Keywords

1 Introduction

The implementation of good management and governance practices has become one of the main focuses within organizations. Therefore, it is necessary for each sector/area to make adaptations to ensure better adaptation to the management policies adopted by the business organization [1]. This work will focus on the adequacy of these management policies within the Robotic Process Automation (RPA) area, one of the IT (Information Technology) areas.

The management of processes in the IT (Information Technology) area seeks to develop policies, standards, norms, and guidelines that ensure everything is done correctly. In this way, it contributes to the guarantee of increasingly reliable and robust processes [2].

The governance and management of IT end up harmonizing and combining the activities that the IT area develops according to the needs and strategic objectives established by the organization. Always looking to develop reliable and available services to achieve business excellence where management processes are implemented [3].

The implementation of a management structure in the RPA area should contribute to a greater effectiveness of all the processes developed, in addition to directing efforts to then achieve the defined results.

Robotic Process Automation (RPA) aims to automate business processes or parts of them with software robots, through the reproduction of human interactions with the graphical user interface [4, 5]. In addition to productivity and improvement of administrative processes, it helps to relieve employees of tedious and repetitive work. Despite being a tool that significantly contributes to improving the quality of life at work, a critical point related to this technology is the rejection by employees for fear of losing their jobs due to the implementation of robots [6].

RPA is about using digital robots and artificial intelligence to eliminate/minimize human errors in repetitive processes and make them faster and more efficient. It is a technology that mimics the way a human interacts with the machine, performing tasks through configured software or another technological aspect, such as one (or more) robots [7].

The implementation of RPA reduces the manual burden within companies, in their various administrative or operational sectors. In this way, it guarantees greater autonomy to the teams, to focus on strategic issues that lead the company to fulfil its objectives [8].

To manage the quality of products, several tools and techniques are used, among them the PDCA cycle, which is also called the Deming cycle. Initially it was created for the process of quality improvement in the production area, however, this is a tool capable of being used in any management process [10].

The PDCA was developed in the 17th century by Francis Bacon when he proposed inductive studies, which went through stages that were later identified in the PDCA cycle.

The application of the PDCA cycle is possible when:

  • Starting a new improvement project;

  • Developing a new or improved design of a process, product, or service;

  • Defining a repetitive work process;

  • Planning data collection and analysis in order to verify and prioritize problems or root causes;

  • Implementing any change;

  • Working toward continuous improvement.

In its currently used version, the PDCA cycle presents steps for the execution of a process, promoting continuous and incremental improvements, as a managerial decision-making tool, promoting the standardization of processes [11]. As the cycle repeats itself, the process is confirmed or adjusted, generating improvements and learning, involving the stages of: Planning (Plan), in which strategies and objectives are defined. Paths to be followed, the re-sources to be used, the attribution of responsibilities, and the definition of objectives in a measurable way; Execution (Do), in which the implementation of the planning occurs, promoting the implementation of the strategy; Control (Check), to study and examine the results, check if the objectives were met, monitor to identify if there were deviations from what was planned; Act, in which the strategy is confirmed or re-thought, lessons about the results of the process are identified, and the standardization of results is carried out, in the search for continuous improvement [12].

The use of the PDCA in the Governance process applied to RPA, Fig. 1, was carried out from the definition of the actions to be carried out in each of the stages of the PDCA, as shown below:

  • Planning:

    • Identify the Objectives of the Business Area

    • Define Your Company's RPA Goals

  • Execution:

    • Definition of Necessary Actions (Internal or External to Your Organization)

  • Control:

    • Definition of Measuring mechanisms of the performance achieved, comparing it with the objectives defined in the planning.

  • Action:

    • Analyze cycle results to complete the process or restart and analyze failures.

Fig. 1.
figure 1

Steps to Implement a Governance Process in RPA.

Full size image

The management process of IT processes is constantly evolving, so the development of management/governance methods must be adapted to the specificities of technology, thus ensuring an improvement in the quality of the projects developed. With this idea as a reference, this work seeks to answer a key question:

  • How is it possible to guarantee the quality of implementation and control of Robotic Process Automation (RPA) projects?

The importance of using management methodologies is directly related to the results achieved by the organization. Regardless of the management model used, planning and monitoring the strategies adopted is the key to achieving the expected results.

The objective of this work is to develop a structure for the implementation and control of Robotic Process Automation projects.

2 Methodology

The methodology for the present work is based on the analysis of a set of data sources considered very important. Through the set of contributions analyzed throughout this work, with investigations of reference authors who investigate this theme or part of it. The set of articles and investigations that were verified and analyzed here were obtained through the database of the online library “B-on”. This platform was selected because it allows reaching the full content of a wide range of scientific publications in relevant and indexed journals, together with publications in international scientific conferences, also indexed in the ISI WOS and/or Scopus systems. “B-on” is one of the most extensive databases, which includes thousands of peer-reviewed journals in a wide range of fields from different scientific fields. Through the online scientific library “B-on”, of the Portuguese Foundation for Science and Technologies, researchers can access the best-known international scientific databases, so this library was used to carry out the research process underlying this work, based on the following three groups (Group 1, Group 2 and Group 3) shown in Table 1.

Table 1. Groups of searched through “B-on”.
Full size table

Four research tests were carried out through the “B-on” by using the three groups and the OR operator as a connector between the Title or the Keywords (KW) of the intended sets. In Table 2 are expressed the number of articles found in each research test.

Table 2. Research tests performed through the “B-on”.
Full size table

After the applied filters (Fig. 2), a reading of the title, the key terms and the resume of each of the articles was carried out to verify which articles were directly related to the research. From the carried-out research, 1729 papers were obtained, applied the filters we verified a total of 948 articles and of which only 18 were framed with the theme.

Fig. 2.
figure 2

Flow diagram of literature search and respective screening adapted from [9].

Full size image

Next, throughout the research process, a set of filters were applied, based on the sets of publications obtained, and the results obtained, in terms of number of publications, are summarized in Table 3.

Table 3. Publications obtained through the B-on, after the application of some filters.
Full size table

The following Sect. 3 the analysis and synthesis of the articles. Here, data about the articles we consider relevant to the subject of this work are presented.

3 Articles Analysis

The following Table 4 presents an analysis of the 18 articles identified related to the subject under study and the phases of the PDCA governance life cycle. We can see the table below.

Table 4. Studies carried out in RPA and the implementation phases of the governance model based on PDCA.
Full size table

3.1 Synthesis Results

After analyzing the previous table, you can verify the following observations:

  • The phases that are most addressed by the investigations found are, respectively, with 61%, at 3 - Define the necessary actions (internal or external to your organization); and 4 - Measure the performance achieved, comparing it with the objectives and applying corrective actions.

  • None of the identified works addresses phase 5 - Summarize the cycle to complete the process in other activities in your IT sectoring their investigation.

  • There is no work that addresses all phases of the PDCA governance life cycle, that is, there is a possibility here for the creation of this work, that is, a model proposal that covers all these phases of the PDCA life cycle governance.

  • The works with the reference [26, 27] were the ones that addressed more phases of the PDCA cycle in their investigations.

  • The works with the reference [13,14,15, 20, 24, 25] were the ones that addressed fewer phases of the PDCA cycle in their investigations.

4 Implementation and Control RPA Projects: Framework Proposal

In this section, the proposed framework will be presented.

Through this proposal for a Robotic Process Automation management model, an organization can implement its exact functions and have the human resources indicated, knowing exactly what each of the functions must perform in its day-to-day work. Determine which are the process indicators and monitor the development of each project in an optimized way.

After identifying the conclusions of the analysis table of the identified works, we move on to the presentation of the proposal for the Robotic Process Automation framework.

4.1 Identify the Objectives of the Business Area

In the first stage, the objective of the RPA area was identified, as observed in the literature review [22, 27] and the definition adopted for this work.

  • Perform routine activities, normally performed by humans, in an automatic, simple and flexible way, making organizations more effective in business processes.

4.2 Define Your Company's RPA Goals

In the second stage, the main goals that guarantee the achievement of the pro-posed objective were defined, according to the works [16, 19, 21, 22, 27, 28].

  • Increase in service productivity;

  • Processing improvements;

  • Reduce service costs;

  • Operational efficiency gains;

  • Greater service profitability.

4.3 Define the Necessary Actions (Internal or External to Your Organization)

In the stage, a definition was created for the organization of the tasks carried out from the analysis of the RPA life cycle and from there different levels of implementation and organization of work were defined.

4.4 Organizational Structure of the Teams

After the articles, the need to create different levels of complexity of the governance process in the area of RPA analysis was defined. Because according to the number of processes, the structure needs a greater organization and specialization of the team in each of the operational and management processes. This was based on the literature and on-site organization of RPA processes in companies using this technology.

Firstly, we identified the various phases of the Robotic Process Automation life cycle (Table 5).

Table 5. RPA lifecycle stages.
Full size table

Each of these phases presented has specific characteristics, which are described below:

  1. 1.

    Analysis – here the main objective is to identify new project opportunities and carry out an analysis of the same project.

  2. 2.

    Requirements gathering – here the main objective is to carry out all the requirements gathering (access/inputs/outputs/details) associated with the project.

  3. 3.

    Design – Project development – here the main objective is to carry out the final design of the solution and the development of the project.

  4. 4.

    Testing phase – here the main objective is, after the end of development, to start testing the project.

  5. 5.

    Deployment & Hyper care – here the main objective is the deployment of the project in production and its follow-up, and final approval of the project.

  6. 6.

    Go-live and Sustentation – here the main objective is to get the project into support, that is, its monitoring, and the accomplishment of some necessary evolution to the project.

After identifying the various phases of the RPA lifecycle, it was proposed, for the implementation of RPA, three levels of Robotic Process Automation state in an organization (Level 1; Level 2; Level 3). Level 1 is the basic level, that is, the moment when an organization is in an initial state of implementation of Robotic Process Automation technology. Level 2 is the intermediate level, that is, the moment when an organization has left Level 1 and is in an intermediate state, with some workload, where there is a need for more functions for the Robotic Process Automaton. Finally, Level 3 is the advanced state, that is, the moment when an organization has left Level 2 and is in an advanced state, with a lot of work, where it has the need to create sub-stations. Teams within the RPA team to do specific tasks.

To this end, specific jobs were identified for each of the team levels, a demonstrated in Tables 6, 7 and, 8.

Table 6. Level 1 functions.
Full size table
Table 7. Level 2 functions.
Full size table
Table 8. Level 3 functions.
Full size table

After identifying the jobs for each of the different levels, we present a set of specific tasks associated with each of the phases of the Robotic Process Automation life cycle, and we classify them for each of the different levels (Level 1; Level 2; Level 3) who are responsible for each of the functions identified for each phase of the life cycle.

Table 9. Accountability for RPA lifecycle tasks.
Full size table

The Table 9 summarizes the organizational structure considering the RPA lifecycle and the roles identified in each of the phases of the cycle. In addition, 3 levels of RPA implementation are presented where the roles for the various identified positions were distributed. Regarding the structures worked on, the RPA team was considered as an internal structure and the client's integration/responsibility as an external structure.

4.5 Governance Frameworks

RPA acts at the tactical and operational level within an organization, for the implementation of efficient indicators it is necessary to develop medium and short-term goals. In order to make clear to the whole team the objectives to be achieved. Thus, one must question the objectives to be achieved and the results that should have been generated as governance in the RPA area is being implemented.

By setting clear goals, it becomes simpler to identify the best KPIs (Key Performance Indicator) for your RPA governance. Due to its form, we present here a set of methodologies that will help each one of the organizations to identify the most suitable KPIs for them. Knowing the frameworks (work models) responsible for providing the metrics and guiding the path to be followed is essential to ensure the effectiveness of the implemented practice.

The main enabling frameworks you have implementing RPA governance are:

  • COBIT (Control Objectives for Information and related Technology) = Work model most used when implementing IT governance.

This framework presents resources that include objective controls, audit maps, executive summary, goal and performance indicators and a guide with management techniques. The management practices of this framework are used to test and guarantee the quality of the IT services provided and it uses its own metrics system.

  • ITIL (Information Technology Infrastructure Library) - defines the set of practices for managing IT services through “libraries” that are part of each management module.

This is a customer-oriented framework and unlike Cobit it is a more focused model for the IT services themselves.

  • PmBOK (Project Management Body of Knowledge) - Focuses on the management of projects in the area, in order to improve the development and performance of information technology professionals.

Therefore, all definitions, sets of actions and processes of PmBOK are described in its manual, which exposes the skills, tools and techniques needed to manage a project.

5 Conclusion

The framework proposal for the implementation and control of RPA projects, which is presented here, is a very important topic because the value resulting from the management of RPA technology projects can compromise the flow of operation of a business area.

This work analyzed the works available in the literature and identified some gaps that served to propose complementary guidelines to the structural framework proposed in this work. The indicated guidelines covered the phases of the PDCA governance cycle, which served as the basis for the design of the model.

Considering the results of the work, the presented structure was developed from the definition of the RPA life cycle. Then, the various functions associated with each of the stages of the RPA life cycle were identified, and the external and internal structure of the organization chart was presented, by RPA implementation levels, given the complexity of this technology. Finally, a proposal of methodologies that help in the creation of RPA KPI's was also presented.

As a suggestion for future work, the implementation and validation of this structure is verified, as well as the elaboration of a research work associated with the identification of KPI's linked to RPA.