Is robotic process automation suitable for startups

Possible uses of Robotic Process Automation in purchasing

Daily recurring tasks require a lot of time and resources. Software robots make it possible to automate such tasks at low cost. The use of robot-controlled process automation generally serves to automate processes that already exist in the company, but can also be used to expand services without having to add additional staff capacities. The following example shows how this technology can be used in operational purchasing.

In addition to negotiating favorable framework conditions, especially purchasing prices, efficient handling of data and process management in purchasing is becoming increasingly important. Purchasing is facing major challenges here, the good handling of which can offer great potential for generating competitive advantages. New technologies in the field of process automation are currently being discussed intensively; their targeted and integrative use can make a high contribution to solving current challenges. Robotic Process Automation (RPA) represents the entry into intelligent process automation. RPA is software that works on a computer or server and maps activities automatically. The software imitates human interaction with the various user interfaces and accesses existing documents and files. RPA thus carries out human tasks within the existing process and thus works with marginal effects on the existing infrastructure. The technology interacts across systems with existing software such as ERP, HR, CRM systems and MS programs. To do this, the robot is placed on the existing IT landscape without the need to change the existing systems and software solutions. Positive effects - in addition to the potential time savings - can be reduced error rates or increased availability intervals, and the use of the system can be used to decouple transaction volume and employee capacity.

Robotic Process Automation works essentially rule-based, i.e. according to rules of the form IF - THEN - ELSE. If the process step requires a human decision, it is essential whether this decision component can be brought into such a rule form. In addition to this important property, there are other criteria for the processes to be automated in order to decide how well they are suitable for automation through RPA. Ideally, the process takes place in a standardized manner based on digitally available, structured data. If the process occurs very frequently, the monetary benefit that results from saving employee capacity is correspondingly greater. The more the quantitative and qualitative characteristics are given in the respective process in purchasing, the more likely it is that it can be automated using RPA. The evaluation of every process within purchasing should be done systematically in order to be able to determine RPA potential holistically.

High potential for automation

We have shown a selection of the core and sub-processes occurring in purchasing in the figure opposite: It can be seen that a large part of the processes has medium or high automation potential. Operational purchasing in particular offers ideal application options for the effective integration of RPA. Many processes in operational purchasing include tasks with a highly repetitive character and a high degree of standardization. Processes with a large proportion of conceptual design, such as in supplier management, are less suitable due to the high proportion of human decisions and interactions.

RPA as an entry-level technology

In addition to RPA, process automation can also be used at other, more intelligent levels, which means that an increasingly higher rate of process automation can be achieved. It is important to understand that this is an integrative approach and that the technologies do not replace one another, but complement one another. The levels describe the degree of automation, intelligence and process complexity. In addition to robotic process automation, cognitive automation and digital assistants are further building blocks in intelligent process automation. Autonomous agents would be able to make autonomous decisions in the future, but are currently not very widespread commercially.

Cognitive Automation (CA) typically uses machine learning; H. by analyzing and evaluating data, an algorithm recognizes patterns in data sets. Using cognitive automation, digitally available data can be read out and structured, such as incoming invoices, e-mails or delivery confirmations. Cognitive automation can therefore be used in addition to RPA and helps to structure incoming information so that it can then be further processed via RPA.

Digital Assistants (DA) are able to understand and extract relevant information from spoken or written texts. They can also answer in the form of speech or text (natural language processing). The name suggests it: digital assistants are located at the interface between the company and internal or external customers, i.e. in the interaction with people. So you have to be able to extract data and information from conversations with users in the front office and often simplify the user interface.

Exemplary use case

How these technologies can be used in a highly efficient and automated interlocking manner in the future is illustrated by an example of operational purchasing: the process from ordering to goods receipt.

So far, the process has not been standardized and varies depending on the request. Accordingly, the order and goods receipt process ties up valuable employee capacity. Up to now, orders have often had to be recorded using a complex and not very intuitive process, which results in incomplete or incorrect orders which have to be resolved by a work cycle. Processing the order and later checking the shipping notification against the order also ties up valuable resources.

Interaction between RPA, CA and DA

This process contains a large number of process steps which, in the interplay of RPA, CA and DA, can run automatically with almost no human involvement by software robots and reduce work cycles.

Through a structured chat with the person making the request, the digital assistant guides you through the ordering process and thus enables a structured and complete order to be available at the end. The Digital Assistant offers an intuitive user interface and accesses relevant information and systems in the background, which means that plausibility checks can be carried out at the same time. These can be freely defined and include, for example, availability checks, cost centers and budgets. After the order has been placed, the digital assistant forwards this information to the RPA robot in a predefined form.

The RPA robot automatically recognizes that there is a new work order for it and checks different ways to serve the order. This can be done, for example, via a framework agreement or individual allocation to a suitable supplier. The order is sent to the supplier as an email by RPA in the form of a PDF.

Finally, the information is compared

If the supplier has received the order confirmation and delivery notification, it is read out using Cognitive Automation and passed on to the RPA robot in a structured data format, for example an XML file.

This information is then automatically compared with the data in ERP - if there are no significant differences, the ordering process can be completed. If there are deviations instead, the RPA robot checks the amount of the difference and whether it is still within the tolerable range. In the event of a discrepancy, the responsible supplier will be informed. The robot stores the order confirmation in ERP and informs the inquiring person about the successful order and the expected delivery date.

Thanks to intelligent process automation, we get a structured process that ties up significantly less valuable employee capacity. Even without automation by DA and CA, RPA already enables a high rate of automation of the process steps to be executed. In addition, the documentation of the individual process steps and thus the transparency of the entire process are significantly increased. Automation through RPA is already an established and economical means and at the same time provides the basis for intelligent process automation.

Versatile application scenarios

The application scenarios in purchasing are varied and, thanks to more intelligent technologies, more complex processes can now be automated. Employee capacities are freed up and higher quality and value-adding activities such as optimization and focus on price negotiations are made possible.


The authors:

Sebastian Künkele, International Performance Research Institute - IPRI gGmbH, Stuttgart
Tobias Beuckes, Horváth & Partner GmbH, Düsseldorf


More about big data