Information modeling takes place in various domains, dimensions and at different levels of detail. Specifically, in the tax compliance domain we encountered a wide field of use cases where information modeling is required. Nowadays, tailored modeling solutions exist that help domain experts to efficiently work on their models. Keeping those models and their backing concepts coherent and aligned is a competing but nevertheless important requirement. Therefore, we see a need for a holistic approach to information modeling that spans across multiple dimensions: abstract to low-level, from semantic to logical as well as from domain-specific to canonical. This industrial experience paper presents such a holistic approach and reports on its process, meta-metamodel, and implementation applied in the tax compliance domain.

The lifecycle data of industrial devices is typically maintained in separate data sources operating in silos. The lack of interoperability between the data sources due to the usage of different APIs, data formats and data models results in time-consuming and error-prone manual data exchange efforts. The notion of digital twins is known to be a solution to the data silo problem and the associated interoperability issues. Despite many studies on digital twins, there is still a need for common architectures that offer means for defining digital twins, ingesting backend data in the digital twins, and enabling interoperable data exchange across the lifecycle of the devices via their digital twins. This paper aims to close this gap by proposing a cloud-based architecture, a common information meta-model for defining the digital twins, and variety of APIs to query the lifecycle data of interest via the digital twins.

Cyber-physical systems and the Internet-of-Things are getting more and more traction in different application areas. Boosted by initiatives such as Industrie 4.0 in Germany or the Industrial Internet Consortium in the US, they are enablers for innovation in industrial automation. To provide the advanced flexibility in production envisioned for future automation systems, Programmable Logic Controllers (PLCs), as one of their main building blocks, also need to become more flexible. However, the conservative nature of this domain prohibits changes in the controller architecture impacting the installed base. Currently there exist various approaches that evolve control architectures to the next level, but none of them address flexible function deployment at the same time with legacy support. In this paper, we present a an architecture for a multi-purpose controller that is inspired by the virtualization trend in cloud systems which moves from heavyweight virtual machines to lightweight containers solutions such as LXC or Docker. Our solution includes the support for multiple PLC execution engines and adds support for the emulation of legacy engines as well. We evaluate this architecture by executing performance measurements that analyze the impact of container technologies to the real-time aspects of PLC engines.

In industrial automation, the control and operation of factories and plants in usually performed by Programmable Logic Controllers (PLCs). As automation tasks find their way into more and more smaller application areas where it is economically not reasonable to employ, engineer and maintain a full, hardware-based PLC, a cloud-based software PLC (Soft-PLC) can serve as a Control-as-a-Service solution. In this paper, we address the question how a Soft-PLC can be augmented with cloud-required aspects such as elastic scalability and multi-tenancy. We introduce an architecture that realizes these properties and allows for the implementation of a multi-tenant, horizontally scalable Soft-PLC. Furthermore, we evaluate the most important quality attributes, i.e. Roundtrip times and scalability of our architecture based on a prototypical implementation and a case study from the building automation domain. Finally, we present lessons learned and give an outlook to further realistic scenarios for cloud-based control in industrial applications.

Architectural Knowledge Management (AKM) has been a major topic in software architecture research since 2004. Open AKM problems include an effective, seamless transition from reusable knowledge found in patterns books and technology blogs to project-specific decision guidance and an efficient, practical approach to knowledge application and maintenance. We extended our previous work with concepts for problem space modeling, focusing on reusable knowledge, as well as solution space management, focusing on project-level decisions. We implemented these concepts in ADMentor, an extension of Sparx Enterprise Architect. AD Mentor features rapid problem space modeling, UML model linkage, question-option-criteria diagram support, meta-information for model tailoring, as well as decision backlog management. We validated ADMentor by modeling and applying 85 cloud application design decisions and 75 workflow management decisions, creating one problem and three sample solution spaces covering control system architectures, and obtaining user feedback on tool and model content.

Today's industrial control systems store large amounts of monitored sensor data in order to optimize industrial processes. In the last decades, architects have designed such systems mainly under the assumption that they operate in closed, plant-side IT infrastructures without horizontal scalability. Cloud technologies could be used in this context to save local IT costs and enable higher scalability, but their maturity for industrial applications with high requirements for responsiveness and robustness is not yet well understood. We propose a conceptual architecture as a basis to designing cloud-native monitoring systems. As a first step we benchmarked three open source time-series databases (OpenTSDB, KairosDB and Databus) on cloud infrastructures with up to 36 nodes with workloads from realistic industrial applications. We found that at least KairosDB fulfills our initial hypotheses concerning scalability and reliability.

Architecture decisions are often not explicitly documented in practice but reside in the architect's mind as tacit knowledge, even though explicit capturing and documentation of architecture decisions has been associated with a multitude of benefits. As part of a research collaboration with ABB, we developed a tool to document architecture decisions. This tool is an add-in for Enterprise Architect and is an implementation of a viewpoint-based decision documentation framework. To validate the add-in, we conducted an exploratory case study with ABB architects. In the study, we assessed the status quo of architecture decision documentation, identified architects' expectations of the ideal decision documentation tool, and evaluated the new add-in. We found that although awareness of decision documentation is increasing at ABB, several barriers exist that limit the use of decisions in practice. Regarding their ideal tool, architects want a descriptive and efficient approach. Supplemental features like reporting or decision sharing are requested. The new add-in, was well-perceived by the architects. As a result of the study, we propose a clearer separation of problem, outcomes, and alternatives for the decision documentation framework.

In a large corporate organization there are sometimes similar software products in certain subdomains with a perceived functional overlap. This promises to be an opportunity for systematic reuse to reduce software development and maintenance costs. In such situations companies have used different domain analysis approaches (e.g., SEI Technical Probe) that helped to assess technical and organizational potential for a software product line approach. We applied existing domain analysis approaches for software product line engineering and tailored them to include a feature analysis as well as architecture evaluation. In this paper, we report our experiences from applying the approach in two subdomains of industrial automation.

Complex information flows in the domain of industrial software systems complicate the creation of performance models to validate the challenging performance requirements. Performance models using annotated UML diagrams or mathematical notations are difficult to discuss with stakeholders from the industrial automation domain, who often have a limited software engineering background. We introduce a novel model transformation from Use Case Maps (UCM) to the Palladio Component Model (PCM), which enables performance modeling based on an intuitive notation for complex information flows. The resulting models can be solved using existing simulators or analytical solvers. We validated the correctness of the transformation with three case study models, and performed a user study. The results showed a performance prediction deviation of less than 10 percent compared to a reference model in most cases.

Model-Driven Engineering (MDE) aims at improving the development of complex computer systems. Within this context textual concrete syntaxes for models are beneficial for many reasons. They foster usability and productivity because of their fast editing style, their usage of error markers, autocompletion and quick fixes. Several frameworks and tools from different communities for creating concrete textual syntaxes for models emerged during recent years. However, open issues are incremental parsing and model updating as well as the handling and synchronisation of partial and federated views with the underlying model. In this paper we present an approach for concrete textual syntaxes that makes use of incremental parsing and transformation techniques.

Efforts for software evolution supersede any other part of the software life cycle. Technological decisions have a major impact on the maintainability, but are not well reflected by existing code or architecture based metrics. The way the persistency of object structures with relational databases is solved affects the maintainability of the overall system. Besides maintainability other quality attributes of the software are of interest, in particular performance metrics. However, a systematic evaluation of the benefits and drawback of different persistency frameworks is lacking. In this paper we systematically evaluate the maintainability and performance of different technological approaches for this mapping. The paper presents a testbed and an evaluation process with specifically designed metrics to evaluate persistency techniques regarding their maintainability and performance. In the second part we present and discuss the results of the case study.