, and pH. We confirm the robust sensing capabilities and favorable biocompatibility of MLIPBS through cross-species validation in lettuce, tomato, and Aloe vera. Additionally, leveraging the LightGBM architecture, we demonstrate that MLIPBS successfully classifies combined stress conditions and varying intensity levels of acid and salt stress, achieving an average accuracy of 90.5%. We further show that the system identifies stress types and intensities within 8 hours of onset, providing an early-warning window at least 48 hours before symptom manifestation. Our study provides reliable wearable tools for stress-resistant crop screening and precision management in smart agriculture.
Zhou et al. (Thu,) studied this question.