Robotic Process Automation


Robotic Process Automation (RPA) is a a type of software that interacts with the computer applications through the use of User Interfaces. The ‘robot’ can capture and interact with the computer applications for processing a transaction, manipulating data, triggering responses and communicating with other digital systems. RPA works at the presentation layer (user interface) and mimics a user.

It can partially or fully automate human activities which are manual, repetitive and rule based. RPA implementation best suits processes which are less complex, volatile and have a high headcount. The robots work alongside human workers creating a digital workforce. They can work 24/7 since they don’t require holidays or need to take leave.

Let us understand the Robotic Process Automation (RPA) and Smart Process Automation (SPA)

RPA must be programmed to perform a task.

SPA based on Artificial Intelligence can train itself to perform complex and subjective work through pattern recognition, process natural language and unstructured data. SPA can adapt and learns the new way as the environments change. Programmers don’t encode computers with instructions.  AI is machine learning and it is like a black box. It can capture information through vision recognition, sound recognition, search (structured and unstructured data), do data analysis (identifying clusters of behavior in data) and can turn that information into something useful through natural language processing (extracting meaningful data from text), reasoning (should it act based on the information) and prediction (predicting behavior). It may also evolve to understand why something is happening………may be in the near future.

The automation has come a long way from macros to Robotic Process Automation which can perform repetitive tasks on the user interface without much of cognitive intelligence. The artificial intelligence (ML) based cognitive automation to perform more complex work is evolving and we are beginning to hear vendors providing this. The real benefit will be seen when the machines learn based on historical data and real time human action. This will lead to very small percentage of work relating to exceptions which cannot be handled by cognitive automation in a repetitive process which will be managed by humans. This is going to change the landscape of outsourcing though it is not impacted much currently as RPA is in the initial phase.

The move is therefore to make the outsourcing intelligent – both from technology and the high knowledge based. This is going to be taking the finance function to the next level with the humans being used for more thinking work rather than repetitive work. Your digital workforce is going to be more smarter than from just programmed robots. Be prepared to adapt to the new environment and collaborate with the digital worker.