Engineering changes (ECs) are new product development activities addressing external or internal challenges, such as market demand, governmental regulations, and competitive reasons. The corresponding EC processes, although perceived as standard, can be very complex and inefficient. There seem to be significant differences between what is the “officially” documented and the executed process. To better understand this complexity, we propose a data-driven approach, based on advanced text analytics and process and data mining techniques. Our approach sets the first steps toward an automatic analysis, extracting detailed events from an unstructured event log, which is necessary for an in-depth understanding of the EC process. The results show that the predictive accuracy associated with certain EC types is high, which assures the method applicability. The contribution of this article is threefold: 1) a detailed model representation of the actual EC process is developed, revealing problematic process steps (such as bottleneck departments); 2) homogeneous, complexity-based EC types are determined (ranging from “standard” to “complex” processes); and 3) process characteristics serving as predictors for EC types are identified (e.g., the sequence of initial process steps determines a “complex” process). The proposed approach facilitates process and product innovation, and efficient design process management in future projects.
- Classification model , clustering , engineering change (EC) , innovation , process mining , process model , text analytics