Robotic Process Automation – What’s that all about?


Robotic Process Automation – What’s that all about?

Robotic process automation or RPA is a hot topic in the business community these days. Why is it so? What is RPA? What are the benefits and pitfalls? How is it used? I shall answer these questions in a three-part blog series:

  1. In the first part, I’ll cover the basics of RPA
  2. In the second part, I will go through a real-world use case
  3. And in the third part, I will talk about implementations and prospects


The basics of RPA

Imagine a technology, that:

  • Lowers costs, incurred by manual work, up to 25-40%;
  • Raises the efficiency of service and reduces the amount of human error;
  • Realizes a return on investment in one year;
  • Does all the above-mentioned without a need to major changes to the current IT landscape.

RPA is this kind of quite disruptive technology.


”Robotic process, what…?”

When using RPA, organizations can configure a software (robot) solution to gather and interpret data from different systems. Based on this data the robot can make transactions, modify the data or communicate with other digital systems. In short the robot does the same tasks as a human user would do. But, if configured correctly, 0a robot does the task faster than a human and basically without errors.

This kind of User Interface (UI) interaction is quite different from traditional IT-integration where API’s are used.

Many organisations still use human labour in big scale general knowledge process work. People do large volumes of transactional work in a system environment. This means copying data by hand from one desktop application to another. In this kind of situation organisations can reach significant results with RPA, because many manual and time consuming tasks can be automated.


”But, everything should be cognitive these days? Is it cognitive?”

Cognitive technologies such as IBM Watson and RPA have been separate from each other, mainly because they are used in such different roles. This will probably change, though. According to Deloitte, RPA suppliers have been adding features such as natural language identification and machine learning to their products. Forrester suggests that RPA should be implemented in a way that a possible link to a cognitive platform can be added easily in the future.

There is also talk about ”teaching” robots. One should separate AI and teaching RPA robots from each other. Former is rather complicated, whereas the latter is just teaching a robot which button to click when using an application.

”Ok, this is all cool, but what do I get from this all?”

Industrial robots revolutionized the production of physical goods by increasing production numbers and improving quality. RPA robots on the other hand are revolutionizing administrative business process, IT support processes, workflows and back office functions. RPA dramatically improves accuracy and throughput time of business processes. At the same time the productivity of work improves. Also, the work satisfaction of personnel is improved as the amount of dull and repetitive tasks decreases.

In process automation, RPA can mimic manual rule based, non-subjective processes, without jeopardizing the current IT-architecture. Robots can execute predetermined functions constantly. This kind of work is easily scalable to respond demand. If we think about back office functions, there is a wealth of potential use cases in for example finance, logistics, customer service and HR.

Think about a back office where robots, using multiple systems simultaneously, are working fully automated. For example, robots can pick up tasks from a queue and complete them using multiple applications.

Excited? Please keep in mind though that one cannot reach full process automation with RPA. One can only automate the UI interaction that was previously done by a human.

Ideal processes for RPA use cases are processes that:

  • Are repetitive;
  • Have access to structured data;
  • Run on windows or web based platforms;
  • Are well documented and standardized;
  • Require the effort of three or more employees;
  • Include entering data, which is prone to human error.


”All that glitters isn’t gold”

I guess you are stoked at this point about the prospects of RPA? It has potential, yes. But it is not a silver bullet that solves all your business process problems. When used properly, RPA is an excellent tactical tool for a specific business process problem, such as using several legacy systems simultaneously in a single process. RPA is a surface level solution that brings new life to existing business processes. But it does not solve any big problems that lie beneath the surface.


”So, what should I make from all this?”

I would say that RPA is an excellent solution when it is used in the right place at the right time. It is an essential tool in the automation tool box of a modern organization.

On the next blog, I shall look at a RPA use case from every day practice. If this blog resonated to you, please follow You-Get in Twitter and Linkedin, so you will not miss the next part of this blog.


Veli-Pekka Mustonen

BPM Consultant, Business Analyst You-Get Finland Oy