Proceedings of the 58th Hawaii International Conference on System Sciences | 2025

Structuring Intelligent Automation: A Hyperautomation Taxonomy

Nicolas Neis Elise Kathol Axel Winkelmann

University of Würzburg University of Würzburg University of Würzburg
[email protected] [email protected] [email protected]

Abstract in the work environment have easy automation

Hyperautomation is an emerging technology potential. Implementing Robotic Process Automation
stack that combines various automation technologies (RPA) ensures the precise and consistent automation
to streamline and optimize business processes. of tasks (Mohamed et al., 2022). In addition,
Companies increasingly adopt these technologies to companies are realizing new levels of automation
reduce costs, increase efficiency, and gain a through the use of AI technologies (Damioli et al.,
competitive advantage. This study focuses on 2021).
combining the different technologies under the Hyperautomation, a prominent trend, integrates
hyperautomation approach. From a management technologies like RPA, AI, and machine learning
perspective, we examine the benefits and challenges of (ML) for comprehensive task and workflow
hyperautomation that focus on increasing company automation (Madakam et al., 2022). It aims to
productivity. Based on a literature review, we develop automate tedious and time-consuming routine tasks
a taxonomy that enables a profound understanding of (Haleem et al., 2021) and strives for continuous
the complex hyperautomation landscape. Findings automation efforts (Jiménez-Ramírez, 2021;
underscore the substantial potential of Ackermann et al., 2024). Beyond routines,
hyperautomation to improve productivity, particularly hyperautomation tackles complex processes,
through cost and time savings and enhanced employee promising increased efficiency, cost savings, and
efficiency. This research addresses the lack of improved outcomes for businesses (Falih et al., 2022).
systematization in previous studies, presenting a Given the growing interest in adopting such
nuanced understanding of hyperautomation's technologies, this article aims to provide a foundation
implications for managerial strategies. for an in-depth analysis of hyperautomation that
focuses on increasing productivity and improving
Keywords: Hyperautomation, Robotic Process organizational work processes. This is particularly
Automation, Productivity, Taxonomy. relevant, as the wrong choice of automation
technology can be detrimental to companies (Engel et
1. Introduction al., 2022; Lacity & Willcocks, 2021). Understanding
what and how to automate is a vital part of
hyperautomation (Ackermann et al., 2024). The
"Imagine simply telling Word to write a letter or
objective of our study can, therefore, be summarized
Excel to create a spreadsheet" (Marr, 2023). Microsoft
with the following research question:
envisions a future where human-computer interaction
is intuitive. Phone call summaries, personalized email RQ: How can we systematize the characteristics
responses, spreadsheet formatting, and various writing that influence productivity in hyperautomation from a
tasks will all be handled seamlessly by Microsoft managerial perspective?
Office (Marr, 2023). Soon, ChatGPT, an artificial
This study thereby addresses the gap in the
intelligence (AI) -based language generation model, is
literature on hyperautomation by providing a holistic
poised to automate meeting recordings and
perspective across the entire technology spectrum, in
translations and create to-do lists (Dahmen et al.,
contrast to previous studies that concentrate mainly on
2023). These innovations aim to relieve companies
specific automation potentials (Aguirre & Rodriguez,
from repetitive tasks, transforming how work is
2017; Denagama et al., 2020). Although the growing
accomplished.
body of research on hyperautomation is increasing,
A key requirement for this success is realizing
there remains a significant deficit in frameworks that
automation potential. Marr's article in Forbes
offer a well-defined and systematic comprehension of
Innovation (2023) notes that about 66% of daily tasks

URI: https://hdl.handle.net/10125/109526
978-0-9981331-8-8 Page 5663
(CC BY-NC-ND 4.0)
hyperautomation, particularly regarding its effects on al. (2022) add that the focus is on intelligent self-
productivity. The absence of such a framework governing automation, starting with simple tasks and
hampers the identification of opportunities, the progressing to advanced capabilities, making
addressing of deficiencies, and the full exploitation of decisions based on data, and laying the groundwork
hyperautomation's potential. This shortfall restricts for process optimization and innovation.
academic advancement and limits practical
Robotic Process Automation. RPA is considered
applications that could enhance efficiency and foster
the fastest-growing enterprise software segment in
innovation in business processes. By developing a
history (Taulli, 2020) and is a pre-version of
taxonomy focusing on productivity, we provide an
hyperautomation (Herm et al., 2021). Addressing
initial analysis and valuable insights for companies
inefficiencies linked to repetitive tasks, RPA
seeking to optimize their operations (Glass & Vessey,
streamlines routine activities, acting as a virtual
1995). With our taxonomy, we present a systematic
workforce replicating human behavior for automating
categorization of research that provides a common
manual, high-volume, and repetitive tasks without
language for domain-specific terms and identifies
requiring programming skills (Aguirre & Rodriguez,
white spots where further research is needed. The
2017; Herm et al., 2023; Syed et al., 2020; Van Der
results of the literature review and the taxonomy
Aalst et al., 2018). Notably, it functions as a non-
derived from it show that hyperautomation has
intrusive, Graphical User Interface-based tool that
significant potential to increase the productivity of
enhances efficiency across various sectors such as call
organizations in various ways.
centers, banking, and IT companies (Madakam et al.,
To address the research question, our study is
2022). Contrary to the misconception of eliminating
structured as follows: Section 2 explores the
human labor, RPA is designed to complement human
fundamentals of hyperautomation with its various
efforts, serving as a virtual assistant (Dan et al., 2022).
technologies. Furthermore, we outline the related
Its immediate benefits impact core business processes,
work. Section 3 covers the methodology, and we
improving productivity, reducing costs, enhancing
explain our literature search and the approach to
speed, and minimizing errors (Bahaweres et al., 2022;
develop the taxonomy. Section 4 presents and
Syed et al., 2020).
evaluates our taxonomy, and Section 5 discusses the
results. Lastly, we present a conclusion and outlook in Artificial Intelligence. Another enabler of
Section 6. hyperautomation is AI, which can potentially disrupt
industries globally (Damioli et al., 2021; Neis et al.,
2. Research Background 2024). AI applications aiming to mimic human
intelligence have significantly increased since the
This section provides the groundwork, explaining early 2000s (Madakam et al., 2022). ML, a subset of
the research foundations of hyperautomation and AI, can enhance decision-making by providing
productivity. Drawing from various disciplines such as predictive insights and enabling a deeper
AI, ML, and organizational theory, we uncover understanding of vast data. Through advanced
fundamental concepts that shape the understanding of algorithms and data analytics, ML systems identify
hyperautomation's potential and its implications for patterns, make accurate predictions, and reveal
productivity. valuable insights, influencing decision-making across
diverse domains such as business, finance, healthcare,
2.1 Hyperautomation and scientific research (Carvalho et al., 2019). Unlike
traditional programming, where machines are
According to Haleem et al. (2021), explicitly instructed to perform specific tasks, in ML,
hyperautomation is an approach that can be divided machines learn from experience and improve their
into six related technologies: RPA, AI/ML, Deep performance over time through data and feedback
Learning (DL), Process Mining (PM), Natural (Bishop, 2013).
Language Processing (NLP), Optical Character Business Process Management. BPM is a
Recognition (OCR), and Digital Twin of an management area that involves modeling,
Organization (DTO). Hyperautomation aims to implementing, executing, monitoring, and evolving
minimize manual tasks, enhancing business execution business processes (Reichert & Weber, 2012). It seeks
efficiency (Gartner, n.d.; Jiménez-Ramírez, 2021). to enhance processes for increased efficiency and
This approach allows automation of complex business effectiveness, employing diverse methods, techniques,
processes, integrating business intelligence systems and software solutions for analyzing, controlling, and
and improving the expertise of both human workers managing organizational activities. As such, BPM
and automated systems (Haleem et al., 2021). Zhao et

Page 5664
takes account of the fundamental reality of constant effects and benefits of RPA and how to measure these
change in the business environment (vom Brocke et effects (Meironke & Kuehnel, 2022; Royhan et al.,
al., 2021). BPM involves orchestrating people, skills, 2023). In a recent study by Royhan et al. (2023), a
applications, documents, and related data to optimize systematic literature review was conducted,
tasks (Falih et al., 2022). By leveraging BPM and examining journal articles published between 2017
automating eligible activities or variations of a and 2021, identifying internal process improvements
process, businesses can streamline operations, reduce and productivity enhancements as key indicators for
manual intervention, and improve overall efficiency successful RPA implementation (Royhan et al., 2023).
and risk management (Falih et al., 2022). Axmann and Harmoko (2020) provide a thorough
overview of the field of RPA and compare it to other
2.2 Productivity technologies in Industry 4.0. They prioritize RPA’s
efficiency and digitalization, highlighting company,
Productivity, the ratio between output produced employee, and customer benefits. They explain which
and the required input, is pivotal for organizational activities RPA could take over and the three types of
success (Björkman, 1992). According to Schreyer and RPA can have: attended mode, unattended mode, and
Pilat (2001), the concept of productivity is critical to hybrid. The different types describe how much human
organizational success and signifies how effectively interaction they need to operate. Furthermore, they
resources like time, labor, and technology are utilized show the benefits of RPA for companies, employees,
to create value. It refers to the way in which an and customers (Axmann & Harmoko, 2020). Other
organization or individual can create value from authors, such as Romao et al. (2019), Aguirre and
available resources. Essentially, it is the measure of Rodriguez (2017), and Fernandez et al. (2020), used
how effectively inputs are transformed into valuable case studies to investigate hyperautomation. In
outputs (Schreyer & Pilat, 2001). Measurement addition, hyperautomation approaches have already
methods vary, encompassing quantitative metrics like been partially and successfully integrated into the
products manufactured per hour and resources used, or organizational environment via technologies such as
qualitative factors like customer satisfaction, and RPA (Agostinelli et al., 2023; Dumas et al., 2023).
concrete measures like return on investment (ROI) and Despite notable contributions, research lacks
labor efficiency (Schreyer & Pilat, 2001). In the digital systematization, necessitating a taxonomy elucidating
age, productivity assessment has evolved, allowing the hyperautomation-organizational productivity
real-time data analysis and data-driven decision- relationship. While recognizing hyperautomation's
making (Hyun et al., 2021). AI and RPA integration in potential impact (Haleem et al., 2021), a detailed
organizations significantly enhance efficiency, output investigation of its mechanisms remains unexplored.
quality, and employee engagement, fostering a
productive work environment (Mukherjee, 2021). 3. Methodology

2.3 Related Work This paper adopted two primary approaches to

address the research question. Firstly, a literature
The hyperautomation, recognized as a crucial review was conducted analyzing the relationship
technology trend, gained prominence, as Gartner between hyperautomation and productivity in
predicted its significance in 2019 (Gartner, 2019). companies. Secondly, a taxonomy development
Among others, Haleem et al. (2021), Jiménez-Ramírez process was undertaken to systematically categorize
(2021), and Madakam et al. (2022) provide a and organize the findings obtained from the literature
comprehensive overview of the field of review.
hyperautomation, covering definitions, concepts,
technologies, and models, along with industry use 3.1 Literature Collection
cases demonstrating global adoption variations.
Hyperautomation, as described by Kerremans et al. The literature review followed a comprehensive
(2020), entails the strategic coordination of diverse approach proposed by vom Brocke et al. (2009). We
automation methodologies across task, process, conducted the literature search using a topic-related
operational, and organizational domains. Lasso- database search, according to Webster & Watson
Rodriguez & Winkler (2023) further elaborate that (2002). Our focus was on the following IS-related
hyperautomation transcends mere task automation, databases: Web of Science, ACM Digital Library,
aspiring to automate entire workplace environments, EBSCOhost Business Source Premier, Science Direct,
thereby achieving comprehensive operational Emerald Insight, and IEEE Xplore. This search covers
optimization. Further articles concentrate on the a contribution index until October 2023.

Page 5665
 (“hyperautomation” OR “robotic process automation”) Scholar and the reference list of each article to identify
AND (product* OR (eff*) additional nine articles. Consequently, the literature
Web of Science 67 23 6 search, according to vom Brocke et al. (2009), shown

Title & Abstract Analysis

Forward & Backward

in Figure 1, yielded 28 relevant articles.

Full Text Analysis

ACM dL 275 1 1
EBSCOhost BSR 121 1 0

Search
Σ 28 3.2 Taxonomy Development
Science Direct 313 4 2
Emerald Insight 251 2 2 Taxonomies facilitate categorizing individual
IEEE Xplore 111 19 8 objects or publications, enabling analysis of the
Σ 1.138 Σ 50 Σ 19
relationships between objects within their specific
context. The approach of Nickerson et al. (2013) was
Figure 1. Overview of literature review procedure applied to create a taxonomy, allowing a deeper
understanding of complex subject areas. The
We aim to provide a comprehensive development process involved three key steps to
understanding of hyperautomation and its potential to establish a structured framework for organizing and
improve worker productivity and workflows as a first categorizing objects and characteristics (cf. Herm et
step. The emphasis is on exploring the impact of al., 2022; Nickerson et al., 2013) relevant to
hyperautomation on productivity improvement. hyperautomation and its impact on productivity.
Therefore, we defined a general query string in which Initially, the meta-characteristic forming the
we integrated RPA, as this technology is seen as the taxonomy's basis was determined. Subsequently,
core enabler of hyperautomation capabilities (e.g., ending conditions were defined to ensure the process's
George et al., 2023; Haleem et al., 2021): completion. The last step involved identifying
((“hyperautomation” OR “robotic process dimensions and characteristics integral to the
automation”) AND (product* OR (eff*)). The initial taxonomy.
search was conducted without filters, resulting in
1,138 potentially relevant publications. As part of an Meta-Characteristics. According to Nickerson
iterative analysis to ensure validity (Fleiss, 1971), the et al. (2013), the first step of taxonomy development
titles were checked, and duplicates, irrelevant articles, is determining the meta-characteristics that indicate
and those with access restrictions were eliminated. We the purpose of the taxonomy. These broad meta-
also excluded articles that were published before 2019, characteristics, namely method, objective, and
as the term hyperautomation was only used from 2019 measurement, serve as the foundational elements for
(Gartner, 2019) onwards, according to the current selecting characteristics within the taxonomy, aligning
understanding that this study follows. Table 1 provides with established practices in other taxonomies (e.g.,
an overview of the inclusion and exclusion criteria Herm et al., 2022). The meta-characteristic method
applied. After abstract analysis, we narrowed the refers to the technique used and the approach
search to 50 potentially relevant publications. addressed. The objective meta-characteristic focuses
on an organization's goals, intentions, or purposes and
emphasizes the specific outcomes or intentions
Inclusion criteria Exclusion criteria
researchers identify for organizations. The
Subject relevance Limited accessibility measurement meta-characteristic deals with process
Written in English All duplicates metrics and productivity measurement.
Article must be peer Publication date prior Ending conditions. Since taxonomy
reviewed 2019 development is an iterative process, defining ending
No restriction on Article does not exceed conditions that determine the process is important.
citation count six pages Nickerson et al. (2013) propose five subjective ending
conditions: concise, robust, comprehensive,
Table 1. Inclusion and exclusion criteria extendible, and explanatory as well as eight objective
ending conditions, which we fully adopted:
The full-text analysis excluded a further 31
articles. It focused mainly on the importance of 1. Uniqueness of the cells.
hyperautomation across a broad spectrum rather than 2. Uniqueness of the characteristics within the
just the impact or implementations of one specific dimensions.
technology. The remaining 19 documents were the 3. Uniqueness of the dimensions.
basis for the forward and backward search, which we 4. Examination of all or a representative sample
conducted using the citation data provided by Google of objects.

Page 5666
 5. Classification of at least one object under were used in the second iteration to develop the
each characteristic. taxonomy further. In a third iteration, in which the
6. No dimensions or characteristics were added entirety of the articles was used, the developed
in the last iteration. taxonomy was validated by meeting the final
7. No merging or splitting of objects in the last conditions.
iteration.
8. No merging or splitting of dimensions or 4. Taxonomy for Hyperautomation
characteristics in the last iteration.
The following section describes our final
These conditions ensure the uniqueness and taxonomy, which we developed according to
completeness of the classification process. We Nickerson et al.'s (2013) iterative approach. We
implemented their conditions except for the object subsequently validated our taxonomy through the final
exclusivity constraint. This modification is consistent assessment of the ending conditions and the evaluation
with methodologies employed by other IS scholars guidelines from Rich (1992) in Section 4.2. The results
(e.g., Herm et al., 2022; Püschel et al., 2016). of the taxonomy development are presented below in
Consequently, our final taxonomy permits specific Table 2. Different dimensions were assigned to each
objects to be classified under multiple characteristics meta-characteristic, which in turn were given
within each dimension, improving the taxonomy's characteristics. The resulting taxonomy contains three
clarity (Herm et al., 2022). meta-characteristics, six dimensions, and 36
Dimensions and Characteristics. Following characteristics. Any further dimension would
Nickerson et al. (2013), an empirical-to-conceptual unnecessarily extend the taxonomy without providing
approach is suggested for collecting and analyzing additional information.
empirical data first and then using this data to develop
conceptual frameworks or models that explain the 4.1 Taxonomy Description
observed phenomena. Our literature search is the basis
for the development of dimensions and characteristics. Method: Technology. This dimension refers to
The patterns, relationships, and relevant factors the technology used for hyperautomation. Halem et al.
obtained from the reviewed literature were converted (2021) divide hyperautomation into six different
into dimensions and characteristics. In addition, the related technologies: RPA, PM, AI/ML, Natural
final conditions helped to determine and select the Language Processing, Optical Character
dimensions and associated characteristics. Three Recognition, and Digital Twin of an Organization.
iterations were conducted. In the first iteration, The taxonomy splits AI and ML and adds deep
publications that provide a comprehensive or holistic learning to the granularity for better differentiation
perspective were initially used to conceptualize the and finer classification. DL represents a branch of AI
taxonomy. Articles focusing on individual dimensions

Table 2. Taxonomy for hyperautomation to increase productivity.

MC Dimension Characteristic
Method

Technology* RPA PM AI ML DL NLP OCR DTO

High Com- Execution Systems Task

Automatability Rule-based Repetitive
manual plexity volume access standardization
Productivity & Customer
Objective

Business goals* Cost savings Quality Insights Availability

effectiveness experience
Technological (Data) Agility &
Governance Cycle time Analytics Availability
goals* Security Scalability
Process
urement

Number of edits Error reduction Process quality improvement

Meas-

metrics
Productivity Internal Customer
Costs Efficiency Errors ROI
measurability processes satisfaction
MC=Meta-Characteristic; *Dimensions are non-exclusive

Page 5667
and can also improve RPA and hyperautomation tasks are also more scalable and agile (Moraes et al.,
(Moraes et al., 2022). 2022; Petkova et al., 2020). Hyperautomation
technologies can guarantee constant availability. The
Method: Automatability. Automatability refers
software is constantly running, does not take vacation
to which tasks in the process can be automated and the
or sick days and responds faster than human
necessary requirements. The criteria for implementing
employees (Meironke & Kuehnel, 2022).
hyperautomation are that the tasks are standardizable
and do not require subjective assessment, creativity, or Measurement: Process metrics. Process metrics
dynamic interpretation. In addition, digital input and measure the effectiveness and efficiency of the
output are required (Haleem et al., 2021; Jiménez- automated process. They can be measured by the
Ramírez, 2021; Wewerka & Reichert, 2023). number of edits the software is able to process
Conversely, the hyperautomation technology stack compared to the manual operation (Aguirre &
can handle a high volume of execution that is Rodriguez, 2017). In addition, the errors between
repetitive, rule-based, highly manual, complex, and these two processes can be compared and
prone to human error (Haleem et al., 2021; Jiménez- improvements in the quality of the process can be
Ramírez, 2021; Wewerka & Reichert, 2023). evaluated (Aguirre & Rodriguez, 2017; Hyun et al.,
Furthermore, the hyperautomation approach requires 2021; Quille et al., 2023).
access to the relevant systems and applications to get
Measurement: Productivity measurability. The
the task done.
different scholarly studies have various success
Objective: Business goals. Business goals are the metrics to measure the increase in productivity. For
higher goals that a company wants to achieve through instance, Aguirre & Rodriguez (2017) cited a study
automation approaches. One of the goals of companies from Capgemini (2016) that compared how many
is to increase profits through cost savings as a result of human workers can be replaced by software and still
automation (Denagama et al., 2020). Companies can get the same amount of work done, how many errors
manage this by focusing on better quality and constant are made by both groups and how many costs can be
availability when introducing automation to realize saved. Internal processes can be improved, leading to
hyperautomation and thus become more productive a reduction in total working time (Denagama et al.,
and effective (Hyun et al., 2021; Pramod, 2022; 2020; Royhan et al., 2023). Other authors compare the
Royhan et al., 2023). Another goal is to gain better return on investment between a group that works with
insight into processes (Mohamed et al., 2022). RPA and another group that does everything manually
Furthermore, companies can improve customer (Aguirre & Rodriguez, 2017). Another way to measure
experience to increase their image (Axmann & productivity is the quality of tasks and documents,
Harmoko, 2020). An illustrative example is Zurich which can lead to higher customer satisfaction
Insurance, a company that has automated its manual (Axmann & Harmoko, 2020). Hyperautomation
billing processes with the help of hyperautomation. In reduces working time as tasks can be completed
2014, 37 out of 81 employees oversaw repetitive and quickly, thus increasing efficiency (Royhan et al.,
patterned tasks. After the implementation of RPA 27 2023).
workers could be relocated to improve service quality
that increases customer satisfaction (Hyun et al., 4.2 Taxonomy Evaluation
2021). In a cross-industry analysis of companies by
Bahaweres et al. (2022) that rely on hyperautomation, The ending conditions were all classified as
it was found that the return on investment averaged fulfilled in the last iteration. The objective ending
159% percent, clearly highlighting the positive conditions can be assessed directly, and we briefly
economic impact and potential cost savings of this illustrate the interpretation of the subjective ending
technology. conditions (Nickerson et al., 2013) as follows:
Objective: Technological goals. Technological
goals describe the technical aspects that a company • It is concise: With six dimensions, our
wants to achieve with hyperautomation. By improving taxonomy is meaningful without being
the handling of data through automation, the transfer unwieldy or overwhelming. We are also
and storage of data can be made more secure, and the within the range of dimensions recommended
general cycle time of a case can be decreased by Nickerson et al. (2013) and Miller (1956).
(Mohamed et al., 2022). Better visibility of data and • It is robust: The interest is given by the
processes also enables better governance and sufficient differentiation of the dimensions
advanced analytics (Quille et al., 2023). Automated and characteristics among the objects.

Page 5668
 • It is comprehensive: All relevant dimensions strengthening the system’s rigor. We refined
of the presented objects of interest were the taxonomy development process using
identified. systematic empirical methods and
• It is extendible: The taxonomy enables easy quantitative analysis to verify the
extensibility through the inclusion of new comprehensiveness of our results.
dimensions and characteristics. • Completeness and logic: The classification
• It is explanatory: The dimensions and must comprehensively cover all relevant
characteristics have sufficient explanatory taxa, demonstrating logical and consistent
value, as illustrated in the descriptive section. relationships within and between groups. Our
approach strictly applies the described
Even though the taxonomy appears intuitively methods, ensuring the taxonomy is complete
reasonable with the fulfillment of the ending and logical by meeting the ending conditions
conditions, we rate the quality of it based on the outlined by Nickerson et al. (2013).
framework proposed by Rich (1992). In the following, • Recognizability: The taxonomy should
we elaborate on the categorization of our taxonomy reflect real-world applicability. This article
into the seven guidelines outlined by Rich (1992), emphasizes practical applicability, guiding
which serve as prerequisites for valid classifications both the selection of literature and the
(cf. Herm et al., 2022; Haas et al., 2014; Neis et al., inclusion of elements within the taxonomy.
2024):
5. Discussion
• Breadth: The classification system must be
structured either as a typology (conceptual With our taxonomy, we present a systematization
framework) or as a taxonomy (hierarchical that can be used to make hyperautomation and the
categorization), thereby guiding the selection resulting productivity improvement comprehensible.
of features and general groupings. We While there is a growing body of knowledge in this
decided to create a taxonomy consisting of domain, science currently focuses on the individual
three meta-characteristics, six dimensions, approaches from the technology stack and neglects the
and 36 characteristics. integrated approach internalized by hyperautomation.
• Meaning: The system should be based on a The individual technology implementations are
philosophical foundation that explains the evaluated on a domain-specific basis to verify their
rationale for the classification and justifies significance, but in the context of a synergistic
the chosen method. Our taxonomy draws on approach to hyperautomation, the focus should not be
existing knowledge from the domains of limited to the task or process level, but a holistic
hyperautomation and productivity. The basis concept for BPM should be developed.
is a representative selection of publications The results presented make it clear that by
that is ensured by a systematic literature implementing hyperautomation technologies,
search (vom Brocke et al., 2009). companies can streamline their processes and reduce
• Depth: Classification should be based on a the need for manual labor, resulting in cost and time
multivariate analysis of objects that captures savings (Quille et al., 2023). Hyperautomation
the complexity of real-world phenomena. We increases productivity by automating routine and
ensured depth by analyzing the interplay of repetitive tasks, allowing employees to focus on more
relevant concepts and developing the value-creating activities such as decision-making,
taxonomy empirically and conceptually problem-solving, and customer service (Haleem et al.,
(Nickerson et al., 2013). 2021). This can lead to significant cost and time
• Theory: Determination of dimensions and savings for the company (Falih et al., 2022). BPM,
characteristics should be based on theoretical RPA, and AI are key tools for process optimization.
knowledge. We utilized 28 research articles By analyzing and improving existing business
as the qualitative foundation to identify processes, BPM increases efficiency, streamlines
dimensions and characteristics for workflows, and reduces manual work (Falih et al.,
systematically categorizing productivity and 2022). RPA, on the other hand, uses software robots to
hyperautomation. automate rule-based tasks, which helps to reduce
• Quantitative measurement: Grouping taxa errors and increase efficiency (Bahaweres et al.,
should rely on numerical methods and 2022).
multivariate analysis, ensuring that theory is Despite these advantages, there are also
supported or refined by quantitative data, disadvantages. For example, RPA lacks optimization

Page 5669
approaches for the design or processes, it works may not uniformly apply to all companies or
simply and cannot identify inefficiencies (Axmann & industries, necessitating further investigation in
Harmoko, 2020). It also has difficulties with non- specific contexts. Although not exhaustive, this study
electronic data and unstructured inputs. Although RPA explores significant relationships between
is cost-effective compared to other technologies, it is hyperautomation and productivity.
not consistently cost-effective (Axmann & Harmoko,
2020). Selecting appropriate processes for automation 6. Conclusion and Further Research
is critical; tasks that require subjective judgment and
creativity are better suited to human workers (Herm et This article addresses the emerging field of
al., 2023). Standardized repetitive, rule-based, and hyperautomation, aiming to enhance organizational
manual tasks are a perfect fit for the capabilities of productivity. Through a comprehensive approach
RPA (Meironke & Kuehnel, 2022) as long as the encompassing diverse technologies, it offers insights
software has access to the required systems. into streamlining business processes. The taxonomy,
IS scholars analyze the use of different with its 3 meta-characteristics, 6 dimensions, and 36
technologies across different sectors, e.g., finance and characteristics, developed from a literature review
industry, and compare different technologies. Yadav underscores hyperautomation's potential for
and Panda (2022) emphasize the importance of significant productivity gains. By systematizing
carefully selecting processes for automation and hyperautomation, this work offers a managerial
identifying the characteristics that make a process perspective on leveraging these technologies. This
optimal for automation. Going one step further in our research bridges a gap in the current literature,
work, we suggest using different technologies across providing a foundation for further exploration. Its
different areas and thus moving more towards structured taxonomy and nuanced insights set the stage
hyperautomation. for optimizing operations in the evolving landscape of
The developed taxonomy enhances our hyperautomation.
understanding of organizational behavior and In our exploration of hyperautomation, we
technology adoption by enhancing productivity perceive numerous opportunities for further research
through hyperautomation. Organizing knowledge and into the cross-technology automation capabilities of
highlighting relationships between different RPA, AI, PM, and other technologies. We
approaches improves access to relevant information acknowledge a crucial knowledge gap regarding the
and identifies gaps in the current literature. The study long-term impacts on employee productivity,
evaluates hyperautomation's impact on company including workplace design and motivation.
productivity across various domains, offering a novel Addressing these gaps provides an opportunity to
taxonomy that categorizes different methods and their improve our understanding of the long-term impact of
interrelationships. The findings provide practical hyperautomation. To advance our research, we plan to
insights for organizations seeking to enhance use this taxonomy and expert interviews to conduct a
productivity through technologies like RPA and AI, pattern analysis using cluster and archetype analysis
suggesting investment and integration for increased aimed at refining and deepening the insights gained
efficiency. Importantly, it emphasizes that RPA and from the literature. In addition, these results will be
AI function as virtual assistants rather than eliminating used as a starting point for the development of a
jobs, guiding the development of policies and framework for hyperautomation implementation
strategies for hyperautomation adoption in various projects. The basis for measuring success will not be
organizational settings. the mere realization of one or more implementations
Naturally, this study is not without limitations. from the hyperautomation technology stack but will
The absence of a universal measuring standard leads include economic factors and, in particular,
to interpretation differences and inconsistencies across productivity as success indicators. The aim of our
studies. The selected articles used different research is the comprehensive exploration of
measurement tools, such as ROI, speed, or error. In hyperautomation and the derivation of resulting
order to obtain a better overview of all automation implications for theory and practice.
options, the technologies should be examined
individually and then transferred to a hyperautomation
concept. Context dependency is another important
Acknowledgment
consideration. The effects of hyperautomation on
productivity vary based on contextual factors like This work has been developed in the project Hyko
industry, company size, processes, and IT and is partly funded by the Bavarian Ministry of
infrastructure. It's important to recognize that results Economic Affairs, Regional Development and Energy

Page 5670
enabled by the European Regional Development Fund conceptualisation of robotic process automation (RPA)
(ERDF) under grant number 1.2-StMWK-F.4- UFR- benefits. Proceedings of the 28th European Conference
002. The authors are responsible for the content of this on Information Systems (ECIS2020).
publication. Dumas, M., La Rosa, M., Leno, V., Polyvyanyy, A., &
Maggi, F. M. (2022). Robotic Process Mining. Process
Mining Handbook, 448, 468-491.
References Engel, C., Ebel, P., & Leimeister, J. M. (2022). Cognitive
automation. Electronic Markets, 32(1), 339–350.
Ackermann, L., Käppel, M., Marcus, L., Moder, L., Dunzer, Falih, N., Supangkat, S. H., & Lubis, F. F. (2022). Robotic
S., Hornsteiner, M., Liessmann, A., Zisgen, Y., Empl, Process Automation in Smart System Platform: A
P., Herm, L., Neis, N., Neuberger, J., Poss, L., Review. 2022 International Conference on ICT for
Schaschek, M., Weinzierl, S., ... & Winkelmann, A. Smart Society (ICISS), 01–05. IEEE.
(2024). Recent Advances in Data-Driven Business Fernandez, D., Aman, A., & Omar, S. (2020). Robotic
Process Management. arXiv e-prints, process automation: A case study of the impacts on
arXiv:2406.01786. employee skills. EasyChair Preprint, 4443.
Aguirre, S., & Rodriguez, A. (2017). Automation of a Fleiss, J. L. (1971). Measuring nominal scale agreement
Business Process Using Robotic Process Automation among many raters. Psychological Bulletin, 76(5), 378.
(RPA): A Case Study. In J. C. Figueroa-García, E. R. Gartner. (2019, October 21). Gartner Top 10 Strategic
López-Santana, J. L. Villa-Ramírez, & R. Ferro- Tech-nology Trends For 2020.
Escobar (Eds.), Applied Computer Sciences in https://www.gartner.com/smarterwithgartner/gartner-
Engineering (Vol. 742, pp. 65–71). Springer top-10-strategic-technology-trends-for-2020
International Publishing. Gartner. (n.d.). Hyperautomation. Gartner Glossary.
Agostinelli, S., Lupia, M., Marrella, A., & Mecella, M. https://www.gartner.com/en/information-
(2022). Reactive synthesis of software robots in RPA technology/glossary/hyperautomation
from user interface logs. Computers in Industry, 142, George, S., George, H. A. S., Baskar, T., & Dr. V.Sujatha,
103721. V. (2023). The Rise of Hyperautomation: A New
Axmann, B., & Harmoko, H. (2020). Robotic Process Frontier for Business Process Automation.
Automation: An Overview and Comparison to Other Glass, R. L., & Vessey, I. (1995). Contemporary
Technology in Industry 4.0. 2020 10th International application-domain taxonomies. IEEE Software,
Conference on Advanced Computer Information 12(4), 63–76
Technologies (ACIT), 559–562. Haleem, A., Javaid, M., Singh, R. P., Rab, S., & Suman, R.
Bahaweres, R. B., Amna, H., & Nurnaningsih, D. (2022). (2021). Hyperautomation for the enhancement of
Improving Purchase to Pay Process Efficiency with automation in industries. Sensors International, 2,
RPA using Fuzzy Miner Algorithm in Process Mining. 100124.
2022 International Conference on Decision Aid Herm, L.-V., Janiesch, C., Helm, A., Imgrund, F., Hofmann,
Sciences and Applications (DASA), 1483–1488. A., & Winkelmann, A. (2023). A framework for
Bishop, C. M. (2013). Model-based machine learning. implementing robotic process automation projects.
Philosophical Transactions of the Royal Society A: Information Systems and E-Business Management,
Mathematical, Physical and Engineering Sciences, 21(1), 1–35.
371(1984), 20120222. Herm, L.-V., Janiesch, C., Reijers, H. A., & Seubert, F.
Björkman, M. (1992). What is Productivity? IFAC (2021). From Symbolic RPA to Intelligent RPA:
Proceedings Volumes, 25(8), 203–210. Challenges for Developing and Operating Intelligent
Carvalho, D. V., Pereira, E. M., & Cardoso, J. S. (2019). Software Robots. In A. Polyvyanyy, M. T. Wynn, A.
Machine Learning Interpretability: A Survey on Van Looy, & M. Reichert (Eds.), Business Process
Methods and Metrics. Electronics, 8(8), 832. Management (Vol. 12875, pp. 289–305). Springer
Dahmen, J., Kayaalp, M. E., Ollivier, M., Pareek, A., International Publishing.
Hirschmann, M. T., Karlsson, J., & Winkler, P. W. Herm, L.-V., Wanner, J., & Janiesch, C. (2022). A
(2023). Artificial intelligence bot ChatGPT in medical Taxonomy of User-centered Explainable AI Studies.
research: The potential game changer as a double- Proceedings of the 2022 Pacific Asia Conference on
edged sword. Knee Surgery, Sports Traumatology, Information Systems. PACIS 2022 Proceedings. 9.
Arthroscopy, 31(4), 1187–1189. Hyun, Y., Lee, D., Chae, U., Ko, J., & Lee, J. (2021).
Damioli, G., Van Roy, V., & Vertesy, D. (2021). The impact Improvement of Business Productivity by Applying
of artificial intelligence on labor productivity. Eurasian Robotic Process Automation. Applied Sciences, 11(22),
Business Review, 11(1), 1–25. 10656.
Dan, G., Claudiu, D., Alexandra, F., Adela, P.-P., Ovidiu, S., Jiménez-Ramírez, A. (2021). Humans, Processes and
Honoriu, V., & Liviu, M. (2022). Multi-Channel Robots: A Journey to Hyperautomation. Business
Chatbot and Robotic Process Automation. 2022 IEEE Process Management: Blockchain and Robotic Process
International Conference on Automation, Quality and Automation Forum, 428, 3–6.
Testing, Robotics (AQTR), 1–6. Kerremans, M., Searle, S., Srivastava, T., & Iijima, K.
Denagama, V., Imesha, M., Bandara, W., Syed, R., & (2020, September 30). Market Guide for Process
Toman, D. (2020). An empirically supported Mining. Marc Kerremans, Samantha Searle, Tushar

Page 5671
 Srivastava, Ki-mihiko Iijima. Analysis Method for Robotic Process Automation.
https://www.gartner.com/doc/reprints?id=1- Sustainability, 15(4), 3702.
28SA9BAA&ct=220118&st=sb Reichert, M., & Weber, B. (2012). Enabling Flexibility in
Lacity, M., & Willcocks, L. (2021). Becoming Strategic with Process-Aware Information Systems. Springer Berlin
Intelligent Automation. MIS Quarterly Executive, Heidelberg.
20(2), 169–182. Rich, P. (1992). The Organizational Taxonomy: Definition
Lasso-Rodriguez, G., & Winkler, K. (2023). and Design. The Academy of Management Review,
Hyperautomation to fulfil jobs rather than executing 17(4), 758–781.
tasks: The BPM manager robot vs human case. Revista Romao, M., Costa, J., & Costa, C. J. (2019). Robotic Process
Română de Informatică Și Automatică, 30(3), 7–22. Automation: A Case Study in the Banking Industry.
Madakam, S., Holmukhe, R. M., & Revulagadda, R. K. 2019 14th Iberian Conference on Information Systems
(2022). The Next Generation Intelligent Automation: and Technologies (CISTI), 1–6.
Hyperautomation. Journal of Information Systems and Royhan, M., Wati, L. N., & Soma, A. M. (2023). Systematic
Technology Management, 19, e202219009. Literature Review Impact of RPA Implementation on
Marr, B. (2023). Microsoft’s Plan To Infuse AI And Company Performance. In S. Saha, D. N. Rahmatika,
ChatGPT Into Everything. Forbes. Y. Li, K. Sandar Kyaw, D. Indriasih, A. Zainudin, A.
https://www.forbes.com/sites/bernardmarr/2023/03/06 Nurmandi, Y. P. Astutie, J. Mariyono, P. Susongko, &
/microsofts-plan-to-infuse-ai-and-chatgpt-into- Y. Arfiani (Eds.), Proceedings of the International
everything/ Conference on Global Innovation and Trends in
Meironke, A., & Kuehnel, S. (2022). How to Measure RPA’s Economics and Business (ICOBIS 2022).
Benefits? A Review on Metrics, Indicators, and Schreyer, P., & Pilat, D. (2001). Measuring productivity.
Evaluation Methods of RPA Benefit Assessment. 5. OECD Economic Studies, 33(2), 127–170.
Miller G.A. (1956) The magic number seven, plus or minus Syed, R., Suriadi, S., Adams, M., Bandara, W., Leemans, S.
two: some limits on our capacity for processing J. J., Ouyang, C., Ter Hofstede, A. H. M., Van De
information. Psychological Review. 101(2), 343–352. Weerd, I., Wynn, M. T., & Reijers, H. A. (2020).
Mohamed, S. A., Mahmoud, M. A., Mahdi, M. N., & Robotic Process Automation: Contemporary themes
Mostafa, S. A. (2022). Improving Efficiency and and challenges. Computers in Industry, 115, 103162.
Effectiveness of Robotic Process Automation in Taulli, T. (2020). The Robotic Process Automation
Human Resource Management. Sustainability, 14(7), Handbook: A Guide to Implementing RPA Systems.
3920. Apress.
Moraes, C. H. V. D., Scolimoski, J., Lambert-Torres, G., Van Der Aalst, W. M. P., Bichler, M., & Heinzl, A. (2018).
Santini, M., Dias, A. L. A., Guerra, F. A., Pedretti, A., Robotic Process Automation. Business & Information
& Ramos, M. P. (2022). Robotic Process Automation Systems Engineering, 60(4), 269–272.
and Machine Learning: A Systematic Review. vom Brocke, J., van der Aalst, W., Grisold, T., Kremser, W.,
Brazilian Archives of Biology and Technology, 65, Mendling, J., Pentland, B., Recker, J., Roeglinger, M.,
e22220096. Rosemann, M., & Weber, B. (2021). Process Science:
Mukherjee, A. (2021). Robotic process automation with The Interdisciplinary Study of Continuous Change
Blue Prism to optimize inventory management. (SSRN Scholarly Paper 3916817).
Neis, N., Gwinner, F., & Haueisen, C. (2024). A taxonomy vom Brocke, J. vom, Simons, A., Niehaves, B., Niehaves,
of artificial intelligence for process mining B., Reimer, K., Plattfaut, R., & Cleven, A. (2009).
enhancement. PACIS 2024 Proceedings. 10. Reconstructing the giant: On the importance of rigour
Nickerson, R. C., Varshney, U., & Muntermann, J. (2013). in documenting the literature search process. 17th
A method for taxonomy development and its European Conference on Information Systems (ECIS),
application in information systems. European Journal Verona, Italien.
of Information Systems, 22(3), 336–359. Webster, J., & Watson, R. T. (2002). Analyzing the past to
Petkova, M., Jekov, B., & Petkova, P. (2020). prepare for the future: Writing a literature review. MIS
Administrative Automatic Solutions in Telecom Quarterly, xiii–xxiii.
Services. 2020 28th National Conference with Wewerka, J., & Reichert, M. (2023). Robotic process
International Participation (TELECOM), 86–89. automation—A systematic mapping study and
Pramod, D. (2022). Robotic process automation for industry: classification framework. Enterprise Information
Adoption status, benefits, challenges and research Systems, 17(2), 1986862.
agenda. Benchmarking: An International Journal, Yadav, N., & Panda, S. P. (2022). A Path Forward for
29(5), 1562-1586. Automation in Robotic Process Automation Projects:
Püschel, Louis; Roeglinger, Maximilian; and Schlott, Helen, Potential Process Selection Strategies. 2022
"What's in a Smart Thing? Development of a Multi- International Conference on Machine Learning, Big
layer Taxonomy" (2016). Proceedings of the 37th Data, Cloud and Parallel Computing (COM-IT-CON),
International Conference on Information Systems, 801–805.
Dublin 2016, ICIS 2016 Proceedings. 6. Zhao, X., Oseni, T., & Medishetty, B. T. (2022). Overview
Quille, R., Almeida, F., Borycz, J., Corrêa, P., Filgueiras, L., of Business Hyper-automation. 2022 IEEE
Machicao, J., Almeida, G., Midorikawa, E., Demuner, International Conference on E-Business Engineering
V., Bedoya, J., & Vajgel, B. (2023). Performance (ICEBE), 100–105.

Page 5672