Home > Articles

What is Robotic Process Automation

What is Robotic Process Automation (RPA)?

If you read the words “robotic” and “process automation” in the title, you must have thought “this must be good.” And you’re right. Still, this is not exactly what you may be thinking. This isn’t about creating cool looking sci-fi robots that are able to bring industrial automation to a new level. This is about process automation, that is, it automates the processes that we tend to do manually in a computer.

That’s right, Robotic Process Automation is about software, taking advantage of the latest developments in artificial intelligence in order to make our lives easier and our jobs more productive. Check it out!

Automating repetitive work

Do you know the word “macro”? It’s an IT term that has been around for decades now and is used to refer to a way of programming your computer to automatically do some simple repetitive tasks, such as replacing a word with another word or even paragraph, opening multiple programs with the press of a button, or even just keep refreshing a web page to keep track of updates.

Still, creating a macro was something done more informally. One strategy used was to use a simple program that could record button presses and then make it repeat the sequence indefinitely. Still, it depended on your ability to create a sequence that would loop well, as the program just repeated the recording sequentially, with no regard to continuity.

RPA takes macros and brings them to the entrepreneurial level. Instead of just being a fancy tool for making some stuff easier, it is now a fully-fledged company-wide automation tool made to take employees out of repetitive, bureaucratic stuff and put them to do more interesting work, while the machines do the dull job.

Machine learning in your work computer

Well, what if some process seems like it can be automated but also involves some kind of decision-making process? Even if it’s something simple, you can’t just automate it using something like a macro, as it isn’t something repetitive that can be done to infinity. However, what if there was a way to put some artificial intelligence in your computer to do this too?

The potential for innovation goes even higher when you add in a bit of machine learning. Machine learning, computer vision, among other AI technologies can add in some simple decision making, by keeping track of the decisions you make while recording the process you want to automate, and also bring the possibility of making it even more efficient if possible.

Using RPA with machine learning also brings a new “recording” process. While with macros and similar technologies you would record your process once and then repeat it whenever needed (or just leave it in a loop), using machine learning means that you will need to repeat the process manually many times and in many different scenarios and situations in order to train the AI correctly.

Difficulties involved

As you may have noticed, while this technology can make processes much more efficient and faster in a company, adopting it and adapting it to the company-wide workflow can be very difficult.

First, there is a lot of testing involved. People that have experience with the technology have to team up with the people currently responsible for the processes to be automated in order to record the key presses and see if they are being repeated correctly and for the correct events.

Even with machine learning this doesn’t get much easier. While you can just leave it to be trained while the employee does the processes as usual, it is hard to know when exactly it has received enough training. You have to take it and put it through some tests in order to see if it’s good enough already. And those tests may also not be enough, as the AI may get confused if it faces new situations or edge cases.

Still, those problems are bound to be solved in time, and newer companies will probably start to take advantage of RPA from the beginning, meaning they will need less employees at the start while being able to be just as productive as other companies.