PEMANFAATAN GENERATIVE AI DALAM MONITORING KESEHATAN TANAMAN

Parwito (Universitas Ratu Samban)
Eko Sumartono (Universitas Dehasen Bengkulu)

JOURNAL OF SCIENCE AND SOCIAL RESEARCH
Vol. 9 No. 3 (2026): June 2026

Abstract: The advancement of generative artificial intelligence (AI) opens new opportunities in precision agriculture, particularly in plant health monitoring and diagnosis. This study aims to systematically review studies discussing the use of Generative AI in plant health monitoring from 2019 to 2024. The method used is a Systematic Literature Review (SLR) following the PRISMA protocol (Preferred Reporting Items for Systematic Reviews and Meta-Analyses). Searches were conducted on Google Scholar, Scopus, Science Direct, and Portal Garuda databases using defined keywords. Of the 247 identified articles, 38 articles met the inclusion criteria and were analyzed in depth. The results show that Generative AI technologies, especially Generative Adversarial Networks (GAN), Diffusion Models, and vision-based Large Language Models (LLMs), have significantly contributed to plant disease detection, image data augmentation, and agronomic recommendation systems. Disease detection accuracy using GAN-based approaches averaged 94.7%, while visual transformer-based models recorded accuracy up to 96.3%. Key barriers include limited labeled datasets, technology gaps in rural areas, and lack of model adaptation to local Indonesian plant varieties. This study concludes that integrating Generative AI in plant health monitoring systems has great potential but requires multidisciplinary collaboration, local dataset development, and appropriate technology adoption policies. Keywords: Generative AI, GAN, Plant Health, Monitoring, Systematic Literature Review, Precision Agriculture   Abstrak: Perkembangan kecerdasan buatan (AI) generatif membuka peluang baru dalam bidang pertanian presisi, khususnya dalam monitoring dan diagnosis kesehatan tanaman. Penelitian ini bertujuan untuk mengkaji secara sistematis berbagai studi yang membahas pemanfaatan Generative AI dalam monitoring kesehatan tanaman dari tahun 2019 hingga 2024. Metode yang digunakan adalah Systematic Literature Review (SLR) dengan mengikuti protokol PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses). Pencarian dilakukan pada basis data Google Scholar, Scopus, Science Direct, dan Portal Garuda menggunakan kata kunci yang telah ditetapkan. Dari 247 artikel yang teridentifikasi, sebanyak 38 artikel memenuhi kriteria inklusi dan dianalisis secara mendalam. Hasil kajian menunjukkan bahwa teknologi Generative AI, terutama Generative Adversarial Networks (GAN), Diffusion Model, dan Large Language Model (LLM) berbasis visi, telah memberikan kontribusi signifikan dalam deteksi penyakit tanaman, augmentasi data citra, dan sistem rekomendasi agronomis. Tingkat akurasi deteksi penyakit menggunakan pendekatan berbasis GAN mencapai rata-rata 94,7%, sedangkan model berbasis transformer visual mencatat akurasi hingga 96,3%. Hambatan utama yang ditemukan meliputi keterbatasan dataset berlabel, kesenjangan teknologi di wilayah pedesaan, serta kurangnya adaptasi model terhadap varietas tanaman lokal Indonesia. Penelitian ini menyimpulkan bahwa integrasi Generative AI dalam sistem monitoring kesehatan tanaman memiliki potensi besar namun memerlukan kolaborasi multidisiplin, pengembangan dataset lokal, dan kebijakan adopsi teknologi yang tepat. Kata Kunci: Generative AI, GAN, Kesehatan Tanaman, Monitoring, Systematic