A PhD student at Eindhoven University of Technology has developed methods to simplify the creation of digital twins, improve their reliability and enhance their ability to support engineers’ decisions.
Digital twins are digital models that replicate the behaviour of machines, vehicles or industrial processes. They’re used to predict problems, optimise performance and plan maintenance. Imagine a virtual copy of a factory machine that can warn you before it breaks, suggest improvements, or test changes safely before applying them in the real world.
Advertisement
While digital twins have great potential, building them can be complicated because different models, sensors, software and services all need to work together, and current development processes are often fragmented and inefficient.
PhD researcher David Manrique Negrín focused on making digital twins more systematic, modular and reusable. Using model-driven engineering, a method that emphasises high-level models to guide software development, he explored integrating and coordinating the many components that make up a digital twin.
Advertisement
Digital twins often combine models created with different tools, methods, or levels of detail. For example, one model might simulate a machine’s physical behaviour, while another model approximates its software or control systems. These models can differ in format, scale and how they represent information, thus making it difficult to connect them and share information between the models.
Manrique Negrín developed a method that standardises how these models communicate with each other, allowing them to exchange information smoothly and operate together as a single system. This approach reduces the effort needed to connect diverse models and ensures that the overall digital twin behaves reliably, giving engineers a clearer and more accurate view of the system.
Advertisement
To help engineers understand, share and reuse models, Manrique Negrín introduced a structured approach to document and explain how models work and how they relate to each other. This method provides context for each model, making it easier to incorporate existing models into new digital twins, save time and improve consistency. By capturing the connections and assumptions behind each model, engineers can reuse work more confidently and avoid errors caused by misunderstanding or misapplying a model.
Beyond combining models, digital twins need orchestration, which is the coordinated execution of all their components, including models, data streams, sensors and services. Manrique Negrín developed a new orchestration approach, called OrchTwin, supported by a domain-specific language known as LOTTS.
This system allows engineers to specify high-level requirements and automatically turn them into working and executable systems. This reduces the need for manual coding, minimises errors and makes it faster and easier to deploy and manage digital twins in practice.
Advertisement
Manrique Negrín tested his methods through prototypes and real-world case studies. He showed that his approaches make digital twins easier to build, more consistent and more robust.
His research bridges the gap between design and implementation, while also paving the way for smarter, scalable digital twins that better support engineers in decision-making and system optimisation.
Advertisement
His work has implications across industries, from manufacturing and energy to mobility and smart infrastructure. Digital twins can help improve efficiency, safety and sustainability by helping engineers to anticipate problems and optimise performance before costly real-world interventions.