Getting RPA projects right

Getting RPA projects right

RPA helps to classify and understand document content, prioritize tasks, make decisions and apply artificial intelligence (AI) to operational data. Central automatisms enable the execution of work processes with the simultaneous use of monitoring and reporting functions. That sounds good, but it's easier said than done. IT-MITTELSTAND therefore asked two experts what is most important in RPA projects.ITM: When medium-sized companies are thinking about introducing RPA: Which processes are particularly suitable for newcomers - and why?Lars Braitinger: Let software robots automate monotonous, recurring processes in almost all departments in companies of all sizes and industries. The software mimics the actions of human employees and thus relieves them of tedious, time-consuming and data-intensive activities. A large number of processes in all departments and industries are suitable for automation: authorities use the technology for digital application or the legal system to process GDPR requests that come in via the website. In our experience, the RPA journey often begins in a company's finance or HR department; other departments usually follow.Tackling RPA projects properly

The additional use of task mining also promotes automation projects, since it captures the daily desktop activities of employees. Through the use of AI, repetitive tasks that have a high potential for automation are identified and visualized, which in turn enables more efficient and productive work.Gunther Rameseder: In order to realize the optimal added value, companies should first use RPA in processes that have large transaction volumes and a high degree of standardization. Processes in invoice processing such as accounts payable or accounts receivable, in purchasing and in sales or customer service are particularly suitable here. However, business processes are often not standardized enough - and companies lack the transparency to identify suitable process steps for RPA. We also see a “need for customizing” in many companies. It is therefore part of the business model to adapt processes individually to the product, a region or the customer, for example. This means: RPA technology must also become more flexible and dynamic in order to be able to continue to deliver scalable added value.ITM: What technical and organizational requirements are necessary for the first projects to be successful?Rameseder: So that processes can be automated with the help of RPA , they must be 100 percent digital and rule-based. This often represents a hurdle: Processes often run through a fragmented and rigid system landscape and are therefore not yet sufficiently standardized.

From an organizational point of view, it is therefore helpful if there is a mandate for process improvement in the company - keywords "Continuous Improvement" or "Operational Excellence". Many companies set up a "Center of Excellence" to scale the technology and methodology across different application areas. From our point of view, this is a very promising way to get the processes under control.

Braitinger: The processes should be manual, repetitive and high-volume so that they can be automated using RPA. It should therefore be rule-based processes that have a low exception rate - i.e. hardly any deviations in different process flows - and whose input data is ideally electronically or machine-readable. But even with input data in paper form, there are now reliable methods of digitization and classification, which means that paper documents, for example, can also be processed. Before the processes are taken over by a software robot, they should be examined more closely and optimized if necessary. Because the software robot cannot fix any problems in the course of the processes, it can only do what it is instructed to do.

In general, the topic of communication is particularly important from an organizational point of view, since automation is still seen as a job killer. However, the aim is to support the workforce in their work. That's why it's important to get them on board, both during the implementation phase and in day-to-day work. Because the employees know their daily processes; they know where their "pain points" are and which tasks they can hand over. Companies should also offer training and education so that employees can continue their education in this area and even develop simple automations themselves. A UiPath study shows that the vast majority (93 percent) of companies are already prepared for training initiatives for employees.ITM: Keyword "integration into the existing IT landscape": How should the interfaces be designed so that there are no problems with long-term operation - e.g. in terms of performance, stability, scalability, security/data protection - are pre-programmed? Rameseder: RPA basically offers flexible application options and can be easily connected via interfaces in the source systems. This means that RPA can be used on almost any system and across industries. However, RPA automates the processes in the frontend - and RPA bots react very sensitively to adjustments in the user interface. This means: Especially with a continuous further development of the processes, bots always have to be maintained and updated at the same time.

To address this, we have developed a scalable automation platform that automates processes via APIs in the backend and is therefore much more stable than conventional platforms. We use AI and process data to not only automate repetitive tasks, but also entire process sections intelligently. Braitinger: More and more companies are moving to the cloud - this should also apply to RPA and automation. With comprehensive automation, security, stability and scalability are guaranteed - regardless of whether it is your own cloud in AWS or Azure or that of the solution provider. In general, RPA and IT teams need to work together. The IT department is responsible for the underlying technological infrastructure. However, she may need training and support to prepare and optimize this infrastructure for automation. In addition, it is important to consider software reuse in automation development and to understand how reused automation code works when applied to new systems and operating environments.

Additionally, because IT teams are knowledgeable about identity and access management, they can help create secure automations that also access systems and processes without compromising security. With a so-called “citizen developer” program, where employees create their own simple automations themselves, companies need to strengthen their governance and compliance. A comprehensive platform supports administrators, for example, in setting up specific guidelines for different user groups for the development of automations.