High-tech manufacturers are increasingly looking for smarter ways to service their systems—driven by the need to maximize equipment effectiveness, manage growing system complexity, and scale service operations efficiently.

The session on intelligent diagnostics explored how combining system knowledge, real-time data, and AI techniques can support both the design of diagnosable systems and the development of digital tools that enhance human decision-making. By embedding intelligence into service strategies, organizations can move toward more proactive, scalable, and effective support models.

Owners of capital equipment ─ such as MRI scanners, production printing machines and EUV lithography systems ─ require high availability and minimal total cost of ownership. Robust diagnostics is essential for achieving dependable and cost-effective systems without relying on hardware redundancy. 
The combination of data analytics and domain-specific system knowledge enables advanced diagnostics to reduce unplanned downtime without increasing architectural complexity or operational costs. This integrated approach supports the entire diagnostics lifecycle ─ from design to implementation and operation of remote, proactive and predictive diagnostic services.  
This presentation will highlight the pivotal role of data- and knowledge-driven diagnostics, emphasizing their role in enhancing the reliability and serviceability of complex capital equipment. 

Servicing complex high-tech systems, such as ASML lithography machines, demands intelligent diagnostics to minimize downtime.​  This talk is given by ASML and TNO-ESI, following a multi-year collaboration developing a methodology to leverage MBSE for diagnostics. The approach combines high level system architecture information with system specific observations to assist service engineers in troubleshooting. We will discuss how this method applies at ASML.  
TNO-ESI will give a demonstration, showcasing the methodology and its capabilities. The methodology is currently being further developed at Canon Production Printing.

Optimizing performance, such as print quality at Canon Production Printing (CPP), is a crucial, yet challenging task for many high-tech system manufacturers. This joint talk by CPP and TNO-ESI outlines the initial results in developing a performance diagnostics methodology. In collaboration with CPP, TNO-ESI has developed a generic approach using probabilistic graphical models to diagnose defectivity patterns on 2D surfaces. 
In this talk, CPP will discuss the potential applications of this methodology and highlight further research activities in diagnostics. TNO will present the key aspects of the performance diagnostic modeling, and reasoning approach.