Sustainability in Robotic Process Automation: Proposing a Universal

Implementation Model

Christian Daase

, Anuraag Pandey, Daniel Staegemann

and Klaus Turowski

Institute of Technical and Business Information Systems, Otto-von-Guericke University, Magdeburg, Germany

Keywords: Robotic Process Automation, Sustainability, Business Automation, Design Science Research.

Abstract: Robotic process automation (RPA) is a key technology for automating mundane, repetitive back-office tasks

that are typically performed by human workers. Because RPA instantiations, known as software robots,

operate partially with the same graphical user interfaces as humans and can only replicate the business

processes for which they were previously designed, they can lack sustainability as they stop working when

sudden changes occur. This paper argues that RPA endeavors should be planned as long-term journeys

through the era of digital transformation. Based on a systematic literature review and interviews with experts

from industries that have successfully implemented software robots, this study summarizes and proposes a

universal model for sustainable RPA implementation. The model consists of three phases, from planning to

development to maintenance and scaling of projects. Although thorough evaluation is required through careful

application of the proposed workflows, a useful addition to the body of knowledge on RPA could be created

as all design decisions were made with the approval of industry experts.

1 INTRODUCTION

In the modern business world, automating various

processes, both physical and digital, is not a new

phenomenon. However, the question of what should

be automated at all and what should better remain in

qualified human hands is still fundamental in business

and information systems engineering (BISE) (van der

Aalst et al., 2018). While in manufacturing, the

outsourcing of entire production lines to robots is

already realized (Scheer, 2019), the office work in the

background is still a field with unleveraged potential

for the automation of repetitive digital tasks

(Siderska, 2020). Robotic process automation (RPA),

taking its beginnings around the year 2015 (Wewerka

and Reichert, 2023), describes the adoption of

software robots that can interact with computer

systems’ user interfaces in a manner like how a

human would and that can imitate the learned

behavior (Daase et al., 2020; Scheer, 2019). Gartner,

a major market research and consulting company,

defines RPA as tools that “perform ‘if, then, else’

statements on structured data” and that map “a

https://orcid.org/0000-0003-4662-7055

https://orcid.org/0000-0001-9957-1003

https://orcid.org/0000-0002-4388-8914

process in the RPA tool language for the software

‘robot’ to follow” (Tornbohm and Dunie, 2017).

Software robots are usually divided into two

types: attended and unattended ones. The basic

difference between both is the role of human workers

during task execution. While attended robots are

actively started, monitored, and interacted with by

responsible employees, unattended robots run

independently in the background on a server and are

triggered either by external events or according to a

schedule (Langmann and Turi, 2022). As a potential

combination, hybrid robots can take on characteristics

of both to enable end-to-end automation for processes

that involve human workers as well as back-end

functionalities (Javed et al., 2021).

As intuitive as the approach to develop software

robots based on the imitation of human actions on

graphical user interfaces may seem, issues arise when

underlying process flows undergo changes and robots

lack the ability to adapt dynamically. Given this

challenge and the fact that RPA has proved its use in

the application in a multitude of fields, such as

banking, insurance, healthcare, telecommunications,

770

Daase, C., Pandey, A., Staegemann, D. and Turowski, K.

Sustainability in Robotic Process Automation: Proposing a Universal Implementation Model.

DOI: 10.5220/0012260200003543

In Proceedings of the 20th International Conference on Informatics in Control, Automation and Robotics (ICINCO 2023) - Volume 1, pages 770-779

ISBN: 978-989-758-670-5; ISSN: 2184-2809

and logistics (Ivančić et al., 2019; Siderska, 2020), it

is critical to find long-term solutions for deploying

RPA. In context of this paper, making RPA solutions

robust and suitable for long-term usage is to be

understood as making them sustainable as further

explained in the second section. This paper

contributes to the understanding and improvement of

RPA projects in various scenarios by highlighting

factors that affect the efficiency directly and

indirectly, by collecting data from experts involved in

the field of industry automation, and by proposing a

model for the sustainable implementation of RPA.

To achieve these goal, empirical research is

conducted by combining a systematic literature

review (SLR) on perspectives from scientific

publications and interviews with experts from the

industry. The research is formalized with three

research questions (RQs). First, the literature on the

application of RPA is examined to identify factors

that determine when an RPA solution is in an

appropriate, meaningful state that is worth to be made

sustainable. Second, organizational factors that

facilitate the successful implementation of software

robots need to be identified to assemble the final

model for sustainable RPA integration. Third, precise

strategic approaches for the design and

implementation process of RPA solutions that target

sustainable implementations have to be analyzed and

synthesized into a unified model.

RQ1: How do RPA adopters assess performance

after the implementation of RPA projects?

RQ2: What are predominant organizational factors

responsible for the success of RPA projects

and how can they be measured?

RQ3: How could a strategic approach for sustainable

design and implementation of software robots

look like, based on currently followed

approaches in the industry?

Subsequently to this introductory section, the term

sustainability in terms of RPA is defined as used in

this paper. In the third section, the adopted

methodology for the construction of a model for

sustainable RPA implementation is presented. In the

fourth and fifth section, respectively, are the results

of the SLR and of the conducted interviews explained

in detail. Section six is focused on the final

composition of the previously outlined model.

Section seven concludes this work with a summary of

the most important findings from the research.

Furthermore, limitations and potential future

directions are outlined.

2 SUSTAINABILITY IN RPA

Before factors and approaches for the sustainable

implementation of software robots can be evaluated,

the term sustainability in this regard needs to be

defined. RPA is primarily designed to automate

structured, repetitive tasks that would otherwise

require manual labor, for example by using virtual

keyboards and user interfaces as a human would do

(Scheer, 2019). By introducing RPA, employees can

be freed to perform tasks that are more value-adding

for the respective organization. This paper argues that

the use of RPA should be long-term, meaning that

even after the customer has received the desired

product, the company should continue to provide

services such as maintenance and post-deployment

support.

Once companies try to scale their automation

endeavors, the number of involved software and

applications increases, which can lead to a situation

where the company suffers a significant loss if

multiple robots fail due to a shared yet

malfunctioning system. Such a failure may seem

inevitable from time to time, since software robots are

usually built on top of existing applications. For

example, the robot will fail as soon as the user

interface of any involved software changes.

Therefore, striving for sustainability with respect to

this research means developing a model that enables

a workflow that minimizes such failures and enables

long-term use of software robots by anticipating

potential problems and providing adequate

maintenance.

3 METHODOLOGY

This work adopts the guidelines on design science

research (DSR) by Hevner et al. (2004) which is a

branch of research in information systems intended to

solve identified business problems by developing a

so-called artifact as a working solution. In this case,

the outcome of this effort is a model for sustainable

RPA integration which is informed by systematic

theoretical groundwork. The organizational problem

to be solved is bridging the gap between the targeted

use of software robots at a given point in time and the

long-term benefits of the same. The six individual

research steps are derived from the DSR methodology

proposed by Peffers et al. (2007). First, the problem

is identified and motivated. This is conducted within

the introduction by outlining the benefits of RPA on

organizational performance and by stating the

Sustainability in Robotic Process Automation: Proposing a Universal Implementation Model

771

challenges regarding technical sustainability of the

robots. Second, objectives of a solution shall be

defined. Since the outcome is a model for a

recommendable development and maintenance

process, the objectives are derived, on the one hand,

from case studies and similar in scientific literature

and from the experience of experts in the field. Third,

the design and development follow. This is conducted

by reasoning and connection the acquired knowledge

from the theoretical groundwork. Up to this point, the

steps carried out form the theoretical basis, while the

remaining three steps represent the practical work that

builds on this (Daase et al., 2022). Forth, the artifact

should be demonstrated in a suitable context. Since

the model encompasses all stages of RPA integration

in a universally applicable form, this demonstration

does only take place through explaining how its

phases would guide the integration process in theory.

Fifth, an evaluation of the artifact is recommended to

gain knowledge about it which could, in turn, be used

to iterate back to the design for a refinement. As the

complexity and scope of the model require a broad

evaluation in various contexts and companies, this

step is postponed to future work. The sixth step, the

communication of the work, is performed by

providing the insights of this research to the scientific

community. Figure 1 summarizes the overall

methodology.

Figure 1: Design science research process.

3.1 Systematic Literature Review

An SLR is a method for assessing and comprehending

all existing research pertinent to a certain research

question, topic, or phenomenon of interest. SLR

assists in recognizing any gaps in the existing

research and recommending topics for additional

study (Kitchenham and Charters, 2007). According to

Hevner et al. (2004), rigor comes from efficiently

using that body of current knowledge for the

development of the artifact. The search was

conducted in the four databases ACM Digital Library,

ScienceDirect, Scopus, and SpringerLink. Based on

the specified research objectives, pilot searches, and

iterative refinements, the search phrase was

constructed. To keep a broad enough knowledge base

as well as a manageable number of articles, the

following query was found to be sufficient:

Q: “robotic process automation” AND

(“sustainable RPA” OR “RPA governance

model” OR “intelligent process automation”

OR “RPA return on investment” OR “RPA

ROI” OR “cognitive robotic process

automation” OR “enterprise automation”)

For ACM Digital Library, Scopus, and

ScienceDirect, the query was searched for in

abstracts, titles and keywords. For SpringerLink, the

individual components were entered into the general

search field and the search was limited to conference

proceedings or journal articles. The time frame was

limited to articles published between 2010 and 2022.

A total of 313 articles were retrieved as a result.

The search then was further refined by applying

inclusion and exclusion criteria during a phase of

reading the titles and abstracts. Articles were rejected

if they were identified as duplicates, either exactly or

semantically, and when they consisted only of an

abstract, a patent, or of an introduction to

proceedings. On the other hand, the articles were

included in the further investigation only if they

covered the topics of software robot development and

implementation as the main topic.

After this phase, 89 articles remained in the

literature pool. In a subsequent phase, when reading

the contents of the publications, articles were

excluded in case potential challenges of RPA

integration were not pointed out, thus omitting a

critical discussion. To pass this phase, articles were

expected to consider key performance indicators for

RPA adoption and to cover impacts on the adopting

organization as well. A total of 23 papers remained

after this detailed examination. The selection criteria

are summarized in Table 1.

Table 1: Inclusion and exclusion criteria for the SLR.

Inclusion criteria Exclusion criteria

Software robots and their

active implementation are

the main topics of the

publication.

Duplicate

The publication considers

key performance

indicators for RPA

adoption.

The article is a patent,

abstract only, or

proceeding’s introduction.

The paper covers impacts

on the organization trying

to utilize software robots.

Critical discussions are

omitted by leaving out

considerations of potential

challenges.

ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics

772

Figure 2 depicts the workflow of the SLR, including

the numbers of remaining articles after each phase.

Figure 2: Visualized SLR workflow.

3.2 Interviews

The intention for conducting interviews was to gain

knowledge about RPA projects carried out in the

industries from the people who were actually

involved. Thus, insights from both theory (academic

literature) and practice (interviews) are integrated in

the research. A group of three experts from different

countries and on different levels of experience in the

field of RPA was composed and as a result, different

viewpoints on how to achieve a sustainable

deployment of RPA could be gained. The details on

the three interviewees are summarized in Table 2.

Table 2: Professional experience of the interviewees.

Interviewee Role Country

Int1 Solution Architect /

Operations Manager

Germany

Int2 RPA Consultant /

UiPath chapter lead

France

Int3 Senior RPA Developer /

Intelligent Automation

Consultant

India

The experts were asked whether they wanted to add

additional steps to be included in an RPA roadmap or

whether they wanted to omit steps. Apart from the

identifiers of the company and the interviewees for

confidentiality reasons, all information from the

interview was transcribed.

4 SLR RESULTS

A part of the IT users has long expressed concerns

that the extensive usage of robotics may lead directly

to an increasing unemployment rate (Frey and

Osborne, 2017). On the other side of the spectrum,

robots have been portrayed as rescuing humanity

from mundane and laborious occupations, allowing

people to concentrate on more valuable work and

more fruitful intellectual pursuits (Lamberton et al.,

2017).

In the literature, special attention is paid to the

digitization of operational and business processes in

businesses of service firms, mainly those in the

financial, banking, insurance, marketing, accounting,

public administration, and logistics sectors

(Madakam et al., 2019; Mendling et al., 2018;

Siderska, 2020; van der Aalst et al., 2018). However,

there is no agreement on a fixed definition of RPA in

the literature identifiable. From the selected

publications, the lowest common denominator among

the authors is that RPA is a recent strategy that uses

robots to automate monotonous digital work. These

bots, for instance, may analyze emails, perform

calculations, open and move files, log into

applications, connect to APIs, create invoices, query

databases, obtain web data, create content, and extract

data from messages (Madakam et al., 2019; Mendling

et al., 2018; Santos et al., 2020).

Software robots have various advantages for

businesses, including enhanced productivity, data

security, shorter cycle times, and better accuracy,

while freeing up staff (Leshob et al., 2018). When

compared to traditional process automation, RPA

promises to be simpler to adopt, relatively

inexpensive, and able to scale, audit, and improve

security and compliance (Fung, 2014; Hallikainen et

al., 2018; van der Aalst et al., 2018). Syed et al.

(2020) give a detailed overview of the basics of RPA,

its benefits, and the related challenges. As per this

study, RPA literature is mostly dominated by position

and white papers, case studies, and experiences

directed at higher-level management.

In their research in the automotive industry,

Wewerka and Reichert (2021) identified bottlenecks

that are related to RPA implementation. The authors

summarize the challenges on five levels which can be

used to guide industries trying to start their RPA

journey: 1) determining the business operation for

automation, 2) comprehending the elements

impacting acceptance and usage, 3) conveying RPA

as a concept to the end-users, 4) designing the bot to

ensure interaction between the bot and its user, 5)

establishing governance and best practices for the bot

Sustainability in Robotic Process Automation: Proposing a Universal Implementation Model

773

development. Similarly, Asatiani and Penttinen

(2016) focus on issues that the OpusCapita Group had

when deploying RPA and the various business

models it encompasses. The study states that

answering certain questions before the actual

implementation is crucial. These questions include

which models should be chosen for the technology,

how to ensure RPA leadership in the long run and

what value the organization can provide to customers

with it. Money (2021) suggests that enterprises

employing or adopting RPA technology should pay

close attention to the data controls, management, and

security elements. The recommendations for special

attention include identity and access management,

data encryption (including credentials), maintaining

policies for data classification, data retention, data

storage, and data location, monitoring of logs and

regular auditing, and vulnerability scanning of all bot

programs prior to promotion into the production

environment.

According to Wewerka et al. (2020), any RPA

initiative will not be successful if it is not accepted by

the users. The authors created a model for evaluating

RPA user acceptance and the factors that affect it. The

findings support the notion that perceived usefulness

is positively influenced by social influence, job

relevance, and result demonstrability. Positive

impacting factors of perceived ease of use include

trust, innovation, delight, and enabling

circumstances. As a conclusion of this paper, factors

that influence user acceptance make the RPA

implementation sustainable. A common topic of

discussion in white papers has questioned the impact

of RPA and to some extent concluded that the

software robots would replace the existing workforce,

such as accountants (Sarilo-Kankaanranta and Frank,

2022). Contrary to this, organizations that have

embarked on the RPA journey still recruit

accountants for their services. Another group of

researchers from New Zealand focused on service-

oriented workforces as their quintessence (Brougham

et al., 2020). This study’s respondents perceived

automation as giving them new chances, possibly

even boosting their current professions, which could

be considered a noteworthy benefit. To complement

the findings from the literature, individual expert

opinions and experiences from the interviews are

presented in the following section. In Table 3, the

findings for RQ1 on performance assessment are

summarized with corresponding questions to be

asked.

Table 3: Summary of RPA performance indicators.

Attributes Corresponding performance question

Reduced

Handling Times

How much human working hours are

saved by through software robots?

Cost Savings How much money can be saved by

deploying software robots?

Lower Error

Rate

How high is the error rate with respect to

total process executions compared to a

human’s performance on the same task?

Employee Skill

Growth

How have the employees’ qualifications

developed since the adoption of RPA?

Bot Usage How heavily are the robots utilized

compared to their maximum capacity?

Standardization To what extent has the diversity of process

execution workflows been reduced?

5 INTERVIEW FINDINGS

The participants of the interviews were questioned

about their experiences regarding the impacts of RPA

on the performance of their companies and their

thoughts and recommendations for sustainable

implementation. While the former answers primarily

RQ2, the latter forms a further basis for answering

RQ3. For comprehensibility, the identifiers from

Table 2 are used. Asked for the impact factors, Int1

responded that especially time savings and the

reduction of errors as well as the constant availability

of services, independent of human employees, are

considered as the main advantages for his business.

The monitoring of these factors was explained with

“We share weekly and monthly performance reports

on each process from different bots to understand the

success/exceptions”. In contrast to these quantifiable

measures, Int2 stated the quality of work, number of

impacted workers, and job satisfaction as a few more

qualitative factors. As the means for assessment,

“customer feedback, stakeholder meetings, and

discussions” were mentioned. The third participant,

Int3, focused his answers on time-related aspects,

stating bot utilization, potential for reusage, general

time savings and delays in execution as additional

factors to the overall quality of work. For assessment,

“project monitoring and stakeholder meetings” were

stated for measuring RPA impacts.

In summary, answering RQ2, seven umbrella

terms for organizational impacts could be agreed on:

(1) productivity and (2) quality improvements, (3)

increased cost effectiveness, (4) employee and (5)

customer/stakeholder satisfaction, (6) adaptability

and reusability of robots, and (7) improved

compliance and comprehensibility of processes. The

responses provided throughout the interview suggest

that the majority of the time, sustainability is

ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics

774

dependent on the demands of the client and any

potential post-deployment problems. When

questioned about the post-deployment support

offered by their organization, Int1 replied: “In some

cases, we have a production support team responsible

for monitoring and resolving any immediate issues

before raising them with the development team. In

other cases, we have trained the client’s team to

handle such issues. For the rest, we ourselves are

responsible for post-deployment support”. Int2

highlighted the importance of a dedicated unit for

software robots in a company: “The center of

excellence takes care of delivery. It keeps updating

documentation and upgrading versions”. Project

sustainability is heavily influenced by an

organization’s capacity to serve its customers and the

warranty period for such assistance. Int3 advised to

have separate teams for post-production support.

First, developers who are responsible for handling

technical problems related to the bots and, second,

controller team responsible for operative

maintenance (scheduling, starting, stopping, and

troubleshooting). As a description for the support

process, Int2 stated: “It is the user that notices that

the bot is malfunctioning. If the error is not a business

one, they open a ticket, and the center of excellence

goes to see the logs. If they notice a change in the user

interface, they modify the code in the validation

environment first. Then the documentation is

updated”. Therefore, it is important to note that a

sustainable implementation should be ongoing. In

addition, effective change management and employee

acceptability are essential for the long-term

implementation of RPA in an organization. To reduce

uncertainty within an organization, RPA awareness

should be raised early, emphasizing the advantages

and potential of the technology.

6 SUSTAINABLE RPA

IMPLEMENTATION MODEL

In this section, a sustainability model for successful

long-term RPA implementation is proposed,

according to the definition presented in section 2. It

provides an overview of the guidelines with regard to

the various stages of RPA implementation.

Businesses that want to build a sustainable operating

model for their automation and gain maximum

enterprise-level benefits typically triangulate aspects

such as identifying scaling factors, acquiring

resources that make the model self-sufficient, and

organizing the automation team for a successful

drive.

The insights from the expert interviews frequently

overlap with the findings from the literature when

studying the stages of a typical RPA project lifecycle.

The model presented here divides RPA projects into

three phases and incorporates many factors that have

been gathered through the theoretical groundwork.

The first phase is based on the project life cycle's

planning and conceptualization phase, which may

alternatively be considered as initialization. In order

to simplify the mapping process, the solution design

and user acceptance phases have been integrated into

phase two. Furthermore, this involves designing,

developing, and testing the software bot. The RPA

phases known as deployment, maintenance, and

scaling are consolidated in a single step termed as

maintenance and scaling. The overall model is

depicted in Figure 3 and subsequently explained in

more detail in the following sections.

Figure 3: Model for sustainable RPA implementation.

Sustainability in Robotic Process Automation: Proposing a Universal Implementation Model

775

6.1 Phase I - Planning and

Conceptualization

The fate of an RPA project is determined by a number

of factors and challenges. Preparation in the form of

skills and awareness before beginning the RPA

journey might be crucial to the project's long-term

sustainability. This is made possible by doing a

thorough study of processes to determine their degree

of automation, identifying problems encountered

throughout the various phases of the project,

analyzing variables to overcome these challenges,

and implementing business process reengineering if

necessary.

Employee Skill Enhancement Program. The

unavailability of trained and skilled employees for

development, support, upgrade, and post deployment

maintenance of RPA technologies pose a potential

bottleneck for management trying to invest into a new

technology. Delays in deployment could result from

a lack of training resources for creating, deploying,

supporting, and upgrading software robots. Employee

training, along with other client services like

customer support and maintenance services, should

be prioritized while using RPA (Ivančić et al., 2019).

The training could consist of several intensities,

beginning with basic instructions in the use of RPA

software to constantly learning and improving

technical skills.

Rationalize Use Cases. The main agenda behind

using RPA is to standardize processes, reduce the

overall costs and time taken to execute a particular

process. Automation enables companies to reduce

employee costs and instead use a robot that performs

the same function much cheaper with lower execution

time. Hence, it is critical to identify and rationalize

business use cases that can be streamlined and then

automated. It is essential to understand how to

compare the current costs of running a process by

humans and the costs of running an RPA program,

which involves different types of expenses such as

licenses, installation, maintenance, support, and

training. The comparison between the two variants of

the process, one with an employee and the other with

automated bots, should be positive. If not, the

organization will suffer losses. In order to use RPA

sustainably to improve business processes with a

positive return on investment, a use case should be

identified as ideal for automation first. For RPA to be

successfully implemented and widely adopted, a

central hub as a center of excellence (CoE) should

exist and made responsible for all RPA-related

endeavors. The CoE should have deep knowledge of

all aspects of the business and be able to assist with

internal expansion plans.

Cost Analysis. RPA vendors guarantee drastic

lowering of costs while raising the quality of work.

Before beginning the automation process,

stakeholders must assess the costs involved, not just

in terms of spendings but also returns. Also, a major

benefit to be considered which augments users’ return

on investment is the repositioning of internal

employees.

Governance Model Creation. IT governance is a

structure for an organization that ensures the

alignment between IT strategy and business strategy.

The governance team is in charge of a specific

number of responsibilities, including but not limited

to the control of risks, the protection of data, and the

recovery of the system after it has faltered. When

integrated with other technologies, RPA creates a

complex structure that necessitates the establishment

of a reliable governance system (Willcocks et al.,

2018). In the conducted interviews, the respondents

included factors such as IT governance model, a well

thought out internal audit and compliance strategy,

error handling, documentation and discussions with

process subject matter experts to really understand

processes. According to Int2, in their respective firm,

the amount of time needed for software robot

deployment was drastically cut down due to the

standardization and centralization of development as

more and more procedures were automated.

Organizations can gain a firm grip on a project by first

laying a solid groundwork of careful planning and

then creating a set of well-thought-out governance

models. This eventually leads to a stable ecosystem

for the organization to grow.

Failover Strategy. Business continuity and system

failure hampers business as usual. It is critical that

executives at the highest levels of the organization

provide support to the IT team, ensuring better

failover and system reliability. IT teams are

completely responsible to identify any potential

impact for systems, storage and backups, as well as

providing the necessary access rights for certain jobs.

An emergency staff member should always be

available in case a malfunctioning robot requires

human intervention.

Organizational Change Observation. Organization

structure might have to be reorganized to include

RPA into the plans (Smeets et al., 2021). All

ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics

776

organizational changes due to automation should be

openly discussed in meetings and made transparent.

Changes in the work process are common and must

be dealt with by all departments in order to compete

at the greatest level.

6.2 Phase II - Design, Development and

Testing

Project managers, developers, solution architects, and

business analysts all collaborate to work efficiently

throughout this phase of an RPA project. Planning

both short-term and long-term objectives facilitates

the implementation of comprehensive customer

solutions. During this phase, the RPA team and the

partner vendor define the appropriate process as well

as the technical and business stack needed to

successfully implement the software robots.

Streamline Processes. The analysis shows that RPA

implementations with high success ratings have much

greater levels of pre-work done in recording and

documenting processes. Another essential aspect is

process preparation with subject matter experts. After

approval from the stakeholders and management, the

RPA project advances to its planning phase. For large

RPA projects, the project should be broken down into

sub parts, as this structure allows the development

team to assign tasks objectively. Here, the team, along

with the project manager or solution architect and

other assigned members, would establish the project’s

scope, including strategic goals, budgeting,

milestones, functions, and timeframes. Processes

should be streamlined depending on the complexity

of the tasks at hand. According to the interviewees,

the success of RPA implementations is dependent on

their ability to be monitored and secured under

governance.

Ensure RPA Security. Data leakage and fraud are

two of the major potential risks for automatically

operating software robots. During the initial process

assessment and analysis phase, a business analyst and

the IT team should conduct workshops to identify

processes and perform risk assessments. Factors of

RPA security, pre-implementation risk assessment,

risk analysis, hazard analysis, and threat analysis can

help to overcome the challenges that a typical RPA

implementation may encounter (van der Aalst et al.,

2018).

Prioritize Processes. Not all processes are suitable

for this type of automation because, as previously

stated, robots used RPA are designed to do repetitive

tasks (Fung, 2014). A pipeline of tasks which satisfy

the criteria of being repetitive, high volume, based on

mostly structured data, and with only few changes

expected in the future is ideally selected for

automation and then arranged based on the level of

difficulty and the team’s bandwidth in terms of

available time resources.

Pilot Rollouts. Based on the findings from the

interviews, it is common for businesses to want to

skip parts of the preparatory work required for RPA

implementations. Often, the expectation is for speedy

and low-cost installation, which results in the system

not performing to its full capacity. This has frequently

resulted in pilot program failures, and it may also

make scaling the company’s RPA journey difficult.

Therefore, pilot rollouts for RPA implementations are

critical for companies. Proof of concept is often used

to refer to preliminary RPA implementations that are

still in the pilot stage. Such implementations can act

as confirmation of the practicality and viability of

RPA technology for the specific application

(Willcocks et al., 2018).

Error Monitoring. Lack of ownership and poorly

defined responsibilities along with inconsistency of

data in different environments all lead to poor

execution or create pitfalls for the development team.

Following predefined security considerations within

the IT team can help ensuring a smooth execution of

development and deployment approaches.

6.3 Phase III - Maintenance and

Scaling

Infrastructure management, a roadmap for

incorporating new technologies for sustainable

progress, risk management, and the establishment of

a center of excellence are all necessary for the long-

term viability of the RPA concept.

Roadmap Maintenance. One of the main inferences

from the interviews was that the use of RPA is

intended to be enhanced with newer technologies

such as optical character recognition, machine

learning, natural language processing, and so on.

Since it is expected to ensure stable long-term usage,

planning in this regard in terms of thinking and

organizing the firm internally to have such

capabilities to improve scaling efforts is vital. This

adds a level of intelligence into the bots which allows

a broader usage of RPA and consequently expands its

applicability and sustainability. Strategic planning by

Sustainability in Robotic Process Automation: Proposing a Universal Implementation Model

777

the management can ensure a smooth transition into

such maturity within the organization.

Risk Management. Allocating IT resources and a

well-structured change management can ensure a

smooth integration of other technologies along with

RPA. In IT, one of the most prevalent issues is a

security breach, but other problems like unintentional

system failures may also occur. One of the most

important things to do at this stage is to create a

fallback option that takes into account all of these

possibilities.

Establishing a Centre of Excellence. When it comes

to implementing and adopting RPA, a well-

established center of excellence serves as a hub. It is

accountable for providing consolidated knowledge

across all company functions and assisting with

internal scalability and it orchestrates the constant

new adding of features, while supervising bug fixing

in older versions. It can also aid in scaling inside the

company and have specialized knowledge in various

business operations. The CoE should further

incorporate stakeholder interaction into its workflow.

Change management and a reliable governance

model are both helpful in this regard.

The overall model is extended by a continuous

improvement cycle, using the proposed components

and then making changes and additions through

introspection to pave the way to a sustainable RPA

realization.

7 CONCLUSION

The proposed model for sustainable RPA is designed

to act as a framework for firms aspiring to implement

RPA. The goal of this paper was to identify factors

that lead to the success of RPA projects as well as

factors that may influence their failure. To this end,

an SLR and interviews were conducted to determine

critical factors that cloud ultimately be used to

compose a model for the sustainable development and

deployment of software robots. Some of the top

factors identified are sufficient pre-work and analysis

in identifying appropriate processes for automation,

establishing an internal center of excellence, and

creating awareness amongst both employees and

customers. The factors uncovered throughout the

research were then amalgamated and classified into

different phases of a project lifecycle. The

empirically derived aspects can be used as guidelines

for organizations that are starting their journey and

also for firms that have successfully scaled RPA.

These criteria can significantly increase the

likelihood of sustainable RPA, which is especially

important given the number of times companies have

reported failing to scale up RPA following their pilot

project. For the purpose of evaluating the model as

stated in the methodology section, one could involve

conducting surveys and soliciting comments on the

methodological framework proposed. Furthermore,

research can be undertaken in the form of case studies

in various sectors and organization sizes, where the

approach is put to the test while engaging all key

stakeholders. When conducting case studies, it would

also be valuable to see how goals and aspirations

change over the course of an RPA journey.

Awareness of the notion might lead to a shift in the

prevailing attitude, which in turn can facilitate

quicker scaling, but it is also important to keep in

mind that contextual factors may play a crucial role.

The list of essential success factors and requirements

outlined in this paper may be expanded in light of the

arrival of a new generation of intelligent automation

technologies. Therefore, it is important to investigate

the necessary modifications to this project

management approach to make it useful for artificial

intelligence and machine learning.

Pointing out the important challenges identified in

the study also highlights that a significant percentage

of RPA initiatives do not move from the pilot phase

to the advanced phase. Case studies can be conducted

to better understand stakeholder ambitions and

customer onboarding. Likewise, all the critical factors

mentioned in the article can be put to the test before

embarking on the digital transformation journey.

Other effects that automation may provide are

also worth investigating. It is important to evaluate

the impacts of freeing up staff members to focus on

more innovative, high-value work. Sustainable RPA

is relatively new and requires more research and

development in order to make it more stable and

economically feasible. The model may be thoroughly

evaluated from a financial point of view, with a study

able to look at the total investment across the entire

project and compare it to the current state of the art.

REFERENCES

Asatiani, A., & Penttinen, E. (2016). Turning robotic

process automation into commercial success – Case

OpusCapita. Journal of Information Technology

Teaching Cases, 6(2), 67–74.

Brougham, D., Haar, J., & Tootell, B. (2020). Service

Sector Employee Insights into the Future of Work and

Technological Disruption, 44, 21–36.

ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics

778

Daase, C., Staegemann, D., Volk, M., Nahhas, A., &

Turowski, K. (2020). Automation of Customer Initiated

Back Office Processes: A Design Science Research

Approach to link Robotic Process Automation and

Chatbots. In ACIS.

Daase, C., Volk, M., Staegemann, D., & Turowski, K.

(2022). Addressing the Dichotomy of Theory and

Practice in Design Science Research Methodologies. In

17th International Conference on Design Science

Research in Information Systems and Technology

(DESRIST), St. Petersburg, FL, USA.

Frey, C. B., & Osborne, M. A. (2017). The future of

employment: How susceptible are jobs to

computerisation? Technological Forecasting and

Social Change, 114, 254–280.

Fung, H. P. (2014). Criteria, Use Cases and Effects of

Information Technology Process Automation (ITPA).

Institute for Robotic Process Automation.

Hallikainen, P., Bekkhus, R., & Pan, S. (2018). How

OpusCapita used internal RPA capabilities to offer

services to clients. MIS Quarterly Executive, 17, 41–52.

Hevner, A., March, S., Park, J., & Ram, S. (2004). Design

Science in Information Systems Research.

Management Information Systems Quarterly, 28, 75-

106.

Ivančić, L., Suša Vugec, D., & Bosilj Vukšić, V. (2019).

Robotic Process Automation: Systematic Literature

Review, Lecture Notes in Business Information

Processing. Business Process Management:

Blockchain and Central and Eastern Europe Forum

(Vol. 361, pp. 280–295). Springer International

Publishing.

Javed, A., Sundrani, A., Malik, N., & Prescott, S. M. (2021).

Robotic Process Automation: Overview, Robotic

Process Automation using UiPath StudioX (pp. 3–8).

Apress.

Kitchenham, B., & Charters, S. (2007). Guidelines for

performing Systematic Literature Reviews in Software

Engineering, 2.

Lamberton, C., Brigo, D., & Hoy, D. (2017). Impact of

Robotics, RPA and AI on the insurance industry:

challenges and opportunities. Journal of Financial

Perspectives, 4(1).

Langmann, C., & Turi, D. (2022). Basics of Robotic

Process Automation (RPA), Robotic Process

Automation (RPA) - Digitization and Automation of

Processes (pp. 5–16). Springer Fachmedien Wiesbaden.

Leshob, A., Bourgouin, A., & Renard, L. (2018). Towards

a Process Analysis Approach to Adopt Robotic Process

Automation. In 2018 IEEE 15th International

Conference on e-Business Engineering (ICEBE) (pp.

46–53). IEEE.

Madakam, S., Holmukhe, R. M., & Kumar Jaiswal, D.

(2019). The Future Digital Work Force: Robotic

Process Automation (RPA). Journal of Information

Systems and Technology Management

, 16, 1–17.

Mendling, J., Decker, G., Hull, R [Richard], Reijers, H. A.,

& Weber, I. (2018). How do Machine Learning,

Robotic Process Automation, and Blockchains Affect

the Human Factor in Business Process Management?

Communications of the Association for Information

Systems, 297–320.

Money, W. H. (2021). Resolving pressure and stress on

governance models from robotic process automation

technologies. In Proceedings of the Conference on

Information Systems Applied Research ISSN.

Peffers, K., Tuunanen, T., Rothenberger, M., & Chatterjee,

S. (2007). A design science research methodology for

information systems research. Journal of Management

Information Systems, 24, 45–77.

Santos, F., Pereira, R., & Vasconcelos, J. B. (2020). Toward

robotic process automation implementation: an end-to-

end perspective. Business Process Management

Journal, 26(2), 405–420.

Sarilo-Kankaanranta, H., & Frank, L. (2022). The

Continued Innovation-Decision Process – A Case

Study of Continued Adoption of Robotic Process

Automation, Lecture Notes in Business Information

Processing. Information Systems (Vol. 437, pp. 737–

755). Springer International Publishing.

Scheer, A.W. (2019). The Development Lines of Process

Automation, The Art of Structuring (pp. 213–220).

Springer International Publishing.

Siderska, J. (2020). Robotic Process Automation — a driver

of digital transformation? Engineering Management in

Production and Services, 12(2), 21–31.

Smeets, M., Erhard, R., & Kaußler, T. (2021). Robotic

Process Automation (RPA) in the Financial Sector.

Springer Fachmedien Wiesbaden. https://doi.org/10.10

07/978-3-658-32974-7

Syed, R., Suriadi, S., Adams, M., Bandara, W., Leemans, S.

J., Ouyang, C., Hofstede, A. H. ter, van de Weerd, I.,

Wynn, M. T., & Reijers, H. A. (2020). Robotic Process

Automation: Contemporary themes and challenges.

Computers in Industry, 115, 103162.

Tornbohm, C., & Dunie, R. (2017). Market Guide for

Robotic Process Automation Software. https://www.

gartner.com/en/documents/3835771

van der Aalst, W. M. P., Bichler, M., & Heinzl, A. (2018).

Robotic Process Automation. Business & Information

Systems Engineering, 60(4), 269–272.

Wewerka, J., Dax, S., & Reichert, M. (2020). A User

Acceptance Model for Robotic Process Automation. In

2020 IEEE 24th International Enterprise Distributed

Object Computing Conference (EDOC) (pp. 97–106).

IEEE.

Wewerka, J., & Reichert, M. (2021). Robotic Process

Automation in the Automotive Industry - Lessons

Learned from an Exploratory Case Study, Lecture

Notes in Business Information Processing. Research

Challenges in Information Science (Vol. 415, pp. 3–19).

Springer International Publishing.

Wewerka, J., & Reichert, M. (2023). Robotic process

automation - a systematic mapping study and

classification framework. Enterprise Information

Systems, 17(2), Article 1986862.

Willcocks, L., Hindle, J., & Lacity, M. (2018). Keys to RPA

success. Executive Res. Rep., Knowl. Capital Partners,

USA, Tech. Rep.

Sustainability in Robotic Process Automation: Proposing a Universal Implementation Model

779