To that end, self-adaptive systems (also called autonomic systems) were proposed as a solution

Full text


Analyzing State-of-the-Practice for Self-Adaptive Systems in Industry using Data Analytics Nadeem Abbas, Danny Weyns, Ilir Jusufi, Bengt Larsson

Core Research Areas: Computer Science (self-adaptive systems), Visual Analytics, Social Sciences Introduction: Computing and communication technologies enable industries to improve productivity and quality while reducing costs and time to market. However, continuous growth in computational power, the interconnectivity of systems, and the integration of ever-changing technologies make software systems nowadays very complex for humans to manage. Back in 2001, IBM already

published a manifesto that urged for innovative solutions to manage the ever-increasing complexity of computing systems [1]. To that end, self-adaptive systems (also called autonomic systems) were proposed as a solution. A self-adaptive system monitors itself and its environment, and adapts its behavior and/or structure to maintain its goals under changing conditions [2]. Over the past two decades, substantial efforts have been made to investigate self-adaptive systems' design and

engineering [3]. With the growing amount of data that systems need to process, the relevance of self- adaptation as a property of software-intensive systems has become more critical than ever before.

Problem Definition and Value: Research has established a vast body of knowledge to support the engineering of self-adaptive systems. While this knowledge is documented in research articles, journal volumes, and books, it is currently not clear whether and to what extent this knowledge is recognized and used by practitioners1 or whether they have developed alternative solutions. To the best of our knowledge, no systematic study has been done to investigate the perception and use of self-adaptation in industry. Thus, there is no clear and documented view on practitioners’ understanding and

knowledge of self-adaptation, for what and how the principles of self-adaptation are applied in practice, and how these approaches support software-intensive systems that are exposed to huge volumes of data. This lack of insights forms a long-lasting open problem that we plan to resolve by doing a large-scale field study. The field study will require significant effort, involving at least 100 practitioners from companies across Europe and a team of about ten researchers. To ensure valid results, the field study requires a solid design, with appropriate methods for data collection and analysis. Therefore, it is crucial that the field study is prepared using a pilot. The primary goal of this seed project proposal is to perform this pilot and optimize the methods used for the large field study.


O1: Design and test research goals, data collection, and analysis methods for the large-scale field study.

O2: Collaborate with researchers specializing in qualitative data collection (questionnaire and interviews), and analysis and visualization methods.

O3: Connect with potential industry partners and identify an industrial Ph.D. student in

association with the LNU industry graduate school Data Intensive Applications (DIA). The PhD project is intended to closely relate to the use of self-adaptation in data-intensive systems. We plan to apply for the KK HÖG2 program to fund the project.

Expected Results: We aim at achieving the following results, which directly map to the above- specified objectives, respectively.

R1: Evaluation results of the pilot - targeting O1. The results will be documented and used to fine-tune the field study’s design and methods. Hence, a final questionnaire comprising open and closed type questions and a guide to conduct interviews with practitioners will be produced based on the results of the pilot project.

R2: Strengthen our ties with the DISA research groups, particularly those that specialize in data analysis and social sciences, and are interested in the pilot study (and potentially the follow-up project afterward) – targeting O2. Hence, we aim for a multidisciplinary approach and plan to investigate whether and how data-driven methods and techniques such as machine learning are used in industry, focusing on self-adaptive software-intensive systems.

R3: Write a research proposal for KK HÖG program in collaboration with an Industrial partner to establish an Industrial PhD student – targeting O3.

1 These are designers, developers, maintainers, and other people involved in engineering industrial software.



In the long run, we expect that the field study (that follows the seed project) will help the industry and academic research community to align their needs and goals and make well-informed decisions. On the one hand, the study findings will enable researchers to understand the open problems with the state-of-the-practice and the challenges practitioners face in adopting self-adaptation. On the other hand, practitioners can exploit the study results to enhance their knowledge to address industrial problems.

Seed Project’s Activities and Time Plan: The main activity of the seed project is to perform a pilot study in collaboration with ten practitioners from Sweden. Each practitioner will be requested to fill an online questionnaire. After analyzing the questionnaire data, three volunteer practitioners will be interviewed to collect additional details, crosscheck and triangulate findings drawn from the questionnaire data. The concrete time plan is as follows:

Activity Researchers Timing Results

A1: Feasibility Study (FS) – Planning and Design Nadeem, Danny Oct 2020 R1, R2 A2: FS – Prepare Online Questionnaire Nadeem, Danny, Bengt Oct 2020 R1, R2, R3

A3: FS – Analyze Questionnaire Data All Nov – Dec 2020 R1, R2

A4: FS – Prepare Interview-Questions Nadeem, Danny, Bengt Dec 2020 R1, R2 A5: FS – Interview Three Practitioners Nadeem, Danny Dec 2020 R1, R2, R3 A6: FS – Analyze Interviews + Questionnaire Data All Dec – Jan 2021 R1, R2

A7: Identify Ph.D. Student Nadeem, Danny Jan – Feb 2021 R3

A8: FS – Report Writing Nadeem, Danny Jan - Feb 2021 R1, R2


Researcher Role

Danny Weyns, Professor, Dept of Computer Science and Media Technology, Linnaeus University.

Principal investigator

Key Responsibilities: Project management, Support Nadeem in feasibility study design and execution, data analysis and reporting Nadeem Abbas, Senior Lecturer, Dept of Computer

Science and Media Technology, Linnaeus University.

Co-principal investigator

Key Responsibilities: Feasibility study design and execution, data analysis and reporting Ilir Jusufi, Senior Lecturer, Dept of Computer

Science and Media Technology, Linnaeus University.


Key Responsibilities: support with data analysis, in particular, visualization Bengt Larsson, Professor, Dept. of Social Sciences,

Linnaeus University.


Key Responsibilities: support with study design and data analysis

Budget: The estimated budget of the seed project is 100K SEK, which is distributed as follows:

• Danny Weyns 30K (20K in 2020 and 10K in 2021)

• Nadeem Abbas 30K (20K in 2020 and 10K in 2021)

• Ilir Jusufi 15K (10K in 2020 and 5K in 2021)

• Bengt Larsson 15K (10K in 2020 and 5K in 2021

• Interviews Transcription 10K (5K in 2020 and 5K in 2021)


1. J. Kephart and D. Chess. The vision of autonomic computing. Computer, 36(1):41-50, 2003.

2. R. De Lemos, H. Giese, H. Muller, et al. Software engineering for self-adaptive systems: A second research roadmap. In Software Engineering for Self-Adaptive Systems II, pages 1-32.

Springer, 2013.

3. D. Weyns, Software engineering of self-adaptive systems: an organised tour and future challenges, Handbook for Software Engineering, 2020.



Relaterade ämnen :