Smart Agriculture: Integrating Artificial Intelligence for Efficient Supply Chain Management
Year: 2026
Type: article
Abstract: Agricultural supply chains are under strain from erratic weather, unstable commodity prices, losses after harvest, and poorly coordinated logistics. Artificial intelligence (AI) offers tools that can support decisions at every stage from sowing to sale, yet the relevant research is scattered across the forecasting, optimization, sensing, and explainability communities. This paper presents a systematic review of thirty-one studies published between 2020 and 2026, selected through a four-stage PRISMA 2020 screening of 523 records drawn from major scholarly databases. Each study was coded by supply chain stage, technique, dataset, baseline, and reported metric, and the extracted results were compared using a common set of error and accuracy measures. The corpus falls into five themes: crop yield and cultivation recommendation, commodity demand and price forecasting, explainable AI (XAI), IoT and remote-sensing data ecosystems, and supply chain network optimization. The comparative analysis shows that hybrid deep-learning designs such as LSTM–Prophet, CNN–BiLSTM, and attention-based LSTM improve on statistical baselines, with a reported error reduction of about 18% against ARIMA and GARCH and a 12.8% accuracy gain against Transformer and genetic-algorithm baselines, while Random Forest remains the most dependable classical learner for crop recommendation at 92.7% accuracy in a field-deployed IoT system. Studies that linked forecasts to optimization reported downstream benefits, including a 57.9% fall in price variance, a 16.5% rise in farmer income, and 30–35% water savings. The main barriers are heterogeneous data, weak transferability across regions, uneven metric reporting, and limited attention to the energy cost of AI. Federated and edge learning, agricultural foundation models, common benchmarks, and Green AI practices are identified as priorities for future work.
Authors: P O, Supritha, P Hegde, Ganapatikrishna
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
Open Access status: hybrid
OpenAlex ID: W7225134407
FWCI:
Cites:
Cited by:
Related to: