The review presents a comprehensive evaluation of 15 published systems centred on the application of Convolutional Neural Network (CNN) in exclusive exhibitions on coffee beans via utilizing Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA). The assessment displayed the accuracy rate, methodical procedure, and relevance of CNN as a tool. Furthermore, an extensive analysis was performed to determine the effectiveness and reliability of CNN which yielded a high-rating accuracy ranging from 90 to 100 percent from chosen studies. The findings highlight CNN's superiority over traditional methods, emphasizing its potential as a pioneering neural network architecture for coffee quality assessment. Nonetheless, limitations of the CNN model are also highlighted alongside the recorded data.