Pertanian & Lingkungan editorial
Open AccessOA2025
Multi-Modal LLMs in Agriculture: A Comprehensive Review
This review examines the application of Multi-Modal Large Language Models (MM-LLMs) in agriculture, addressing 11 research questions to outline current capabilities, limitations, and future directions for optimizing agricultural processes.
Ranjan Sapkota; Rizwan Qureshi; M. U. Hadi; Syed Zohaib Hassan; F. Sadak; Maged Shoman; Muhammad Sajjad; Fayaz Ali Dharejo; Achyut Paudel; Jiajia Li; Zhichao Meng; John M. Shutske; Manoj Karkeeยท IEEE Transactions on Automation Science and Engineeringยท 2025ยท DOI 10.1109/TASE.2025.3612154
The core problem
Multi-Modal Large Language Models (MM-LLMs) have rapidly emerged across scientific fields, yet their applicability in agriculture remains only partially explored. This paper conducts an in-depth review of MM-LLMs in agriculture, focusing on how these models can be developed and implemented to optimize agricultural processes, increase efficiency, and reduce costs. Recent studies have explored MM-LLM capabilities in agricultural information processing and decision-making, but significant gaps persist. These include domain-specific challenges such as variable data quality and availability, integration with existing agricultural systems, and the creation of robust training datasets that accurately represent complex agricultural environments. Moreover, a comprehensive understanding of the capabilities, challenges, and limitations of MM-LLMs in agricultural information processing and application is still missing. To bridge this gap, the survey reviews the progress of MM-LLMs and their utilization in agriculture, with an additional focus on 11 key research questions (RQs)โ4 general and 7 agriculture-focused. By addressing these RQs, the review outlines current opportunities, challenges, limitations, and a future roadmap for MM-LLMs in agriculture.
Innovation
The authors conducted a systematic survey of the literature on MM-LLMs and their application in agriculture. The review is structured around 11 research questions: 4 general RQs that cover the fundamentals of MM-LLMs (e.g., architectures, training paradigms, and evaluation metrics) and 7 agriculture-specific RQs that address domain challenges such as data quality, system integration, and dataset creation. The methodology involved identifying relevant studies, categorizing them by application area, and synthesizing findings to provide a benchmark for the current state and emerging trends. An organized and regularly updated list of papers on MM-LLMs in agriculture is maintained at https://github.com/JiajiaLi04/Multi-Modal-LLMs-in-Agriculture. The review also includes a note to practitioners, emphasizing the need to optimize agricultural practices through MM-LLMs.
Introduction
Multi-Modal Large Language Models (MM-LLMs) have rapidly emerged across scientific fields, yet their applicability in agriculture remains only partially explored. This paper conducts an in-depth review of MM-LLMs in agriculture, focusing on how these models can be developed and implemented to optimize agricultural processes, increase efficiency, and reduce costs. Recent studies have explored MM-LLM capabilities in agricultural information processing and decision-making, but significant gaps persist. These include domain-specific challenges such as variable data quality and availability, integration with existing agricultural systems, and the creation of robust training datasets that accurately represent complex agricultural environments. Moreover, a comprehensive understanding of the capabilities, challenges, and limitations of MM-LLMs in agricultural information processing and application is still missing. To bridge this gap, the survey reviews the progress of MM-LLMs and their utilization in agriculture, with an additional focus on 11 key research questions (RQs)โ4 general and 7 agriculture-focused. By addressing these RQs, the review outlines current opportunities, challenges, limitations, and a future roadmap for MM-LLMs in agriculture.
Why it matters
The review discusses the opportunities and challenges of deploying MM-LLMs in agriculture. Key opportunities include improved efficiency, cost reduction, and enhanced decision-making through multi-modal data fusion. Challenges encompass domain-specific issues such as data scarcity, variability, and the need for specialized datasets. The authors outline a future roadmap that includes developing robust training datasets, integrating MM-LLMs with existing agricultural systems, and addressing ethical and practical considerations. The note to practitioners emphasizes that MM-LLMs have the potential to revolutionize how farms are managed and operated. The survey concludes that continued research and an organized, regularly updated list of papers (available at the provided GitHub repository) are essential for advancing the field.
Who should read this
Pembaca Research Digest dan praktisi rumpun ini.
Opening member contentโฆ