Artificial intelligence and internet of things interventions for improving antioxidant levels in heritage crops: a perspective review

Abstract Heritage crops are recognized for their genetic diversity and naturally elevated antioxidant profiles, including polyphenols, flavonoids, and carotenoids. However, their nutritional value is highly sensitive to environmental fluctuations, and traditional agricultural practices often fail to balance the trade-off between yield and bioactive compound accumulation. This perspective review targets the convergence of vision-based artificial intelligence (AI) and the Internet of Things (IoT) as a solution to optimize these phytochemical traits. We synthesize recent advancements to demonstrate how AI-driven systems can shift from ’passive monitoring’ to the ’predictive modulation’ of plant metabolism, specifically managing ’eustress’ to enhance antioxidant biosynthesis. Key findings indicate that integrating computer vision with hyperspectral imaging allows for the pre-symptomatic detection of abiotic stress, while IoT-based sensing networks facilitate precise environmental control. Furthermore, we propose a framework integrating edge computing with cloud-based analytics to overcome data heterogeneity and latency challenges. This work outlines a forward-looking strategy for deploying AI to ensure the sustainable production of antioxidant-rich heritage crops, contributing to global nutritional security.