Abstract

The design and operation of complex engineering systems demand reliable models that accurately describe their dynamic behavior, forming the foundation of model-based engineering. With the surge in data availability and advancements in machine learning, the integration of these technologies into model-based engineering presents both significant opportunities and challenges.

Hybrid models are critical for enhancing diagnostics capabilities, such as fault detection, isolation, and root cause analysis, thereby supporting predictive maintenance.

Moreover, controllers that shape the dynamic behavior of engineering systems in terms of performance and robustness also stand to benefit from these hybrid technologies.

We will illustrate the transformative impact of combining models, data, and learning on various applications, including high-tech equipment, healthcare, and mobility. Examples will span from robots and semiconductor equipment to mechanical ventilators in hospitals and autonomous vehicles. Additionally, we will address the challenges encountered in this integration and share valuable lessons learned from practical implementations.