Researchers have developed Function+Data Flow (FDF), a visual domain-specific language that enables the composition and reuse of machine learning models. To evaluate its usability, they conducted an empirical study where participants used FDF in DesCartes Builder to implement a real-time digital twin prototype. The results showed that FDF and DesCartes Builder are accessible and reliable for a broad range of users, especially domain experts. However, the study also identified areas for improvement in both the tool and the underlying framework. Based on these findings, the researchers proposed H-FDF, a hierarchical extension of FDF that supports iterative and modular pipelines. --- Why it matters: This matters to AI engineers because it provides a promising direction for transforming AI-based digital twin engineering into a disciplined modeling practice, enabling more efficient and reliable development of real-time digital twins. Source: https://arxiv.org/abs/2608.18480