Abstract: To improve the adaptability of the work plan to the machine state, in this study, we establish a job-shop scheduling model driven by quality data. To do so, we first use expert domain knowledge and relevancy analysis to mine the association rules hidden in the quality data and identify the machine-state fluctuations. Then, we construct a global scheduling performance indicator based on the loss time function to enable the scheduling model to make cycle optimization adjustments in response to machine-state fluctuations and to ensure the feasibility of the small-time-scale scheduling scheme. Finally, we verifie the effectiveness of the model by successfully changing the scheduling in a manufacturing enterprise in response to changes in the machine state.