Plant wearables in smart agriculture: advancements, technical challenges, opportunities, and a cyber-physical roadmap in volatile organic compound detection

The agricultural sector faces escalating challenges from climate-induced abiotic stressors, herbivory, and pathogens. Optimizing crop productivity requires advanced technologies for real-time monitoring. While volatile organic compounds (VOCs) serve as early-onset biomarkers for plant stress, conventional laboratory methods lack high-temporal resolution, and portable electronic noses suffer from limited selectivity. However, transitioning to in situ plant wearables is severely bottlenecked by critical research gaps: severe signal baseline drift caused by fluctuating ambient humidity crosstalk, stomatal shielding from non-porous substrates that disrupts natural leaf transpiration, and a computational void in deploying artificial intelligence for autonomous material stoichiometry optimization and real-time deconvolution of overlapping VOC profiles at the resource-constrained edge. To critically address these gaps, this review systematizes the transition toward Cyber-Physical Plant Systems (CPPS) via wearable VOC-sensing architectures. We evaluate the AI-material nexus, demonstrating how metaheuristic optimization and Explainable AI (XAI) frameworks like SHAP and LIME can autonomously optimize sensor materials and validate biogenic fingerprints. A central contribution of this work is a transdisciplinary and strategic four-phase technical roadmap integrating materials science, edge-computing, and Industry 5.0 principles. This framework outlines the engineering of transpirational-neutral architectures and low-power Edge-AI hardware to resolve multi-stressor profiles. By bridging the gap between plant metabolism and digital intelligence, this review provides a strategic pathway aligning with Agriculture 5.0 to build autonomous bio-information systems for enhanced agricultural resilience.