Work

Evaluation of real‐time milk tank monitoring using IoT system and error analysis in manual measurements

Year: 2025

Type: article

Abstract: Background Accurate measurement of milk volume in bulk tanks is essential to ensure traceability and safety in dairy product transactions. However, human errors in reading graduated rulers and technical failures can compromise accuracy. With the advancement of digital technologies, smart solutions can promote greater efficiency and reliability in agriculture. Aims This study aimed to quantify potential errors in interpreting graduated rulers for milk tank measurement (Study I) and to evaluate the effectiveness of a real‐time monitoring system using Internet of Things (IoT) technology (Study II). Methods Six evaluators performed visual perception tests using the graduated ruler at different volumes (Study I). In Study II, an IoT monitoring system was installed in a 1000‐L tank for automated measurement. Data were analysed using mixed‐effects linear regressions (lmer function, lme4 package, R), testing whether the intercept differed from 0 and the slope from 1. Major findings In the ruler model, evaluator variation explained 7.86% of the total variance. Bland–Altman analysis revealed bias between actual and observed volumes, with one clear human error included to illustrate the method's limitations. In the sensor model, variation attributed to the device was negligible (<0.0001%), and no bias was detected. Both models showed low mean square error (<1%), with mean errors of 0.09% for the ruler and 1% for the sensor. Concordance indices and R 2 confirmed high precision and accuracy for both approaches. Industrial implications This study highlights reliable methods for milk volume determination, reducing human error and improving efficiency in the dairy supply chain. Furthermore, it fosters agricultural digitalisation by encouraging modern technological practices that can support rural development and innovation in the sector.

Source: International Journal of Dairy Technology

DOI: 10.1111/1471-0307.70072