Frank de Lange

From Models to Trust: MBSE Adoption Challenges

Bio.

Bio: Frank de Lange leads a team of eight within ASML’s System Engineering department. He is responsible for developing Systems Engineering capabilities and defining Product Safety at system level. His competence portfolio includes system decomposition, requirements engineering, configuration management, design process modelling and system architecture modelling.

Before joining ASML in 1991, Frank worked in Mozambique, where he led the privatization of a sheet-metal workshop, and in the Netherlands as production lead at a Power-Take-Off shaft factory. At ASML, he began as a manufacturing and service engineer before taking on project leadership roles in issue resolution, availability improvement, and reticle stage design. He moved to the Systems Engineering department in 2001 as project lead overlay and, from 2006, led productivity and overlay teams.

Frank earned a master’s degree in mechanical engineering, specializing in precision mechanics, from TU Delft in 1986.
He represents ASML in the NXTGEN Hightech programme.

Abstract.

Model-Based Systems Engineering is recognized as an enabler for managing increasing system complexity and enabling earlier, data-driven decision-making. However, organizations transitioning from document-driven to model-based approaches encounter practical challenges that are often underrepresented in the literature. These challenges are not primarily technical but revolve around trust, data ownership, and value perception.
In industrial practice, systems rarely start from a clean slate; instead, they evolve from legacy architectures with heterogeneous data sources. As a result, MBSE adoption takes place in a hybrid environment where spreadsheets, Product Lifecycle Management systems, and models coexist. A key observation is that expectations regarding data quality shift dramatically once information is integrated into shared system models. “Shared” models are collaboratively created, maintained, and used by stakeholders across multiple disciplines, rather than being confined to a single local team. While incomplete or partially validated data is often accepted within local tools, the same data is expected to be near-complete and accurate when presented in such a shared context. This “expectation shift” creates a barrier to adoption, as the model becomes a visible reference point in which previously tolerated uncertainty is no longer acceptable and can undermine confidence across the engineering community.
Furthermore, ownership of information becomes unclear when moving from local artifacts to shared models. When stakeholders expect the model to be complete and correct, but no clear accountability exists for validating and maintaining the underlying data, trust in the model remains limited. Additionally, user expectations for consuming model data are heavily influenced by reporting environments such as Power BI and PowerPoint, which provide high-quality visualization and analysis capabilities. Although these reporting capabilities can be leveraged, their effectiveness depends on the availability of consistent, well-linked data chains across systems, highlighting the importance of data integration and governance.
This contribution presents practical experiences from ongoing MBSE adoption efforts in an industrial context. Rather than proposing a complete solution, it identifies key patterns observed in practice: (1) the absence of robust mechanisms for model assurance and data quality control, (2) the risk of uncontrolled model growth without clearly defined modeling objectives, (3) the importance of treating models as structures that connect and organize information from multiple sources, supporting engineers in communication and decision-making, rather than as tools in which all data must be created and maintained directly and (4) the necessity of combining MBSE models with external reporting solutions to deliver actionable insights.
The paper concludes with a set of open questions aimed at the INCOSE community, focusing on pragmatic approaches to handling incomplete models, establishing sustainable data ownership, and bridging the gap between modeling and decision-making. By sharing these insights, this work aims to contribute to a realistic understanding of MBSE adoption and stimulate cross-industry discussion.