Smart System Approach for Prediction of Microbial Quality of Milk Using Machine Learning

Abstract

Developing a real-time system to assess milk qualityusingResazurindyeandmachinelearning. The system combines an Arduino Uno, color sensor, and LCD display to detect color changesinmilksamples.Thesechangesindicatemicrobial activity and quality. By integrating Python and machinelearningalgorithms,thesystemprocesses and analyzes data, providing accurate and immediate milk quality assessments. This automated solution enhances efficiency and reliability, reducing the need for manual inspections and improving overall dairy quality control.The system\'s ability to deliver accurate, real-time information underscores its potential for impactful use in scientific and industrial settings, highlighting its significance in the field of chemical analysis.

Introduction

This project integrates IoT sensors with AI-driven analysis to monitor and predict chemical properties in real time. Using components like Arduino Uno, pH sensor, and color sensor, it primarily focuses on milk quality assessment. A Python-based machine learning model (Random Forest) analyzes the sensor data to detect adulteration and classify milk as Fresh, Acceptable, or Spoiled, enhancing accuracy, reducing human effort, and increasing reliability.

B. Objectives

1. Real-time chemical monitoring and prediction.

2. Automation using IoT and AI technologies.

3. Detection of chemical anomalies with machine learning.

C. Literature Survey Highlights

• Random Forest model shows 96% accuracy in milk quality prediction.

• Deep learning autoencoders used for extracting sensory data features.

• Distance-Weighted KNN (DW-KNN) achieves 99.53% accuracy, outperforming KNN for milk quality detection.

D. Proposed Methodology

1. System Functionality:

• pH Sensor: Measures acidity.

• Color Sensor (TCS3200): Detects color changes due to microbial activity.

• Arduino Uno: Collects and processes sensor data, outputs it via LCD and serial monitor.

• Python & Machine Learning (Random Forest): Predicts milk quality using RGB data.

2. Hardware Used:

• Arduino Uno: Main microcontroller.

• LCD Display: Displays real-time sensor readings.

• pH Sensor: Measures acidity or alkalinity.

• Color Sensor: Measures RGB values for milk analysis.

• Power Supply: Provides 5V regulated power.

• Resazurin Sodium: Indicates spoilage via color change.

3. Software Used:

• Embedded C: Controls hardware via Arduino IDE.

• Python: Runs machine learning model for prediction.

• Arduino IDE: Programs the Arduino to collect RGB data and transmit it to Python.

E. System Working

• Starts with a power supply.

• Sensors (pH and color) detect key chemical features.

• Arduino processes and sends data to LCD and Python script.

• AI model analyzes data and predicts milk quality.

• System raises alerts when values fall outside normal range.

• Applicable to food safety, pharmaceuticals, and environmental monitoring.

F. Results & Discussion

• Different milk products show distinct RGB values and pH:

  • Raw Milk: RGB ~ (240-255, 235-250, 225-240)

  • Pasteurized Milk: RGB ~ (245-255, 240-255, 230-245)

  • Ghee: RGB ~ (190-210, 150-180, 80-110)

• Quality is determined by temperature, pH, and color:

  • Example (Pasteurized Milk):

    • At 4–7°C, pH 6.6, RGB 255 → Excellent

    • At 30–44°C, pH 6.3, RGB 180 → Bad

Conclusion

The integration of an Arduino UNO with a TCS3200 color sensor provides an effective and cost-efficient method for monitoring milk quality based on color changes. By capturing real-time RGB values from the milksample, thesystem can detect the degree of spoilage, offering a non- invasive solution for quality assessment. The collected data can be further processed using machine learning algorithms, such as Random Forest, to accurately classify milk as fresh, acceptable, or spoiled. This system offers a promising approach for ensuring milk safety and quality in dairy production, with the potential for integration into automated milk processing environments, improving both efficiency and consumer confidence in milk products. In the future, the system can be integrated into automated dairy farms and processing plants for continuous, real-time monitoring of milk quality. By using sensors at multiple stages of milk production, the system can help track the quality from the farm to the consumer, ensuring optimal freshness and safety throughout the supply chain.

References

[1] K. A. &. D. V. S. Sharma, \"Internet of Things (IoT)-enabled smart sensor-based real-time monitoring system for food quality and safety.,\" Journal of Food Quality, 2021, 2021. [2] M. &. O. D. León, \"Machine Learning Applied to Milk Sample Classification.,\" Journal of Food Quality Monitoring, 45(2), 120-130., 2022. [3] S. Shinde, \"Milk quality prediction and yogurt fermentation analysis using Machine Learning.,\" International Journal of Food Science and Technology, 58(4), 450-460., 2023. [4] A. Samad, \"Enhancing Milk Quality Detection with Machine Learning: A Comparative Analysis of KNN and Distance-Weighted KNN Algorithms,,\" International Journal of Dairy Technology, 59(1), 100-110., 2024. [5] A. Çeli, \"Using Machine Learning Algorithms to Detect Milk Quality,\" Eurasian Journal of Food Science and Technology. , 2022. [6] S. M. M. Saranya, \"IoT Based Smart Milk Monitoring System using Sensors,\" International Journal of Innovative Research in Computer and Communication Engineering, vol. 7, no. 4, pp. 2165-2171, 2019., 2019. [7] M. A. K. a. M. K. Singh, \"AI-Based Quality Monitoring System for Food Safety Applications,\" International Journal of Scientific & Technology Research, Vols. vol. 9, no. 2, pp. 2541-2546, 2020., 2020.

Copyright

Copyright © 2025 Jenefar S, Rajakumar B, Balan V, Kaviya M, Mouriya A.V, Swathi M. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.