As software modeling repositories expand, the need for semantic retrieval of conceptual models becomes increasingly important for effective model management and reuse. In this paper, we propose a classification-based semantic retrieval scenario where the classifier acts as the core component enabling automated and intelligent model organization. Each metamodel, expressed in Ecore, is transformed into a semantic attributed graph that captures entities, attributes, associations, and multiplicities. To effectively exploit semantic information, we employ a fine-tuned BERT embedding as the primary representation method. Additionally, we use a TF-IDF-based embedding as a baseline to compare semantic and non-semantic approaches and to demonstrate the impact of semantic representations. These embeddings, together with node and edge labels, are used as input features to a graph neural network (GNN) that classifies conceptual models into 24 predefined categories. Since the models are represented as semantic graphs, our experiments confirm that GNNs significantly outperform traditional baseline classifiers. Specifically, the BERT-based GNN achieved the highest accuracy (around 96%), while the TF-IDF-based GNN also performed strongly (above 95%). In contrast, non-graph baseline models such as MLP (84.3%), SVM (78.6%), and C5.0 (85.4%) exhibited notably lower performance. These findings validate the proposed classification-based semantic retrieval scenario and highlight the advantages of graph-based semantic learning for automated, scalable, and semantically aware retrieval.
Biodiversity monitoring is concerned with keeping track of different species in an ecosystem over time, with respect to their abundance, distribution and diversity. Environmental digital twins used for biodiversity monitoring share characteristics with industrial digital twins, but face additional challenges in connecting data and models: Biodiversity data is often not livestreamed, interventions are slow and require human interaction, and the scientific knowledge about species and their habitats constantly evolves. Today, environmental digital twins offer little automation support, or any support to help scientists link species observations to assumptions about biodiversity. This paper presents an application of structural self-adaptation, originally developed in the industrial domain, to environmental digital twins. We show how structural self-adaptation enables to autonomously adapt monitored assumptions to changes in the available data sources, and further discuss how digital twins can adapt to changes in the domain knowledge. A first evaluation is given based on underwater cameras in the Oslo Fjord.
Digital Twin-enabled Systems (DTeS) are increasingly relevant in practice, particularly in the context of large, complex, and long-lived industrial facilities, buildings, and infrastructure. These systems are typically composed by orchestrating existing artefacts using a wide range of technologies and artefacts originating from diverse engineering disciplines. Due to continuous adaptation and evolution of the underlying systems, DTeS are subject to significant change over time, making both their development and long-term maintenance highly challenging.This paper presents TwinLang, a domain-specific language (DSL) designed to support the modelling and management of DTeS throughout their lifecycle. In TwinLang, Digital Twins can be modelled as code, enabling abstraction, reusability, and integration with software engineering workflows. The DSL supports model transformations as a central concept, which enables refinement from abstract representations to concrete implementations. These transformations also facilitate the projection of models into different views, supporting tasks such as deployment, monitoring, and visualisation. TwinLang is realised as an internal DSL in Scala, leveraging the language’s expressiveness and strong type system. To illustrate and evaluate our approach, we present the development of a Digital Twin for our AI Lab as a concrete use case.
Modern logistics systems aim to leverage digital technologies and may integrate autonomous components to increase efficiency and flexibility. In so-called mixed-fleet systems, human workers, manually operated machines, and autonomous machines collaboratively work towards a common goal. The subsystems are loosely coupled and can be reconfigured flexibly, leading to a change in behavior. Modeling this behavior requires flexible designs for model-driven systems engineering. Serviceoriented architectures can help focus on defining the expected behavior, independently of the involved actors. Additionally, eventbased communication mechanisms can decouple interactions between subsystems. This paper explores the use of SysML v2 for modeling a service-oriented architecture of mixed-fleet systems. Based on an available set of requirements, suitable SysML v2 modeling elements are identified that can describe services, events, and service choreographies. We use the described concepts to create a SysML v2 model of a mixed-fleet harbor logistics use case. Based on this model, we demonstrate how business processes can be composed of reusable services and how requirements and verification can be integrated to ensure correctness of behavior. The results show that SysML v2 meets key requirements for service-oriented architectures and enables separating service definitions, actors, and verification elements. Reusable modeling patterns were applied to support scalability, and enable traceability within the model across actors and services. Furthermore, domain-specific constraints and requirements were composed into modeling elements using formal mechanisms to ensure that they are not only documented, but actively connected to the model.
While Digital Twins are traditionally associated with domains of human-made systems, Earth Science Digital Twins (ESDTs) have been rising in popularity. An ESDT represents a twin of a natural entity or ecosystem at a local, regional, or global scale, such as a river, a forest, an ocean, or the Earth itself. In this work, we conduct a systematic literature review of the existing ESDT systems. We identify the main ESDT domains as hydrology, geology, ecology, Earth, and atmosphere. We analyse the purposes of ESDTs, describe the software development tools that are often employed in ESDT development, and discuss the challenges and limitations that are specific to ESDTs. We conclude that data collection and computational limitations are among the most prominent, and maintaining the real-time mirror of the physical entity remains a challenge.
We present XModeler(ML) v3, a multi-level modeling tool that is based on the Flexible Multi-Level Modeling and Execution Language (FMMLx). Multi-level modeling enables the representation of domain knowledge at higher levels of classification, thus reducing conceptual redundancy and improving the integrity, adaptability, and comprehensibility of domain models. This furthermore allows for the specification of domain-specific languages using more generic domain-specific languages, rather than starting from scratch every time. The XModeler(ML) provides a shared representation of model and code and can be used to instantiate and execute models at runtime. With version 3, we now offer the opportunity to specify executable UML diagrams in XModeler(ML) using a specifically developed dialect of FMMLx called UML++. In this demonstration, we showcase how XModeler(ML) v3 integrates multi-level and two-level modeling modes. We describe practical modeling scenarios where this integrated approach offers clear advantages and demonstrate how users can define, execute, and evolve executable models seamlessly within the tool.
In model-driven network ecosystems, high-level specifications such as YANG data models define APIs that engineers extend with custom code and handlers. When the model evolves, it can become misaligned with its API implementation, which typically requires manual, costly, and error-prone realignment, highlighting the need for automated repair. This paper presents our ongoing work on an automated repair approach to resynchronize evolving YANG models with their API implementations, using a four-stage pipeline: (i) localize compatible-vs-breaking diffs between the current and updated models, (ii) enrich each diff record with contextual information, (iii) instantiate precise transformation rules for every change, and (iv) apply those rules to generate a repaired version of the code, complete with a reviewable change log for engineer validation. To explore the feasibility of this approach, we have developed an initial prototype-YANG-APR-that targets five common model evolution scenarios: datatype change, node rename, endpoint URL rename, endpoint removal, and endpoint addition. We illustrate the approach through two representative cases: a breaking datatype modification and a non-breaking endpoint addition. These early results show promising potential to facilitate model-implementation coevolution, providing a foundation for continued development and broader evaluation.
As graphical modeling editors evolve, legacy tools face growing obsolescence, prompting the need for effective migration strategies. This paper introduces GEM, a model-driven migration framework that leverages a pivot model to enable reusable and extensible transformations across heterogeneous platforms. By preserving detailed visualization semantics and simplifying migration through an intermediary representation, GEM facilitates modernization of legacy editors. The paper also reviews existing pivot model approaches, outlines evaluation criteria for migration frameworks, and presents a prototype implementation to validate the concept.
Large-scale industrial projects often experience significant delays and cost overruns due to the inherent complexity of modern industrial operations. These challenges necessitate robust process control mechanisms to optimize lead times and expenditures. To enhance operational efficiency and ensure compliance with international standards such as ISO, the formalization of business processes has become imperative. This formalization serves as a fundamental step toward effective activity monitoring within complex organizations. When properly modeled, business processes can be further optimized through verification and simulation techniques. This study examines the research conducted by PragmaDev in collaboration with Airbus under the OneWay project framework. The primary objectives of the project included the verification and simulation of business processes to assess cost implications and lead-time efficiency. Additionally, the study explored methodologies for managing model variability and developing digital twin architectures to enhance industrial process optimization.
Trustworthy and human-centric adaptation remains a central challenge for autonomous vehicles (AVs), which operate in dynamic and uncertain environments. This paper proposes a real-time, multi-modal, and self-adaptive framework that operationalizes Human-Centric Requirements Engineering (HCRE) by treating contextual signals as non-functional requirements (NFRs), where NFRs denote stakeholder-oriented soft goals such as trust, cognitive comfort, and perceived safety that must be continuously fulfilled alongside functional driving tasks. The framework integrates driver emotions, behaviors, traffic conditions, and vehicle dynamics within an interpretable neural architecture to deliver proactive behavior recommendations aligned with drivers' needs, and unlike prior approaches that rely on static rules or thresholds, it continuously elicits, monitors, and fulfills latent human-centric goals through transparent and context-aware adaptation. Trained and evaluated on the AIDE dataset, the system achieves high accuracy across perception modules (83-93%) and 89.32% exact match accuracy for integrated behavior recommendations, satisfies real-time constraints with an average inference latency of 106.84 ms, and maintains interpretability through explicit mappings from multimodal input to adaptive output. The results demonstrate the feasibility of embedding HCRE principles, particularly dynamic NFR fulfillment, into the core of AV control architectures, thereby enabling emotionally responsive and stakeholder-aligned autonomous systems.
The rise of large language models (LLMs) has led to new opportunities and challenges, particularly in model-driven engineering (MDE), where they are utilized to automate various modeling tasks. However, current research highlights some limitations that hamper the LLMs’ effectiveness, e.g., producing formatting errors in the generated modeling artifacts or semantic hallucination in domain modeling. Modeling business process management (BPM) applications presents significant challenges to end-users, ranging from adhering to OMG standards to ensuring data privacy. Although some approaches have started investigating the intersection between LLMs and BPM modeling, a completely automated process is still far from being realized. The main blocking issue is represented by the scarcity of modeling data, which is needed to train both LLMs and traditional modeling assistants. Being built on our previous work, we adopt an MDE framework called BP-MASTER-LLM, which is devoted to supporting the specification of BPM models by combining a set of automated modeling tools, i.e., model-event recorder and modeling assistant. Our findings reveal that the prominent LLMs can generate synthetic content even though human models and operations represent a better source of training data for the tested modeling assistant; nonetheless, if instructed properly, LLMs can represent a solution to overcome the data scarcity issue.
Language design, in Linguistics and in Programming, consists of three main categories: Syntax, semantics, and pragmatics. Pragmatics deals with the usage of meaningful combinations. The categories are interconnected, and neither can be fully understood in isolation. Pragmatic considerations of MLM are mainly devoted to the engineering advantages of MLM, over the “two-level” standard modeling approach. The main pragmatic argument is attributed to the reduction in accidental complexity, which results from elimination of duplication of instance-of links. However, the duplication occurs not in the model itself, but in the combined model-data space. Pragmatic considerations referring to the quality of MLM models appear implicitly in discussions about MLM modeling of the challenge examples presented in the sequence of MULTI workshops. In this paper, we discuss pragmatic criteria in MLM modeling. The paper examines semantic considerations, combination patterns, and MLM idioms (recurring fragments). The discussion is based on the experience gained while implementing an MLM model for the warehouse challenge example of the MULTI workshops in our MLM-USE application.
In the design of Electrical Vehicles (EVs) there are often many trade-offs to consider during system design. When considering safety aspects, these needs are further increased due to counter-acting incentives in design. Model-based tools and methods present a means of providing analytical capabilities through modeling of complex systems, possibly assisting in the analysis and design of EVs. In this paper we detail the experiences of using model-based methods in the design and analysis of a thermal management system for battery systems in solar cars. We use the context of the challenger class vehicles in the Bridgestone World Solar Challenge (BWSC) and discuss insights from applying a model-based method. The paper further reports on feedback from an active solar car project at a Swedish University. Our findings indicate that a model-based approach for safety analysis can provide meaningful tools for students to reason about safety during early system design stages.
As society transitions from traditional engineered physical systems to software-driven ecosystems, the demand for adaptive, cost-effective, and rapidly evolving technologies continues to rise. Digital Twins (DTs) have recently emerged as a key architectural and conceptual tool to address these needs by decoupling high-level system control and decision making from physical operations. By integrating deductive engineering models with inductive, data-driven insights, DTs offer powerful capabilities for simulation, prediction, and adaptive system management. However, despite their promise, current DT implementations often remain fragmented, domain-specific, and expensive to build and maintain. This paper proposes a research agenda for a structured and principled approach to the engineering of digital twins (EDT), grounded in established software and systems engineering practices. Our goal is to empower the creation of digital twins that are scalable, reusable, and trustworthy. Building on the capabilities outlined by the Digital Twin Consortium, we offer a complementary perspective by identifying key research challenges associated with their integration in the context of digital twin engineering. In addition, we examine the engineering lifecycle of DT systems and present a comprehensive research agenda for the EDT community.
Model validation is a critical challenge in ensuring the consistency of a SysML model, particularly in collaborative or large scale projects. This paper introduces the application of openCAESAR to model validation of SysML system model. This paper presents an approach for validating Sysml models using the openCAESAR platform and the Ontological Modeling Language (OML). By leveraging openCAESAR’s powerful consistency checking and inference capabilities, we can validate SysML models rigorously. The proposed approach has been applied to a sample SysML model from the Satellite DX Research Working Group at the Japan Aerospace Exploration Agency (JAXA). Initial experiments demonstrated that the OWL 2 Description Logic (DL) reasoner integrated within openCAESAR can effectively validate the semantic correctness of a system model against its meta-model.
In cross-organizational software engineering, every model managed may contain intellectual property (IP). That IP must be protected from potential competitors, who may also be fellow collaborators. To that end, organizations use access control systems to manage access to the parts of a model or the model as a whole as part of their model management efforts. However, access control systems often need to be shared, thereby potentially compromising the IP protection in place. In this paper, we introduce DEBAC, an approach for using the rules of role-based access control to derive encryption rules. We are able to circumvent the sharing of access control systems by using instead the encryption of model deltas (i.e., arbitrary changes to a model). To evaluate our approach, we generate use cases from open source GitHub repositories (committers being regarded as collaborating organizations) and show evidence for the functionality and applicability of the DEBAC approach. With the DEBAC approach, we can derive the encryption rules to be able to store any model on any untrusted infrastructure while still being able to preserve IP.
Model-driven optimization promises the availability of meta-heuristic approaches, such as evolutionary algorithms, for optimization problems to domain experts, without requiring deeper knowledge of the underlying technologies. Models directly serve as the problem specification, while model transformations form the evolutionary operators for exploring the search space. Evolutionary algorithms, however, are usually designed to operate on low-level representations (or encodings) such as bit strings or vectors of integers or real numbers. As models generally consist of multiple high-level constructs and are much more complex than the usual low-level encodings, a performance penalty is to be expected when operating directly on such models. On the other hand, models provide additional domain knowledge and, as such, allow for the design of more sophisticated evolutionary operators. In this work we compare the performance of evolutionary algorithms operating on high- and low-level representations of solutions on two problems, the class responsibility assignment problem as well as the multi-objective knapsack problem. The results show a clear advantage of high-level model representations in terms of solution quality, producing higher-quality results for all evaluated instances, which generally come at the cost of a higher total duration. The extent of the differences and their reasons are found to greatly depend on the complexity of the model and its instances.
Collaborative Modeling (CM) tools increasingly offer advanced editing features, such as live co-editing, embedded chat, and automated reasoning. Yet, the impact of these features on the user experience (UX) of modeling is less understood. In this paper, we explore the relationship between CM and UX by developing a mapping between existing CM and UX ontologies. We propose CMUX Explorer, an ontology-driven assessment tool that enables tool builders to identify CM features with UX impact. The conversational front-end allows for natural-language interactions, which CMUX Explorer translates to SPARQL queries, executes them against the ontology, and returns ranked, interpretable recommendations to guide design decisions. Importantly, our approach supports continuous, semi-automated improvement of the ontology, leading to more precise recommendations over time. Through this work, we pave a path towards computer-aided design of UX-aware collaborative modeling tools.
The dissemination of multi-level modeling (MLM) depends not least on graduates who learned to appreciate the benefits of this modeling paradigm. To this end, effective approaches to teaching MLM are of vital relevance. Knowledge about teaching MLM is scattered and limited. Therefore, there is a need to capture and further develop this knowledge. This article presents a multi-level model that supports the entire cycle from the preparation, implementation, and evaluation of courses on conceptual modeling in general and on MLM in particular. It is executable and, thus, also supports course monitoring and management. The multi-level model should allow for further instantiations to represent courses on other topics.
With the ongoing adoption of Model-Based Systems Engineering (MBSE) as a best practice in several domains there are ongoing efforts to improve the maturity of MBSE methods. In the vision for MBSE several organizations and actors foresee a landscape where there is a governing system model acting as an Authoritative Source of Truth (ASoT). The main argument for the ASoT is to have a single source where all information can be accessed and written to inside of an organization, supporting consistency and management of data. Recent works have proposed solutions that implement tool chains to support the ASoT. In this paper we take a critical look at the concept of ASoT and discuss several issues that inhibit ASoT creation and maintenance. We discuss challenges from three perspectives; systems theory, technical challenges, and industrial barriers.
