Food Grading Device for Sugarcane Juice Analysis

Abstract

Near-Infrared (NIR) Spectroscopy is a value addition as a non-destructive analytical technique, having a Wide range of multifaceted applications, Our Thesis explains threadbare, the development of an economical, portable NIR-based device designed to evaluate the maturity of sugarcane juice the device measures the light absorption patterns of sugarcane juice samples at six specific wavelengths, with the help of NIR sensors AS7263. The acquired spectral data is then processed by a machine learning algorithm, specifically Linear Regression, to categorize sugarcane juice maturity into its various levels. The Device integrates an ESP32 microcontroller with the AS7263 sensor, utilizing Micropython for programming. The ESP32 microcontroller facilitates seamless communication, connection and data transmission, being Wi-Fi enabled to interface and interact with the Firebase platform, an exclusively designed web app enables real-time data visualization and analysis. Linear Regression ML model analysis demonstrates the device's capability to effectively differentiate between various maturity levels of sugarcane juice and predict the maturity and purity based on the brix value, this is demonstrated through linear regression algorithm, achieving a regression accuracy of 90-93%. This highlights the potential of machine learning algorithms and techniques in revolutionizing the sugarcane industry with regards to its quality control and optimizing production processes. The integration of the ESP32 microcontroller with the AS7263 NIR sensor is the high mark for efficient operation of the device in regulating the data acquisition and transmission processes. Micropython programming enables flexibility and ease of development, facilitating rapid prototyping and customization of device functionalities. The NIR-based device is a cost-effective 8 | Page solution for sugarcane juice analysis, offering the end users an efficient and reliable tool for quality assessment and process optimization. It may be safely concluded that by harnessing the power of NIR spectroscopy, machine learning, and IoT technologies, the proposed device is the way forward in the agricultural sector which substantially promotes efficient, sustainable practices in sugarcane production.