Data‐Driven Automated Chemometric FTIR Analysis for Rapid Carbohydrate Quantification in Milk Powder: Validation Against Conventional Methods

Accurate quantification of carbohydrates in dairy products is essential for regulatory compliance and process control. Conventional chromatographic and colorimetric techniques, although reliable, are often limited by long analysis times, complex sample preparation, and high operational costs. In this study, data‐driven automated chemometric FTIR analysis developed in the Google Colab computational environment was evaluated as an alternative for carbohydrate quantification in milk powder. The approach is based on automated spectral preprocessing by baseline correction, Savitzky‐Golay smoothing, and normalization, feature extraction within the 899–955 cm ⁻¹ region associated with lactose, glucose, and galactose, and univariate regression analysis. Calibration was performed using mixtures of solely standards in an inert matrix (KBr), ranging from 0% to 100%, and validated against high‐performance anion‐exchange chromatography with pulsed amperometric detection (HPAEC‐PAD) and the dinitrosalicylic acid (DNS) method. The FTIR approach demonstrated strong analytical performance for carbohydrate quantification, particularly for lactose, with a coefficient of determination R ² ≥ 0.98, accuracy values of 114.95%, and high precision (%RSD < 0.1), comparable to reference methods under controlled conditions. It was observed that during the analysis of commercial milk powder samples, lactose quantification remained accurate; however, matrix effects and spectral overlap influenced glucose quantitative performance in the presence of starch. These results indicate that FTIR spectroscopy combined with automated spectral processing is better suited as a rapid screening and decision‐support tool for dairy quality control. Furthermore, the chemically informed preprocessing strategy established in this study provides an analytical foundation for the future development of advanced predictive models for multicomponent food analysis.

Practical Applications

Automated data‐processing‐assisted FTIR spectroscopy provides a rapid, non‐destructive approach for monitoring carbohydrate content in milk powder with minimal sample preparation. This method enables faster decision‐making in quality control compared to conventional techniques and can be implemented as a routine screening tool in dairy processing environments. It is particularly useful for verifying product consistency and supporting process monitoring, while complementing established analytical methods.