Biologically-Grounded AI for Plant Stress Detection and Real-Time Crop Intelligence

Artificial intelligence in agriculture has rapidly evolved, enabling the detection of plant stress through advanced imaging, spectral analysis, and sensor-based monitoring. Despite these technical strides, prevailing approaches predominantly rely on pattern recognition, often divorced from the underlying physiological mechanisms that drive plant responses to stress. This disconnect poses challenges around interpretability, robustness across diverse environments, and actionable reliability for field applications. While the use of multimodal sensing and autonomous systems is on the rise, a critical gap persists in the integration of explicit biological knowledge—specifically, how plants’ physiological and biochemical responses to various stresses manifest across observational datasets. Recent studies have begun to highlight measurable changes in pigment content, water relations, tissue organization, and metabolism as key indicators of stress, yet comprehensive frameworks that translate these biological phenomena into AI-driven insights remain rare. This Research Topic aims to bridge the gap between plant physiological understanding and next-generation artificial intelligence applied to plant stress detection. The objective is to foster methodologies that embed knowledge of plant function directly into AI models, thereby enhancing the scientific credibility, interpretability, and practical relevance of crop intelligence systems. Key questions include: How can AI systems be constructed to recognize not just statistical patterns, but to identify causal links between sensor data and underlying plant stress mechanisms? What representations or models best facilitate real-time, biologically-informed decision support for growers? To gather further insights in the integration of physiological knowledge with AI-driven crop monitoring, this Research Topic welcomes contributions that tackle the underlying biology of plant stress in tandem with advanced sensing and computation. Submissions may address, but are not limited to, the following themes: • Physiology-informed machine learning and model development for plant stress detection • Linking spectral, thermal, and imaging data to biochemical and cellular processes • Mechanistic interpretation of remote sensing signals in relation to plant health • Cross-scale modeling approaches from cellular to canopy level • Integration of biochemical, structural, and phenotypic data into AI frameworks • Early detection strategies rooted in biological understanding • Real-time monitoring and decision support guided by functional plant responses • Interpretable AI models enhancing scientific validity and practical agricultural outcomes We encourage submissions from researchers in plant biology, remote sensing, artificial intelligence, and agronomy who are advancing interpretable, biologically-grounded approaches to monitoring and managing crop health. If relevant, article types may include original research, reviews, methods, perspectives, and case studies. Topic editors: Mohamed Farag Taha, Alwaseela Abdalla Mohamed Hassan, Mustafa Mhamed, Alfadhl Alkhaled Source: https://www.frontiersin.org/research-topics/79758/biologically-grounded-ai-for-plant-stress-detection-and-real-time-crop-intelligence