Plant Wearable Sensors for Data-Driven Precision Agriculture: From Multilevel Sensing to Crop-State Assessment and Adaptive Management

Plant wearable sensors can resolve physiological responses that soil, weather, and canopy observations capture only indirectly. This critical narrative review evaluates optical, electrochemical, electrical, mechanical, thermal, and acoustic platforms across microenvironmental, growth, biochemical, and stress-related measurements. Crop-state assessment is defined as estimation of management-specific physiological variables, with independent reference measurements and growth-stage-dependent interpretation. The central trade-off is between continuous operation and physiological specificity: physical sensors offer relatively simple monitoring architectures, whereas biochemical sensors provide more selective information but impose greater calibration and maintenance demands. Neither analytical sensitivity nor short-term plant attachment establishes field readiness. We identify requirements for drift control, temporally aligned multimodal fusion, and data assimilation with explicit observation and uncertainty models. Physiological sentinel nodes are proposed as a sampling strategy whose within-plant and field representativeness must be tested before spatial extrapolation. Progress toward adaptive management requires staged validation from organ-level measurements to replicated field trials that demonstrate benefits beyond existing monitoring systems.