Automation and Optimization of Food Process Using CNN and Six-Axis Robotic Arm

Area of Science:

• Robotics and Automation
• Artificial Intelligence
• Food Science

Background:

• Traditional coffee roasting lacks precise control over the degree of roasting.
• Automation in food processing can enhance efficiency and product consistency.

Purpose of the Study:

• To develop and evaluate the Food Process Robot Intelligent System (FPRIS) for automated coffee roasting.
• To investigate the system's ability to control the Degree of Roasting (DoR) using integrated sensors and AI.

Main Methods:

• Integration of a 3D-printed six-axis robotic arm with AI and Computer Vision (CV).
• Utilizing a Convolutional Neural Network (CNN) for real-time coffee bean classification and roaster control.
• Combining gas and image sensor data for assessing coffee bean quality and DoR.

Main Results:

• FPRIS demonstrated precise control over the DoR by correlating sensor data with roast intensity.
• Consistent trends observed: increased weight loss and Gas sensor Initial Difference (GID), decreased Sum of Pixel Grayscale Values (SPGVs) with higher roast intensity.
• The system effectively navigated obstacles and empty spaces within the roaster.

Conclusions:

• FPRIS offers enhanced precision and efficiency in automated coffee roasting.
• The system shows potential for broader applications in food processing automation.
• Further research can establish FPRIS as a universal automation solution for the food industry.