[Paper Review] Empowering Agrifood System with Artificial Intelligence: A Survey of the Progress, Challenges and Opportunities

Tao Chen, Liang Lv+13 more|arXiv (Cornell University)|May 3, 2023

Smart Agriculture and AI4 citations

TL;DR

This survey comprehensively reviews the application of artificial intelligence (AI) in agrifood systems, covering data acquisition, AI techniques like deep learning and traditional machine learning, and their use in agriculture, animal husbandry, and fisheries for tasks such as crop classification, yield prediction, and quality assessment. The key contribution is a structured analysis of AI's transformative potential, challenges like data heterogeneity and model interpretability, and opportunities in AIoT and foundation models for sustainable, efficient agrifood systems.

ABSTRACT

With the world population rapidly increasing, transforming our agrifood systems to be more productive, efficient, safe, and sustainable is crucial to mitigate potential food shortages. Recently, artificial intelligence (AI) techniques such as deep learning (DL) have demonstrated their strong abilities in various areas, including language, vision, remote sensing (RS), and agrifood systems applications. However, the overall impact of AI on agrifood systems remains unclear. In this paper, we thoroughly review how AI techniques can transform agrifood systems and contribute to the modern agrifood industry. Firstly, we summarize the data acquisition methods in agrifood systems, including acquisition, storage, and processing techniques. Secondly, we present a progress review of AI methods in agrifood systems, specifically in agriculture, animal husbandry, and fishery, covering topics such as agrifood classification, growth monitoring, yield prediction, and quality assessment. Furthermore, we highlight potential challenges and promising research opportunities for transforming modern agrifood systems with AI. We hope this survey could offer an overall picture to newcomers in the field and serve as a starting point for their further research. The project website is https://github.com/Frenkie14/Agrifood-Survey.

Motivation & Objective

• To provide a systematic review of AI's role in enhancing productivity, efficiency, safety, and sustainability in agrifood systems.
• To analyze data acquisition methods—including satellite, UAV, and sensor-based systems—used in AI-driven agrifood applications.
• To evaluate the performance and limitations of traditional machine learning and deep learning models in agrifood tasks such as crop monitoring, yield prediction, and disease detection.
• To identify key challenges, including data quality, model robustness, and interpretability, and to highlight emerging opportunities in AIoT and foundation models.
• To guide researchers and practitioners by offering a comprehensive overview and identifying future research directions in AI for agrifood systems.

Proposed method

• Systematically categorizes data sources in agrifood systems, including remote sensing (satellites, UAVs), GPS, and ground-based sensors, for monitoring crop and livestock conditions.
• Reviews traditional machine learning techniques such as support vector machines (SVMs) and their application in coarse-grained agricultural classification using remote sensing data.
• Analyzes deep learning models including convolutional neural networks (CNNs) for image-based disease detection in crops and animals, and long short-term memory (LSTM) networks for spatiotemporal modeling of climate and soil data.
• Examines multimodal AI frameworks that combine spatial and temporal data (e.g., CNN-LSTM hybrids) to improve yield prediction and growth monitoring.
• Explores interpretable AI techniques to address the black-box nature of deep learning, enabling model transparency and user trust in agrifood decision-making.
• Proposes an AIoT closed-loop framework integrating edge devices, cloud/fog computing, and foundation models to enable on-device inference through knowledge distillation and automated data labeling.

Experimental results

Research questions

• RQ1 How can AI techniques such as deep learning and traditional machine learning be effectively applied to improve monitoring, classification, and prediction in agriculture, animal husbandry, and fisheries?
• RQ2 What are the major data acquisition and processing challenges in agrifood systems, and how do different data sources (e.g., satellites, UAVs, sensors) support diverse AI applications?
• RQ3 What are the key limitations of current AI models in agrifood systems, particularly regarding robustness to distribution shifts and interpretability of predictions?
• RQ4 How can AI and the Internet of Things (AIoT) be integrated to create scalable, real-time, and traceable agrifood systems with closed-loop feedback mechanisms?
• RQ5 What role can foundation models and transfer learning play in enabling efficient deployment of AI on edge devices in resource-constrained agrifood environments?

Key findings

• Deep learning models such as CNNs and LSTMs significantly outperform traditional ML methods in tasks like crop disease detection and yield prediction by leveraging hierarchical feature representation from large-scale data.
• AI techniques applied to remote sensing data enable accurate crop classification and growth monitoring, with UAV-based imaging showing high spatial resolution benefits for precision agriculture.
• Interpretable AI methods are essential for building trust in agrifood systems, as they reveal decision-making logic and help identify key input factors influencing predictions.
• Robust AI models that generalize across varying environmental conditions (e.g., weather, location) are critical for real-world deployment and improved reliability in agrifood applications.
• The integration of AI with IoT creates a closed-loop evolution framework where foundation models pre-trained on edge-collected data can be distilled for on-device inference, enabling scalable and real-time agrifood monitoring.
• Text detection and recognition models enhance food traceability by enabling automated reading of product labels, supporting supply chain transparency and food safety.