DevOps in Industry 4.0: A Systematic Mapping

DevOps is the acronym for the integration of development and operations, which allows the improvement of communication and collaboration. Likewise, its objective is to help organizations develop software products and services quickly without sacrificing quality or cost. Additionally, its focus is on development, integration, delivery, release, testing, and continuous monitoring through the automation of tasks . Industry 4.0 solutions seek to digitize the industries’ processes and services


I. INTRODUCTION
The software development process is a formal agreement for the construction of systems with the following stages: analysis, design, implementation, testing, putting into production, and maintenance [1].Thus, a development product must go through these stages, from its conception and requirements to its obsolescence, after the transformations throughout its use history.
In the 90s, software development was characterized by imprecise and changing requirements, in which delivery times required greater speed.In this scenario, there was an attempt to implement classical methodologies and models [2,3]; however, these did not provide the expected results.Thus, demonstrating the inefficiency of these methodologies to provide an efficient response to this typology of software products [4].In this context, agile approaches emerge as an alternative to traditional methods without compromising product quality, which iteratively focuses on software development, with changing requirements, rapid return on investment (ROI), deliveries functional in early stages, and high customer participation.Among them, we have: Scrum, Feature Driven Development, Test Driven Development, Extreme Programming [1], [5].
As software development processes have been improved, their automation has also begun to improve.This is how the DevOps concept appears, the combination of the words Development and Operations, which highlights the quality of products and services in correlation with technology to achieve virtualized environments, cloud technologies, greater automation, and tools for configuration management [6].In addition to motivating collaborative work, communication and integration of people to enhance continuous delivery and software products easier to deliver and maintain through automation [13].
On the other hand [7], the term Industry 4.0 emerges as a reference to what is considered to be the fourth industrial revolution, a synergy of digitization and extreme interconnection of productive activities.In addition to this, the impact on society is described, which is not merely technological.The integration of production technologies and Information and Communication Technologies (ICT) in the form of cyber-physical production systems presents infinite possibilities and opportunities to change how the industry adds value.Furthermore, the notable increase in communication, processing, and interaction capabilities with the environment suggests new products, services, business models, needs, and challenges, which ultimately question how the solutions proposed in this context are designed, developed, and maintained.
Considering this, it is pertinent to relate the literature on DevOps and Industry 4.0.
Such reflection can reveal opportunities to propose more efficient processes, lead to a notable increase in flexibility and agility, adaptability, speed in time to market, shorter delivery times, among others [8].In this sense, the objective of this article is to show the results of a systematic mapping that presents an updated vision about the links, benefits, and possibilities of the adoption of DevOps in Industry 4.0 solutions reported in the literature.
The article is organized as follows: Section 2 describes the research protocol used to perform the systematic mapping.Section 3 contains the analysis of the results obtained in the mapping.Section 4 presents the main discussion and observations of the results and the limitations and implications of adopting DevOps in Industry 4.0 solutions.Finally, Section 5 presents the conclusions and recommendations for future works.

II. METHODOLOGY
A systematic mapping of the literature is a methodological process that allows collecting and categorizing existing information on a research topic.This systematic mapping was made following the guidelines presented in the following studies: [9,10 11, 12].The systematic mapping was carried out in three stages: (i) Planning, where the research questions, keywords, synonyms, and search strings, which led to the inclusion and exclusion criteria, were defined; (ii) Execution, and (iii) Documentation of the results.For space reasons, this process is not included in this article; however, it can be found in its entirety at https://n9.cl/7otaz for consultation or eventual replication.

III. RESULTS
The contribution that each of the works made to the research questions proposed for this work (Selected Works in the following link: https://n9.cl/4lno3)can be consulted at https://n9.cl/pbktu.Also, the convention used to name the Selected Works will be an S for Study, followed by the consecutive number.Parentheses will be used to differentiate from the general references of this work.

A. Question 1: What is the Current State of Implementation of DevOps Practices in Industry 4.0?
According to the literature, there is a specific interest in the adoption of DevOps in Industry 4.0.Thirty-three percent of the works showed the benefits in the development of current software and its transfer to organizations in digital transformation processes or Industry 4.0 (S3, S7, S8, S9, S10, S11, S13, S15, and S16 ).The maturity and learnings that the software industry has had in culture, agile frameworks, and the use of agile principles and practices is something that Industry 4.0 should not ignore.Both in (S11) and (S10), it is highlighted that Industry 4.0 needs an immediate adoption of new proposals for the implementation of solutions and forms of deployment.In (S11) and (S13), on the one hand, the cascade model in general manufacturing systems integration projects is analyzed, as well as DevOps adapted to this type of development.The first does not adjust to the new challenges, defiances, and modern solutions; the second proposes applying practices such as continuous deployment and rapid feedback.
The different enabling technological developments in Industry 4.0, such as blockchain, big data, and IoT, can be managed and implemented with DevOps, bringing benefits similar to those identified in the software industry.In addition, it can be gained in the generation mechanism of approach to all areas involved with the development and management of production, with the active participation of the company's different roles, developers, and stakeholders.Explicitly in (S7), it is said that in many organizations, the digital transformation is related to Industry 4.0.This work highlights that to address these new software development and delivery  In this same work, it is proposed to break the segmentation between areas and their different functions to promote more cross-functional roles; for example, support engineers involved in the technical development process and developers involved in the operation and maintenance process and establish a unified support team.
As a whole, the article that contributed the most to answer this question was (S3).It explicitly addresses the concepts of DevOps for Industry 4.0 and practices such as the continuous process in operation, observation, and development.The work also describes how it is incorporated in industrial environments and presents the advantages or benefits that DevOps practices, methods, and culture would bring when integrated into industrial production environments, in which software is one of its components.

B. Question 2: What are the Benefits of Implementing DevOps Practices in Industry 4.0?
Seventy-four percent of the works reported benefits in the adoption of DevOps practices in Industry 4.0 Solutions (S3, S4, S5, S6, S7, S9, S10, S11, S14, S15, S16, S18, S20, S21, S22, S23, S24, S25, and S27), including automation practices, continuous process in operation, continuous observation and development, agile manufacturing, cloud manufacturing, collaborative manufacturing, on-site reliability engineering, dynamic infrastructure and operations, delivery continuous, software deployment in a short life cycle with continuous monitoring, communication between the development team, agile practices, Scrum, and Lean, among others.Furthermore, works like (S3, S4, and S7) confirmed that there is a change of thought when integrating DevOps in Industry 4.0; bringing notorious benefits such as processes with more flexibility to change due to continuous adaptation, the culture of iterative deliveries, continuous analysis and monitoring processes, early identification of requirements and in this way a continuous improvement of the product.Additionally, it was reported that when implementing DevOps, there is a motivation to have a high degree of automation in the processes, identifying what can be automated and what cannot, with particular emphasis on repetitive processes in companies to ensure the quality of these and in many cases their efficiency (S16 and S20).Also, with the implementation of DevOps in companies, a change in perspective was identified when producing a good or a service; it means that agile frameworks motivate the creation of value based on the user and constant feedback from clients or users so that things are correctly done (S17).
In (S22), an architecture proposal that offers a roadmap for companies with an Industry 4.0 approach using DevOps principles was developed.In this study, the authors claim that DevOps represents the latest agile, lean, and automation thinking advocated by today's O&M management.Additionally, (S25) examines the implementation of DevOps in practice through an exploratory case study based on interviews with 11 industries and professionals from nine organizations.As a result, an empirical taxonomy of DevOps implementation is described, in which the interaction of developers with local operations, outsourced operations, DevOps teams, and DevOps bridge teams can work as a reference for Industry 4.0.In this same line of work is (S27), where infrastructure as code is described as a DevOps tactic to manage and provision infrastructure through machine-readable definition files instead of physical hardware configuration or interactive configuration tools and from a very pertinent maintenance and evolution perspective in Industry 4.0.Finally, in (S26), the reconsideration of the traditional models of an regarding its IT areas is proposed.It suggests that these should follow certain fundamental design principles, with minimal functionality and fewer dependencies, portability, shared knowledge, predictable contracts, and maximized human value.According to its authors, the last three points encapsulate the very definition of DevOps.In the end, this work presents through a case the valuable lessons learned from the mistakes in the adoption of DevOps.

4.0?
Of the studies analyzed, 81% (S1, S2, S3, S5, S6, S7, S8, S9, S12, S13, S14, S17, S18, S19, S20, S21, S22, S23, S24, S25, S26, and S27) propose or describe DevOps practices that could be effective in companies with a vision of Industry 4.0; however, it is pointed out that it will depend a lot on the sector and the type of operations and developments that the company has; and although they are different adoption frameworks, it is possible to identify some similarities between them.In this context, we want to highlight that the implementation of DevOps implies a change at a cultural level in the company; this is one of the strongest barriers, especially when there are highly static processes or tasks that have been pre-established for a long time (S4 and S26).Likewise, people should be motivated to think that there may be a better way to get things done, improve the process or develop within the company to obtain a greater benefit.Additionally, having a continuous delivery of value, agile development, product management, reliability engineering on site, infrastructure and dynamic operations, continuous delivery, and teams that develop and deploy software in a short life cycle, with execution and resource management is also vital.This implies a change from project management approaches to techniques and theories influenced by operations management, agility, industrial engineering, and organizational culture (S7).
Revista Facultad de Ingeniería (Rev.Fac.Ing.) Vol. 30 (57), e13314.July-September 2021.Tunja-Boyacá, Colombia.L-ISSN: 0121-1129, e-ISSN: 2357-5328.DOI: https://doi.org/10.19053/01211129.v30.n57.2021.13314learning from mistakes, managing complexity, and technical possibilities.In addition, it is vital to identify which enabling technologies for Industry 4.0 are wanted to be implemented, as well as the processes and tasks that can be automated.Likewise, it is needed to measure the organization's capabilities and cost; in short, be realistic with the adoption of DevOps (S5).This work presents a literature review that identifies preparation models for Industry 4.0; it proposes dimensions such as technology, people, strategy, leadership, processes, and innovation.This work also highlights organizations' strengthening and technological preparation and the evaluation of technological capabilities as the most critical dimensions.

Adoption of DevOps Practices in Industry 4.0?
The tools used in the software development life cycle can be considered in Industry 4.0 management environments.Thirty-seven percent of the selected works describe it that way (S2, S3, S4, S8, S12, S13, S15, S16, S17, and S19).It should be noted that many of the systems that are currently used in this context focus on collecting data from both the physical and digital environment with diverse technologies: industrial IoT devices, robotic tools, autonomous robots, Big Data systems, digital twins that collect and analyze large amounts of data, artificial intelligence systems, cognitive systems and systems that implement augmented reality; all of which requires understanding other requirements beyond technologies.For example, connectivity, interoperability, and an understanding of data, some for making decisions in real-time and others in the medium and long term (S3, S15, S16, S17, and S20).Therefore, the entire production and value chain must be analyzed to find a way to integrate DevOps and the technologies that support it in the product life cycle, and in the case of Industry 4.0, to generate a synergy between production and development of systems.
In (S15), the authors claim that IT is fundamental for innovation in almost any industry, and those that are used correctly and effectively will have a significant leadership position; otherwise, they will not be able to survive.The authors of this work illustrate Blockchain technology as an enabler of Industry 4.0.This research suggests that this will play a fundamental role in the digital transformation of industries and applications in general.It will allow secure trust frameworks, generate an agile value chain production and a closer integration with technologies, including cloud computing and the IoT.With a case study, the researchers demonstrated the ability to apply engineering principles, combined with a DevOps approach to development and iterative management and integration of security.Other works that present an enabler of Industry 4.0 are (S16) and (S20 in which it is stated that the manufacture of products must respond to greater optimization to satisfy the needs of the global market. Consequently, an agile, dynamic production process that allows competitiveness will be required.Responding to the need for digitization of factories, an example of such digitization is the digital twin.These works describe the implementation of a digital twin and how it can even be an enabler for DevOps applications in cyber-physical production systems.According to the authors, DevOps aims to merge development and operations to provide a continuous and agile process; the production process can be fully integrated and automated with continuous improvement since a digital twin improves flexibility due to its adaptability and perfect interaction between the physical system and its virtual model.In short, DevOps is relevant in the process of implementing this type of system.Plus, it closes the gap between development and operations throughout the entire product lifecycle.Another enabling technology is the industrial big data analysis shown in (S17) and (S3).In (S17) specifically, the adoption of DevOps is presented to allow adaptation and continuous improvement in industrial manufacturing by making industrial data available to stakeholders; this, given that trends towards smart manufacturing, cyber-physical production systems, and Industry 4.0 result in a large number of industrial internet of things (IoT) sensors and data to be managed.Finally, in (S3), a software platform called Titan is presented, accompanied by a model that incorporates industrial production environments with DevOps, which can integrate different types of sensors that use data schemas, formats, and protocols in addition to what allows infrastructure In (S8), it is said that automation and monitoring practices must be considered for the adoption of DevOps, which is based on agile thinking in the context of solutions for Industry 4.0.It also emphasizes breaking down the division of different roles and establishing a unified incident support team.In the same sense, in (S9), DevOps is presented as the latest Agile and Lean vision for the current management of operations and maintenance.It is the new mode of operation and maintenance proposed in new technologies such as cloud computing and automated operation, emphasizing automation and continuous operation and maintenance of everything.
DevOps concepts such as agile development, product management, on-site reliability engineering, dynamic infrastructure and operations, continuous delivery, and teams that develop and deploy must be integrated into the software.