Semantically Enriched Interfaces for Product Configuration

This paper highlights the challenges of product customization in the context of mass customization, particularly the combinatorial explosion of product variants that arises when trying to offer lot size one production. A knowledge graph can be used to store the required product knowledge including all potential products that can be manufactured. This paper proposes a method of reducing the complexity of the resulting configuration problem using automated reasoning. The approach decreases the number of individuals and by extension the number of product variants involved in the problem by materializing constraint results and grouping equivalent specifications. As opposed to many existing approaches, the contribution fully solves the underlying constraint satisfaction problem inside the knowledge graph without the need of an external solution. The approach still benefits from facilitated data interchange and the simple implementation of various downstream tasks.