A look into the (fascinating) world of RPA software

Within an organisation, employees often spend many hours doing repetitive work. Think, for example, of copying and pasting from one system into another. Or checking whether all of a customer's details match their application form. How many times has something been overlooked? And have you once again had to figure out where it went wrong? Robotic Process Automation (RPA) is the lifesaver here: it takes over repetitive tasks and works without error too. Wondering how to envisage it? We asked Solution Developer Willem-Jan.

"It is as if there is an invisible robot inside your computer that takes over all the checking and copy-paste work from you."

From Repetitive to Truly Human Work with RPA

Very simply, RPA boils down to using software to automate common, repetitive actions. An example here is scanning a PDF file.

''Suppose a company receives invoices from contracted employees in PDF files via email. Using RPA software, you can pick up every incoming invoice email. Using a checklist, the robot software checks whether the sender is known. It then reads the PDF, then uses the extracted data to check whether, for example, the employee name and IBAN are correct,'' explains Willem-Jan during the interview.

''If the details are correct, the payment can be prepared in the payment system. An employee then only has to approve the payment. It is as if there is an invisible robot in your computer that takes over all the checking and copy-paste work," says Willem-Jan. ''The robot also recognises errors in the PDF files. For example, if there is a different account number that cannot be linked to the name of the person in the document. The software then transfers the incorrect PDF file to an employee, who then has to check the file and adjust it.''

Sound complicated? Watch our colleague Ernst's video series where he explains exactly how RPA software works.

’’The RPA software recognises errors, for example if data in the PDF file is incorrect. It then transfers the document to an employee to correct the data.’’

More is possible with RPA from UiPath

UiPath is one of our RPA software vendors. In their software, they combine several robotisation techniques such as artificial intelligence (AI), which offers more application opportunities. AI is a smart technology that involves computers applying human intelligence to perform tasks independently. UiPath develops these programs to automate repetitive and boring digital tasks that would normally be performed by a human. For instance, the program includes image recognition, natural language processing, and learning. As a result, this software makes it possible to perform more complex tasks. This advanced software can, for example, interpret texts, participate in chats or conversations, and understand unstructured data.

A simple example is a chatbot with a bit of intelligence, in which case a conversation with it is like a human conversation. A restaurant's chatbot requests all the information from the customer. A link with RPA checks availability and the reservation is confirmed.

Robotics and image recognition also work together

First, a brief reminder of what image recognition is. Image recognition (IR) interprets useful data from pixels, such as photos or videos. If you then use this data to take action automatically, this is called computer vision. Computer vision is an area of artificial intelligence (AI) that trains computers to interpret and understand the visual world. In addition, it is also possible to identify people, objects, text, and places by means of algorithms and machine learning.

Image recognition and robotisation, such as AI and RPA, are increasingly being combined. The aim of this is to extract important information from imagery. Which you, or the robot (automatically) can anticipate.

An example of robotisation in a printing house

First, employees manually checked the products in a box and then stuck the correct order information on it. They then had to enter these orders into the system one by one. A time-consuming task. Now a camera replaces the employee. Through image recognition, it recognises the product in the box. AI then automatically ensures that the correct order information is stuck on the box and that the order enters the system. RPA then ensures that everything is properly registered in all systems. This eliminates the need for an employee to check and enter orders, giving them more time for other important work. An increased production process is created with the same manpower.

''In an automated warehouse, you use robots and automated warehouse systems instead of people.''

Despite the fact that it is fully automated, the human touch is still required. Humans have to check that everything is going smoothly. When there are anomalous or special cases, as mentioned in the example of scanning the PDF file, an employee will be involved. Therefore, full automation will never happen, and humans will always be relevant.

How do we envisage it?

We are increasingly applying robotisation in everyday life, in fact more often than you think. Think of scanning the QR code of the Corona Check app, or sending an automatic reply email when you are on vacation. We use it not to replace us, but to support us. It takes over monitoring and repetitive tasks from us, leaving us more time for real human work. For example, more time for personal contact with our customers, or for thinking strategically. These techniques can even help us gain insight into data, and we can then use this insight to develop our strategic approach.

Despite the fact that these technologies are still developing, the future of robotisation looks bright.