Robotic Process Automation

What RPA isn't

Have you ever worked on a task that includes excessive Excel clicking, multiple copy-pasting, or migrating data from one application to another – merely fingertips and a strong will at your disposal?

We bet you have.

And we bet you figured that there must be a better way. A way that doesn’t rely solely on you.

Well, there is.

Robotic Process Automation (or RPA for friends) is a new-age technology that does this kind of boring job for you.

But only to a certain extent.

 

So, what isn't RPA?

Sometimes, to better understand the scope of a term, it’s important to know what it doesn’t cover. So, let’s start with the most common misconception.

RPA isn’t an artificial intelligence

Software robots cannot think or recognize voices. They cannot make decisions on their own. They cannot self-learn or become smarter in any way.

The best they can do is follow clear rules and predefined structures. Which is a good thing because they can’t mess up. They act under the instructions they get from their human friends. Nothing more. And nothing less.

Their motto is: “Show me how to do the task once – and I’ll do it the same way every time until you tell me otherwise.”

Anyhow, Artificial Intelligence can be added to Robotic Automation. In that case, we are talking about Intelligent Process Automation (IPA) which includes computer vision, cognitive automation, and/or machine learning.

More about IPA’s superpowers, read here.

RPA isn’t a physical robot

Yes, there is “robot” in “robotic process automation” but in a software, not hardware way. The only space software robots take up and their home is a functional windows machine – the only precondition to their existence.  

Walking and talking robots call for entirely different know-how. RPA (Robotic Process Automation) isn’t part of IoT (the Internet of Things). Since we know nothing about creating physical robots, we are going back to other things software robots aren’t in 3, 2…

RPA isn’t BPM, ERP, or process optimization

Business Process Management is a holistic approach to optimizing and automating business processes from start to finish. In the meantime, RPA is a technology that deals with smaller, repetitive tasks and a tool that extends and enhances BPM’s essential purpose. Altogether, BPM covers strategy and tactics; RPA supports execution and operations.

Enterprise Resource Planning unites business processes and it does automate them. However, RPA can improve those same processes involving ERP and add speed to the enterprise architecture. More precisely, RPA is a perfect tool to assist as the middleman between legacy software and ERP applications.

Process Optimization isn’t directly linked to RPA but is rather the main idea and goal behind developing software robots. It is a discipline of adapting processes to perfect their features. Before developing software robots, it sure helps to optimize a process. Then the process can be automated most efficiently.

Although Robotic Process Automation, Business Process Automation and Process Optimization have some overlap, it's important to remember that they are not synonymous.

This infographic might help you understand the differences.

 

RPA isn’t a traditional IT

RPA executes a process that is already in place, replacing human beings working in the existing infrastructure. Because of the non-invasive characteristic of software robots, they don’t disturb any business logic that is governed by existing applications.

Due to Leslie from McKinsey, when organizations consider proof of concept for RPA, they look at the business case and compare it to an IT solution. Often that’s pretty inconvenient for IT. In one case, the return on investment for RPA was about 200% in the first year and they could implement it within 3 months. The IT solution did the same thing but with a 3-year payback period and it was going to take 9 months to implement.

 

And what RPA can’t do?

Prerequisites and limitations are part of every technology. RPA doesn’t have many, but there are still some you need to consider before getting software robots.

RPA can’t make judgment-based decisions

Standardization. Structure. Stability. Three S-words that need to be true to a certain level for a process to prove its automation suitability.

Uncertain decision reasoning. No clear process steps. Unreadable electronic input types. Three things that are a dealbreaker for RPA.

As you already know, RPA isn’t intelligent in any aspect, and it can only handle rule-based processes that have standard inputs. Automating anything other than basic if-then decision-making branches is out of their scope.

RPA can’t be profitable with infrequent processes

Automating a simple structured process that rarely happens is a bad idea. Automating a complex and exhausting process that rarely happens is also a bad idea, with bad meaning it's benefits won't be worth the money invested in it. Simply, automating any kind of process that rarely happens can’t be profitable.

So, to cover development costs and to get the most out of RPA, search for the repetitive processes with a high volume of transactions and long average process duration. They will bring you the ROI you look for.

RPA can’t become smarter overnight

To achieve intelligent process automation, developers should join forces with data scientists to smarten up the robots. Luckily, bonsai.tech is a symbiosis of both types of people – RPA developers who can develop software robots and artificial intelligence experts who can make them smart. Sounds like a simple say-please-and-we-will-deliver situation to us.

RPA can’t even become RPA overnight

The more complex the process, the more time for the robot’s development is needed. Some processes are simple and can be automated in a month’s notice. But more often, business processes are complex, with many if-then branches taking place in specific applications. And automating them simply requires more time.

 

RPA process assessment

How to know which processes shouldn’t be automated? Easy. With the right tool.

This simple excel file does the trick.

Use it to detect unfeasible processes, processes that can be automated only at a high cost, or processes that will bring unsatisfying results. Of course, in the end, you may decide that you are willing to sacrifice one thing for the sake of others (e.g. paying more now will bring higher savings later).

Get relevant information before you make such a call.  

Fill in the process assessment template.

Comments are closed