Electrophysiological Assessment of Tomato Plants Under Environmental Stress Stimuli by Using Machine-Learning Augmentation and Circuit Component Analysis

Plants play a vital role in sustaining life by not only providing food but also medicines and many other natural resources. They constantly experience multiple environmental stresses; therefore, early detection of these stresses is essential for improving crop productivity and sustainability. Smart agriculture systems that combine sensing, data analytics and Machine Learning (ML) are emerging as promising tools for real-time crop stress monitoring and management. In this present study, 24 tomato plants are subjected to five different stress conditions, i.e., Sulfuric acid (H 2 SO 4 ), Sodium Chloride (NaCl), urea, dryness and excess water along with a control group of healthy plants to investigate their electrophysiological responses. This is done to observe how stress affects the plants by comparing them with normally grown plants. A minimally invasive signal acquisition circuit is designed to monitor electrophysiological signals and capture variations associated with stress. Supervised ML models are used to classify the stress conditions, achieving a maximum accuracy of 92%. These results demonstrate that electrophysiological signals monitoring, when combined with data-driven analysis, offers an effective and non-destructive means for early stress detection in plants and improves crop management.