smartphone imagery, which supports timely interventions and mitigates potential losses [84].

In the domain of quality inspection and food safety, machine learning and deep learning approaches are increasingly being adopted to evaluate the quality of agricultural products. Lu et al. (2017) established an ML-enhanced methodology for the precise quantification of rice starch, refining protocols for grain quality assessment [94]. Zhou et al. (2017) designed an SVM-based classifier utilizing spectral data to differentiate between green tea varieties, achieving accuracies of 96% on prediction sets [92]. Subsequent studies have further employed regression techniques alongside the SVM to construct robust models for evaluating crop quality [95,96,97].

In supply chain optimization, AI can enhance various segments of the agricultural supply chain [84]. Logistics and distribution are widely recognized as critical bottlenecks in maintaining product integrity [98], benefit from AI-powered analyses of multi-source data coupled with intelligent routing algorithms, transportation efficiency is improved while operational costs are reduced [99,100]. In inventory management, AI-enhanced collaborative forecasting facilitates dynamic replenishment and allocation strategies driven by real-time sales and demand data. Furthermore, supplier relationship management is strengthened through AI-facilitated coordination between demand and production along with risk prediction capabilities that ensure supply stability [101]. Sales and marketing operations benefit significantly from demand forecasting models that effectively minimize both overstocking and shortages, while simultaneously maximizing profitability. Moreover, machine learning-based decision support systems, such as the clustering-assisted framework introduced by Boshkoska et al. (2019), facilitate knowledge sharing and identify critical barriers within supply chain networks, thereby enhancing strategic decision-making [102].

The accuracy of machine learning models in prediction tasks is highly dependent on the availability of high-quality and large-scale datasets. However, agricultural data is often fragmented and non-standardized [83,88] and though it has demonstrated impressive performance on specific datasets, the generalizability and robustness of these models across diverse geographical regions and climatic conditions remain inadequately validated. This highlights a persistent “data gap” between current research efforts and practical agricultural applications.

3.2.4. Heterogeneous Application of Digital Technology in the Agricultural Products Supply Chain of Different Types

A detailed comparison of the heterogeneous application of digital technology is provided in

Table 1. Heterogeneous application of digital technology in the agricultural products supply chain of different types.

		High-Value, Highly Traceable Products (e.g., Premium Beef, Fine Wine)	Bulk, Storable Commodities (e.g., Wheat, Maize)	Fresh, Perishable Produce (e.g., Strawberries, Leafy Greens)
Digital Technologies	IoT	Application focus: End-to-end precision monitoring. RFID ear-tags or collars track activity, body temperature, transport shocks and tilt to safeguard animal welfare and meat quality [64].	Application focus: Storage-environment surveillance. Dense networks of temperature-, humidity-, pest- and gas-sensors in granaries prevent mold and quality loss [43,59].	Application focus: Cold-chain monitoring from field to shelf [61]; real-time temperature, humidity and location data; UAV remote sensing of crop health.
		Value: Enhances brand narrative and credibility, enabling targeted price premiums [62].	Value: Reduces post-harvest storage losses and underpins national food security [44].	Value: Core objective is “freshness preservation” [23,24,25]; markedly lowers spoilage rates and extends shelf-life.
	Blockchain	Application focus: Anti-counterfeiting and full provenance [71,72]. Each item receives a unique digital identity; immutable, time-stamped records cover breed, ranch, abattoir, processing, inspection and logistics.	Application focus: Trade-finance and process automation [44]. Digitizes and automates letters of credit, bills of lading and quality certificates through smart contracts, shortening transaction cycles for bulk agri-trade [42].	Application focus: Rapid traceability and liability assignment. When food-safety incidents occur, problematic batches are located within minutes, enabling swift recall and limiting losses and panic [74,77].
		Value: Eliminates information asymmetry [80], builds ultimate trust and anchors brand equity.	Value: Raises trade efficiency, cuts transaction costs and mitigates fraud risk [71].	Value: Protects consumer safety, safeguards brand reputation and satisfies regulatory mandates [51].
	AI	Application focus: Quality grading and outcome prediction. Computer vision automatically grades marbling, color and carcass weight; AI predicts optimal fattening periods and market-ready dates.	Application focus: Yield and price forecasting. Integrates satellite imagery, meteorological and soil data to predict global output [89,90]; mines massive market datasets to forecast price trends [31,86].	Application focus: Demand forecasting and intelligent scheduling. Analyses retail data, weather and holiday effects to predict daily demand; optimizes planting plans and logistics routes to curb waste.
		Value: Standardizes production and maximizes profit [85].	Value: Guides governmental and corporate macro-level stocking and policy decisions [83].	Value: Core aim is “loss reduction”; achieves supply–demand matching and mitigates “produce-dies-at-farm-gate” phenomena [84,87].
Main cost drivers		Tag + middleware + certification	Sensor network; platform subscription	Logger hardware; data analytics; cold-chain retrofit.
Key enabling; risk factors		Premium price readily offsets tag cost; needs brand power	Savings from mold/weight-loss cover fee; benefit grows with volume.	Shelf-life extend/waste reduce; smallholders need group-buying to reach MOQ.

. It outlines the key distinctions observed across different supply chain types, offering a clear overview of the findings.

3.2.5. Synthesis and Comparative Analysis: Towards an Integrated Technological Ecosystem

Although IoT, blockchain, and AI each offer distinct capabilities, their convergence generates a synergistic technological ecosystem. The future of digital transformation in agricultural supply chains lies in their strategic integration (as shown in Figure 5): IoT functions as the sensory nervous system, AI as the analytical brain, and blockchain as the trust-bearing backbone that ensures integrity and automated execution. This integration not only mitigates the inherent limitations of each standalone technology but also establishes a compounded value proposition essential for modernizing agricultural supply chains.

IoT devices constitute the sensory foundation of the ecosystem, continuously capturing high-volume, real-time data on variables such as crop health, soil moisture, and logistical conditions. Blockchain provides an immutable record for IoT-generated data, certifying its origin and preventing tampering. This combination establishes end-to-end transparency and verifiable provenance, enabling trust among stakeholders without prior relationships through cryptographic assurance rather than interpersonal reliance. AI and ML algorithms process this data to generate predictive and prescriptive insights. This transition from passive monitoring to proactive management significantly enhances both process efficiency and environmental sustainability. Together, they form a holistic technological foundation capable of supporting the multi-stakeholder collaboration required for achieving comprehensive sustainability goals.

Although this integration shows great potential, the majority of existing studies have primarily focused on technical feasibility and efficiency gains within pilot projects, conclusions were often drawn from controlled or idealized commercial settings. Critical issues such as the economic viability of large-scale deployment, ROI across farms of different sizes, and the long-term durability of sensors in harsh agricultural environments remain inadequately supported by robust, large-scale empirical evidence. Future research should therefore prioritize longitudinal case studies to validate how a combination of these technologies yields long-term benefits.

3.3. Multi-Agent Collaborative Mechanisms in the Digital Transformation of Agricultural Product Supply Chains

3.3.1. Motivations for Multi-Agent Collaboration

Supply chain cooperation is defined as a fundamental agreement in which partners integrate resources for their mutual benefit [103]. This agreement is inherently an institutional arrangement that, through the design of rules, trust, and power structures, determines how cooperation is conducted, risks are shared, and the value generated is ultimately distributed. The emergence of multi-agent collaboration in the agricultural product supply chain is subject to multiple conditions. Over the past few decades, the necessity for inter-supply chain collaboration has gradually become evident in order to efficiently respond to customer demands.

Currently, collaboration among nodes in the agricultural product supply chain is characterized by short cooperation cycles and loose relationships [104]. Divergent perceptions among stakeholders regarding the complexity and uncertainty of agricultural product quality and safety have resulted in participants being unable to maximize benefits through collaborative efforts. Moreover, the absence of an effective supply chain collaboration mechanism has led to a dual dilemma—insufficient value creation on the supply side and a lack of stability on the demand side [105]. This has long hindered the high-quality development of agricultural product supply chains, as reduced product added value and unstable cooperation cause persistent supply disruptions, thereby urgently necessitating enhanced collaboration to ensure stability.

However, merely increasing the degree of coordination is not a panacea. Although this approach may reduce uncertainty within the network [106], the withdrawal of one entity could still trigger failures in others or even cause the collapse of the entire supply chain. Therefore, it is imperative for supply chains to possess sufficient flexibility to manage potential risks [107,108], which has given rise to a supply chain paradigm that leverages collaborative advantages. For example, Zurek et al. (2020) analyzed the resilience of fruit and vegetable systems in the face of water-related risks, finding that the resilience of individual actors does not necessarily translate into overall system resilience; overlapping and mutually reinforcing individual resilience strategies can, in fact, enhance the resilience of the entire system [109]. Similarly, research by Wang & Xu (2022) demonstrates a negative correlation between the risk levels in agricultural product supply chains and the extent of collaborative governance among multiple stakeholders—the greater the collaboration, the lower the inherent risk of the supply chain [110]. Regarding cost implications, Shen et al. (2011) confirmed that the introduction of collaborative forecasting can reduce the operational costs within agricultural product supply chains [101].

Consequently, governments, agricultural enterprises, and a broader range of stakeholders must leverage their respective advantages to collaboratively build an “inclusive” governance framework [111]. Supported by innovative practices, tools, and digital solutions, diverse stakeholder groups can mobilize data, innovation, and technology to direct agricultural development, promote investment, and facilitate behavioral change [112].

The intrinsic characteristics of agricultural products also impose specific requirements for multi-agent collaboration, the structure and attributes of these products influence the forms of cooperation [113]. For instance, maintaining consistent quality throughout the journey from production sites to end consumers requires a high level of collaboration among the entities within the agricultural supply chain [114]. This necessity is especially pronounced in cold supply chains, where product quality is critically impacted. Ensuring the quality of agricultural products along the chain cannot rely solely on governments and agricultural enterprises, it requires the active involvement of multiple stakeholders [115,116]. Furthermore, the success of agricultural product marketing depends on joint efforts [111]. The evolutionary strategy proposed by Xie & Lei (2021) illustrates that collaborative cooperation among actors can promote the stable development of the supply chain [117], thereby underscoring the importance of multi-agent collaboration. In the realm of environmental and crop improvement, cooperation and data sharing among farmers, animal husbandry specialists, physiologists, geographical information system experts, and research institution programmers are essential to fully harness the potential of big data, which in turn can drive advancements in agriculture and crop enhancement [118].

3.3.2. Main Actors and Collaboration Patterns

(1) Main Actors

Farmers: Farmers occupy a fundamental position in the agricultural product supply chain, providing the basic raw materials necessary for food production and engaging in agricultural activities such as planting, cultivation, and harvesting [119]. Farmers often prioritize vegetables with historically high economic returns to maximize short-term profits [120], a strategy driven by their weak bargaining power and susceptibility to market volatility when transacting with larger buyers or retailers. Many farmers rely on government support and incentives—including subsidies, crop insurance programs, price stabilization measures, and tax benefits. When subsidies are implemented at the grassroots level, deviations in their execution may arise. Local elites or large enterprises may capture or divert these subsidies and associated benefits, thereby restricting small-scale farmers’ access to the intended support [104]. In selling their products [121], they determine a fair market price through cost analysis and cost-sharing mechanisms [122]. Consequently, they often find themselves in a disadvantaged position with respect to access to information and economic power. The participation of famers is crucial in innovation processes oriented toward sustainable development [123].

Agricultural Enterprises: Agricultural enterprises that are effectively taxed serve as organizations involved in the production, processing, and distribution of agricultural products within the supply chain. In the research conducted by Cao & Tao (2025), agricultural enterprises were defined as organizations primarily engaged in the production of agricultural raw materials (excluding processing activities), acting as the central producers in the agricultural product supply chain and bearing primary responsibility for ensuring product quality [124]. As core supply chain enterprises, they typically possess greater market power, superior information channels, and stronger financial capacity [114]. They acquire agricultural products from farmers and consider their own shares and associated costs when setting market prices [121]. This can lead to an imbalance of power, manifesting in practices such as imposing self-favoring contract terms [111] or transferring the costs of sustainability compliance to upstream producers. Large-scale agricultural enterprises can exert an impact on policy formulation through lobbying, job creation, and tax contributions, while simultaneously remaining subject to governmental oversight [125]. In some cases, they may even give rise to regulatory capture, thereby weakening the authority of the government.

Government: The government plays a crucial regulatory and guiding role in the agricultural product supply chain. By establishing quality standards, enhancing supervision, and conducting random inspections, it encourages agricultural enterprises to self-regulate and reduces instances of non-compliance [126]. Moreover, the government intervenes in supply chain governance through measures such as providing subsidies and reducing carbon taxes, thereby promoting low-carbon applications within the agricultural product supply chain [121,122]. Koo-Oshima et al. (2023) noted that organizations such as the United Nations Food and Agriculture Organization and the International Commission on Irrigation and Drainage collaborate through data mobilization, innovation, and technology to guide agricultural development, reflecting the government’s leading role in fostering collaboration and advancement in the agricultural sector [112].

Non-Governmental Organizations (NGOs): As key non-profit entities in the agricultural field, non-governmental organizations play an important role in promoting self-regulation among supply chain participants, ensuring product quality, and supervising production processes [127,128]. They influence the quality of agricultural products and the safety of regulatory procedures by establishing industry standards, monitoring compliance, and providing resources and training to agricultural enterprises [129]. Additionally, NGOs are crucial in exposing corporate misconduct and can help curb improper practices through targeted on-site inspections and third-party testing [130].

Consumers: Consumers play an important role in ensuring the quality of the agricultural product supply chain. Their reports can break the “information silos” of agricultural enterprises, serving as a vital tool for quality governance [124]. When combined with government enforcement mechanisms, consumer reports can significantly reduce opportunistic behavior by agricultural enterprises [131]. Research by Godrich et al. (2022) in Australia during the COVID-19 pandemic revealed that consumers preferred locally sourced food, demonstrating the influence of consumer decisions on the distribution of the agricultural product supply chain [132].

(2) Collaboration Patterns

Supply chain collaboration refers to the process whereby multiple node enterprises within the supply chain work closely together to achieve shared goals and benefits; the collection of main actors at each node is collectively referred to as stakeholders [133]. The agricultural product supply chain system involves multiple key stakeholders, including agricultural enterprises, government, consumers, and non-governmental organizations, forming a highly complex and interdependent network [134]. Stakeholders establish specific business relationships capable of generating competitive advantages based on the principles of mutual trust, openness, risk sharing, and reward sharing, thereby enhancing overall performance [135]. Barratt et al. (2004) categorizes supply chain cooperation into vertical and horizontal forms: vertical cooperation pertains to collaboration between upstream and downstream segments of the supply chain, while horizontal cooperation involves resource sharing among entities at the same industry level [136]. Based on this classification, this article briefly divides the multi-actor collaboration models within the agricultural product supply chain into the following three types.

Cross-Regional Collaboration: Due to the growing intensity of international agricultural supply chains, agricultural companies—operating as upstream and downstream members—are now present in diverse countries (or regions) worldwide. By performing economic functions characteristic of these national (regional) boundaries, a cross-regional agri-product supply chain network has emerged [137]. Depending on their attributes and stakeholders, cross-border agricultural supply chains can be categorized into various dominant types, such as those driven by cross-border e-commerce platforms and logistics parks [138]. Gao & Zong (2024) examined a case in which leading Chinese supply chain enterprises participated in benefit sharing within cross-regional agri-product supply chains [139]. They systematically elaborated on the key elements and theoretical logic of benefit sharing among supply chain members and constructed a “premise-action-result” theoretical model for the multi-party benefit sharing mechanism in cross-regional agri-product supply chains. Aerni et al. (2015) discovered that by strengthening triangular cooperation and enhancing institutional capacity, it is possible to rectify the mismatch between the external supply of individual capacity building and the actual demands for institutional capacity building in the three regions studied, thereby enabling the national agricultural innovation system to be more demand-oriented and responsive to the needs of domestic smallholder farmers [131].

Industry-University-Research Collaboration: Xu & Yao (2022) developed a collaboration model between smallholders and agricultural service providers, which offers balanced, multi-objective supplier scheduling schemes for farmers and serves as an effective decision-making tool for sustainable agriculture [140]. This research introduces new perspectives and methodologies for optimizing supply chain scheduling to enhance the efficiency of agri-product supply chains. Falcão et al. (2024) employed farmer associations and educational institutions as intermediaries to distribute questionnaires, engaging multiple stakeholders in the assessment and management of soil issues [141]. This approach achieved an integration of production and research for agricultural development and provided a practical example of multi-party cooperation to address real-world agricultural challenges. Settle & Garba’s (2011) study revealed that after participating in seasonal farmer field schools, farmers not only experienced gains in yield and income but also made significant progress in reducing chemical pesticide use and improving the application of fertilizers and organic amendments [142]. This indicates that when farmers engage in specific educational activities, the sustainability of agricultural production is enhanced. However, in current agricultural extension practices, national innovation agendas are often led by non-governmental organizations, with smallholders mostly acting as passive recipients. This limits the potential for deep collaboration and inclusive stakeholder engagement; hence, establishing a “farmer-organization-government” tripartite collaboration mechanism is crucial to ensure that technology promotion remains aligned with the actual needs of farmers [143]. Catarino et al. (2021) in cooperation with stakeholders, designed three scenarios and evaluated self-sufficiency, sustainability, and vulnerability at the farm, community, and regional levels under different conditions [144]. Their findings indicated that the majority of performance indicators at the regional level improved under collaborative scenarios, thereby providing scenario-based practical evidence and decision-making references for the sustainable development of agri-product supply chains. Kim (2024) argued that standardized metrics, interdisciplinary collaboration, and user-friendly interfaces can enhance platform accessibility, thus promoting progress in agricultural research [145]. Moreover, the government should assume a guiding role in agriculture by highlighting the benefits of sustainable practices and disseminating detailed information on sustainable agricultural strategies to bridge the knowledge gap among smallholders [146].

Platformized Collaboration Model: A real-time, transparent platform can foster active stakeholder participation [124]. An et al. (2024) noted that during the design and promotion of platform technologies, engaging farmers in equitable dialogue, proactive interactive cooperation, and effectively integrating smallholder feedback can enhance the technology’s acceptability and significantly boost agricultural productivity [147]. Their findings underscore the importance of farmer involvement in technological innovation and application. Agyekumhene et al. (2020) posited that embedding farmer participation in the technology design process can effectively reduce the difficulty for farmers in using digital devices [148]. Such farmer empowerment mechanisms highlight the positive impact of farmers on communication and collaboration within the agri-product supply chain, thereby enhancing on-chain collaborative capabilities. Reinforcing traditional farmers through government-led infrastructure development and training, along with providing technological empowerment via platforms, is also a manifestation of platformized collaboration. Chen et al. (2024) demonstrated that a collaborative model involving the government, platforms, and farmers exerts a positive influence on strengthening the relationship between farmers and rural live-stream e-commerce [149]. The study further explored the impact of government support, platform support, and social learning on farmers’ acceptance of agricultural live-stream e-commerce platforms and recommended that the government devise differentiated promotion strategies to help e-commerce platforms leverage their technologies to provide efficient services for farmers. Dawson et al. (2024) analyzed the process by which integrated production data, user needs, and third-party services enabled agricultural conglomerate cooperatives to co-create a decentralized digital platform, thereby achieving resource synergy and empowering farmers [150]. Basel et al. (2023) introduced a multi-stakeholder decision support system—WATERMED, an intelligent agricultural tool optimization platform that serves different users (such as farmers, irrigation communities, or governments) and plays a pivotal role in strengthening water governance strategies through multilateral decision-making [27].

3.3.3. Factors Influencing Multistakeholder Collaborative Willingness

(1) External environment

The agricultural product supply chain is confronted with six major environmental turbulences—policy regulation, market competition, climate risk, price fluctuations, geopolitical issues, and economic cycles [151]—with market uncertainty and the policy environment serving as the core external variables influencing collaboration. An empirical study conducted by Molist et al. (2025) on the Catalonian tomato supply chain indicates that economic feasibility is the primary consideration for stakeholders engaging in short supply chain collaborations, while environmental and social benefits must be converted into economic value to stimulate collaborative motivation [152]. In the context of dramatic economic fluctuations, moderate economic pressure can generate a demand for cooperation [153], and government subsidies can mitigate collaboration risks, significantly enhancing the sustainability of such collaborations [121]: for instance, the Indonesian government has effectively promoted stakeholder collaboration by disseminating models and providing organic agriculture subsidies, thereby lowering the participation threshold for small farmers in high value-added chains [154].

(2) Trust and perceived reliability

Trust forms the cornerstone of collaborative relationships, which can be maintained through three dimensions: contractual trust, competence trust, and goodwill trust [155]. Research by Pezeshki et al. (2013) has confirmed that incorporating trust into the benefit distribution mechanism enhances efficiency compared to traditional models [156]. Non-profit organizations can bolster partner trust by empowering resources and endorsing legitimacy, thereby resolving the “strategic paradox” within sustainable value chains [157].

The actual reliability of, and the perceived reliability among, node enterprises in the supply chain also affect the collaborative relationships [153]. When members are unable to ascertain the reliability of their relationships, forming collaborations becomes challenging. When the level of trust within the chain is sufficiently high, particularly under conditions of demand uncertainty, trust can reduce coordination costs and curb speculative behaviors [158].

(3) Power Structures and Perceptions of Fairness

Power asymmetry constitutes a critical structural factor influencing the willingness of multiple actors to engage in collaboration. For instance, when dominant large retailers force small and medium cooperatives to accept short-term contracts—can undermine cooperative stability. This imbalance in power dynamics is a structural factor contributing to “weak multistakeholder collaborative relationships” [111]. A survey conducted by Morton et al. (2017) among farmers in the Midwestern United States found that a disconnect exists between their risk perception and willingness to act due to information asymmetry and unfair social norms, leading them to favor maintaining the status quo [159]. Similarly, when farmers are unfamiliar with the service models offered by agricultural service institutions, their rights to engage in modern agricultural production are, to a certain extent, undermined [140].

Groot-Kormelinck et al. (2021), through a case study of the Uruguayan food chain, revealed that the alignment between quality standards and contract types determines the perceived fairness of power distribution, which in turn affects the sustainability of collaboration [160]. Widadie et al.’s (2022) study on the Indonesian vegetable supply chain demonstrates that clarity in contractual arrangements and the depth of resource support can directly improve supply chain standardization, thereby affecting participants’ willingness to engage [154].

This asymmetry does not merely refer to a straightforward “who dominates whom” situation. Instead, it represents a dynamic and multi-tiered power structure. The Figure 6 below clearly illustrates the primary power relationships and their directions among all parties:

(4) Interest coordination and risk sharing

Different profit distribution and risk-sharing contracts exert heterogeneous effects on collaborative behavior. Income stability, payment cycles, and supply chain adaptability form the “iron triangle” that underpins farmers’ willingness to engage in cooperation [120]. If the supply chain adopts a tightly integrated interest-linkage mechanism—such as contract transactions or horizontal cooperation—it will better facilitate collaboration among parties and enhance income stability [104]. As a common-interest mechanism, resource sharing that can foster collaboration among actors, boost operational flexibility, and generate synergistic benefits through resource integration [161]. In an omni-channel model, when the return-sharing factor is relatively high, profit-sharing contracts can enhance both the profitability of individual supply chain members and the overall performance of the supply chain [162].

However, to leverage profit distribution to secure supply chain partnerships, it is essential to ensure that the distribution ratios are acceptable to all members along the chain. Research by Ge & Zhang (2014) indicates that the stability of a sharing strategy depends on the degree of alignment between the risk compensation coefficient and resource investments, the system tends toward a cooperative equilibrium when shared gains exceed an individual’s reservation utility by 20% [163]. When profit distribution deviates from the proportions of input, the likelihood of weaker members exiting increases significantly [161].

These factors do not operate in isolation. For instance, information sharing can strengthen resource complementarities in an omni-channel model [120], and policy subsidies may indirectly boost trust by enhancing the fairness of profit distribution [121].

Therefore, contract designs for multi-agent collaborations must balance channel conflicts. Distributing residual profits in a three-tier chain comprising manufacturers, distributors, and retailers through contractual negotiations can enhance overall welfare [164]. In the realm of policymaking, governments should fully leverage their regulatory role by establishing clear mechanisms to address power asymmetries and prevent dominant firms from abusing their market power. Instruments such as fair trade certification, legally binding equitable contracts, and digital platforms owned by producer cooperatives can serve as countervailing forces to ensure a more equitable distribution of value [159]. Future research should further elucidate the coupling pathways and the evolving weight of these factors to provide theoretical support for differentiated collaboration strategies.

3.4. Digital-Driven Sustainable Development Path for Agricultural Supply Chains

This section summarizes the development potential of agriculture in economic, environmental, and social sustainability, and proposes corresponding guidelines.

3.4.1. Strengthening Technological Applications to Build a Solid Development Foundation

In the agricultural sector, digital technologies influence the entire value chain [165]. By strategically applying digital technologies—which include intelligent data management, automation of production processes, and the establishment of comprehensive traceability systems—the agricultural supply chain can significantly enhance its environmental sustainability performance [42]. Specifically, emerging technologies such as the IoT, blockchain, and big data serve as core driving forces for the sustainable development of agricultural supply chains. Precision agriculture technologies can optimize resource utilization efficiency; blockchain-based traceability systems can reinforce full-chain monitoring [166], ensuring data security and reliability while increasing supply chain transparency and traceability [167]; big data analytics can mine consumer demand to aid precise decision-making [168], improve supply chain agility, optimize inventory and marketing strategies, and steer the supply chain toward greater intelligence and efficiency [169]; and IoT-enabled smart monitoring, which facilitates real-time information collection and sharing, can help reduce losses and improve labor conditions for practitioners [170]. These digital transformation measures systematically lower the ecological footprint of agricultural production, directly promoting biodiversity conservation while providing the technological foundation for constructing an environmentally friendly food system.

3.4.2. Integrating Supply Chain Resources and Optimizing Resource Allocation

Sustainable development in agriculture necessitates a keen focus on the interconnection between economic sustainability and the current feasibility of agricultural enterprises. The efficient allocation of organizational, technological, human, financial, and intangible resources forms the foundation for sustainability [166]. Proper coordination of resources within the agricultural supply chain has a positive impact on fostering strategic relationships, enhancing resource utilization efficiency, and improving ecological performance [8]. For instance, market demand can be forecasted through data analysis to plan production and logistics effectively, thereby reducing inefficiencies and saving costs, utilizing appropriate supply chain resources during planning to address gaps between demand and supply [171,172]. Shukla & Tiwari (2017) suggest that decision-making regarding agricultural products should be data-driven while fully recognizing the significant role that smallholder farmers play in agricultural development [172]. Incorporating smallholder farmers into the supply chain and establishing robust institutional arrangements can contribute to increased productivity [173], rural development, and land conservation [36], thus advancing the social sustainability goals of the agricultural supply chain. Another major challenge for sustainable agriculture is to implement agricultural processes more efficiently at lower costs, thereby providing safer and better working conditions for both the environment and all stakeholders. Government initiatives, investments, and meeting substantial infrastructure needs are critical means to achieve this [174]. Once resource synergies are established, the overall competitiveness of the supply chain will be significantly enhanced.

3.4.3. Enhancing Data Analysis Capabilities to Strengthen Supply Chain Resilience

Sustainability driven by agriculture should accommodate increased responsiveness and resilience, thereby enhancing security [11]. Data analysis capability is vital for the sustainable development of the agricultural supply chain. By establishing a digital integrated management platform, traditional information silos can be dismantled, enabling information sharing and collaborative operations across all links of the supply chain [166]. Utilizing digital technologies can bolster the supply chain’s risk resistance—for example, by constructing a stable, efficient, and low-cost cold chain logistics network to ensure the quality and safety of agricultural products. The comprehensive application of descriptive, predictive, and prescriptive analysis methods allows for deeper exploration of data value, providing a scientific basis for supply chain decision-making, optimizing operations [175,176], improving resource utilization efficiency, and ultimately achieving sustainability goals.

3.4.4. Enhancing Information Exchange and Improving Supply Chain Visibility and Collaboration

Supply chain visibility is crucial for the sustainable development of the agricultural supply chain, as it centers on creating an information-sharing mechanism that spans the entire chain. According to Schoenthaler (2018), a visibility system facilitates the real-time transmission of key information both within and outside the organization, thereby enabling dynamic monitoring of the entire process from service acquisition to final delivery [177]. Ahoa et al. (2018) relying on the Supply Chain Operations Reference (SCOR) model, developed an analytical framework that systematically collects and analyzes information at core stages such as planning, procurement, production, and delivery, thereby establishing a mechanism for operational visibility along the chain [178]. When combined with data processing technologies, this approach can effectively enhance the efficiency of information collaboration and the overall responsiveness of the supply chain [166].

3.4.5. Establishing a Circular Economy Concept and Optimizing Industrial Processes

Minimizing waste and valorizing by-products are key to effective management and enhancing sustainability in the food industry [179]. The construction of sustainable agricultural supply chains must go beyond traditional agricultural production by developing a systematic transformation strategy based on circular economy theory, thereby fully leveraging its role in optimizing resource flows and reducing waste [180]. The solutions should not be confined solely to agricultural production but should cover the entire supply chain, including food processing, packaging, distribution, and consumption. Key nodes—such as post-harvest preservation [178], energy efficiency in processing stages, degradability of packaging materials [181], the carbon footprint of distribution networks, and waste recycling at the consumption end [171,182,183]—must be optimized in a coordinated manner [184,185]. The circular economy concept also plays a role in reducing waste by influencing consumer behavior; for example, research by Drago et al. (2020) indicates that smart packaging can extend the shelf life of perishable foods by 20–40% [186], while Gualandris et al. (2023) have demonstrated that using dynamic labels (such as “best before” instead of “expiration date”) can effectively reduce household food waste by 22% [187].

4. Discussion

4.1. Digital-Enabled Shortening of Agri-Food Supply Chains

4.1.1. Environmental and Economic Dividends

Logistics-related CO₂ reduction: Digital direct-to-consumer channels eliminate intermediate consolidation hubs, shortening haulage distances. Platforms that combine front-distribution centers with pooled direct-delivery schedules simultaneously raise drop density and cut per-parcel emissions.

Lower physical losses: Conventional wholesale routes involve three to four handlings; digital sourcing reduces this to one or two. The shortened pipeline also decreases calendar time from field to fork, markedly curbing spoilage of perishable produce.

Farm-gate premium and consumer savings: Order-based procurement and real-time price transparency shield farmers from asymmetric-information down-pricing while trimming intermediary mark-ups, yielding a Pareto improvement that raises farm-gate prices and lowers retail prices.

Triggering regional circularity: High-frequency demand data generated by short chains attract satellite industries—small-scale processors, composting facilities, cold-chain maintenance services—thereby activating a peri-urban circular economy.

4.1.2. Implementation Challenges

Many smallholders lack the digital literacy, technical skills or infrastructure required to engage with e-commerce platforms. Inefficient last-mile logistics, payment-security concerns and consumer distrust of online food purchases remain significant barriers. Moreover, existing regulatory frameworks often fail to accommodate these emerging business models, particularly in rural and developing regions.

In summary, digitally enabled chain shortening offers a triple dividend—loss reduction, carbon mitigation and income gains—but fully realizing these benefits hinges on parallel innovation in cold-chain infrastructure, inclusive governance and risk-sharing mechanisms. Future research should identify the dynamic equilibrium among short-chain performance, scale thresholds and policy intervention, using longitudinal case studies and econometric models to quantify optimal chain length across commodities and regions.

4.2. Key Challenges and Adoption Barriers

While digital transformation presents immense opportunities for agricultural supply chains, its implementation is fraught with multifaceted challenges that extend beyond technological availability. Understanding these barriers is crucial for policymakers, technology developers, and supply chain actors to design effective intervention strategies.

4.2.1. Economic and Financial Barriers

The high upfront cost of digital infrastructure (e.g., sensors, IoT gateways, communication modules) is a primary deterrent, especially for smallholder farmers and small-to-medium agri-enterprises (SMAs) [62]. The lack of clarity regarding the ROI and the long payback period further exacerbate financial hesitancy [53,109]. Many farmers operate on thin profit margins and are risk-averse, making them reluctant to invest in technologies whose economic benefits are not immediately tangible or guaranteed. Furthermore, access to credit and favorable financing options for digital adoption remains limited in rural areas of developing countries.

4.2.2. Technical and Infrastructural Barriers

The effective deployment of digital technologies hinges on reliable supporting infrastructure, which is often lacking in agricultural heartlands. Unstable or absent internet connectivity, inadequate electricity supply, and poor road networks severely limit the functionality of data-intensive applications [66,67]. Even when infrastructure exists, issues of technological interoperability arise. The absence of universal data standards leads to “data silos”, where systems from different vendors cannot communicate seamlessly, hindering chain-wide integration [76]. Additionally, the complexity of some digital tools creates a usability challenge, as farmers may lack the technical proficiency to operate and maintain them effectively.

4.2.3. Social and Human Capital Barriers

A significant digital literacy gap continues to exist among numerous farmers and workers throughout the agricultural supply chain. Resistance to change, coupled with skepticism towards new technologies—often rooted in deeply entrenched traditional practices [81]. This reluctance is further compounded by escalating concerns regarding data privacy and ownership. Farmers are increasingly wary about who collects their data, how it is utilized, and who stands to benefit from it, which fosters distrust toward technology providers and larger agribusiness firms [105,109]. Moreover, the process of digital transformation may exacerbate existing social inequalities. There exists a tangible risk that large, well-capitalized farms will capture the majority of benefits associated with these advancements, thereby widening the gap between them and smallholders who lack the resources necessary for adopting such technologies—a phenomenon commonly referred to as the “digital divide” [113]. If these barriers remain unaddressed, they could impede progress in digitalization and undermine the sustainability of the supply chain.

4.2.4. Governance and Collaboration Barriers

The agricultural supply chain is characterized by fragmentation and involves numerous stakeholders with often misaligned interests [104]. A major barrier is the power asymmetry between large retailers/processors and small-scale producers. Without fair governance mechanisms, digital platforms can be used to reinforce the dominance of powerful actors rather than creating equitable value for all [159]. The lack of trusted collaboration models and the absence of clear rules for data sharing and benefit distribution create a reluctance to engage in the transparency that digital systems require [154]. Furthermore, outdated or restrictive policies and regulations that lag behind technological innovation can create legal uncertainty in the use of drones, data rights confirmation and smart contract generation.

Having established the common barries failure and their underlying causes. As Box 1 elucidates, these examples reveal recurring themes that must be addressed.

In conclusion, overcoming these barriers requires a coordinated, multi-stakeholder approach that combines technological innovation with financial incentives, capacity-building programs, infrastructure development, and the establishment of fair governance frameworks that ensure inclusive and sustainable digitalization.

Box 1. Learning from Digital Roll-out Failures.

• Local food platform cases in Australia during the COVID-19 period [132]

Failure/obstacle scenario: Forecast lag left the chain unable to absorb the “spike-then-cancel” wave, inflicting financial losses.

Lesson: The demand forecasting model needs to be constrained by real-time logistics; otherwise, it will inversely magnify food waste and user churn.

2.

Italian smart greenhouse IoT pilot [53]

Failure/obstacle scenario: 30% of the sensors failed within six months, mainly due to high humidity/pesticide corrosion and lack of calibration.

Lesson: Agricultural IoT devices require IP67 or higher protection and regular calibration protocols; otherwise, data quality will deteriorate rapidly.

3.

The UK’s fresh fruit and vegetable system addresses water risks [109]

Failure/obstacle scenario: Relying solely on imports and lacking redundant production areas, which leads to supply chain disruptions during climate change.

Lesson: The “resilience” of the supply chain cannot rely on a single node; it requires a redundant layout of multiple regions and varieties.

4.3. Feasible Hypotheses of the TCS Framework

Based on the research framework and the above-mentioned study of the previous literature, we propose the following propositions to guide future empirical research:

P1(Technology Enablement as Moderator): The positive impact of digital technology adoption on sustainability performance is positively moderated by the maturity of multi-agent collaboration mechanisms. That is, the full potential of technology can only be realized under effective institutional collaboration.

P2 (Centrality of Institutions): Multi-agent collaboration mechanisms play a key mediating role between technology enablement and sustainability performance. Technology does not directly create sustainable value; rather, it achieves value by reshaping the rules of interaction, power structures, and benefit distribution among participants.

P3 (Co-evolutionary Nature): There exists a dual feedback loop among technology, collaboration, and sustainability. Sustainability goals can steer the direction of technological innovation (e.g., developing lower-power sensors), while emerging technologies can foster new collaborative models (e.g., platform-based cooperation), thereby redefining the meaning and achievable pathways of sustainability.

P4 (Context Dependency): The form and effectiveness of the above co-evolutionary pathways are significantly moderated by external contextual factors, such as the policy environment, power structure within the supply chain, and product perishability.

This propositions translate a descriptive framework into an operationalizable and measurable theoretical statement, thereby facilitating a shift in the field from narrative review to systematic theory construction and empirical validation.

4.4. Limitations and Future Research Directions

4.4.1. Limitations

This study has several limitations that point to valuable directions for future research:

Empirical Validation of the Theoretical Framework: The proposed integrated TCS framework requires further empirical testing across diverse regions, crop types, and agricultural supply chains of varying operational scales to verify its applicability and explanatory power.

Potential Bias in Literature Sources: As this research is based on the existing literature and case studies, it may be subject to publication bias or context-specific limitations. Future studies could employ quantitative methods or comparative multi-case analyses to enhance the generalizability and robustness of the findings.

An urgent need for in-depth microeconomic analysis: Future research should adopt more rigorous econometric methods (such as cost–benefit analysis and real option models), simulation, or longitudinal studies to quantify the costs, benefits, and risks of digital transformation, thereby providing more accurate economic evidence for investors and decision-makers.

4.4.2. Future Research Directions

Future research should: Develop measurable constructs for TCS performance, and employ structural equation modeling (SEM) or case-based comparative analysis to test the causal relationships within the TCS framework. For example:

RQ1: Which governance models support effective multi-stakeholder collaboration? What factors lead to its sustainability or abandonment?

RQ2: How does digitalization affect the resilience and income of small-scale farmers? Through simulation modeling, what is the minimum viable scale (e.g., number of farmers, hectares) to become economically sustainable? After small-scale farmers adopted IoT-based precision irrigation, did the fluctuation (coefficient of variation) of crop yield per unit area significantly decrease, and what was the relationship between the extent of the decrease and the size of the plot?

RQ3: Short Food Supply Chains (SFSCs) through what path does it reshape the relationship between producers and consumers? Does the price premium and market assurance derived from SFSCs effectively lower the perceived risk and economic barriers for smallholders? Is this adoption driven primarily by economic incentive, or also by the construction of a new “producer identity” and social recognition within the short-chain community?

5. Conclusions

This paper systematically reviews the domestic and international literature and summarizes the research progress of the digital transformation in the agricultural supply chain from three dimensions: driving forces, technology, and pathways, leading to the following key conclusions:

5.1. Multidimensionality of Driving Forces Analysis

Existing studies generally attribute the digital transformation of agricultural supply chains to both endogenous and exogenous factors. Endogenous drivers include the need to enhance supply chain efficiency (e.g., reducing losses and optimizing inventories), the pressure of cost control (such as insufficient coordination between logistics and information flows), and stringent requirements for traceability systems. Exogenous forces comprise policy drivers (for instance, China’s “Digital Countryside” strategy), evolving consumer demands (such as the need for transparency and traceability), and technological iterations (such as the proliferation of 5G and edge computing). Notably, there are significant differences in drivers between developing and developed countries; the former focus more on infrastructural enhancements, while the latter concentrate on deeper technological applications.

5.2. Integration Characteristics of the Technology System

The technological framework for digital transformation can be characterized by a three-tier structure: the perception layer, the transmission layer, and the application layer. In the perception layer, IoT and sensor technologies have been extensively employed in applications such as monitoring agricultural production (e.g., temperature and humidity control) and managing storage environments. The immutability provided by blockchain technology in the transmission layer effectively addresses trust issues within the supply chain. At the application layer, big data analytics and artificial intelligence technologies have shown remarkable performance in demand forecasting (e.g., using machine learning models based on historical sales data) and route optimization (e.g., planning drone delivery paths). Existing research indicates a significant positive correlation between the degree of technology integration and supply chain performance, although barriers to technology adoption (such as the digital literacy of small- and medium-sized farmers) remain a major constraint.

5.3. Mutually Beneficial Synergy Through Multi-Stakeholder Collaboration

Multi-stakeholder collaboration is at the heart of value creation in the digital transformation of agricultural supply chains. For long-term and robust development, it is essential to abandon the traditional model of isolated breakthroughs and build an ecosystem in which governments, enterprises, farmers, consumers, and research institutions collaborate. The core logic lies in fully leveraging the differentiated resources provided by various stakeholders (e.g., government policies and infrastructure, technological capital from enterprises, and production data from farmers) to achieve resource complementarity; enhancing chain-wide value addition through data sharing and process optimization (for example, consumer demand data feeding back to the production side to reduce losses, and government subsidies encouraging enterprises to disseminate technology further); and distributing the risks associated with technology implementation costs and market uncertainties among multiple parties, thereby boosting the resilience and risk-handling capabilities of the supply chain.