https://www.mdu.se/

After more than thirty-five years of research and development in software architecture, several fundamental challenges remain unsolved. First, despite the importance of having a well-defined architecture description aligned with the system, inconsistencies and misalignments are still prevalent. Second, although numerous languages exist to describe architectures, none have achieved widespread use or recognition as a de facto standard. Third, while architecture is dynamic and evolving, with architectural decisions often made by non-architect stakeholders, there are no universally accepted methodologies to capture emergent aspects and incorporate them into the architecture.In this paper, we explore the emerging concept of architecture as code. Inspired by the success of infrastructure as code, which enables infrastructure management in a codified, automated, and repeatable manner, architecture as code aims to bring similar benefits to software architecture. To the best of our knowledge, this is the first scientific paper to study this concept in depth within the context of software architecture, providing a comprehensive description and analysis of its characteristics. We also investigate how architecture as code is implemented and applied in practice.

Digital twins serve as virtual representations of systems, enabling capabilities such as

intelligent monitoring, real-time control, decision-making, and predictive analytics. The

Asset Administration Shell (AAS) is the pivotal Industry 4.0 standard for digital twin

engineering. In parallel, the Systems Modeling Language version 2 (SysMLv2) has

emerged as the main modeling standard for systems engineering, providing a

formalized and semantically rich approach to system modeling. With its growing

adoption, models developed in the early engineering phases are expected to be widely

available in an interchangeable format, encouraging their reuse in digital twins. Instead

of manually re-creating models for digital twinning, existing system models can be

leveraged.

However, SysMLv2 lacks direct integration with digital twin standards such as AAS,

necessitating a dedicated approach for seamless interoperability between the early

engineering phases and the operation of digital twins.

This paper investigates the conceptual alignment between SysMLv2 and AAS

specifications, examines structural and behavioral modeling aspects, and proposes a

systematic approach to mapping SysMLv2 to AAS, ensuring the automatic generation

of AAS models from SysMLv2 models.

To realize this approach, we employ model-driven engineering techniques leveraging

the Eclipse Modeling Framework and model transformations based on the Query View

Transformation language.

The proposed model transformation incorporates query mechanisms for structured

element extraction, information and structural integrity preservation, ensuring semantic

consistency and seamless data integration between the two investigated standards.

We develop and validate the model transformation following an iterative test-driven

development approach using a set of 24 SysMLv2 examples, sourced from the official

SysMLv2 repository.

The 6th edition of the workshop on Automated and verifiable Software sYstem DEvelopment (ASYDE) provided a forum to share and discuss innovative contributions to research and practice related to novel software engineering approaches to automated and verifiable development of software systems.

This paper introduces the Hazard and Threat Ontology, a hybrid ontology designed to illustrate safety hazards and security threats in complex systems of systems. Hazard Ontology and Combined Security Ontology are two ontologies with extensive terminology and complementary methodologies. They allow us to develop a hybrid approach that enables safety and security experts to analyze complex systems thoroughly. Combining these ontologies enhances the depth and scope of experts' analysis and decision-making process, and several tangible benefits are associated with using a hybrid approach across different industrial sectors. In this paper, an industrial use case illustrates the practical utility of the Hazard and Threat Ontology. Our approach facilitates the identification of hazards and threats, providing actionable insights into how to mitigate them. Consequently, assets and personnel can be protected, downtime can be reduced, and operational resilience can be enhanced.

A crucial element of Industry 4.0, is the utilization of smart devices that generate log files. Log files are key components containing data on system operations, faults (unexpected glitches or malfunctions), errors (mistakes or incorrect actions), and failures (complete breakdowns or non-functionality). This paper presents a systematic mapping study analyzing research conducted on log files for smart-troubleshooting in Industry 4.0. To the best of our knowledge, this is the study that aims to identify research trends, log file attributes, techniques, and challenges involved in log file analysis for smart-troubleshooting. From an initial set of 941 potentially relevant peer-reviewed publications, 74 primary studies were selected and analyzed using a meticulous data extraction, analysis, and synthesis process. The results of the study demonstrate that the majority of research has focused on developing algorithms for log analysis, with machine learning being the most commonly used approach. The smart-troubleshooting encompasses a range of activities and tools that are essential for collecting failure data generated by diverse interconnected devices, conducting analyses, and aligning them with troubleshooting instructions and software remedies. Moreover, the study identifies the need for further research in the areas of real-time log analysis, anomaly detection, and the integration of log analysis with other Industry 4.0 technologies. In conclusion, our study provides insights into the current state of research in log analysis for smart-troubleshooting in Industry 4.0 and identifies areas for future research. The use of smart devices generating log files in Industry 4.0 highlights the importance of log file analysis for troubleshooting purposes. Further research is needed to address the challenges and opportunities in this field to integrate log analysis with other Industry 4.0 technologies for performing more efficient and effective troubleshooting.

Context: Machine learning is increasingly used in a wide set of applications ranging from recommendation engines to autonomous systems through business intelligence and smart assistants. Designing and developing machine learning systems is a complex process that can be eased by leveraging effective design decisions tackling the most important challenges and by having a good system and software architecture. Goal: The research goal of this work is to identify common challenges, best design practices, and main software architecture design decisions of machine learning enabled systems from the point of view of researchers and practitioners. Method: We performed a mixed method including a systematic literature review and expert interviews. We started with a systematic literature review. From an initial set of 3038 studies, we selected 41 primary studies, which we analysed according to a data extraction, analysis, and synthesis process. In addition, we conducted 12 expert interviews that involved researchers and professionals with machine learning expertise from 9 different countries. Findings: We identify 35 design challenges, 42 best practices and 27 design decisions when architecting machine learning systems. By eliciting main design challenges, we contribute to best practices and design decisions. In addition, we identify correlations among design challenges, decisions and best practices. Conclusions: We believe that practitioners and researchers can benefit from this first and comprehensive analysis of current software architecture design challenges, best practices, and design decisions. 

In the system development of modern railroad vehicles, engineers frequently use a plethora of diverse notations to specify various systems, subsystems, and their associated concerns. The use of diverse notations introduces complex challenges linked with their management and integration. Conventional practices, which rely on manual revisions and translations, prove to be both time-intensive and cost-prohibitive. In addition, they carry substantial risks of human error, thereby potentially introducing faults into the system. Such practices are deemed inadequate for the railway industry, which is safety-critical in its nature and places paramount importance on the assurance of reliability and data integrity. To address these challenges, we developed a regular expression-based system facilitating the automatic translation of semi-structured texts into structured data, with a particular focus on ensuring data integrity and reliability. We have defined the system capitalizing on the insights and practical experience of our industrial partner, Alstom Rail Sweden AB, and validated it within their development process. The validation demonstrated the practicality of the system in a real-world context and highlighted valuable lessons learned throughout the process. Building on these insights, we applied model-driven engineering principles to generalize the system, providing an automated solution to the data extraction challenge from tender documents in the railway domain.