• Eliciting a comprehensive of requirements to qualify and manage semantic drift in model-driven digital twins. • Developing assessment metrics, grounded in the requirements, to evaluate digital twin frameworks in their ability to address semantic drift. • Evaluating a selection of language and data specific frameworks supporting model-driven digital twins. • Reporting on the limitations of existing frameworks as evaluated by our metrics. Digital Twin (DT) technology is a key enabler of Industry 4.0, integrating diverse digital models to optimize processes, enhance decisions, and support predictive maintenance. However, as physical systems evolve, delays in synchronization can cause semantic drift, leading to discrepancies between digital and real-world entities. Effective semantic drift management needs DT frameworks that support all modeling layers, i.e., data, model, metamodel, and ontology, and provide different model management mechanisms. In this paper, we make three key contributions. First, we define a comprehensive set of requirements to address semantic drift effectively, drawing on both our understanding of the phenomenon and a real-world use case on mobility. These requirements capture the essential characteristics needed to maintain synchronization between the digital and physical domains as systems evolve. Second, we introduce a set of metrics designed to evaluate the ability of DT platforms to tackle semantic drift, grounded in both the defined requirements and insights of the use-case. Finally, we apply these metrics to evaluate four existing DT platforms demonstrating their utility for identifying the limitations of current frameworks in handling semantic drift. Through this evaluation, we highlight the strengths and weaknesses of these platforms, providing a foundation for future improvements in DT infrastructure.

Network Digital Twins (NDTs) offer a structured framework for modeling, predicting, and optimizing wireless networks. This paper presents a modular NDT implementation based on the GreyCat platform, integrating graph-based data models and external functional algorithms for indoor radio coverage prediction. For the first time, we implement an NDT system aligned with ITU-T Recommendation Y.3090, covering both basic and functional model instantiation from modular and interoperable abstract structures. We generated a practical dataset using a software-defined radio (SDR)-based OpenAirInterface5G setup, with a gNB and commercial UE deployed in a controlled environment. This real-world dataset was used to benchmark Gaussian Process Regression (GPR) and Convolutional Neural Network (CNN) models for predicting RSRP-based radio coverage. Our results show that CNN outperforms GPR in under-sampled conditions, and we demonstrate how the modular architecture supports flexible model integration and benchmarking. This work represents a significant step toward practical, data-driven NDT deployments for wireless systems.

Model-Driven Engineering (MDE) is a Software Engineering approach that relies on models to create applications. Models capture knowledge at several levels of abstraction (e.g. statechart models, class models, etc.). often created by different stakeholders that, at the end, need to combine them all together (e.g. through transformations, mappings, etc.) to produce the desired application. Collaboration is thus a key aspect. However, MDE mostly happens in non collaborative environments usually because of technological restrictions of the tools and languages used to describe the models and the transformations manipulating such models. This article discuses strategies for building real-time collaborative modelling environments. We propose a problem space that helps MDE practitioners to understand all the combinations of metamodels-models-representations that take place in MDE, as well as how to combine collaborative real-time algorithms and transformational approaches for each of them. Finally, we discuss implications from the topology perspective that MDE practitioners should consider when architecturing their collaborative environments.

Human-Computer Interaction (HCI) addresses the study, planning and design of the interaction between people and computers through User Interfaces (UIs). The co-design and co-development of these UIs involve different stakeholders as Developers, Functional Analysts, Usability Experts among others, all of them responsible for different UI elements (respectively implementation, functional requirements, usability). Collaboration between stakeholders has been identified as a key factor for UI design and development. This article proposes a multi-viewpoint approach to support collaboration in a UI co-design and co-development process. The article discusses the process itself and the different viewpoints manipulated by all the involved stakeholders. We propose an implementation of our approach in a model-based multi-viewpoint prototype. We show the feasibility of this approach implementing it on a case-study and a multi-viewpoint prototype.

User Interfaces (UI) design is a complex and multi-faceted problem. It depends on devices, users and their environments and involves various stakeholders with different backgrounds. Moreover, user requirements are difficult to evaluate precisely upfront and require iterative design cycles based on trial and error. All this variability is complex and should be managed efficiently to ensure moderate design costs. One solution is to implement in the UI design process Model-Driven Engineering (MDE) and Software Product Lines (SPL). However, current SPL approaches do not consider problems related to specific UI design models, notably the many concerns underlying them. We propose an SPL approach that supports the separation of concerns through multi-step partial configuration of UI features. The approach is implemented in our existing MDE UI generation framework.

Early model-based approaches for Human-Computer Interaction (HCI) clearly depicted models and frameworks for generating User Interfaces (UI) but considered model transformations as black-boxes. In the 2000's, these approaches were criticized due to the poor quality of the produced UI. One of the main reasons of this poor quality can be easily observed in state of the art UI transformations: they are the heart of designers' know-how but are maintained by a minority of specialists. Meanwhile, mainstream UI design methods have shown a growing number of heterogeneous stakeholders that collaborate to produce modern and qualitative UI. We claim that these stakeholders must comprehend and interact with transformations and thus we need to make the transformation language affordable to these stakeholders. Indeed, such a simplification should hide transformations complexity and burden for any stakeholder, finally focusing on a specific part of the design domain: a Domain Specific Language (DSL) for transformations or Domain Specific Transformation Language (DSTL). We provide in this paper a method and a supporting tool for systematizing and finally executing DSTL for model-driven UI development. We depict that framework on a proof of concept implementation for an HCI-specific stakeholder: the usability expert.