Understanding complex process sensor signal may be the source for deep Learning: A smart IoT delving into a tomato weight sensor

Abstract

With the rapid advancements of Internet of Things (IoT) technology, understanding sensor signals has become the key to acquiring plant physiological data and realizing the concept of “speaking plants”. This study developed a real-time and intelligent monitoring system using IoT-enabled weight sensors to continuously track the total weight of the shoot and root systems of tomato plants. A newly introduced variable, the photosynthetic leaf area index (LAIₚ), which plays a crucial role in capturing solar radiant heat, was used to balance energy components in the modeling of tomato transpiration processes. In this framework, a canopy transpiration model was developed and trained using data collected by sensors to enhance the online accuracy of tomato transpiration predictions. Supported by this model, in-depth analysis of weight sensor signals enabled the non-destructive, data-driven, and real-time extraction of LAI and LAIₚ. The parameter LAIₚ, the radiation absorbing portion of LAI, also serves as a measure for photosynthetic efficiency when related to the concurrently assessed growth rate. The results demonstrated strong consistency between simulated and measured values of LAI and LAIₚ (LAI simulation: R² = 0.99, NRMSE = 0.04; LAIₚ simulation: R² = 0.98, NRMSE = 0.04). This new WEP (Weight Effectiveness Photosynthesis) based modeling provides the required biofeedback for robust crop management, expanding the validity of model precision in cases when non-optimal environmental conditions dominate due to economic considerations. This study reveals that multiple critical insights can be obtained from complex process signals monitored by a single sensor, thereby providing a new perspective on multi-parameter estimates of plant growth and management.

Plant phenotyping relevance

トマトの重量センサー信号とモデルにより、LAIおよび光合成関連LAIを非破壊・リアルタイム推定する方法を開発しており、植物表現型の取得・抽出が研究の中心である。

Code and data availability

公開状態または取得可能な本文経路を確認できませんでした。

No evidence-backed public reproduction asset is currently recorded.