Robotic Process Automation (RPA) in Lieu of a New ERP System

Robotic Process Automation (RPA) stands for the software-supported operation of software solutions via their user interface [1]. According to Gartner, the technology is one of the fastest growing markets in enterprise software [2]. The primary goal that RPA aims to achieve is the automated execution of tasks that previously required human intervention. RPA software thus imitates the interaction of a human with a software system, e.g. when reading out, changing and re-entering data.

The trend can also be described as a bridging technology [3], as Robotic Process Automation is not fully integrated into the existing system landscape, but only uses the top architectural layer (namely the user interface) [4]. Data can be transferred between different software applications or just between different modules or input screens of an application [1]. For selected tasks, RPA can replace humans or – when transferring data between two software systems – the programming interface.

Use and structure of classic RPA software

Robotic Process Automation is typically used for repetitive tasks that are characterized by simple, clearly structured rules and a high level of standard­ization. The starting point for this is usually processes that offer the advantages of rapid processing and data transmission times. In such an RPA use case, the application is installed on a client. First, the necessary set of rules is created and used to determine which software user actions to automate.

A macro recorder, which records user interactions with the programs and creates a template for automation, aids in this determination. The recordings are then turned into configurable rules according to which the software robot should ultimately work. Basic dependencies and conditions can also be modeled here. This configuration is made possible by an editor. The management console is responsible for load distribution and monitoring of the individual robots [1, 3].

RPA software is characterized by its lack of platform dependence. No individual system interfaces are necessary, because access to the systems involved in the process takes place via the respective presentation tier [4]. This also means that the underlying logic of the existing systems remains untouched. RPA software does not create new applications and does not store transaction data.

Basics of RPA

RPA solutions can generally be divided into two operating modes: attended and unattended. With attended RPA, also known as Robotic Desktop Automation, the program runs on a client. A supervised mode allows individuals to trigger bot activities in order to perform, automate, and actively monitor parts of a process. Unattended RPA software, on the other hand, runs on the server and acts autonomously. The processes are processed by the software in a fully autonomous manner. Different systems can also be controlled. In some cases, several robots interact with each other and divide tasks among themselves [5].

What opportunities and possibilities arise from the use of RPA?

Nowadays, against the current background of competition and cost pressures, the importance of efficient processes in companies is constantly increasing. Robotic Process Automation promises the idea of increasing efficiency by integrating software bots instead of employees to carry out work processes. In this way, manual steps can be automated by processing, preparing or storing structured information from various sources in (e.g. ERP) systems according to specific rules. Three main opportunities arise from the use of RPA:

Efficiency and productivity gains

The use of RPA technologies is useful when the automation of simple processes can demonstrably lead to gains in efficiency and productivity. Operational efficiency is sought in the form of time, cost and personnel savings, reduction in manual tasks and labor effort, and increase in productivity [1, 6].

Increased data quality

By using Robotic Process Automation, common transaction errors such as incorrect data entry, skipped process steps or errors caused by the application of rules are reduced. In addition, the number of human errors can be reduced [7].

Cost-effective and technology-independent alternatives

Robotic Process Automation can be implemented, configured, and maintained more easily and cost-effectively compared to large enterprise systems and other forms of automation [8]. At the same time, it offers users a simple and intuitive interface. Well-known applications and sources, from the mainframe to Excel and from CRM or ERP to web applications, can all be used by the software bots. Implementation is possible without disruption to the underlying IT infrastructure, provided that the underlying forensic software is compatible with the RPA software [7].

The limits of RPA

However, Robotic Process Automation also has its limitations. A major obstacle to successful RPA deployment is currently the mandatory requirement for structured data. AI and other complementary technologies could enable the extension of RPA benefits to unstructured or semi-structured data [6]. In addition, RPA cannot carry out dynamic decision-making processes [1], meaning that possible RPA processes are limited to rule-based processes and should ideally be based on simple business rules.

There is only limited automated response to unforeseen process sequences and errors (complex error handling is difficult to implement). When it comes to the advantages of Robotic Process Automation, the low technical barriers and lack of programming knowledge required of users are repeatedly emphasized. However, implementing RPA exclusively on the presentation tier also has certain disadvantages. Since the software only interacts with the programs once they are in a human-readable format, RPA eliminates the need for the multi-tier network communication relied upon by traditional applications.

Modules, input forms and applications are only read via the user interface. If the form of presentation is changed even marginally during an application update, errors can arise in the execution of the process. A high error rate can occur, particularly in web-based application systems that are characterized by design and display changes. It is also unclear to what extent the accuracy of the automated processes should be checked and where the responsibility for this lies.

A changed arrangement of buttons or a new menu entry on portals such as Amazon is difficult to control. Furthermore, in the context of automation, the surface-based automation approach results in a low data throughput rate when compared to decoupled interfaces. The process steps and inputs always depend on the interface. This severely limits the scalability of possible applications.

So which process is right for RPA?

Although the implementation of bots can lead to impressive results, Robotic Process Automation does not meet the requirement of being able to automate highly complex processes end-to-end and can thus only partially contribute to a complete automation of such processes. Companies run the risk of selecting unsuitable processes for bot automation due to unrealistic expectations. Therefore, selecting the right processes can be considered the biggest challenge of RPA automation. The following characteristics can be used as criteria for selecting a suitable RPA process [6]:

• Strongly rule-based processes: RPA requires a prescribed rule for each eventuality, and this rule must be unambiguously clear. The decision logic of the selected process must be logic-based and and almost completely predictable for all cases.
• Repetitive tasks: Adequate transaction volume helps maximize the benefits derived from implementing software bots in an organization. Generally, routine and repetitive tasks constitute an ideal choice for automation.
• Process maturity: Established processes that are stable and understood by all process participants are suitable for RPA.
• Structured and digital input data: All input data must be digital and in a well-structured format.
• Transaction-oriented tasks: RPA is suitable for transactional tasks that are performed by humans and can be replaced by software robots. In particular, transaction errors (e.g. incorrect data) can be reduced, and many transactional activities can be carried out at once.
• Degree of standardization: Processes with a high degree of standardization are best suited for the application of RPA. The more complex a process, the more complex it will be to implement using RPA.
• Few exceptional cases: Processes designed for RPA should not have to deal with exceptional cases. The more special rules are handled by bots, the likelier it is that process automation and optimization efforts are delayed or canceled. In this case, increased manual intervention is also necessary.
• Process documentation: Process descriptions are essential to teach bots specific keystroke and click sequence behaviors. If processes are well documented, bot implementation will be more successful.
• Number of systems in the process: There is great potential for the use of RPA in processes that extend across multiple systems and have grown unnecessarily complex or time-consuming due to media disruptions. Accordingly, the requirements must then be met by all systems involved. The strength of RPA lies in its usefulness as a bridging technology.

What can RPA do in the context of ERP systems?

Companies often use outdated, monolithic ERP systems. Programming interfaces to connect to modern digital applications or switch to a modern, open and service-oriented ERP system is a task that often exceeds the budget of the IT departments. In this case, Robotic Process Automation offers a way out of the dilemma and automates or accelerates processes without interfering with or changing the existing ERP systems.

• RPA as an integrator: If an ERP system does not offer an interface for an existing problem, content can be exchanged with a third-party system via GUI remote control (RPA) [1]. Reasons for a lack of interface could be an outdated technology platform or a lack of development resources.
• RPA to improve ergonomics: A lack of usability due to outdated systems and synchronization with third-party systems can in many cases obstruct productivity. However, people often stick with such systems because the costs and effort involved in changing systems outweigh any expected gains in processing speed. This is where RPA can offer a short-term solution by compen­sating for the deficits in the installed base [1].
• RPA as a testing tool: New release and version changes in ERP systems are based on a test phase lasting several weeks, in which a copy of the available productive data is used to check whether the new system works correctly for the specified requirements. However, by using RPA, test guides with click sequences can be created, which can then replace special testing software at this point [1].
• RPA as a bridge between the real and digital world: Processes based on (paper) forms can be ideally automated in the ERP system using RPA. Particularly when forms involve structured data, RPA can be used to extract, interpret and assign said data to one or more business processes in the ERP system. In the case of paper documents, they are first scanned and then converted into a machine-readable form via OCR technology. [1]

Conclusion and outlook

As long as rigid and monolithic IT and ERP systems are in operational use at companies, Robotic Process Automation can be used as a solution to compensate for openness and flexibility deficits in the installed base. The user gains short-term flexibility, but the problems of the existing systems are only shifted elsewhere, rather than solved. RPA is suitable for automating repetitive and standardized processes. However, the most important prerequisite here is the availability of well-structured and digitalized data.

In many practical applications, however, this structure is missing. The best example of this is emails, which have digital content but are generally unstructured. This means that the combination of RPA with other cognitive components – such as an OCR component – offers significantly greater automation potential than the possible further development of Robotic Process Automation itself into a cognitive solution.

The potential of Robotic Process Automation for long-term process improvement should be seen as severely limited. The automation of processes and bridging front-end media disruptions results in a variety of dependencies and conditions, which have been summarized in this article. The path to a future-proof enterprise architecture (and the processes and systems that comprise it) requires open, adaptable systems with modern architecture, which are characterized by a high level of interoperability between various system levels. This ensures that interaction and coupling with other systems, as well as the automation of processes, are sustainably supported.

You might also be interested in

Conducting Experiments in Hybrid Learning Factories The example of the InTraLab Potsdam

Industrial production is undergoing rapid transformation through digitalization, automation and cyber-physical systems, creating new competence requirements for employees. Learning factories provide experiential environments for developing these competences. This article presents the Industrial Transformation Lab (InTraLab) as a hybrid learning factory combining physical demonstrators and digital simulations.

Learning Factories for the Future of Manufacturing in Brazil Advancing manufacturing through technology and skills development

Manufacturing firms in developing countries face challenges in closing productivity gaps while adopting Industry 4.0 technologies. Learning factories are one helpful approach to countering these challenges. One such example is the learning factory Fábrica do Futuroin São Paulo, Brazil, which has engaged students, supported competence development, and collaborated with industry in applied research, functioning as a hub for advanced manufacturing initiatives.

AI Implementation in Industrial Quality Control A design science approach bridging technical and human factors Erdi Ünal , Kathrin Nauth , Pavlos Rath-Manakidis, Jens Pöppelbuß , Felix Hoenig, Christian Meske

Artificial intelligence (AI) offers significant potential to enhance industrial quality control, yet successful implementation requires careful consideration of ethical and human factors. This article examines how automated surface inspection systems can be deployed to augment human capabilities while ensuring ethical integration into workflows. Through design science research, twelve stakeholders from six organizations across three continents are interviewed and twelve sociotechnical design requirements are derived. These are organized into pre-implementation and implementation/operation phases, addressing human agency, employee participation, and responsible knowledge management. Key findings include the critical importance of meaningful employee participation during pre-implementation, and maintaining human agency through experiential learning, building on existing expertise. This research contributes to ethical AI workplace implementation by providing guidelines that preserve human ...

Industry 4.0 Science | Volume 42 | 2026 | Edition 1 | Pages 120-127 | DOI 10.30844/I4SE.26.1.112

Applied AI for Human-Centric Assembly Workplace Design An ethics-informed approach Tadele Belay Tuli , Michael Jonek , Sascha Niethammer, Henning Vogler, Martin Manns

Artificial intelligence (AI) can enhance smart assembly by predicting human motion and adapting workplace design. Using probabilistic models such as Gaussian Mixture Models (GMMs), AI systems anticipate operator actions to improve coordination with robots. However, these predictive systems raise ethical concerns related to safety, fairness, and privacy under the EU AI Act, which classifies them as high-risk. This paper presents a conceptual method integrating probabilistic motion modeling with ethical evaluation via Z-Inspection®. An industrial case study using the Smart Work Assistant (SWA) demonstrates how multimodal sensing (motion, gaze) and interpretable models enable anticipatory assistance. The approach moves from ethics evaluation to ethics-informed work design, yielding transferable principles and a configurable assessment matrix that supports compliance-by-design in collaborative assembly.

Industry 4.0 Science | Volume 42 | 2026 | Edition 1 | Pages 60-68 | DOI 10.30844/I4SE.26.1.58

Increasing Resilience in Logistics with IT Investigating supply chain risk management information systems Alexander Baur, Jasmin Hauser, Dieter Uckelmann

The blockage of the Suez Canal in 2021, caused by the accident involving the container ship Ever Given, clearly illustrates the need to design global supply chains in such a way that they can respond quickly to disruptions. In a volatile, uncertain, complex, and ambiguous (VUCA) environment, conventional logistics processes that focus on efficiency, and supply chain management methods in particular, are increasingly reaching their limits. Resilience, achieved through a combination of robustness and agility, is essential to ensure responsiveness. This article analyzes how risk management information systems (RMIS) can increase resilience. The analysis covers data availability, data transparency, modeling and simulation of risk scenarios, and the development of appropriate emergency action plans. Despite existing challenges in designing IT infrastructure, the measures mentioned have the potential to increase resilience in logistics.

Industry 4.0 Science | Volume 41 | Edition 4 | Pages 36-42

The Core Principles of the Digital Twin Transformingorder processes and the automation pyramid Wilmjakob Herlyn

The digital twin [DT] is considered a key technology of Industry 4.0. Its basic concept is now being successfully applied in practice, as demonstrated by the commissioning of Mercedes' Factory56 in 2022. New identification technologies, tracking systems and communication solutions faciliate new ways of controlling production and managing material flows, particularly at the shop floor level. With precise technical data permanently available not only for products, but also for material availability and order fulfillment status, production processes can be managed more dynamically and efficiently. This is precisely where the concept of the DT comes into play, enabling the immediate use and evaluation of this data.Its relevance continues to grow, especially in the context of make-to-order production, the rising variety of product configurations, and the globalization of production and supply networks. This article introduces the basic concept of the DT and illustrates how it connects to ...

Industry 4.0 Science | Volume 41 | 2025 | Edition 3 | Pages 92-101