Don't be a robot, give it to a robot (RPA)

RPA is an acronym for words Robotic Process Automation and, as the name suggests, is an automation method. In other words, it is one of the latest tools for automation, especially routine processes, and actions, which take unnecessary time and we do them over and over again..

Surely you have met before, for example with a macro in Excel, which has facilitated (or probably facilitates) your work. Whether by some automatic formatting, saving / backing up files or updating data, etc.

And now imagine that you have a tool that in most tasks can simulate (repeat) the same behaviour as the user does. From the click itself, through copying texts / files, opening programs, logging in / out, to manipulating data in databases. In short, almost anything a user does on their computer.

What is RPA?

It is a robot but be careful. You cannot touch this robot! It is "just" a set of commands, source code. And do not confuse this robot with artificial intelligence. In RPA, the robot does exactly what it is programmed to do. Neither less nor more. Therefore, if he gets into a situation for which he is not prepared and in which the user needs to decide how to proceed, the robot "falls" into error, because it does not have precise instructions on how to proceed.

This is good to know because it can be an advantage as well as a disadvantage. But about that later…

What is such a robot suitable for?

In general, it is a more efficient, cheaper, and more accurate execution of routine processesthat take people unnecessarily time. These processes usually do not require human thinking. This "kills" the main advantage over robots. As I indicated above. In RPA, the robot cannot "think" because it does not contain any AI (Artificial Intelligence).. At least for now…

Robotic Process Automation is ideal in situations where we have a process that has clear rules, repeats itself and, above all, have structured data that it works with.That is crucial. In order for the robot to work, it still needs data in the same structure at the input. Although this will probably change over time when AI will be part of RPA. But about that, maybe another time…

When to use a robot and when not?

There is one equivalent. It is called FTE (Full-time equivalent) and it determines the workload of one worker. Why am I writing about it? Because this is one of the key criteria for whether a robot process to develop or not.

If I take it from a broader point of view, the robot only makes sense if it pays off financially (for example, by saving FTE). Then, of course, there are many other factors, such as whether the effort involved in developing, testing, and deploying the robot is acceptable. In other words, whether it pays for us to spend time and employee’s capacity on it. This is a demanding process, from detailed process analysis, through evaluation (or minor optimization) to development and testing. If all goes well, the pilot operation itself and the initial checks of the robot's behaviour in the production environment (i.e., "sharply" in real systems) is a matter of tens of hours. I am not even talking about tens (hundreds) of hours before.

Here, anyone who knows at least a little bit about how things work in a corporate will argue that it is super fast. Normally, development and change are done on a quarterly cycle, so a few tens of hours is a great time and ultra-fast implementation.

And I totally agree with that. It is full of processes that do not deserve a comprehensive system solution in the infrastructure but take unnecessarily a lot of time and also often "bother" those workers. RPA in these cases offers a relatively quick solution.

Why should we use a robot?

It is necessary to add that if you want to do a robot yourself, which will answer emails instead of you, for example, then you will most likely be unlucky. The largest RPA tools(see the article to be published later) offer their solutions, especially to corporate clients. So, if you do not have an unnecessary few hundred thousand to buy licenses, you have to wait.

However, there are two types of robots. The so-called Front-office (FO) and Back-office (BO))… The name itself suggests how they differ. One (Front-office) works with the user. The user launches it himself. Whether to do something for him or to help him in a given situation. For example, when you need to get some client information from several different systems at the same time. While the other(Back-office) is a "substitute" for a human worker. He is actually a virtual colleague who works completely alone and without supervision.And that is magic, because it is the tool that makes boring and tedious work easier for human workers and creates space for more interesting tasks.

I immediately remembered the first time I started the BO robot. I sat in front of the monitor and watched the robot work. Alone… No breaks… No chatter or delays…. One case after case…

Then I just closed my laptop and went home. I came in the morning and our newest "colleague" was done. The work, which originally burdened 5 people, tens of hours, the robot had done in a few hours. And colleagues suddenly had time to do more necessary tasks.

This may be the answer to the question "why should we use a robot".