Research

As AI models become increasingly powerful yet complex, there is a growing need for transparency in manufacturing applications where reliability and trust are critical. 

At ASML, we develop explainable and physics-informed AI frameworks that enhance the interpretability and usability of deep learning models on the shop floor. By visualizing model reasoning and connecting AI outputs with physical phenomena, we improve fault diagnosis, process monitoring, and predictive maintenance in systems such as machining, welding, and additive manufacturing. This research bridges the gap between black-box AI and real-world decision-making in dynamic manufacturing environments. 

Modern manufacturing demands robotic systems that can operate autonomously across diverse environments and material conditions.

At ASML, we design and implement AI-driven robotic systems that adapt to a wide range of tasks—such as machine tending, machining chip removal, and fabric handling—while responding intelligently to uncertainty and variability in real-world settings. Our approach integrates visual perception, multi-modal sensing, and learning-based decision-making to enable autonomous robots to manipulate complex objects with flexibility and precision.

Digital twins enable real-time synchronization between physical manufacturing systems and their virtual counterparts, enhancing visibility, interactivity, and data-driven decision-making on the shop floor. 

At ASML, we develop immersive and interactive digital twin frameworks that integrate machine data, sensor streams, and AI models to create intuitive and scalable representations of manufacturing processes. Our digital twins support applications such as real-time process monitoring, predictive visualization of additive geometries, and cyber-physical alignment using tools like Unity 3D. These efforts aim to improve system transparency, adaptability, and human-machine collaboration in smart manufacturing environments.