MDEO-Cloud: A Modular Cloud-Based Platform for Model-Driven Optimization
Abstract
Model-driven optimization (MDO) uses domain-specific modeling languages (DSMLs) to describe and solve optimization problems, aiming to reduce complexity for domain experts who lack optimization expertise. However, existing research prototypes remain fragmented and tightly coupled to the Eclipse ecosystem, making them difficult to access and reproduce. We present MDEO-Cloud, a cloud-based platform for model-driven optimization that provides a unified, web-accessible environment for collaborative research and development. Built on a plugin-based architecture, the platform supports extensibility through language and contribution plugins, enabling seamless integration of new optimization approaches. We demonstrate the platform’s capabilities through a scalable execution engine that features federated distributed computing and present a family of DSMLs supporting blended graphical-textual modeling, which we evaluate through a user study. We hope that MDEO-Cloud can serve as a common foundation for the MDO research community, facilitating collaboration and accelerating progress in the field.