Machine learning (ML)-enabled systems are being increasingly adopted by companies aiming to enhance their products and operational processes. This paper aims to deliver a comprehensive overview of the current status quo of engineering ML-enabled systems and lay the foundation to steer practically relevant and problem-driven academic research. We conducted an international survey to collect insights from practitioners on the current practices and problems in engineering ML-enabled systems. We received 188 complete responses from 25 countries. We conducted quantitative statistical analyses on contemporary practices using bootstrapping with confidence intervals and qualitative analyses on the reported problems using open and axial coding procedures. Our survey results reinforce and extend existing empirical evidence on engineering ML-enabled systems, providing additional insights into typical ML-enabled systems project contexts, the perceived relevance and complexity of ML life cycle phases, and current practices related to problem understanding, model deployment, and model monitoring. Furthermore, the qualitative analysis provides a detailed map of the problems practitioners face within each ML life cycle phase and the problems causing overall project failure. The results contribute to a better understanding of the status quo and problems in practical environments. We advocate for the further adaptation and dissemination of software engineering practices to enhance the engineering of ML-enabled systems. • International survey gathering insights from 188 practitioners across 25 countries. • Overview of current practices and challenges in engineering ML-enabled systems. • Inferential quantitative analysis reporting the status quo with confidence intervals. • Qualitative analysis mapping ML life cycle challenges and causes of project failure.

Computing systems form the backbone of many areas in our society, from manufacturing to traffic control, healthcare, and financial systems. When software plays a vital role in the design, construction, and operation, these systems are referred to as software-intensive systems. Self-adaptation equips a software-intensive system with a feedback loop that either automates tasks that otherwise need to be performed by human operators or deals with uncertain conditions. Such feedback loops have found their way to a variety of practical applications; typical examples are an elastic cloud to adapt computing resources and automated server management to respond quickly to business needs. To gain insight into the motivations for applying self-adaptation in practice, the problems solved using self-adaptation and how these problems are solved, and the difficulties and risks that industry faces in adopting self-adaptation, we performed a large-scale survey. We received 184 valid responses from practitioners spread over 21 countries. Based on the analysis of the survey data, we provide an empirically grounded overview the of state of the practice in the application of self-adaptation. From that, we derive insights for researchers to check their current research with industrial needs, and for practitioners to compare their current practice in applying self-adaptation. These insights also provide opportunities for applying self-adaptation in practice and pave the way for future industry-research collaborations.

The engineering of production systems requires capabilities for the coordinated reconfiguration between production variants, i.e., Product-Process-Resource (PPR) variants with multi-disciplinary dependencies. However, traditional approaches to coordinate reconfiguration, e.g., scripted workflows, consider production dependencies implicitly and require validation regarding multi-disciplinary PPR dependencies. This vision paper explores knowledge representation to coordinate production re-configuration with Industry 4.0 components. For validating pre-and post-conditions of a flexible, coordinated reconfiguration process, we introduce the PPR Asset Network with Reconfiguration (PAN+R) approach, which builds on the PPR Asset Network (PAN) to represent PPR model variants with their dependencies and states in transition between variants. We initially evaluate the PAN+R approach with a use case on a work cell for joining car parts. We conclude with a research agenda towards coordinating production reconfiguration with human and machine agents.

In Production System Engineering (PSE), domain experts from different disciplines reuse assets such as products, production processes, and resources. Therefore, PSE organizations aim at establishing reuse across engineering disciplines. However, the coordination of multi-disciplinary reuse tasks, e.g., the re-validation of related assets after changes, is hampered by the coarse-grained representation of tasks and by scattered, heterogeneous domain knowledge. This paper introduces the Multi-disciplinary Reuse Coordination (MRC) artifact to improve task management for multi-disciplinary reuse. For assets and their properties, the MRC artifact describes sub-tasks with progress and result states to provide references for detailed reuse task management across engineering disciplines. In a feasibility study on a typical robot cell in automotive manufacturing, we investigate the effectiveness of task management with the MRC artifact compared to traditional approaches. Results indicate that the MRC artifact is feasible and provides effective capabilities for coordinating multi-disciplinary re-validation after changes.

Self-adaptation equips a software system with a feedback loop that automates tasks that otherwise need to be performed by operators. Such feedback loops have found their way to a variety of practical applications, one typical example is an elastic cloud. Yet, the state of the practice in self-adaptation is currently not clear. To get insights into the use of self-adaptation in practice, we are running a large-scale survey with industry. This paper reports preliminary results based on survey data that we obtained from 113 practitioners spread over 16 countries, 62 of them work with concrete self-adaptive systems. We highlight the main insights obtained so far: motivations for self-adaptation, concrete use cases, and difficulties encountered when applying self-adaptation in practice. We conclude the paper with outlining our plans for the remainder of the study.

The timely discovery, sharing and integration of architectural knowledge (AK) have become critical aspects in enabling the software architects to make meaningful conceptual and technical design decisions and trade-offs. In large-scale organizations particular obstacles in making AK available to architects are a heterogeneous pool of internal and external knowledge sources, poor interoperability between AK management tools and limited support of computational AK reasoning. Therefore we introduce the Continuous Architectural Knowledge Integration (CAKI) approach that combines the continuous integration of internal and external AK sources together with enhanced semantic reasoning and personalization capabilities dedicated to large organizations. Preliminary evaluation results show that CAKI potentially reduces AK search effort by concurrently yielding more diverse and relevant results.

Collective intelligence systems (CIS), such as wikis, social networks and content sharing platforms, have dramatically improved knowledge creation and sharing at society level. There is a trend to exploit the stigmergic mechanisms of CIS also at organization/corporate level. However, despite the wide adoption of CIS, there is a lack of consolidated systematic knowledge of the architectural principles and practices that underlie CIS. Software architects lack guidance to design CIS for the application context of individual organizations. To address these challenges, we contribute with an architecture framework for CIS, aligned with ISO/IEC/IEEE 42010. The CIS-AF framework provides guidance for architects to describe key CIS elements and systematically model a CIS that is well-suited for an organization's context and goals. The framework is grounded in an in-depth analysis of existing CIS, workshops and interviews with key stakeholders, and experiences from developing a prototypical CIS. We evaluated the architecture framework in two cases in industry setting where CIS have been designed and implemented using the framework. Results show that the framework effectively supports stakeholders with providing a shared vocabulary of CIS concepts, guiding them to systematically apply the stigmergic principles of CIS, and supporting them with kick starting CIS in their organizations.

The development of video games comprises engineering teams within various disciplines, e.g., software engineering, game production, and creative arts. Game jams are a promising approach for (software+) development projects to foster on new product development. This paper evaluates the concept of game jam, a community design/development activity, and its positive effects on new software product development with tight schedules in time-oriented, competitive environments. Game jams have received more public attention in recent times, but the concept itself has not been formally discussed so far. A game jam is a composition of design and development strategies: new product development, participatory design, lightweight construction, rapid experience prototyping, product-value focusing, aesthetics and technology, concurrent development and multidisciplinarity. Although game jams are normally used for rapid prototyping of small computer games, the constellation of the mentioned elements provides a powerful technique for rapidly prototyping new product ideas and disruptive innovations.