The quality of products produced on the production line is particularly important for manufacturers. However, how to trace the cause of product quality problems is a common problem in the field of intelligent manufacturing, and has not received good attention and research. Solving this problem requires the person in charge of production to be very familiar with the production links in a specific production scenario and have good reasoning ability, which is often challenging and a key factor to the factory’s production efficiency. In this paper, we designed a system for a manufacturer that uses the knowledge graph to reason and analyze the product quality in production. To solve the difficulty of tracing the cause of the problem in a specific production scenario, we carefully analyze and mine the original Manufacturing Execution System (MES) data set of that manufacturer, establish a knowledge graph model, and find out the shallow causes and deep correlations of quality problems through the custom rule reasoning of the Jena reasoning engine. The reasoning system overcomes the low accuracy and high time-consuming of traditional manual inspection and provides effective help for factories to improve production efficiency, save production costs, and reduce defective product rates.