Prototype of SwiftletNest Moisture ContentMeasurement Using Resistance Sensor and Machine Learning

Prototype of SwiftletNest Moisture ContentMeasurement Using Resistance Sensor and Machine Learning

swiftlet nest;moisture content;IoT; Machine Learning; Neural Network; PRORESKA

Swiftletnests are highly valued for their health and cosmetic benefits, with moisture content crucial in determining their quality. Traditional moisture measurement methods are often slow and can potentially damage the samples. This study introduces PRORESKA, an innovative system utilizing resistance sensors and Machine Learning (ML) for non-destructive,andreal-time moisture measurement. The system incorporates a voltage divider circuit to establish a correlation between resistance data and moisture content. Three mathematical models (linear, exponential, and modulated exponential) and a neural network wereemployed to predict moisture content. Validation tests conducted on paper and swiftletnests indicated that the neural network model, enhanced through transfer learning, achieved superior accuracy. The results demonstrated a strong correlation between predicted and actual moisture content (R² = 0.9759), with the neural network model attaining a mean squared error (MSE) of 0.01. This method holds significant potential to improve the efficiency and cost-effectiveness of moisture measurement for swiftletnests and similar applications.

Ratu Anggriani Tangke Parung1, Hanna Arini Parhusip2*, Suryasatriya Trihandaru

https://jurnal.iaii.or.id/index.php/RESTI/article/view/5923/983

Master of Data Science, Faculty of Science and Mathematics, Satya Wacana Christian University

28-10-2024

FAJAR BAGUS W

PDF

ENGLISH

TEXT