Title: Leveraging Artificial Intelligence for Postharvest Aflatoxin Management in Ugandan Groundnuts- A Structural Equation Modeling Approach
Authors: Lillian Tamale,, Denis Ssebuggwawo, Drake Patrick Mirembe, Jude T. Lubega
Volume: 9
Issue: 6
Pages: 61-69
Publication Date: 2025/06/28
Abstract:
This study evaluated the influence of artificial intelligence-driven constructs on postharvest aflatoxin management in groundnuts using structural equation modeling (SEM) guided by the DeLone and McLean Information Systems Success Model with a sample of 268 participants. The model looked at how advanced feature extraction, real-time monitoring and decision support, and healthy groundnut detection directly and indirectly affected the control of aflatoxin. The results showed that real-time monitoring and decision support had a strong and positive effect on postharvest management (? = 0.591, p < 0.001), emphasizing how important AI-driven real-time information is for better decision-making. Although healthy groundnut detection and advanced feature extraction showed positive effects, their direct impacts on management were marginally insignificant (p = 0.062 and p = 0.094, respectively). However, both predictors strongly influenced healthy groundnut detection, which emphasizes the value of intelligent sensing and feature-based classification. The SEM showed very good fit results (?² = 0.00, RMSEA = 0.000, CFI = 1.000, TLI = 1.000), and the model accounted for 71.5% of the variation in the variables it measured. These findings point out the transformative value of AI systems in improving aflatoxin postharvest management through advanced monitoring and detection technologies.