AUTHOR=Pompeo Jean , Yu Ziwen , Zhang Chi , Wu Songzi , Zhang Ying , Gomez Celina , Correll Melanie 

TITLE=Assessing the stability of indoor farming systems using data outlier detection

JOURNAL=Frontiers in Plant Science

VOLUME=Volume 15 - 2024

YEAR=2025

URL=https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2024.1270544

DOI=10.3389/fpls.2024.1270544

ISSN=1664-462X

ABSTRACT=This study investigates the quality of air temperature data collected from a small-scale Controlled Environment Agriculture (CEA) system using low-cost IoT sensors during lettuce cultivation at four different temperatures. The methodology includes a generalized linear model regression analysis to examine the correlation between cumulative agricultural operations (Agr.Ops) and z-scores of air temperature residuals, assessing system stability. Outliers were identified and analyzed to determine their impact on system performance. Residual distribution and curve fitting revealed a log-normal distribution as the best fit for the sensor data. Regression analysis showed a strong inverse relationship between Agr.Ops and residual z-scores, suggesting that Agr.Ops contribute to outlier presence and impact system stability. The study highlights that system stability in CEA is influenced by the quality of data, with outliers indicating potential issues such as sensor noise, drift, or other uncertainties. The findings suggest that cumulative Agr.Ops affect system stability differently across trials, with some showing increased resistance to these operations over time. The alternative decomposition method used in this study effectively identified outliers and provided insights into system functionality. Future research should focus on improving surrogate data models, refining sensor selection, and ensuring data redundancy to enhance system reliability. The proposed method offers a promising approach for monitoring and managing uncertainties in indoor farming systems to improve their stability and efficiency.