Development of an Embedded System for Lipid Content Detection in Peanuts Using Near-Infrared Sensors and Machine Learning

• Grant number: 24/16073-1
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: January 01, 2025
• End date: December 31, 2025
• Field of knowledge: Agronomical Sciences - Agricultural Engineering
• Principal Investigator: Flávio José de Oliveira Morais
• Grantee: José Vitor Romualdo
• Host Institution: Faculdade de Ciências e Engenharia. Universidade Estadual Paulista (UNESP). Campus de Tupã. Tupã , SP, Brazil

Abstract

Artificial Intelligence (AI) and the Internet of Things (IoT) are currently highly significant fields, and the emerging area of Artificial Intelligence of Things (AIoT) represents the convergence of these two domains. As IoT becomes more prevalent, the concept of AI can extend to embedded systems, increasing the problem-solving capabilities without relying on internet connectivity or high-cost hardware. Many microcontrollers can execute machine learning models developed from libraries that can be implemented in their integrated development environment. Additionally, some system-on-chip (SoC) devices can process images from an integrated sensor, applying machine learning models to their intended use. Peanuts are one of the most commercially important crops due to their rich content of macro and micronutrients, and they are widely used for oil extraction due to their high lipid content. Accurate and non-destructive detection of this lipid content is crucial in the oil extraction industry. Various methods have been developed to achieve precise and non-destructive detection. One such method involves using Near-Infrared (NIR) sensors, which, through software, generate a visible image for analyzing the lipid content in the grains. This project proposes the development of a low-cost, ultra-low-power embedded system that utilizes the concept of Artificial Intelligence of Things (AIoT). The system will execute a machine learning model and be capable of detecting lipid content in peanut grains using a Near-Infrared (NIR) sensor