Prototype of a Coffee Bean Weight Measuring Device Using a Webcam with the Convolutional Neural Network (CNN) Method at Roetin Coffee Shop

Nurul Budi (Universitas Negeri Yogyakarta)
Faris Yusuf Baktiar (Universitas Negeri Yogyakarta)

Journal of Robotics, Automation, and Electronics Engineering
Vol. 4 No. 1 (2026): March 2026

The advancement of Artificial Intelligence (AI) and Computer Vision has enabled new opportunities for automation within the coffee industry, particularly in weight measurement of coffee beans, which is still performed manually and becomes inefficient at large scale. This study proposes an automatic weight estimation system using images captured by a webcam and processed through a Convolutional Neural Network (CNN) employing MobileNet as a lightweight regression model. The developed system analyzes visual features to estimate weight autonomously, offering an efficient, contactless alternative to conventional weighing tools and supporting stock monitoring for coffee industries and small enterprises.