Arnold Reuser

AI-Supported Architectural Reasoning for High-Tech
Systems-of-Systems

Bio.

Bio: Arnold Reuser is a Systems Engineer with 20+ years of experience leading digital transformations across diverse industries, business domains, and architectural layers.  At ASML, he focuses on Systems-of-Systems Engineering for Data & Digital Platforms, with a particular emphasis on improving customer ecosystem availability. He holds a Master’s degree in Pure Mathematics (cum laude) and a Bachelor’s degree in Applied Computer Science (cum laude).

Abstract.

The Dutch high-tech industry is globally recognized for its capability to engineer complex, multidisciplinary systems. However, this strength is increasingly challenged by the transition from individual systems toward Systems-of-Systems (SoS), in which independent systems, organizations, data platforms, and digital ecosystems must collaborate to create emergent capabilities. This complexity is further amplified by AI-enabled components, distributed value chains, shorter system lifecycles, and the growing need to coordinate design decisions across technical, enterprise, and ecosystem levels.
This evolution is also reflected in national and international systems engineering initiatives. The HTSM Systems Engineering Roadmap emphasizes the need for improved integration, digitalization, and next-generation systems engineering capabilities in the Dutch high-tech ecosystem. Similarly, the INCOSE Systems Engineering Vision 2035 highlights the transition toward digitally integrated, model-based, and AI-enabled engineering environments. Building on this direction, the SERC/INCOSE AI4SE–SE4AI roadmap further underscores AI’s role both as an enabler for systems engineering and as a source of new engineering challenges .
Yet despite these initiatives and ambitions, current engineering practice remains constrained by fragmented toolchains, disconnected viewpoints, and reliance on tacit architectural reasoning. As a result, critical decisions about system architecture, enterprise context, interfaces, trade-offs, risks, and design alternatives are often made informally and are difficult to trace. Although architecture description standards such as ISO/IEC/IEEE 42010 emphasize the importance of structured architectural viewpoints and concerns System, enterprise, and ecosystem layers are still often weakly integrated in practice. Consequently, inconsistencies, stakeholder misalignment, and system-level risks are frequently discovered too late, leading to redesign, rework and impacts on project time and cost. These limitations become even more visible in SoS contexts, where changes increasingly occur across a broader context rather than only within physical or technical subsystems.
Against this background, this contribution argues for a shift from architecture as static documentation toward architecture as an active, AI-supported reasoning process. More specifically, we propose the concept of an integrated architectural reasoning environment in which AI acts as a collaborative design partner for system architecting teams. In this environment, system architecture, enterprise context, engineering tool ecosystems, and heterogeneous artefacts are connected through structured representations and AI-supported reasoning services.
Within such an environment, the proposed AI capabilities include knowledge extraction from engineering artefacts, cross-view consistency analysis, trade-off reasoning, design-space exploration, risk and issue detection, and explainable insight generation. At the same time, the human architecting team remains in control: architects define goals and constraints, evaluate alternatives, and make final decisions.
Taken together, the expected impact is improved alignment across stakeholders, more consistent and traceable architectural decisions, earlier detection of risks and conflicts, improved quality and increased engineering productivity through reduced fragmentation and rework and improved reuse of artefacts.